- Method `get_remote_file_content` is object's method now

Signed-off-by: Aditya Bisoi <abisoi@redhat.com>

(cherry picked from commit 4c7611291d (centos_master))

MANIFEST.in

@@ -2,6 +2,7 @@ include AUTHORS
 include COPYING
 include GPL
 include pungi.spec
+include setup.cfg
 include tox.ini
 include share/*
 include share/multilib/*

contrib/tmpfiles.d/pungi-clean-cache.conf

@@ -0,0 +1,2 @@
+# Clean up pungi cache
+d /var/cache/pungi/createrepo_c/ - - - 30d

doc/conf.py

@@ -18,12 +18,12 @@ import os
 # If extensions (or modules to document with autodoc) are in another directory,
 # add these directories to sys.path here. If the directory is relative to the
 # documentation root, use os.path.abspath to make it absolute, like shown here.
-#sys.path.insert(0, os.path.abspath('.'))
+# sys.path.insert(0, os.path.abspath('.'))
 
 # -- General configuration ------------------------------------------------
 
 # If your documentation needs a minimal Sphinx version, state it here.
-#needs_sphinx = '1.0'
+# needs_sphinx = '1.0'
 
 # Add any Sphinx extension module names here, as strings. They can be
 # extensions coming with Sphinx (named 'sphinx.ext.*') or your custom
@@ -31,207 +31,201 @@ import os
 extensions = []
 
 # Add any paths that contain templates here, relative to this directory.
-templates_path = ['_templates']
+templates_path = ["_templates"]
 
 # The suffix of source filenames.
-source_suffix = '.rst'
+source_suffix = ".rst"
 
 # The encoding of source files.
-#source_encoding = 'utf-8-sig'
+# source_encoding = 'utf-8-sig'
 
 # The master toctree document.
-master_doc = 'index'
+master_doc = "index"
 
 # General information about the project.
-project = u'Pungi'
-copyright = u'2016, Red Hat, Inc.'
+project = "Pungi"
+copyright = "2016, Red Hat, Inc."
 
 # The version info for the project you're documenting, acts as replacement for
 # |version| and |release|, also used in various other places throughout the
 # built documents.
 #
 # The short X.Y version.
-version = '4.3'
+version = "4.5"
 # The full version, including alpha/beta/rc tags.
-release = '4.3.6'
+release = "4.5.0"
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.
-#language = None
+# language = None
 
 # There are two options for replacing |today|: either, you set today to some
 # non-false value, then it is used:
-#today = ''
+# today = ''
 # Else, today_fmt is used as the format for a strftime call.
-#today_fmt = '%B %d, %Y'
+# today_fmt = '%B %d, %Y'
 
 # List of patterns, relative to source directory, that match files and
 # directories to ignore when looking for source files.
-exclude_patterns = ['_build']
+exclude_patterns = ["_build"]
 
 # The reST default role (used for this markup: `text`) to use for all
 # documents.
-#default_role = None
+# default_role = None
 
 # If true, '()' will be appended to :func: etc. cross-reference text.
-#add_function_parentheses = True
+# add_function_parentheses = True
 
 # If true, the current module name will be prepended to all description
 # unit titles (such as .. function::).
-#add_module_names = True
+# add_module_names = True
 
 # If true, sectionauthor and moduleauthor directives will be shown in the
 # output. They are ignored by default.
-#show_authors = False
+# show_authors = False
 
 # The name of the Pygments (syntax highlighting) style to use.
-pygments_style = 'sphinx'
+pygments_style = "sphinx"
 
 # A list of ignored prefixes for module index sorting.
-#modindex_common_prefix = []
+# modindex_common_prefix = []
 
 # If true, keep warnings as "system message" paragraphs in the built documents.
-#keep_warnings = False
+# keep_warnings = False
 
 
 # -- Options for HTML output ----------------------------------------------
 
 # The theme to use for HTML and HTML Help pages.  See the documentation for
 # a list of builtin themes.
-html_theme = 'default'
+html_theme = "default"
 
 # Theme options are theme-specific and customize the look and feel of a theme
 # further.  For a list of options available for each theme, see the
 # documentation.
-#html_theme_options = {}
+# html_theme_options = {}
 
 # Add any paths that contain custom themes here, relative to this directory.
-#html_theme_path = []
+# html_theme_path = []
 
 # The name for this set of Sphinx documents.  If None, it defaults to
 # "<project> v<release> documentation".
-#html_title = None
+# html_title = None
 
 # A shorter title for the navigation bar.  Default is the same as html_title.
-#html_short_title = None
+# html_short_title = None
 
 # The name of an image file (relative to this directory) to place at the top
 # of the sidebar.
-#html_logo = None
+# html_logo = None
 
 # The name of an image file (within the static path) to use as favicon of the
 # docs.  This file should be a Windows icon file (.ico) being 16x16 or 32x32
 # pixels large.
-#html_favicon = None
+# html_favicon = None
 
 # Add any paths that contain custom static files (such as style sheets) here,
 # relative to this directory. They are copied after the builtin static files,
 # so a file named "default.css" will overwrite the builtin "default.css".
-html_static_path = ['_static']
+html_static_path = ["_static"]
 
 # Add any extra paths that contain custom files (such as robots.txt or
 # .htaccess) here, relative to this directory. These files are copied
 # directly to the root of the documentation.
-#html_extra_path = []
+# html_extra_path = []
 
 # If not '', a 'Last updated on:' timestamp is inserted at every page bottom,
 # using the given strftime format.
-#html_last_updated_fmt = '%b %d, %Y'
+# html_last_updated_fmt = '%b %d, %Y'
 
 # If true, SmartyPants will be used to convert quotes and dashes to
 # typographically correct entities.
-#html_use_smartypants = True
+# html_use_smartypants = True
 
 # Custom sidebar templates, maps document names to template names.
-#html_sidebars = {}
+# html_sidebars = {}
 
 # Additional templates that should be rendered to pages, maps page names to
 # template names.
-#html_additional_pages = {}
+# html_additional_pages = {}
 
 # If false, no module index is generated.
-#html_domain_indices = True
+# html_domain_indices = True
 
 # If false, no index is generated.
-#html_use_index = True
+# html_use_index = True
 
 # If true, the index is split into individual pages for each letter.
-#html_split_index = False
+# html_split_index = False
 
 # If true, links to the reST sources are added to the pages.
-#html_show_sourcelink = True
+# html_show_sourcelink = True
 
 # If true, "Created using Sphinx" is shown in the HTML footer. Default is True.
-#html_show_sphinx = True
+# html_show_sphinx = True
 
 # If true, "(C) Copyright ..." is shown in the HTML footer. Default is True.
-#html_show_copyright = True
+# html_show_copyright = True
 
 # If true, an OpenSearch description file will be output, and all pages will
 # contain a <link> tag referring to it.  The value of this option must be the
 # base URL from which the finished HTML is served.
-#html_use_opensearch = ''
+# html_use_opensearch = ''
 
 # This is the file name suffix for HTML files (e.g. ".xhtml").
-#html_file_suffix = None
+# html_file_suffix = None
 
 # Output file base name for HTML help builder.
-htmlhelp_basename = 'Pungidoc'
+htmlhelp_basename = "Pungidoc"
 
 
 # -- Options for LaTeX output ---------------------------------------------
 
 latex_elements = {
-# The paper size ('letterpaper' or 'a4paper').
-#'papersize': 'letterpaper',
-
-# The font size ('10pt', '11pt' or '12pt').
-#'pointsize': '10pt',
-
-# Additional stuff for the LaTeX preamble.
-#'preamble': '',
+    # The paper size ('letterpaper' or 'a4paper').
+    #'papersize': 'letterpaper',
+    # The font size ('10pt', '11pt' or '12pt').
+    #'pointsize': '10pt',
+    # Additional stuff for the LaTeX preamble.
+    #'preamble': '',
 }
 
 # Grouping the document tree into LaTeX files. List of tuples
 # (source start file, target name, title,
 #  author, documentclass [howto, manual, or own class]).
 latex_documents = [
-  ('index', 'Pungi.tex', u'Pungi Documentation',
-   u'Daniel Mach', 'manual'),
+    ("index", "Pungi.tex", "Pungi Documentation", "Daniel Mach", "manual"),
 ]
 
 # The name of an image file (relative to this directory) to place at the top of
 # the title page.
-#latex_logo = None
+# latex_logo = None
 
 # For "manual" documents, if this is true, then toplevel headings are parts,
 # not chapters.
-#latex_use_parts = False
+# latex_use_parts = False
 
 # If true, show page references after internal links.
-#latex_show_pagerefs = False
+# latex_show_pagerefs = False
 
 # If true, show URL addresses after external links.
-#latex_show_urls = False
+# latex_show_urls = False
 
 # Documents to append as an appendix to all manuals.
-#latex_appendices = []
+# latex_appendices = []
 
 # If false, no module index is generated.
-#latex_domain_indices = True
+# latex_domain_indices = True
 
 
 # -- Options for manual page output ---------------------------------------
 
 # One entry per manual page. List of tuples
 # (source start file, name, description, authors, manual section).
-man_pages = [
-    ('index', 'pungi', u'Pungi Documentation',
-     [u'Daniel Mach'], 1)
-]
+man_pages = [("index", "pungi", "Pungi Documentation", ["Daniel Mach"], 1)]
 
 # If true, show URL addresses after external links.
-#man_show_urls = False
+# man_show_urls = False
 
 
 # -- Options for Texinfo output -------------------------------------------
@@ -240,19 +234,25 @@ man_pages = [
 # (source start file, target name, title, author,
 #  dir menu entry, description, category)
 texinfo_documents = [
-  ('index', 'Pungi', u'Pungi Documentation',
-   u'Daniel Mach', 'Pungi', 'One line description of project.',
-   'Miscellaneous'),
+    (
+        "index",
+        "Pungi",
+        "Pungi Documentation",
+        "Daniel Mach",
+        "Pungi",
+        "One line description of project.",
+        "Miscellaneous",
+    ),
 ]
 
 # Documents to append as an appendix to all manuals.
-#texinfo_appendices = []
+# texinfo_appendices = []
 
 # If false, no module index is generated.
-#texinfo_domain_indices = True
+# texinfo_domain_indices = True
 
 # How to display URL addresses: 'footnote', 'no', or 'inline'.
-#texinfo_show_urls = 'footnote'
+# texinfo_show_urls = 'footnote'
 
 # If true, do not generate a @detailmenu in the "Top" node's menu.
-#texinfo_no_detailmenu = False
+# texinfo_no_detailmenu = False

doc/configuration.rst

@@ -194,6 +194,17 @@ Options
     Tracking Service Kerberos authentication. If not defined, the default
     Kerberos principal is used.
 
+**cts_oidc_token_url**
+    (*str*) -- URL to the OIDC token endpoint.
+    For example ``https://oidc.example.com/openid-connect/token``.
+    This option can be overridden by the environment variable ``CTS_OIDC_TOKEN_URL``.
+
+**cts_oidc_client_id*
+    (*str*) -- OIDC client ID.
+    This option can be overridden by the environment variable ``CTS_OIDC_CLIENT_ID``.
+    Note that environment variable ``CTS_OIDC_CLIENT_SECRET`` must be configured with
+    corresponding client secret to authenticate to CTS via OIDC.
+
 **compose_type**
     (*str*) -- Allows to set default compose type. Type set via a command-line
     option overwrites this.
@@ -581,6 +592,16 @@ Options
     with everything. Set this option to ``False`` to ignore ``noarch`` in
     ``ExclusiveArch`` and always consider only binary architectures.
 
+**pkgset_inherit_exclusive_arch_to_noarch** = True
+    (*bool*) -- When set to ``True``, the value of ``ExclusiveArch`` or
+    ``ExcludeArch`` will be copied from source rpm to all its noarch packages.
+    That will than limit which architectures the noarch packages can be
+    included in.
+
+    By setting this option to ``False`` this step is skipped, and noarch
+    packages will by default land in all architectures. They can still be
+    excluded by listing them in a relevant section of ``filter_packages``.
+
 **pkgset_allow_reuse** = True
     (*bool*) -- When set to ``True``, *Pungi* will try to reuse pkgset data
     from the old composes specified by ``--old-composes``. When enabled, this
@@ -920,6 +941,10 @@ Options
     comps file can not be found in the package set. When disabled (the
     default), such cases are still reported as warnings in the log.
 
+    With ``dnf`` gather backend, this option will abort the compose on any
+    missing package no matter if it's listed in comps, ``additional_packages``
+    or prepopulate file.
+
 **gather_source_mapping**
     (*str*) -- JSON mapping with initial packages for the compose. The value
     should be a path to JSON file with following mapping: ``{variant: {arch:
@@ -1607,8 +1632,23 @@ OSBuild Composer for building images
     * ``release`` -- release part of the final NVR. If neither this option nor
       the global ``osbuild_release`` is set, Koji will automatically generate a
       value.
-    * ``repo`` -- a list of repository URLs from which to consume packages for
+    * ``repo`` -- a list of repositories from which to consume packages for
       building the image. By default only the variant repository is used.
+      The list items may use one of the following formats:
+
+      * String with just the repository URL.
+
+      * Dictionary with the following keys:
+
+        * ``baseurl`` -- URL of the repository.
+        * ``package_sets`` -- a list of package set names to use for this
+            repository. Package sets are an internal concept of Image Builder
+            and are used in image definitions. If specified, the repository is
+            used by Image Builder only for the pipeline with the same name.
+            For example, specifying the ``build`` package set name will make
+            the repository to be used only for the build environment in which
+            the image will be built. (optional)
+
     * ``arches`` -- list of architectures for which to build the image. By
       default, the variant arches are used. This option can only restrict it,
       not add a new one.
@@ -1641,13 +1681,13 @@ OSBuild Composer for building images
         * ``tenant_id`` -- Azure tenant ID to upload the image to
         * ``subscription_id`` -- Azure subscription ID to upload the image to
         * ``resource_group`` -- Azure resource group to upload the image to
-        * ``location`` -- Azure location to upload the image to
+        * ``location`` -- Azure location of the resource group (optional)
         * ``image_name`` -- Image name of the uploaded Azure image (optional)
 
       * **GCP upload options** -- upload to Google Cloud Platform.
 
         * ``region`` -- GCP region to upload the image to
-        * ``bucket`` -- GCP bucket to upload the image to
+        * ``bucket`` -- GCP bucket to upload the image to (optional)
         * ``share_with_accounts`` -- list of GCP accounts to share the image
           with
         * ``image_name`` -- Image name of the uploaded GCP image (optional)
@@ -1764,6 +1804,8 @@ repository with a new commit.
     * ``tag_ref`` -- (*bool*, default ``True``) If set to ``False``, a git
       reference will not be created.
     * ``ostree_ref`` -- (*str*) To override value ``ref`` from ``treefile``.
+    * ``runroot_packages`` -- (*list*) A list of additional package names to be
+      installed in the runroot environment in Koji.
 
 Example config
 --------------

doc/index.rst

@@ -19,7 +19,7 @@ Contents:
     scm_support
     messaging
     gathering
+    koji
     comps
     contributing
     testing
-    multi_compose

doc/koji.rst

@@ -0,0 +1,105 @@
+======================
+Getting data from koji
+======================
+
+When Pungi is configured to get packages from a Koji tag, it somehow needs to
+access the actual RPM files.
+
+Historically, this required the storage used by Koji to be directly available
+on the host where Pungi was running. This was usually achieved by using NFS for
+the Koji volume, and mounting it on the compose host.
+
+The compose could be created directly on the same volume. In such case the
+packages would be hardlinked, significantly reducing space consumption.
+
+The compose could also be created on a different storage, in which case the
+packages would either need to be copied over or symlinked. Using symlinks
+requires that anything that accesses the compose (e.g. a download server) would
+also need to mount the Koji volume in the same location.
+
+There is also a risk with symlinks that the package in Koji can change (due to
+being resigned for example), which would invalidate composes linking to it.
+
+
+Using Koji without direct mount
+===============================
+
+It is possible now to run a compose from a Koji tag without direct access to
+Koji storage.
+
+Pungi can download the packages over HTTP protocol, store them in a local
+cache, and consume them from there.
+
+The local cache has similar structure to what is on the Koji volume.
+
+When Pungi needs some package, it has a path on Koji volume. It will replace
+the ``topdir`` with the cache location. If such file exists, it will be used.
+If it doesn't exist, it will be downloaded from Koji (by replacing the
+``topdir`` with ``topurl``).
+
+::
+
+    Koji path                            /mnt/koji/packages/foo/1/1.fc38/data/signed/abcdef/noarch/foo-1-1.fc38.noarch.rpm
+    Koji URL    https://kojipkgs.fedoraproject.org/packages/foo/1/1.fc38/data/signed/abcdef/noarch/foo-1-1.fc38.noarch.rpm
+    Local path                  /mnt/compose/cache/packages/foo/1/1.fc38/data/signed/abcdef/noarch/foo-1-1.fc38.noarch.rpm
+
+The packages can be hardlinked from this cache directory.
+
+
+Cleanup
+-------
+
+While the approach above allows each RPM to be downloaded only once, it will
+eventually result in the Koji volume being mirrored locally. Most of the
+packages will however no longer be needed.
+
+There is a script ``pungi-cache-cleanup`` that can help with that. It can find
+and remove files from the cache that are no longer needed.
+
+A file is no longer needed if it has a single link (meaning it is only in the
+cache, not in any compose), and it has mtime older than a given threshold.
+
+It doesn't make sense to delete files that are hardlinked in an existing
+compose as it would not save any space anyway.
+
+The mtime check is meant to preserve files that are downloaded but not actually
+used in a compose, like a subpackage that is not included in any variant. Every
+time its existence in the local cache is checked, the mtime is updated.
+
+
+Race conditions?
+----------------
+
+It should be safe to have multiple compose hosts share the same storage volume
+for generated composes and local cache.
+
+If a cache file is accessed and it exists, there's no risk of race condition.
+
+If two composes need the same file at the same time and it is not present yet,
+one of them will take a lock on it and start downloading. The other will wait
+until the download is finished.
+
+The lock is only valid for a set amount of time (5 minutes) to avoid issues
+where the downloading process is killed in a way that blocks it from releasing
+the lock.
+
+If the file is large and network slow, the limit may not be enough finish
+downloading. In that case the second process will steal the lock while the
+first process is still downloading. This will result in the same file being
+downloaded twice.
+
+When the first process finishes the download, it will put the file into the
+local cache location. When the second process finishes, it will atomically
+replace it, but since it's the same file it will be the same file.
+
+If the first compose already managed to hardlink the file before it gets
+replaced, there will be two copies of the file present locally.
+
+
+Integrity checking
+------------------
+
+There is minimal integrity checking. RPM packages belonging to real builds will
+be check to match the checksum provided by Koji hub.
+
+There is no checking for scratch builds or any images.

doc/multi_compose.rst

@@ -1,107 +0,0 @@
-.. _multi_compose:
-
-Managing compose from multiple parts
-====================================
-
-There may be cases where it makes sense to split a big compose into separate
-parts, but create a compose output that links all output into one familiar
-structure.
-
-The `pungi-orchestrate` tools allows that.
-
-It works with an INI-style configuration file. The ``[general]`` section
-contains information about identity of the main compose. Other sections define
-individual parts.
-
-The parts are scheduled to run in parallel, with the minimal amount of
-serialization. The final compose directory will contain hard-links to the
-files.
-
-
-General settings
-----------------
-
-**target**
-   Path to directory where the final compose should be created.
-**compose_type**
-   Type of compose to make.
-**release_name**
-   Name of the product for the final compose.
-**release_short**
-   Short name of the product for the final compose.
-**release_version**
-   Version of the product for the final compose.
-**release_type**
-   Type of the product for the final compose.
-**extra_args**
-   Additional arguments that will be passed to the child Pungi processes.
-**koji_profile**
-   If specified, a current event will be retrieved from the Koji instance and
-   used for all parts.
-
-**kerberos**
-   If set to yes, a kerberos ticket will be automatically created at the start.
-   Set keytab and principal as well.
-**kerberos_keytab**
-   Path to keytab file used to create the kerberos ticket.
-**kerberos_principal**
-   Kerberos principal for the ticket
-
-**pre_compose_script**
-   Commands to execute before first part is started. Can contain multiple
-   commands on separate lines.
-**post_compose_script**
-   Commands to execute after the last part finishes and final status is
-   updated. Can contain multiple commands on separate lines. ::
-
-      post_compose_script =
-          compose-latest-symlink $COMPOSE_PATH
-          custom-post-compose-script.sh
-
-   Multiple environment variables are defined for the scripts:
-
-    * ``COMPOSE_PATH``
-    * ``COMPOSE_ID``
-    * ``COMPOSE_DATE``
-    * ``COMPOSE_TYPE``
-    * ``COMPOSE_RESPIN``
-    * ``COMPOSE_LABEL``
-    * ``RELEASE_ID``
-    * ``RELEASE_NAME``
-    * ``RELEASE_SHORT``
-    * ``RELEASE_VERSION``
-    * ``RELEASE_TYPE``
-    * ``RELEASE_IS_LAYERED`` – ``YES`` for layered products, empty otherwise
-    * ``BASE_PRODUCT_NAME`` – only set for layered products
-    * ``BASE_PRODUCT_SHORT`` – only set for layered products
-    * ``BASE_PRODUCT_VERSION`` – only set for layered products
-    * ``BASE_PRODUCT_TYPE`` – only set for layered products
-
-**notification_script**
-   Executable name (or path to a script) that will be used to send a message
-   once the compose is finished. In order for a valid URL to be included in the
-   message, at least one part must configure path translation that would apply
-   to location of main compose.
-
-   Only two messages will be sent, one for start and one for finish (either
-   successful or not).
-
-
-Partial compose settings
-------------------------
-
-Each part should have a separate section in the config file.
-
-It can specify these options:
-
-**config**
-   Path to configuration file that describes this part. If relative, it is
-   resolved relative to the file with parts configuration.
-**just_phase**, **skip_phase**
-   Customize which phases should run for this part.
-**depends_on**
-   A comma separated list of other parts that must be finished before this part
-   starts.
-**failable**
-   A boolean toggle to mark a part as failable. A failure in such part will
-   mark the final compose as incomplete, but still successful.

doc/scm_support.rst

@@ -41,6 +41,14 @@ which can contain following keys.
 * ``command`` -- defines a shell command to run after Git clone to generate the
   needed file (for example to run ``make``). Only supported in Git backend.
 
+* ``options`` -- a dictionary of additional configuration options. These are
+  specific to different backends.
+
+  Currently supported values for Git:
+
+  * ``credential_helper`` -- path to a credential helper used to supply
+    username/password for remotes that require authentication.
+
 
 Koji examples
 -------------

pungi/checks.py

@@ -227,9 +227,19 @@ def validate(config, offline=False, schema=None):
     DefaultValidator = _extend_with_default_and_alias(
         jsonschema.Draft4Validator, offline=offline
     )
-    validator = DefaultValidator(
-        schema,
-    )
+
+    if hasattr(jsonschema.Draft4Validator, "TYPE_CHECKER"):
+        # jsonschema >= 3.0 has new interface for checking types
+        validator = DefaultValidator(schema)
+    else:
+        validator = DefaultValidator(
+            schema,
+            {
+                "array": (tuple, list),
+                "regex": six.string_types,
+                "url": six.string_types,
+            },
+        )
     errors = []
     warnings = []
     for error in validator.iter_errors(config):
@@ -377,6 +387,7 @@ def _extend_with_default_and_alias(validator_class, offline=False):
                 instance[property]["branch"] = resolver(
                     instance[property]["repo"],
                     instance[property].get("branch") or "HEAD",
+                    instance[property].get("options"),
                 )
 
         for error in _hook_errors(properties, instance, schema):
@@ -444,15 +455,18 @@ def _extend_with_default_and_alias(validator_class, offline=False):
                 context=all_errors,
             )
 
-    def is_array(checker, instance):
-        return isinstance(instance, (tuple, list))
+    kwargs = {}
+    if hasattr(validator_class, "TYPE_CHECKER"):
+        # jsonschema >= 3
+        def is_array(checker, instance):
+            return isinstance(instance, (tuple, list))
 
-    def is_string_type(checker, instance):
-        return isinstance(instance, six.string_types)
+        def is_string_type(checker, instance):
+            return isinstance(instance, six.string_types)
 
-    type_checker = validator_class.TYPE_CHECKER.redefine_many(
-        {"array": is_array, "regex": is_string_type, "url": is_string_type}
-    )
+        kwargs["type_checker"] = validator_class.TYPE_CHECKER.redefine_many(
+            {"array": is_array, "regex": is_string_type, "url": is_string_type}
+        )
 
     return jsonschema.validators.extend(
         validator_class,
@@ -464,7 +478,7 @@ def _extend_with_default_and_alias(validator_class, offline=False):
             "additionalProperties": _validate_additional_properties,
             "anyOf": _validate_any_of,
         },
-        type_checker=type_checker,
+        **kwargs
     )
 
 
@@ -507,6 +521,13 @@ def make_schema():
                     "file": {"type": "string"},
                     "dir": {"type": "string"},
                     "command": {"type": "string"},
+                    "options": {
+                        "type": "object",
+                        "properties": {
+                            "credential_helper": {"type": "string"},
+                        },
+                        "additionalProperties": False,
+                    },
                 },
                 "additionalProperties": False,
             },
@@ -588,6 +609,7 @@ def make_schema():
             "release_discinfo_description": {"type": "string"},
             "treeinfo_version": {"type": "string"},
             "compose_type": {"type": "string", "enum": COMPOSE_TYPES},
+            "label": {"type": "string"},
             "base_product_name": {"type": "string"},
             "base_product_short": {"type": "string"},
             "base_product_version": {"type": "string"},
@@ -665,7 +687,11 @@ def make_schema():
             "pkgset_allow_reuse": {"type": "boolean", "default": True},
             "createiso_allow_reuse": {"type": "boolean", "default": True},
             "extraiso_allow_reuse": {"type": "boolean", "default": True},
-            "pkgset_source": {"type": "string", "enum": ["koji", "repos"]},
+            "pkgset_source": {"type": "string", "enum": [
+                "koji",
+                "repos",
+                "kojimock",
+            ]},
             "createrepo_c": {"type": "boolean", "default": True},
             "createrepo_checksum": {
                 "type": "string",
@@ -794,6 +820,14 @@ def make_schema():
                 "type": "string",
                 "enum": ["lorax", "buildinstall"],
             },
+            # In phase `buildinstall` we should add to compose only the
+            # images that will be used only as netinstall
+            "netinstall_variants": {
+                "$ref": "#/definitions/list_of_strings",
+                "default": [
+                    "BaseOS",
+                ],
+            },
             "buildinstall_topdir": {"type": "string"},
             "buildinstall_kickstart": {"$ref": "#/definitions/str_or_scm_dict"},
             "buildinstall_use_guestmount": {"type": "boolean", "default": True},
@@ -811,8 +845,11 @@ def make_schema():
             "pdc_insecure": {"deprecated": "Koji is queried instead"},
             "cts_url": {"type": "string"},
             "cts_keytab": {"type": "string"},
+            "cts_oidc_token_url": {"type": "url"},
+            "cts_oidc_client_id": {"type": "string"},
             "koji_profile": {"type": "string"},
             "koji_event": {"type": "number"},
+            "koji_cache": {"type": "string"},
             "pkgset_koji_tag": {"$ref": "#/definitions/strings"},
             "pkgset_koji_builds": {"$ref": "#/definitions/strings"},
             "pkgset_koji_scratch_tasks": {"$ref": "#/definitions/strings"},
@@ -830,6 +867,10 @@ def make_schema():
                 "type": "boolean",
                 "default": True,
             },
+            "pkgset_inherit_exclusive_arch_to_noarch": {
+                "type": "boolean",
+                "default": True,
+            },
             "pkgset_scratch_modules": {
                 "type": "object",
                 "patternProperties": {
@@ -1032,6 +1073,9 @@ def make_schema():
                                         "config_branch": {"type": "string"},
                                         "tag_ref": {"type": "boolean"},
                                         "ostree_ref": {"type": "string"},
+                                        "runroot_packages": {
+                                            "$ref": "#/definitions/list_of_strings",
+                                        },
                                     },
                                     "required": [
                                         "treefile",
@@ -1188,14 +1232,36 @@ def make_schema():
                                 },
                                 "arches": {"$ref": "#/definitions/list_of_strings"},
                                 "release": {"type": "string"},
-                                "repo": {"$ref": "#/definitions/list_of_strings"},
+                                "repo": {
+                                    "type": "array",
+                                    "items": {
+                                        "oneOf": [
+                                            {
+                                                "type": "object",
+                                                "additionalProperties": False,
+                                                "required": ["baseurl"],
+                                                "properties": {
+                                                    "baseurl": {"type": "string"},
+                                                    "package_sets": {
+                                                        "type": "array",
+                                                        "items": {"type": "string"},
+                                                    },
+                                                },
+                                            },
+                                            {"type": "string"},
+                                        ]
+                                    },
+                                },
                                 "failable": {"$ref": "#/definitions/list_of_strings"},
                                 "subvariant": {"type": "string"},
                                 "ostree_url": {"type": "string"},
                                 "ostree_ref": {"type": "string"},
                                 "ostree_parent": {"type": "string"},
                                 "upload_options": {
-                                    "oneOf": [
+                                    # this should be really 'oneOf', but the minimal
+                                    # required properties in AWSEC2 and GCP options
+                                    # overlap.
+                                    "anyOf": [
                                         # AWSEC2UploadOptions
                                         {
                                             "type": "object",
@@ -1234,7 +1300,6 @@ def make_schema():
                                                 "tenant_id",
                                                 "subscription_id",
                                                 "resource_group",
-                                                "location",
                                             ],
                                             "properties": {
                                                 "tenant_id": {"type": "string"},
@@ -1250,7 +1315,7 @@ def make_schema():
                                         {
                                             "type": "object",
                                             "additionalProperties": False,
-                                            "required": ["region", "bucket"],
+                                            "required": ["region"],
                                             "properties": {
                                                 "region": {"type": "string"},
                                                 "bucket": {"type": "string"},

pungi/compose.py

@@ -17,6 +17,7 @@
 __all__ = ("Compose",)
 
 
+import contextlib
 import errno
 import logging
 import os
@@ -38,6 +39,7 @@ from dogpile.cache import make_region
 from pungi.graph import SimpleAcyclicOrientedGraph
 from pungi.wrappers.variants import VariantsXmlParser
 from pungi.paths import Paths
+from pungi.wrappers.kojiwrapper import KojiDownloadProxy
 from pungi.wrappers.scm import get_file_from_scm
 from pungi.util import (
     makedirs,
@@ -57,14 +59,101 @@ except ImportError:
     SUPPORTED_MILESTONES = ["RC", "Update", "SecurityFix"]
 
 
+def is_status_fatal(status_code):
+    """Check if status code returned from CTS reports an error that is unlikely
+    to be fixed by retrying. Generally client errors (4XX) are fatal, with the
+    exception of 401 Unauthorized which could be caused by transient network
+    issue between compose host and KDC.
+    """
+    if status_code == 401:
+        return False
+    return status_code >= 400 and status_code < 500
+
+
 @retry(wait_on=RequestException)
-def retry_request(method, url, data=None, auth=None):
+def retry_request(method, url, data=None, json_data=None, auth=None):
+    """
+    :param str method: Reqest method.
+    :param str url: Target URL.
+    :param dict data: form-urlencoded data to send in the body of the request.
+    :param dict json_data: json data to send in the body of the request.
+    """
     request_method = getattr(requests, method)
-    rv = request_method(url, json=data, auth=auth)
+    rv = request_method(url, data=data, json=json_data, auth=auth)
+    if is_status_fatal(rv.status_code):
+        try:
+            error = rv.json()
+        except ValueError:
+            error = rv.text
+        raise RuntimeError("%s responded with %d: %s" % (url, rv.status_code, error))
     rv.raise_for_status()
     return rv
 
 
+class BearerAuth(requests.auth.AuthBase):
+    def __init__(self, token):
+        self.token = token
+
+    def __call__(self, r):
+        r.headers["authorization"] = "Bearer " + self.token
+        return r
+
+
+@contextlib.contextmanager
+def cts_auth(pungi_conf):
+    """
+    :param dict pungi_conf: dict obj of pungi.json config.
+    """
+    auth = None
+    token = None
+    cts_keytab = pungi_conf.get("cts_keytab")
+    cts_oidc_token_url = os.environ.get("CTS_OIDC_TOKEN_URL", "") or pungi_conf.get(
+        "cts_oidc_token_url"
+    )
+
+    try:
+        if cts_keytab:
+            # requests-kerberos cannot accept custom keytab, we need to use
+            # environment variable for this. But we need to change environment
+            # only temporarily just for this single requests.post.
+            # So at first backup the current environment and revert to it
+            # after the requests call.
+            from requests_kerberos import HTTPKerberosAuth
+
+            auth = HTTPKerberosAuth()
+            environ_copy = dict(os.environ)
+            if "$HOSTNAME" in cts_keytab:
+                cts_keytab = cts_keytab.replace("$HOSTNAME", socket.gethostname())
+            os.environ["KRB5_CLIENT_KTNAME"] = cts_keytab
+            os.environ["KRB5CCNAME"] = "DIR:%s" % tempfile.mkdtemp()
+        elif cts_oidc_token_url:
+            cts_oidc_client_id = os.environ.get(
+                "CTS_OIDC_CLIENT_ID", ""
+            ) or pungi_conf.get("cts_oidc_client_id", "")
+            token = retry_request(
+                "post",
+                cts_oidc_token_url,
+                data={
+                    "grant_type": "client_credentials",
+                    "client_id": cts_oidc_client_id,
+                    "client_secret": os.environ.get("CTS_OIDC_CLIENT_SECRET", ""),
+                },
+            ).json()["access_token"]
+            auth = BearerAuth(token)
+            del token
+
+        yield auth
+    except Exception as e:
+        # Avoid leaking client secret in trackback
+        e.show_locals = False
+        raise e
+    finally:
+        if cts_keytab:
+            shutil.rmtree(os.environ["KRB5CCNAME"].split(":", 1)[1])
+            os.environ.clear()
+            os.environ.update(environ_copy)
+
+
 def get_compose_info(
     conf,
     compose_type="production",
@@ -94,38 +183,19 @@ def get_compose_info(
     ci.compose.type = compose_type
     ci.compose.date = compose_date or time.strftime("%Y%m%d", time.localtime())
     ci.compose.respin = compose_respin or 0
+    ci.compose.id = ci.create_compose_id()
 
-    cts_url = conf.get("cts_url", None)
+    cts_url = conf.get("cts_url")
     if cts_url:
-        # Requests-kerberos cannot accept custom keytab, we need to use
-        # environment variable for this. But we need to change environment
-        # only temporarily just for this single requests.post.
-        # So at first backup the current environment and revert to it
-        # after the requests.post call.
-        cts_keytab = conf.get("cts_keytab", None)
-        authentication = get_authentication(conf)
-        if cts_keytab:
-            environ_copy = dict(os.environ)
-            if "$HOSTNAME" in cts_keytab:
-                cts_keytab = cts_keytab.replace("$HOSTNAME", socket.gethostname())
-            os.environ["KRB5_CLIENT_KTNAME"] = cts_keytab
-            os.environ["KRB5CCNAME"] = "DIR:%s" % tempfile.mkdtemp()
-
-        try:
-            # Create compose in CTS and get the reserved compose ID.
-            ci.compose.id = ci.create_compose_id()
-            url = os.path.join(cts_url, "api/1/composes/")
-            data = {
-                "compose_info": json.loads(ci.dumps()),
-                "parent_compose_ids": parent_compose_ids,
-                "respin_of": respin_of,
-            }
-            rv = retry_request("post", url, data=data, auth=authentication)
-        finally:
-            if cts_keytab:
-                shutil.rmtree(os.environ["KRB5CCNAME"].split(":", 1)[1])
-                os.environ.clear()
-                os.environ.update(environ_copy)
+        # Create compose in CTS and get the reserved compose ID.
+        url = os.path.join(cts_url, "api/1/composes/")
+        data = {
+            "compose_info": json.loads(ci.dumps()),
+            "parent_compose_ids": parent_compose_ids,
+            "respin_of": respin_of,
+        }
+        with cts_auth(conf) as authentication:
+            rv = retry_request("post", url, json_data=data, auth=authentication)
 
         # Update local ComposeInfo with received ComposeInfo.
         cts_ci = ComposeInfo()
@@ -133,22 +203,9 @@ def get_compose_info(
         ci.compose.respin = cts_ci.compose.respin
         ci.compose.id = cts_ci.compose.id
 
-    else:
-        ci.compose.id = ci.create_compose_id()
-
     return ci
 
 
-def get_authentication(conf):
-    authentication = None
-    cts_keytab = conf.get("cts_keytab", None)
-    if cts_keytab:
-        from requests_kerberos import HTTPKerberosAuth
-
-        authentication = HTTPKerberosAuth()
-    return authentication
-
-
 def write_compose_info(compose_dir, ci):
     """
     Write ComposeInfo `ci` to `compose_dir` subdirectories.
@@ -162,17 +219,20 @@ def write_compose_info(compose_dir, ci):
 
 
 def update_compose_url(compose_id, compose_dir, conf):
-    authentication = get_authentication(conf)
     cts_url = conf.get("cts_url", None)
     if cts_url:
         url = os.path.join(cts_url, "api/1/composes", compose_id)
         tp = conf.get("translate_paths", None)
         compose_url = translate_path_raw(tp, compose_dir)
+        if compose_url == compose_dir:
+            # We do not have a URL, do not attempt the update.
+            return
         data = {
             "action": "set_url",
             "compose_url": compose_url,
         }
-        return retry_request("patch", url, data=data, auth=authentication)
+        with cts_auth(conf) as authentication:
+            return retry_request("patch", url, json_data=data, auth=authentication)
 
 
 def get_compose_dir(
@@ -183,11 +243,19 @@ def get_compose_dir(
     compose_respin=None,
     compose_label=None,
     already_exists_callbacks=None,
+    parent_compose_ids=None,
+    respin_of=None,
 ):
     already_exists_callbacks = already_exists_callbacks or []
 
     ci = get_compose_info(
-        conf, compose_type, compose_date, compose_respin, compose_label
+        conf,
+        compose_type,
+        compose_date,
+        compose_respin,
+        compose_label,
+        parent_compose_ids,
+        respin_of,
     )
 
     cts_url = conf.get("cts_url", None)
@@ -342,6 +410,8 @@ class Compose(kobo.log.LoggingBase):
         else:
             self.cache_region = make_region().configure("dogpile.cache.null")
 
+        self.koji_downloader = KojiDownloadProxy.from_config(self.conf, self._logger)
+
     get_compose_info = staticmethod(get_compose_info)
     write_compose_info = staticmethod(write_compose_info)
     get_compose_dir = staticmethod(get_compose_dir)
@@ -637,7 +707,7 @@ class Compose(kobo.log.LoggingBase):
                 separators=(",", ": "),
             )
 
-    def traceback(self, detail=None):
+    def traceback(self, detail=None, show_locals=True):
         """Store an extended traceback. This method should only be called when
         handling an exception.
 
@@ -649,7 +719,7 @@ class Compose(kobo.log.LoggingBase):
         tb_path = self.paths.log.log_file("global", basename)
         self.log_error("Extended traceback in: %s", tb_path)
         with open(tb_path, "wb") as f:
-            f.write(kobo.tback.Traceback().get_traceback())
+            f.write(kobo.tback.Traceback(show_locals=show_locals).get_traceback())
 
     def load_old_compose_config(self):
         """

pungi/createiso.py

@@ -5,11 +5,14 @@ from __future__ import print_function
 import os
 import six
 from collections import namedtuple
+from kobo.shortcuts import run
 from six.moves import shlex_quote
 
 from .wrappers import iso
 from .wrappers.jigdo import JigdoWrapper
 
+from .phases.buildinstall import BOOT_CONFIGS, BOOT_IMAGES
+
 
 CreateIsoOpts = namedtuple(
     "CreateIsoOpts",
@@ -118,23 +121,73 @@ def make_jigdo(f, opts):
     emit(f, cmd)
 
 
+def _get_perms(fs_path):
+    """Compute proper permissions for a file.
+
+    This mimicks what -rational-rock option of genisoimage does. All read bits
+    are set, so that files and directories are globally readable. If any
+    execute bit is set for a file, set them all. No writes are allowed and
+    special bits are erased too.
+    """
+    statinfo = os.stat(fs_path)
+    perms = 0o444
+    if statinfo.st_mode & 0o111:
+        perms |= 0o111
+    return perms
+
+
 def write_xorriso_commands(opts):
+    # Create manifest for the boot.iso listing all contents
+    boot_iso_manifest = "%s.manifest" % os.path.join(
+        opts.script_dir, os.path.basename(opts.boot_iso)
+    )
+    run(
+        iso.get_manifest_cmd(
+            opts.boot_iso, opts.use_xorrisofs, output_file=boot_iso_manifest
+        )
+    )
+    # Find which files may have been updated by pungi. This only includes a few
+    # files from tweaking buildinstall and .discinfo metadata. There's no good
+    # way to detect whether the boot config files actually changed, so we may
+    # be updating files in the ISO with the same data.
+    UPDATEABLE_FILES = set(BOOT_IMAGES + BOOT_CONFIGS + [".discinfo"])
+    updated_files = set()
+    excluded_files = set()
+    with open(boot_iso_manifest) as f:
+        for line in f:
+            path = line.lstrip("/").rstrip("\n")
+            if path in UPDATEABLE_FILES:
+                updated_files.add(path)
+            else:
+                excluded_files.add(path)
+
     script = os.path.join(opts.script_dir, "xorriso-%s.txt" % id(opts))
     with open(script, "w") as f:
         emit(f, "-indev %s" % opts.boot_iso)
         emit(f, "-outdev %s" % os.path.join(opts.output_dir, opts.iso_name))
         emit(f, "-boot_image any replay")
         emit(f, "-volid %s" % opts.volid)
+        # isoinfo -J uses the Joliet tree, and it's used by virt-install
+        emit(f, "-joliet on")
+        # Support long filenames in the Joliet trees. Repodata is particularly
+        # likely to run into this limit.
+        emit(f, "-compliance joliet_long_names")
 
         with open(opts.graft_points) as gp:
             for line in gp:
                 iso_path, fs_path = line.strip().split("=", 1)
-                emit(f, "-map %s %s" % (fs_path, iso_path))
+                if iso_path in excluded_files:
+                    continue
+                cmd = "-update" if iso_path in updated_files else "-map"
+                emit(f, "%s %s %s" % (cmd, fs_path, iso_path))
+                emit(f, "-chmod 0%o %s" % (_get_perms(fs_path), iso_path))
 
         if opts.arch == "ppc64le":
             # This is needed for the image to be bootable.
             emit(f, "-as mkisofs -U --")
 
+        emit(f, "-chown_r 0 /")
+        emit(f, "-chgrp_r 0 /")
         emit(f, "-end")
     return script
 

pungi/gather.py

@@ -1118,7 +1118,6 @@ class Pungi(PungiBase):
         self.logger.info("Finished gathering package objects.")
 
     def gather(self):
-
         # get package objects according to the input list
         self.getPackageObjects()
         if self.is_sources:

pungi/gather_dnf.py

@@ -15,17 +15,20 @@
 
 
 from enum import Enum
-from itertools import count
+from functools import cmp_to_key
+from itertools import count, groupby
 import logging
 import os
 import re
 
 from kobo.rpmlib import parse_nvra
+import rpm
 
 import pungi.common
 import pungi.dnf_wrapper
 import pungi.multilib_dnf
 import pungi.util
+from pungi import arch_utils
 from pungi.linker import Linker
 from pungi.profiler import Profiler
 from pungi.util import DEBUG_PATTERNS
@@ -245,12 +248,36 @@ class Gather(GatherBase):
         # from lookaside. This can be achieved by removing any package that is
         # also in lookaside from the list.
         lookaside_pkgs = set()
-        for pkg in package_list:
-            if pkg.repoid in self.opts.lookaside_repos:
-                lookaside_pkgs.add("{0.name}-{0.evr}".format(pkg))
 
-        if self.opts.greedy_method == "all":
-            return list(package_list)
+        if self.opts.lookaside_repos:
+            # We will call `latest()` to get the highest version packages only.
+            # However, that is per name and architecture. If a package switches
+            # from arched to noarch or the other way, it is possible that the
+            # package_list contains different versions in main repos and in
+            # lookaside repos.
+            # We need to manually filter the latest version.
+            def vercmp(x, y):
+                return rpm.labelCompare(x[1], y[1])
+
+            # Annotate the packages with their version.
+            versioned_packages = [
+                (pkg, (str(pkg.epoch) or "0", pkg.version, pkg.release))
+                for pkg in package_list
+            ]
+            # Sort the packages newest first.
+            sorted_packages = sorted(
+                versioned_packages, key=cmp_to_key(vercmp), reverse=True
+            )
+            # Group packages by version, take the first group and discard the
+            # version info from the tuple.
+            package_list = list(
+                x[0] for x in next(groupby(sorted_packages, key=lambda x: x[1]))[1]
+            )
+
+            # Now we can decide what is used from lookaside.
+            for pkg in package_list:
+                if pkg.repoid in self.opts.lookaside_repos:
+                    lookaside_pkgs.add("{0.name}-{0.evr}".format(pkg))
 
         all_pkgs = []
         for pkg in package_list:
@@ -263,16 +290,21 @@ class Gather(GatherBase):
 
         if not debuginfo:
             native_pkgs = set(
-                self.q_native_binary_packages.filter(pkg=all_pkgs).apply()
+                self.q_native_binary_packages.filter(pkg=all_pkgs).latest().apply()
             )
             multilib_pkgs = set(
-                self.q_multilib_binary_packages.filter(pkg=all_pkgs).apply()
+                self.q_multilib_binary_packages.filter(pkg=all_pkgs).latest().apply()
             )
         else:
-            native_pkgs = set(self.q_native_debug_packages.filter(pkg=all_pkgs).apply())
-            multilib_pkgs = set(
-                self.q_multilib_debug_packages.filter(pkg=all_pkgs).apply()
+            native_pkgs = set(
+                self.q_native_debug_packages.filter(pkg=all_pkgs).latest().apply()
             )
+            multilib_pkgs = set(
+                self.q_multilib_debug_packages.filter(pkg=all_pkgs).latest().apply()
+            )
+
+        if self.opts.greedy_method == "all":
+            return list(native_pkgs | multilib_pkgs)
 
         result = set()
 
@@ -392,9 +424,7 @@ class Gather(GatherBase):
         """Given an name of a queue (stored as attribute in `self`), exclude
         all given packages and keep only the latest per package name and arch.
         """
-        setattr(
-            self, queue, getattr(self, queue).filter(pkg__neq=exclude).latest().apply()
-        )
+        setattr(self, queue, getattr(self, queue).filter(pkg__neq=exclude).apply())
 
     @Profiler("Gather._apply_excludes()")
     def _apply_excludes(self, excludes):
@@ -500,12 +530,21 @@ class Gather(GatherBase):
                                 name__glob=pattern[:-2]
                             ).apply()
                         else:
-                            pkgs = self.q_binary_packages.filter(
-                                name__glob=pattern
-                            ).apply()
+                            kwargs = {"name__glob": pattern}
+                            if "." in pattern:
+                                # The pattern could be name.arch. Check if the
+                                # arch is valid, and if yes, make a more
+                                # specific query.
+                                name, arch = pattern.split(".", 1)
+                                if arch in arch_utils.arches:
+                                    kwargs["name__glob"] = name
+                                    kwargs["arch__eq"] = arch
+                            pkgs = self.q_binary_packages.filter(**kwargs).apply()
 
                 if not pkgs:
-                    self.logger.error("No package matches pattern %s" % pattern)
+                    self.logger.error(
+                        "Could not find a match for %s in any configured repo", pattern
+                    )
 
                 # The pattern could have been a glob. In that case we want to
                 # group the packages by name and get best match in those
@@ -616,7 +655,6 @@ class Gather(GatherBase):
             return added
 
         for pkg in self.result_debug_packages.copy():
-
             if pkg not in self.finished_add_debug_package_deps:
                 deps = self._get_package_deps(pkg, debuginfo=True)
                 for i, req in deps:
@@ -784,7 +822,6 @@ class Gather(GatherBase):
                 continue
 
             debug_pkgs = []
-            pkg_in_lookaside = pkg.repoid in self.opts.lookaside_repos
             for i in candidates:
                 if pkg.arch != i.arch:
                     continue
@@ -792,7 +829,7 @@ class Gather(GatherBase):
                     # If it's not debugsource package or does not match name of
                     # the package, we don't want it in.
                     continue
-                if i.repoid in self.opts.lookaside_repos or pkg_in_lookaside:
+                if self.is_from_lookaside(i):
                     self._set_flag(i, PkgFlag.lookaside)
                 if i not in self.result_debug_packages:
                     added.add(i)

pungi/metadata.py

@@ -306,11 +306,6 @@ def write_tree_info(compose, arch, variant, timestamp=None, bi=None):
     if variant.type in ("addon",) or variant.is_empty:
         return
 
-    compose.log_debug(
-        "on arch '%s' looking at variant '%s' of type '%s'"
-        % (arch, variant, variant.type)
-    )
-
     if not timestamp:
         timestamp = int(time.time())
     else:

pungi/phases/buildinstall.py

@@ -297,7 +297,7 @@ class BuildinstallPhase(PhaseBase):
                     "Unsupported buildinstall method: %s" % self.buildinstall_method
                 )
 
-            for (variant, cmd) in commands:
+            for variant, cmd in commands:
                 self.pool.add(BuildinstallThread(self.pool))
                 self.pool.queue_put(
                     (self.compose, arch, variant, cmd, self.pkgset_phase)
@@ -364,9 +364,17 @@ BOOT_CONFIGS = [
     "EFI/BOOT/BOOTX64.conf",
     "EFI/BOOT/grub.cfg",
 ]
+BOOT_IMAGES = [
+    "images/efiboot.img",
+]
 
 
 def tweak_configs(path, volid, ks_file, configs=BOOT_CONFIGS, logger=None):
+    """
+    Put escaped volume ID and possibly kickstart file into the boot
+    configuration files.
+    :returns: list of paths to modified config files
+    """
     volid_escaped = volid.replace(" ", r"\x20").replace("\\", "\\\\")
     volid_escaped_2 = volid_escaped.replace("\\", "\\\\")
     found_configs = []
@@ -374,7 +382,6 @@ def tweak_configs(path, volid, ks_file, configs=BOOT_CONFIGS, logger=None):
         config_path = os.path.join(path, config)
         if not os.path.exists(config_path):
             continue
-        found_configs.append(config)
 
         with open(config_path, "r") as f:
             data = original_data = f.read()
@@ -394,8 +401,13 @@ def tweak_configs(path, volid, ks_file, configs=BOOT_CONFIGS, logger=None):
         with open(config_path, "w") as f:
             f.write(data)
 
-        if logger and data != original_data:
-            logger.info("Boot config %s changed" % config_path)
+        if data != original_data:
+            found_configs.append(config)
+            if logger:
+                # Generally lorax should create file with correct volume id
+                # already. If we don't have a kickstart, this function should
+                # be a no-op.
+                logger.info("Boot config %s changed" % config_path)
 
     return found_configs
 
@@ -434,31 +446,32 @@ def tweak_buildinstall(
     if kickstart_file and found_configs:
         shutil.copy2(kickstart_file, os.path.join(dst, "ks.cfg"))
 
-    images = [
-        os.path.join(tmp_dir, "images", "efiboot.img"),
-    ]
-    for image in images:
-        if not os.path.isfile(image):
-            continue
+    images = [os.path.join(tmp_dir, img) for img in BOOT_IMAGES]
+    if found_configs:
+        for image in images:
+            if not os.path.isfile(image):
+                continue
 
-        with iso.mount(
-            image,
-            logger=compose._logger,
-            use_guestmount=compose.conf.get("buildinstall_use_guestmount"),
-        ) as mount_tmp_dir:
-            for config in BOOT_CONFIGS:
-                config_path = os.path.join(tmp_dir, config)
-                config_in_image = os.path.join(mount_tmp_dir, config)
+            with iso.mount(
+                image,
+                logger=compose._logger,
+                use_guestmount=compose.conf.get("buildinstall_use_guestmount"),
+            ) as mount_tmp_dir:
+                for config in found_configs:
+                    # Put each modified config file into the image (overwriting the
+                    # original).
+                    config_path = os.path.join(tmp_dir, config)
+                    config_in_image = os.path.join(mount_tmp_dir, config)
 
-                if os.path.isfile(config_in_image):
-                    cmd = [
-                        "cp",
-                        "-v",
-                        "--remove-destination",
-                        config_path,
-                        config_in_image,
-                    ]
-                    run(cmd)
+                    if os.path.isfile(config_in_image):
+                        cmd = [
+                            "cp",
+                            "-v",
+                            "--remove-destination",
+                            config_path,
+                            config_in_image,
+                        ]
+                        run(cmd)
 
     # HACK: make buildinstall files world readable
     run("chmod -R a+rX %s" % shlex_quote(tmp_dir))
@@ -533,7 +546,14 @@ def link_boot_iso(compose, arch, variant, can_fail):
         img.volume_id = iso.get_volume_id(new_boot_iso_path)
     except RuntimeError:
         pass
-    compose.im.add(variant.uid, arch, img)
+    # In this phase we should add to compose only the images that
+    # will be used only as netinstall.
+    # On this step lorax generates environment
+    # for creating isos and create them.
+    # On step `extra_isos` we overwrite the not needed iso `boot Minimal` by
+    # new iso. It already contains necessary packages from incldued variants.
+    if variant.uid in compose.conf['netinstall_variants']:
+        compose.im.add(variant.uid, arch, img)
     compose.log_info("[DONE ] %s" % msg)
 
 

pungi/phases/createiso.py

@@ -369,7 +369,7 @@ class CreateisoPhase(PhaseLoggerMixin, PhaseBase):
         if self.compose.notifier:
             self.compose.notifier.send("createiso-targets", deliverables=deliverables)
 
-        for (cmd, variant, arch) in commands:
+        for cmd, variant, arch in commands:
             self.pool.add(CreateIsoThread(self.pool))
             self.pool.queue_put((self.compose, cmd, variant, arch))
 

pungi/phases/extra_isos.py

@@ -76,7 +76,7 @@ class ExtraIsosPhase(PhaseLoggerMixin, ConfigGuardedPhase, PhaseBase):
                 for arch in sorted(arches):
                     commands.append((config, variant, arch))
 
-        for (config, variant, arch) in commands:
+        for config, variant, arch in commands:
             self.pool.add(ExtraIsosThread(self.pool, self.bi))
             self.pool.queue_put((self.compose, config, variant, arch))
 
@@ -420,6 +420,12 @@ def get_iso_contents(
         original_treeinfo,
         os.path.join(extra_files_dir, ".treeinfo"),
     )
+    tweak_repo_treeinfo(
+        compose,
+        include_variants,
+        original_treeinfo,
+        original_treeinfo,
+    )
 
     # Add extra files specific for the ISO
     files.update(
@@ -431,6 +437,45 @@ def get_iso_contents(
     return gp
 
 
+def tweak_repo_treeinfo(compose, include_variants, source_file, dest_file):
+    """
+    The method includes the variants to file .treeinfo of a variant. It takes
+    the variants which are described
+    by options `extra_isos -> include_variants`.
+    """
+    ti = productmd.treeinfo.TreeInfo()
+    ti.load(source_file)
+    main_variant = next(iter(ti.variants))
+    for variant_uid in include_variants:
+        variant = compose.all_variants[variant_uid]
+        var = productmd.treeinfo.Variant(ti)
+        var.id = variant.id
+        var.uid = variant.uid
+        var.name = variant.name
+        var.type = variant.type
+        ti.variants.add(var)
+
+    for variant_id in ti.variants:
+        var = ti.variants[variant_id]
+        if variant_id == main_variant:
+            var.paths.packages = 'Packages'
+            var.paths.repository = '.'
+        else:
+            var.paths.packages = os.path.join(
+                '../../..',
+                var.uid,
+                var.arch,
+                'os/Packages',
+            )
+            var.paths.repository = os.path.join(
+                '../../..',
+                var.uid,
+                var.arch,
+                'os',
+            )
+    ti.dump(dest_file, main_variant=main_variant)
+
+
 def tweak_treeinfo(compose, include_variants, source_file, dest_file):
     ti = load_and_tweak_treeinfo(source_file)
     for variant_uid in include_variants:
@@ -446,7 +491,6 @@ def tweak_treeinfo(compose, include_variants, source_file, dest_file):
         var = ti.variants[variant_id]
         var.paths.packages = os.path.join(var.uid, "Packages")
         var.paths.repository = var.uid
-
     ti.dump(dest_file)
 
 

pungi/phases/gather/__init__.py

@@ -23,6 +23,7 @@ import threading
 from kobo.rpmlib import parse_nvra
 from kobo.shortcuts import run
 from productmd.rpms import Rpms
+from pungi.phases.pkgset.common import get_all_arches
 from six.moves import cPickle as pickle
 
 try:
@@ -90,7 +91,7 @@ class GatherPhase(PhaseBase):
 
         # check whether variants from configuration value
         # 'variant_as_lookaside' are correct
-        for (requiring, required) in variant_as_lookaside:
+        for requiring, required in variant_as_lookaside:
             if requiring in all_variants and required not in all_variants:
                 errors.append(
                     "variant_as_lookaside: variant %r doesn't exist but is "
@@ -99,7 +100,7 @@ class GatherPhase(PhaseBase):
 
         # check whether variants from configuration value
         # 'variant_as_lookaside' have same architectures
-        for (requiring, required) in variant_as_lookaside:
+        for requiring, required in variant_as_lookaside:
             if (
                 requiring in all_variants
                 and required in all_variants
@@ -235,7 +236,7 @@ def reuse_old_gather_packages(compose, arch, variant, package_sets, methods):
     if not hasattr(compose, "_gather_reused_variant_arch"):
         setattr(compose, "_gather_reused_variant_arch", [])
     variant_as_lookaside = compose.conf.get("variant_as_lookaside", [])
-    for (requiring, required) in variant_as_lookaside:
+    for requiring, required in variant_as_lookaside:
         if (
             requiring == variant.uid
             and (required, arch) not in compose._gather_reused_variant_arch
@@ -468,9 +469,7 @@ def gather_packages(compose, arch, variant, package_sets, fulltree_excludes=None
         )
 
     else:
-
         for source_name in ("module", "comps", "json"):
-
             packages, groups, filter_packages = get_variant_packages(
                 compose, arch, variant, source_name, package_sets
             )
@@ -575,7 +574,6 @@ def trim_packages(compose, arch, variant, pkg_map, parent_pkgs=None, remove_pkgs
     move_to_parent_pkgs = _mk_pkg_map()
     removed_pkgs = _mk_pkg_map()
     for pkg_type, pkgs in pkg_map.items():
-
         new_pkgs = []
         for pkg in pkgs:
             pkg_path = pkg["path"]
@@ -647,8 +645,14 @@ def _make_lookaside_repo(compose, variant, arch, pkg_map, package_sets=None):
             compose.paths.work.topdir(arch="global"), "download"
         )
         + "/",
-        "koji": lambda: pungi.wrappers.kojiwrapper.KojiWrapper(
-            compose
+        "koji": lambda: compose.conf.get(
+            "koji_cache",
+            pungi.wrappers.kojiwrapper.KojiWrapper(compose).koji_module.config.topdir,
+        ).rstrip("/")
+        + "/",
+        "kojimock": lambda: pungi.wrappers.kojiwrapper.KojiMockWrapper(
+            compose,
+            get_all_arches(compose),
         ).koji_module.config.topdir.rstrip("/")
         + "/",
     }

pungi/phases/gather/methods/method_hybrid.py

@@ -47,9 +47,15 @@ class FakePackage(object):
 
     @property
     def files(self):
-        return [
-            os.path.join(dirname, basename) for (_, dirname, basename) in self.pkg.files
-        ]
+        paths = []
+        # createrepo_c.Package.files is a tuple, but its length differs across
+        # versions. The constants define index at which the related value is
+        # located.
+        for entry in self.pkg.files:
+            paths.append(
+                os.path.join(entry[cr.FILE_ENTRY_PATH], entry[cr.FILE_ENTRY_NAME])
+            )
+        return paths
 
     @property
     def provides(self):

pungi/phases/image_build.py

@@ -25,6 +25,7 @@ from productmd.rpms import Rpms
 # results will be pulled into the compose.
 EXTENSIONS = {
     "docker": ["tar.gz", "tar.xz"],
+    "iso": ["iso"],
     "liveimg-squashfs": ["liveimg.squashfs"],
     "qcow": ["qcow"],
     "qcow2": ["qcow2"],
@@ -39,6 +40,7 @@ EXTENSIONS = {
     "vdi": ["vdi"],
     "vmdk": ["vmdk"],
     "vpc": ["vhd"],
+    "vhd-compressed": ["vhd.gz", "vhd.xz"],
     "vsphere-ova": ["vsphere.ova"],
 }
 
@@ -344,7 +346,9 @@ class CreateImageBuildThread(WorkerThread):
             # let's not change filename of koji outputs
             image_dest = os.path.join(image_dir, os.path.basename(image_info["path"]))
 
-            src_file = os.path.realpath(image_info["path"])
+            src_file = compose.koji_downloader.get_file(
+                os.path.realpath(image_info["path"])
+            )
             linker.link(src_file, image_dest, link_type=cmd["link_type"])
 
             # Update image manifest

pungi/phases/live_images.py

@@ -117,7 +117,7 @@ class LiveImagesPhase(
 
                     commands.append((cmd, variant, arch))
 
-        for (cmd, variant, arch) in commands:
+        for cmd, variant, arch in commands:
             self.pool.add(CreateLiveImageThread(self.pool))
             self.pool.queue_put((self.compose, cmd, variant, arch))
 
@@ -232,7 +232,7 @@ class CreateLiveImageThread(WorkerThread):
                 "Got %d images from task %d, expected 1."
                 % (len(image_path), output["task_id"])
             )
-        image_path = image_path[0]
+        image_path = compose.koji_downloader.get_file(image_path[0])
         filename = cmd.get("filename") or os.path.basename(image_path)
         destination = os.path.join(cmd["dest_dir"], filename)
         shutil.copy2(image_path, destination)

pungi/phases/livemedia_phase.py

@@ -182,7 +182,9 @@ class LiveMediaThread(WorkerThread):
             # let's not change filename of koji outputs
             image_dest = os.path.join(image_dir, os.path.basename(image_info["path"]))
 
-            src_file = os.path.realpath(image_info["path"])
+            src_file = compose.koji_downloader.get_file(
+                os.path.realpath(image_info["path"])
+            )
             linker.link(src_file, image_dest, link_type=link_type)
 
             # Update image manifest

pungi/phases/osbuild.py

@@ -27,6 +27,35 @@ class OSBuildPhase(
             arches = set(image_conf["arches"]) & arches
         return sorted(arches)
 
+    @staticmethod
+    def _get_repo_urls(compose, repos, arch="$basearch"):
+        """
+        Get list of repos with resolved repo URLs. Preserve repos defined
+        as dicts.
+        """
+        resolved_repos = []
+
+        for repo in repos:
+            if isinstance(repo, dict):
+                try:
+                    url = repo["baseurl"]
+                except KeyError:
+                    raise RuntimeError(
+                        "`baseurl` is required in repo dict %s" % str(repo)
+                    )
+                url = util.get_repo_url(compose, url, arch=arch)
+                if url is None:
+                    raise RuntimeError("Failed to resolve repo URL for %s" % str(repo))
+                repo["baseurl"] = url
+                resolved_repos.append(repo)
+            else:
+                repo = util.get_repo_url(compose, repo, arch=arch)
+                if repo is None:
+                    raise RuntimeError("Failed to resolve repo URL for %s" % repo)
+                resolved_repos.append(repo)
+
+        return resolved_repos
+
     def _get_repo(self, image_conf, variant):
         """
         Get a list of repos. First included are those explicitly listed in
@@ -38,7 +67,7 @@ class OSBuildPhase(
         if not variant.is_empty and variant.uid not in repos:
             repos.append(variant.uid)
 
-        return util.get_repo_urls(self.compose, repos, arch="$arch")
+        return OSBuildPhase._get_repo_urls(self.compose, repos, arch="$arch")
 
     def run(self):
         for variant in self.compose.get_variants():
@@ -183,16 +212,27 @@ class RunOSBuildThread(WorkerThread):
             # image_dir is absolute path to which the image should be copied.
             # We also need the same path as relative to compose directory for
             # including in the metadata.
-            image_dir = compose.paths.compose.image_dir(variant) % {"arch": arch}
-            rel_image_dir = compose.paths.compose.image_dir(variant, relative=True) % {
-                "arch": arch
-            }
+            if archive["type_name"] == "iso":
+                # If the produced image is actually an ISO, it should go to
+                # iso/ subdirectory.
+                image_dir = compose.paths.compose.iso_dir(arch, variant)
+                rel_image_dir = compose.paths.compose.iso_dir(
+                    arch, variant, relative=True
+                )
+            else:
+                image_dir = compose.paths.compose.image_dir(variant) % {"arch": arch}
+                rel_image_dir = compose.paths.compose.image_dir(
+                    variant, relative=True
+                ) % {"arch": arch}
             util.makedirs(image_dir)
 
             image_dest = os.path.join(image_dir, archive["filename"])
 
-            src_file = os.path.join(
-                koji.koji_module.pathinfo.imagebuild(build_info), archive["filename"]
+            src_file = compose.koji_downloader.get_file(
+                os.path.join(
+                    koji.koji_module.pathinfo.imagebuild(build_info),
+                    archive["filename"],
+                ),
             )
 
             linker.link(src_file, image_dest, link_type=compose.conf["link_type"])
@@ -209,7 +249,7 @@ class RunOSBuildThread(WorkerThread):
 
             # Update image manifest
             img = Image(compose.im)
-            img.type = archive["type_name"]
+            img.type = archive["type_name"] if archive["type_name"] != "iso" else "dvd"
             img.format = suffix
             img.path = os.path.join(rel_image_dir, archive["filename"])
             img.mtime = util.get_mtime(image_dest)

pungi/phases/ostree.py

@@ -168,7 +168,9 @@ class OSTreeThread(WorkerThread):
                 ("unified-core", config.get("unified_core", False)),
             ]
         )
-        packages = ["pungi", "ostree", "rpm-ostree"]
+        default_packages = ["pungi", "ostree", "rpm-ostree"]
+        additional_packages = config.get("runroot_packages", [])
+        packages = default_packages + additional_packages
         log_file = os.path.join(self.logdir, "runroot.log")
         mounts = [compose.topdir, config["ostree_repo"]]
         runroot = Runroot(compose, phase="ostree")

pungi/phases/pkgset/common.py

@@ -38,12 +38,17 @@ from pungi.phases.createrepo import add_modular_metadata
 
 def populate_arch_pkgsets(compose, path_prefix, global_pkgset):
     result = {}
-    exclusive_noarch = compose.conf["pkgset_exclusive_arch_considers_noarch"]
+
     for arch in compose.get_arches():
         compose.log_info("Populating package set for arch: %s", arch)
         is_multilib = is_arch_multilib(compose.conf, arch)
         arches = get_valid_arches(arch, is_multilib, add_src=True)
-        pkgset = global_pkgset.subset(arch, arches, exclusive_noarch=exclusive_noarch)
+        pkgset = global_pkgset.subset(
+            arch,
+            arches,
+            exclusive_noarch=compose.conf["pkgset_exclusive_arch_considers_noarch"],
+            inherit_to_noarch=compose.conf["pkgset_inherit_exclusive_arch_to_noarch"],
+        )
         pkgset.save_file_list(
             compose.paths.work.package_list(arch=arch, pkgset=global_pkgset),
             remove_path_prefix=path_prefix,

pungi/phases/pkgset/pkgsets.py

@@ -23,11 +23,15 @@ import itertools
 import json
 import os
 import time
+import pgpy
+import rpm
 from six.moves import cPickle as pickle
+from functools import partial
 
 import kobo.log
 import kobo.pkgset
 import kobo.rpmlib
+from kobo.shortcuts import compute_file_checksums
 
 from kobo.threads import WorkerThread, ThreadPool
 
@@ -150,9 +154,15 @@ class PackageSetBase(kobo.log.LoggingBase):
         """
 
         def nvr_formatter(package_info):
-            # joins NVR parts of the package with '-' character.
-            return "-".join(
-                (package_info["name"], package_info["version"], package_info["release"])
+            epoch_suffix = ''
+            if package_info['epoch'] is not None:
+                epoch_suffix = ':' + package_info['epoch']
+            return (
+                f"{package_info['name']}"
+                f"{epoch_suffix}-"
+                f"{package_info['version']}-"
+                f"{package_info['release']}."
+                f"{package_info['arch']}"
             )
 
         def get_error(sigkeys, infos):
@@ -203,16 +213,31 @@ class PackageSetBase(kobo.log.LoggingBase):
 
         return self.rpms_by_arch
 
-    def subset(self, primary_arch, arch_list, exclusive_noarch=True):
+    def subset(
+        self, primary_arch, arch_list, exclusive_noarch=True, inherit_to_noarch=True
+    ):
         """Create a subset of this package set that only includes
         packages compatible with"""
         pkgset = PackageSetBase(
             self.name, self.sigkey_ordering, logger=self._logger, arches=arch_list
         )
-        pkgset.merge(self, primary_arch, arch_list, exclusive_noarch=exclusive_noarch)
+        pkgset.merge(
+            self,
+            primary_arch,
+            arch_list,
+            exclusive_noarch=exclusive_noarch,
+            inherit_to_noarch=inherit_to_noarch,
+        )
         return pkgset
 
-    def merge(self, other, primary_arch, arch_list, exclusive_noarch=True):
+    def merge(
+        self,
+        other,
+        primary_arch,
+        arch_list,
+        exclusive_noarch=True,
+        inherit_to_noarch=True,
+    ):
         """
         Merge ``other`` package set into this instance.
         """
@@ -251,7 +276,7 @@ class PackageSetBase(kobo.log.LoggingBase):
                 if i.file_path in self.file_cache:
                     # TODO: test if it really works
                     continue
-                if exclusivearch_list and arch == "noarch":
+                if inherit_to_noarch and exclusivearch_list and arch == "noarch":
                     if is_excluded(i, exclusivearch_list, logger=self._logger):
                         continue
 
@@ -318,6 +343,11 @@ class FilelistPackageSet(PackageSetBase):
         return result
 
 
+# This is a marker to indicate package set with only extra builds/tasks and no
+# tasks.
+MISSING_KOJI_TAG = object()
+
+
 class KojiPackageSet(PackageSetBase):
     def __init__(
         self,
@@ -334,6 +364,7 @@ class KojiPackageSet(PackageSetBase):
         extra_tasks=None,
         signed_packages_retries=0,
         signed_packages_wait=30,
+        downloader=None,
     ):
         """
         Creates new KojiPackageSet.
@@ -371,7 +402,7 @@ class KojiPackageSet(PackageSetBase):
         :param int signed_packages_wait: How long to wait between search attemts.
         """
         super(KojiPackageSet, self).__init__(
-            name,
+            name if name != MISSING_KOJI_TAG else "no-tag",
             sigkey_ordering=sigkey_ordering,
             arches=arches,
             logger=logger,
@@ -388,6 +419,8 @@ class KojiPackageSet(PackageSetBase):
         self.signed_packages_retries = signed_packages_retries
         self.signed_packages_wait = signed_packages_wait
 
+        self.downloader = downloader
+
     def __getstate__(self):
         result = self.__dict__.copy()
         del result["koji_wrapper"]
@@ -478,7 +511,8 @@ class KojiPackageSet(PackageSetBase):
 
         response = None
         if self.cache_region:
-            cache_key = "KojiPackageSet.get_latest_rpms_%s_%s_%s" % (
+            cache_key = "%s.get_latest_rpms_%s_%s_%s" % (
+                str(self.__class__.__name__),
                 str(tag),
                 str(event),
                 str(inherit),
@@ -500,17 +534,36 @@ class KojiPackageSet(PackageSetBase):
 
         return response
 
+
+
     def get_package_path(self, queue_item):
         rpm_info, build_info = queue_item
 
         # Check if this RPM is coming from scratch task. In this case, we already
         # know the path.
         if "path_from_task" in rpm_info:
-            return rpm_info["path_from_task"]
+            return self.downloader.get_file(rpm_info["path_from_task"])
 
         pathinfo = self.koji_wrapper.koji_module.pathinfo
         paths = []
 
+        if "getRPMChecksums" in self.koji_proxy.system.listMethods():
+
+            def checksum_validator(keyname, pkg_path):
+                checksums = self.koji_proxy.getRPMChecksums(
+                    rpm_info["id"], checksum_types=("sha256",)
+                )
+                if "sha256" in checksums.get(keyname, {}):
+                    computed = compute_file_checksums(pkg_path, ("sha256",))
+                    if computed["sha256"] != checksums[keyname]["sha256"]:
+                        raise RuntimeError("Checksum mismatch for %s" % pkg_path)
+
+        else:
+
+            def checksum_validator(keyname, pkg_path):
+                # Koji doesn't support checksums yet
+                pass
+
         attempts_left = self.signed_packages_retries + 1
         while attempts_left > 0:
             for sigkey in self.sigkey_ordering:
@@ -523,8 +576,11 @@ class KojiPackageSet(PackageSetBase):
                 )
                 if rpm_path not in paths:
                     paths.append(rpm_path)
-                if os.path.isfile(rpm_path):
-                    return rpm_path
+                path = self.downloader.get_file(
+                    rpm_path, partial(checksum_validator, sigkey)
+                )
+                if path:
+                    return path
 
             # No signed copy was found, wait a little and try again.
             attempts_left -= 1
@@ -537,16 +593,18 @@ class KojiPackageSet(PackageSetBase):
             # use an unsigned copy (if allowed)
             rpm_path = os.path.join(pathinfo.build(build_info), pathinfo.rpm(rpm_info))
             paths.append(rpm_path)
-            if os.path.isfile(rpm_path):
-                return rpm_path
+            path = self.downloader.get_file(rpm_path, partial(checksum_validator, ""))
+            if path:
+                return path
 
         if self._allow_invalid_sigkeys and rpm_info["name"] not in self.packages:
             # use an unsigned copy (if allowed)
             rpm_path = os.path.join(pathinfo.build(build_info), pathinfo.rpm(rpm_info))
             paths.append(rpm_path)
-            if os.path.isfile(rpm_path):
+            path = self.downloader.get_file(rpm_path)
+            if path:
                 self._invalid_sigkey_rpms.append(rpm_info)
-                return rpm_path
+                return path
 
         self._invalid_sigkey_rpms.append(rpm_info)
         self.log_error(
@@ -576,7 +634,9 @@ class KojiPackageSet(PackageSetBase):
             inherit,
         )
         self.log_info("[BEGIN] %s" % msg)
-        rpms, builds = self.get_latest_rpms(tag, event, inherit=inherit)
+        rpms, builds = [], []
+        if tag != MISSING_KOJI_TAG:
+            rpms, builds = self.get_latest_rpms(tag, event, inherit=inherit)
         extra_rpms, extra_builds = self.get_extra_rpms()
         rpms += extra_rpms
         builds += extra_builds
@@ -681,6 +741,15 @@ class KojiPackageSet(PackageSetBase):
         :param include_packages: an iterable of tuples (package name, arch) that should
                                  be included.
         """
+        if len(self.sigkey_ordering) > 1 and (
+            None in self.sigkey_ordering or "" in self.sigkey_ordering
+        ):
+            self.log_warning(
+                "Stop writing reuse file as unsigned packages are allowed "
+                "in the compose."
+            )
+            return
+
         reuse_file = compose.paths.work.pkgset_reuse_file(self.name)
         self.log_info("Writing pkgset reuse file: %s" % reuse_file)
         try:
@@ -697,6 +766,12 @@ class KojiPackageSet(PackageSetBase):
                         "srpms_by_name": self.srpms_by_name,
                         "extra_builds": self.extra_builds,
                         "include_packages": include_packages,
+                        "inherit_to_noarch": compose.conf[
+                            "pkgset_inherit_exclusive_arch_to_noarch"
+                        ],
+                        "exclusive_noarch": compose.conf[
+                            "pkgset_exclusive_arch_considers_noarch"
+                        ],
                     },
                     f,
                     protocol=pickle.HIGHEST_PROTOCOL,
@@ -791,6 +866,8 @@ class KojiPackageSet(PackageSetBase):
             self.log_debug("Failed to load reuse file: %s" % str(e))
             return False
 
+        inherit_to_noarch = compose.conf["pkgset_inherit_exclusive_arch_to_noarch"]
+        exclusive_noarch = compose.conf["pkgset_exclusive_arch_considers_noarch"]
         if (
             reuse_data["allow_invalid_sigkeys"] == self._allow_invalid_sigkeys
             and reuse_data["packages"] == self.packages
@@ -798,6 +875,10 @@ class KojiPackageSet(PackageSetBase):
             and reuse_data["extra_builds"] == self.extra_builds
             and reuse_data["sigkeys"] == self.sigkey_ordering
             and reuse_data["include_packages"] == include_packages
+            # If the value is not present in reuse data, the compose was
+            # generated with older version of Pungi. Best to not reuse.
+            and reuse_data.get("inherit_to_noarch") == inherit_to_noarch
+            and reuse_data.get("exclusive_noarch") == exclusive_noarch
         ):
             self.log_info("Copying repo data for reuse: %s" % old_repo_dir)
             copy_all(old_repo_dir, repo_dir)
@@ -812,6 +893,67 @@ class KojiPackageSet(PackageSetBase):
             return False
 
 
+class KojiMockPackageSet(KojiPackageSet):
+
+    def _is_rpm_signed(self, rpm_path) -> bool:
+        ts = rpm.TransactionSet()
+        ts.setVSFlags(rpm._RPMVSF_NOSIGNATURES)
+        sigkeys = [
+            sigkey.lower() for sigkey in self.sigkey_ordering
+            if sigkey is not None
+        ]
+        if not sigkeys:
+            return True
+        with open(rpm_path, 'rb') as fd:
+            header = ts.hdrFromFdno(fd)
+            signature = header[rpm.RPMTAG_SIGGPG] or header[rpm.RPMTAG_SIGPGP]
+            if signature is None:
+                return False
+            pgp_msg = pgpy.PGPMessage.from_blob(signature)
+            return any(
+                signature.signer.lower() in sigkeys
+                for signature in pgp_msg.signatures
+            )
+
+    def get_package_path(self, queue_item):
+        rpm_info, build_info = queue_item
+
+        # Check if this RPM is coming from scratch task.
+        # In this case, we already know the path.
+        if "path_from_task" in rpm_info:
+            return rpm_info["path_from_task"]
+
+        # we replaced this part because pungi uses way
+        # of guessing path of package on koji based on sigkey
+        # we don't need that because all our packages will
+        # be ready for release
+        # signature verification is still done during deps resolution
+        pathinfo = self.koji_wrapper.koji_module.pathinfo
+
+        rpm_path = os.path.join(pathinfo.topdir, pathinfo.rpm(rpm_info))
+        if os.path.isfile(rpm_path):
+            if not self._is_rpm_signed(rpm_path):
+                self._invalid_sigkey_rpms.append(rpm_info)
+                self.log_error(
+                    'RPM "%s" not found for sigs: "%s". Path checked: "%s"',
+                    rpm_info, self.sigkey_ordering, rpm_path
+                )
+                return
+            return rpm_path
+        else:
+            self.log_warning("RPM %s not found" % rpm_path)
+            return None
+
+    def populate(self, tag, event=None, inherit=True, include_packages=None):
+        result = super().populate(
+            tag=tag,
+            event=event,
+            inherit=inherit,
+            include_packages=include_packages,
+        )
+        return result
+
+
 def _is_src(rpm_info):
     """Check if rpm info object returned by Koji refers to source packages."""
     return rpm_info["arch"] in ("src", "nosrc")

pungi/phases/pkgset/sources/__init__.py

@@ -15,8 +15,10 @@
 
 from .source_koji import PkgsetSourceKoji
 from .source_repos import PkgsetSourceRepos
+from .source_kojimock import PkgsetSourceKojiMock
 
 ALL_SOURCES = {
     "koji": PkgsetSourceKoji,
     "repos": PkgsetSourceRepos,
+    "kojimock": PkgsetSourceKojiMock,
 }

pungi/phases/pkgset/sources/source_koji.py

@@ -193,17 +193,13 @@ class PkgsetSourceKoji(pungi.phases.pkgset.source.PkgsetSourceBase):
     def __call__(self):
         compose = self.compose
         self.koji_wrapper = pungi.wrappers.kojiwrapper.KojiWrapper(compose)
-        # path prefix must contain trailing '/'
-        path_prefix = self.koji_wrapper.koji_module.config.topdir.rstrip("/") + "/"
-        package_sets = get_pkgset_from_koji(
-            self.compose, self.koji_wrapper, path_prefix
-        )
-        return (package_sets, path_prefix)
+        package_sets = get_pkgset_from_koji(self.compose, self.koji_wrapper)
+        return (package_sets, self.compose.koji_downloader.path_prefix)
 
 
-def get_pkgset_from_koji(compose, koji_wrapper, path_prefix):
+def get_pkgset_from_koji(compose, koji_wrapper):
     event_info = get_koji_event_info(compose, koji_wrapper)
-    return populate_global_pkgset(compose, koji_wrapper, path_prefix, event_info)
+    return populate_global_pkgset(compose, koji_wrapper, event_info)
 
 
 def _add_module_to_variant(
@@ -232,7 +228,7 @@ def _add_module_to_variant(
             continue
         typedir = koji_wrapper.koji_module.pathinfo.typedir(build, archive["btype"])
         filename = archive["filename"]
-        file_path = os.path.join(typedir, filename)
+        file_path = compose.koji_downloader.get_file(os.path.join(typedir, filename))
         try:
             # If there are two dots, the arch is in the middle. MBS uploads
             # files with actual architecture in the filename, but Pungi deals
@@ -270,9 +266,14 @@ def _add_module_to_variant(
                 "Module %s does not have metadata for arch %s and is not filtered "
                 "out via filter_modules option." % (nsvc, arch)
             )
-        mod_stream = read_single_module_stream_from_file(
-            mmds[filename], compose, arch, build
-        )
+        try:
+            mod_stream = read_single_module_stream_from_file(
+                mmds[filename], compose, arch, build
+            )
+        except Exception as exc:
+            # libmodulemd raises various GLib exceptions with not very helpful
+            # messages. Let's replace it with something more useful.
+            raise RuntimeError("Failed to read %s: %s", mmds[filename], str(exc))
         if mod_stream:
             added = True
         variant.arch_mmds.setdefault(arch, {})[nsvc] = mod_stream
@@ -395,7 +396,13 @@ def _is_filtered_out(compose, variant, arch, module_name, module_stream):
 
 
 def _get_modules_from_koji(
-    compose, koji_wrapper, event, variant, variant_tags, tag_to_mmd, exclude_module_ns
+    compose,
+    koji_wrapper,
+    event,
+    variant,
+    variant_tags,
+    tag_to_mmd,
+    exclude_module_ns,
 ):
     """
     Loads modules for given `variant` from koji `session`, adds them to
@@ -670,7 +677,7 @@ def _get_modules_from_koji_tags(
         )
 
 
-def populate_global_pkgset(compose, koji_wrapper, path_prefix, event):
+def populate_global_pkgset(compose, koji_wrapper, event):
     all_arches = get_all_arches(compose)
 
     # List of compose tags from which we create this compose
@@ -764,7 +771,12 @@ def populate_global_pkgset(compose, koji_wrapper, path_prefix, event):
 
         if extra_modules:
             _add_extra_modules_to_variant(
-                compose, koji_wrapper, variant, extra_modules, variant_tags, tag_to_mmd
+                compose,
+                koji_wrapper,
+                variant,
+                extra_modules,
+                variant_tags,
+                tag_to_mmd,
             )
 
         variant_scratch_modules = get_variant_data(
@@ -791,17 +803,23 @@ def populate_global_pkgset(compose, koji_wrapper, path_prefix, event):
 
     pkgsets = []
 
+    extra_builds = force_list(compose.conf.get("pkgset_koji_builds", []))
+    extra_tasks = force_list(compose.conf.get("pkgset_koji_scratch_tasks", []))
+
+    if not pkgset_koji_tags and (extra_builds or extra_tasks):
+        # We have extra packages to pull in, but no tag to merge them with.
+        compose_tags.append(pungi.phases.pkgset.pkgsets.MISSING_KOJI_TAG)
+        pkgset_koji_tags.append(pungi.phases.pkgset.pkgsets.MISSING_KOJI_TAG)
+
     # Get package set for each compose tag and merge it to global package
     # list. Also prepare per-variant pkgset, because we do not have list
     # of binary RPMs in module definition - there is just list of SRPMs.
     for compose_tag in compose_tags:
         compose.log_info("Loading package set for tag %s", compose_tag)
+        kwargs = {}
         if compose_tag in pkgset_koji_tags:
-            extra_builds = force_list(compose.conf.get("pkgset_koji_builds", []))
-            extra_tasks = force_list(compose.conf.get("pkgset_koji_scratch_tasks", []))
-        else:
-            extra_builds = []
-            extra_tasks = []
+            kwargs["extra_builds"] = extra_builds
+            kwargs["extra_tasks"] = extra_tasks
 
         pkgset = pungi.phases.pkgset.pkgsets.KojiPackageSet(
             compose_tag,
@@ -813,10 +831,10 @@ def populate_global_pkgset(compose, koji_wrapper, path_prefix, event):
             allow_invalid_sigkeys=allow_invalid_sigkeys,
             populate_only_packages=populate_only_packages_to_gather,
             cache_region=compose.cache_region,
-            extra_builds=extra_builds,
-            extra_tasks=extra_tasks,
             signed_packages_retries=compose.conf["signed_packages_retries"],
             signed_packages_wait=compose.conf["signed_packages_wait"],
+            downloader=compose.koji_downloader,
+            **kwargs
         )
 
         # Check if we have cache for this tag from previous compose. If so, use
@@ -880,7 +898,6 @@ def populate_global_pkgset(compose, koji_wrapper, path_prefix, event):
             )
         for variant in compose.all_variants.values():
             if compose_tag in variant_tags[variant]:
-
                 # If it's a modular tag, store the package set for the module.
                 for nsvc, koji_tag in variant.module_uid_to_koji_tag.items():
                     if compose_tag == koji_tag:
@@ -903,7 +920,7 @@ def populate_global_pkgset(compose, koji_wrapper, path_prefix, event):
                 MaterializedPackageSet.create,
                 compose,
                 pkgset,
-                path_prefix,
+                compose.koji_downloader.path_prefix,
                 mmd=tag_to_mmd.get(pkgset.name),
             )
         )

pungi/runroot.py

@@ -94,7 +94,7 @@ class Runroot(kobo.log.LoggingBase):
         log_file = os.path.join(log_dir, "program.log")
         try:
             with open(log_file) as f:
-                for line in f:
+                for line in f.readlines():
                     if "losetup: cannot find an unused loop device" in line:
                         return True
                     if re.match("losetup: .* failed to set up loop device", line):

pungi/scripts/cache_cleanup.py

@@ -0,0 +1,63 @@
+import argparse
+import os
+import re
+import time
+
+from pungi.util import format_size
+
+
+LOCK_RE = re.compile(r".*\.lock(\|[A-Za-z0-9]+)*$")
+
+
+def should_be_cleaned_up(path, st, threshold):
+    if st.st_nlink == 1 and st.st_mtime < threshold:
+        # No other instances, older than limit
+        return True
+
+    if LOCK_RE.match(path) and st.st_mtime < threshold:
+        # Suspiciously old lock
+        return True
+
+    return False
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("CACHE_DIR")
+    parser.add_argument("-n", "--dry-run", action="store_true")
+    parser.add_argument("--verbose", action="store_true")
+    parser.add_argument(
+        "--max-age",
+        help="how old files should be considered for deletion",
+        default=7,
+        type=int,
+    )
+
+    args = parser.parse_args()
+
+    topdir = os.path.abspath(args.CACHE_DIR)
+    max_age = args.max_age * 24 * 3600
+
+    cleaned_up = 0
+
+    threshold = time.time() - max_age
+    for dirpath, dirnames, filenames in os.walk(topdir):
+        for f in filenames:
+            filepath = os.path.join(dirpath, f)
+            st = os.stat(filepath)
+            if should_be_cleaned_up(filepath, st, threshold):
+                if args.verbose:
+                    print("RM %s" % filepath)
+                cleaned_up += st.st_size
+                if not args.dry_run:
+                    os.remove(filepath)
+        if not dirnames and not filenames:
+            if args.verbose:
+                print("RMDIR %s" % dirpath)
+            if not args.dry_run:
+                os.rmdir(dirpath)
+
+    if args.dry_run:
+        print("Would reclaim %s bytes." % format_size(cleaned_up))
+    else:
+        print("Reclaimed %s bytes." % format_size(cleaned_up))

pungi/scripts/config_dump.py

@@ -171,32 +171,11 @@ def main():
     group.add_argument(
         "--offline", action="store_true", help="Do not resolve git references."
     )
-    parser.add_argument(
-        "--multi",
-        metavar="DIR",
-        help=(
-            "Treat source as config for pungi-orchestrate and store dump into "
-            "given directory."
-        ),
-    )
 
     args = parser.parse_args()
 
     defines = config_utils.extract_defines(args.define)
 
-    if args.multi:
-        if len(args.sources) > 1:
-            parser.error("Only one multi config can be specified.")
-
-        return dump_multi_config(
-            args.sources[0],
-            dest=args.multi,
-            defines=defines,
-            just_dump=args.just_dump,
-            event=args.freeze_event,
-            offline=args.offline,
-        )
-
     return process_file(
         args.sources,
         defines=defines,

pungi/scripts/create_extra_repo.py

@@ -0,0 +1,441 @@
+# coding=utf-8
+
+import argparse
+import os
+import subprocess
+import tempfile
+from shutil import rmtree
+from typing import (
+    AnyStr,
+    List,
+    Dict,
+    Optional,
+)
+
+import createrepo_c as cr
+import requests
+import yaml
+from dataclasses import dataclass, field
+
+from .create_packages_json import (
+    PackagesGenerator,
+    RepoInfo,
+    VariantInfo,
+)
+
+
+@dataclass
+class ExtraVariantInfo(VariantInfo):
+
+    modules: List[AnyStr] = field(default_factory=list)
+    packages: List[AnyStr] = field(default_factory=list)
+
+
+class CreateExtraRepo(PackagesGenerator):
+
+    def __init__(
+            self,
+            variants: List[ExtraVariantInfo],
+            bs_auth_token: AnyStr,
+            local_repository_path: AnyStr,
+            clear_target_repo: bool = True,
+    ):
+        self.variants = []  # type: List[ExtraVariantInfo]
+        super().__init__(variants, [], [])
+        self.auth_headers = {
+            'Authorization': f'Bearer {bs_auth_token}',
+        }
+        # modules data of modules.yaml.gz from an existing local repo
+        self.local_modules_data = []
+        self.local_repository_path = local_repository_path
+        # path to modules.yaml, which generated by the class
+        self.default_modules_yaml_path = os.path.join(
+            local_repository_path,
+            'modules.yaml',
+        )
+        if clear_target_repo:
+            if os.path.exists(self.local_repository_path):
+                rmtree(self.local_repository_path)
+            os.makedirs(self.local_repository_path, exist_ok=True)
+        else:
+            self._read_local_modules_yaml()
+
+    def _read_local_modules_yaml(self):
+        """
+        Read modules data from an existin local repo
+        """
+        repomd_file_path = os.path.join(
+            self.local_repository_path,
+            'repodata',
+            'repomd.xml',
+        )
+        repomd_object = self._parse_repomd(repomd_file_path)
+        for repomd_record in repomd_object.records:
+            if repomd_record.type != 'modules':
+                continue
+            modules_yaml_path = os.path.join(
+                self.local_repository_path,
+                repomd_record.location_href,
+            )
+            self.local_modules_data = list(self._parse_modules_file(
+                modules_yaml_path,
+            ))
+            break
+
+    def _dump_local_modules_yaml(self):
+        """
+        Dump merged modules data to an local repo
+        """
+        if self.local_modules_data:
+            with open(self.default_modules_yaml_path, 'w') as yaml_file:
+                yaml.dump_all(
+                    self.local_modules_data,
+                    yaml_file,
+                )
+
+    @staticmethod
+    def get_repo_info_from_bs_repo(
+            auth_token: AnyStr,
+            build_id: AnyStr,
+            arch: AnyStr,
+            packages: Optional[List[AnyStr]] = None,
+            modules: Optional[List[AnyStr]] = None,
+    ) -> List[ExtraVariantInfo]:
+        """
+        Get info about a BS repo and save it to
+            an object of class ExtraRepoInfo
+        :param auth_token: Auth token to Build System
+        :param build_id: ID of a build from BS
+        :param arch: an architecture of repo which will be used
+        :param packages: list of names of packages which will be put to an
+                         local repo from a BS repo
+        :param modules: list of names of modules which will be put to an
+                        local repo from a BS repo
+        :return: list of ExtraRepoInfo with info about the BS repos
+        """
+
+        bs_url = 'https://build.cloudlinux.com'
+        api_uri = 'api/v1'
+        bs_repo_suffix = 'build_repos'
+
+        variants_info = []
+
+        # get the full info about a BS repo
+        repo_request = requests.get(
+            url=os.path.join(
+                bs_url,
+                api_uri,
+                'builds',
+                build_id,
+            ),
+            headers={
+                'Authorization': f'Bearer {auth_token}',
+            },
+        )
+        repo_request.raise_for_status()
+        result = repo_request.json()
+        for build_platform in result['build_platforms']:
+            platform_name = build_platform['name']
+            for architecture in build_platform['architectures']:
+                # skip repo with unsuitable architecture
+                if architecture != arch:
+                    continue
+                variant_info = ExtraVariantInfo(
+                    name=f'{build_id}-{platform_name}-{architecture}',
+                    arch=architecture,
+                    packages=packages,
+                    modules=modules,
+                    repos=[
+                        RepoInfo(
+                            path=os.path.join(
+                                bs_url,
+                                bs_repo_suffix,
+                                build_id,
+                                platform_name,
+                            ),
+                            folder=architecture,
+                            is_remote=True,
+                        )
+                    ]
+                )
+                variants_info.append(variant_info)
+        return variants_info
+
+    def _create_local_extra_repo(self):
+        """
+        Call `createrepo_c <path_to_repo>` for creating a local repo
+        """
+        subprocess.call(
+            f'createrepo_c {self.local_repository_path}',
+            shell=True,
+        )
+        # remove an unnecessary temporary modules.yaml
+        if os.path.exists(self.default_modules_yaml_path):
+            os.remove(self.default_modules_yaml_path)
+
+    def get_remote_file_content(
+            self,
+            file_url: AnyStr,
+    ) -> AnyStr:
+        """
+        Get content from a remote file and write it to a temp file
+        :param file_url: url of a remote file
+        :return: path to a temp file
+        """
+
+        file_request = requests.get(
+            url=file_url,
+            # for the case when we get a file from BS
+            headers=self.auth_headers,
+        )
+        file_request.raise_for_status()
+        with tempfile.NamedTemporaryFile(delete=False) as file_stream:
+            file_stream.write(file_request.content)
+        return file_stream.name
+
+    def _download_rpm_to_local_repo(
+            self,
+            package_location: AnyStr,
+            repo_info: RepoInfo,
+    ) -> None:
+        """
+        Download a rpm package from a remote repo and save it to a local repo
+        :param package_location: relative uri of a package in a remote repo
+        :param repo_info: info about a remote repo which contains a specific
+                          rpm package
+        """
+        rpm_package_remote_path = os.path.join(
+            repo_info.path,
+            repo_info.folder,
+            package_location,
+        )
+        rpm_package_local_path = os.path.join(
+            self.local_repository_path,
+            os.path.basename(package_location),
+        )
+        rpm_request = requests.get(
+            url=rpm_package_remote_path,
+            headers=self.auth_headers,
+        )
+        rpm_request.raise_for_status()
+        with open(rpm_package_local_path, 'wb') as rpm_file:
+            rpm_file.write(rpm_request.content)
+
+    def _download_packages(
+            self,
+            packages: Dict[AnyStr, cr.Package],
+            variant_info: ExtraVariantInfo
+    ):
+        """
+        Download all defined packages from a remote repo
+        :param packages: information about all packages (including
+                         modularity) in a remote repo
+        :param variant_info: information about a remote variant
+        """
+        for package in packages.values():
+            package_name = package.name
+            # Skip a current package from a remote repo if we defined
+            # the list packages and a current package doesn't belong to it
+            if variant_info.packages and \
+                    package_name not in variant_info.packages:
+                continue
+            for repo_info in variant_info.repos:
+                self._download_rpm_to_local_repo(
+                    package_location=package.location_href,
+                    repo_info=repo_info,
+                )
+
+    def _download_modules(
+            self,
+            modules_data: List[Dict],
+            variant_info: ExtraVariantInfo,
+            packages: Dict[AnyStr, cr.Package]
+    ):
+        """
+        Download all defined modularity packages and their data from
+            a remote repo
+        :param modules_data: information about all modules in a remote repo
+        :param variant_info: information about a remote variant
+        :param packages: information about all packages (including
+                         modularity) in a remote repo
+        """
+        for module in modules_data:
+            module_data = module['data']
+            # Skip a current module from a remote repo if we defined
+            # the list modules and a current module doesn't belong to it
+            if variant_info.modules and \
+                    module_data['name'] not in variant_info.modules:
+                continue
+            # we should add info about a module if the local repodata
+            # doesn't have it
+            if module not in self.local_modules_data:
+                self.local_modules_data.append(module)
+            # just skip a module's record if it doesn't have rpm artifact
+            if module['document'] != 'modulemd' or \
+                    'artifacts' not in module_data or \
+                    'rpms' not in module_data['artifacts']:
+                continue
+            for rpm in module['data']['artifacts']['rpms']:
+                # Empty repo_info.packages means that we will download
+                # all packages from repo including
+                # the modularity packages
+                if not variant_info.packages:
+                    break
+                # skip a rpm if it doesn't belong to a processed repo
+                if rpm not in packages:
+                    continue
+                for repo_info in variant_info.repos:
+                    self._download_rpm_to_local_repo(
+                        package_location=packages[rpm].location_href,
+                        repo_info=repo_info,
+                    )
+
+    def create_extra_repo(self):
+        """
+        1. Get from the remote repos the specific (or all) packages/modules
+        2. Save them to a local repo
+        3. Save info about the modules to a local repo
+        3. Call `createrepo_c` which creates a local repo
+           with the right repodata
+        """
+        for variant_info in self.variants:
+            for repo_info in variant_info.repos:
+                repomd_records = self._get_repomd_records(
+                    repo_info=repo_info,
+                )
+                packages_iterator = self.get_packages_iterator(repo_info)
+                # parse the repodata (including modules.yaml.gz)
+                modules_data = self._parse_module_repomd_record(
+                    repo_info=repo_info,
+                    repomd_records=repomd_records,
+                )
+                # convert the packages dict to more usable form
+                # for future checking that a rpm from the module's artifacts
+                # belongs to a processed repository
+                packages = {
+                    f'{package.name}-{package.epoch}:{package.version}-'
+                    f'{package.release}.{package.arch}':
+                        package for package in packages_iterator
+                }
+                self._download_modules(
+                    modules_data=modules_data,
+                    variant_info=variant_info,
+                    packages=packages,
+                )
+                self._download_packages(
+                    packages=packages,
+                    variant_info=variant_info,
+                )
+
+        self._dump_local_modules_yaml()
+        self._create_local_extra_repo()
+
+
+def create_parser():
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        '--bs-auth-token',
+        help='Auth token for Build System',
+    )
+    parser.add_argument(
+        '--local-repo-path',
+        help='Path to a local repo. E.g. /var/repo/test_repo',
+        required=True,
+    )
+    parser.add_argument(
+        '--clear-local-repo',
+        help='Clear a local repo before creating a new',
+        action='store_true',
+        default=False,
+    )
+    parser.add_argument(
+        '--repo',
+        action='append',
+        help='Path to a folder with repofolders or build id. E.g. '
+             '"http://koji.cloudlinux.com/mirrors/rhel_mirror" or '
+             '"601809b3c2f5b0e458b14cd3"',
+        required=True,
+    )
+    parser.add_argument(
+        '--repo-folder',
+        action='append',
+        help='A folder which contains folder repodata . E.g. "baseos-stream"',
+        required=True,
+    )
+    parser.add_argument(
+        '--repo-arch',
+        action='append',
+        help='What architecture packages a repository contains. E.g. "x86_64"',
+        required=True,
+    )
+    parser.add_argument(
+        '--packages',
+        action='append',
+        type=str,
+        default=[],
+        help='A list of packages names which we want to download to local '
+             'extra repo. We will download all of packages if param is empty',
+        required=True,
+    )
+    parser.add_argument(
+        '--modules',
+        action='append',
+        type=str,
+        default=[],
+        help='A list of modules names which we want to download to local '
+             'extra repo. We will download all of modules if param is empty',
+        required=True,
+    )
+
+    return parser
+
+
+def cli_main():
+    args = create_parser().parse_args()
+    repos_info = []
+    for repo, repo_folder, repo_arch, packages, modules in zip(
+        args.repo,
+        args.repo_folder,
+        args.repo_arch,
+        args.packages,
+        args.modules,
+    ):
+        modules = modules.split()
+        packages = packages.split()
+        if repo.startswith('http://'):
+            repos_info.append(
+                ExtraVariantInfo(
+                    name=repo_folder,
+                    arch=repo_arch,
+                    repos=[
+                        RepoInfo(
+                            path=repo,
+                            folder=repo_folder,
+                            is_remote=True,
+                        )
+                    ],
+                    modules=modules,
+                    packages=packages,
+                )
+            )
+        else:
+            repos_info.extend(
+                CreateExtraRepo.get_repo_info_from_bs_repo(
+                    auth_token=args.bs_auth_token,
+                    build_id=repo,
+                    arch=repo_arch,
+                    modules=modules,
+                    packages=packages,
+                )
+            )
+    cer = CreateExtraRepo(
+        variants=repos_info,
+        bs_auth_token=args.bs_auth_token,
+        local_repository_path=args.local_repo_path,
+        clear_target_repo=args.clear_local_repo,
+    )
+    cer.create_extra_repo()
+
+
+if __name__ == '__main__':
+    cli_main()

pungi/scripts/create_packages_json.py

@@ -0,0 +1,514 @@
+# coding=utf-8
+"""
+The tool allow to generate package.json. This file is used by pungi
+# as parameter `gather_prepopulate`
+Sample of using repodata files taken from
+https://github.com/rpm-software-management/createrepo_c/blob/master/examples/python/repodata_parsing.py
+"""
+
+import argparse
+import gzip
+import json
+import logging
+import lzma
+import os
+import re
+import tempfile
+from collections import defaultdict
+from itertools import tee
+from pathlib import Path
+from typing import (
+    AnyStr,
+    Dict,
+    List,
+    Any,
+    Iterator,
+    Optional,
+    Tuple,
+    Union,
+)
+
+import binascii
+from urllib.parse import urljoin
+
+import requests
+import rpm
+import yaml
+from createrepo_c import (
+    Package,
+    PackageIterator,
+    Repomd,
+    RepomdRecord,
+)
+from dataclasses import dataclass, field
+from kobo.rpmlib import parse_nvra
+
+logging.basicConfig(level=logging.INFO)
+
+
+def _is_compressed_file(first_two_bytes: bytes, initial_bytes: bytes):
+    return binascii.hexlify(first_two_bytes) == initial_bytes
+
+
+def is_gzip_file(first_two_bytes):
+    return _is_compressed_file(
+        first_two_bytes=first_two_bytes,
+        initial_bytes=b'1f8b',
+    )
+
+
+def is_xz_file(first_two_bytes):
+    return _is_compressed_file(
+        first_two_bytes=first_two_bytes,
+        initial_bytes=b'fd37',
+    )
+
+
+@dataclass
+class RepoInfo:
+    # path to a directory with repo directories. E.g. '/var/repos' contains
+    # 'appstream', 'baseos', etc.
+    # Or 'http://koji.cloudlinux.com/mirrors/rhel_mirror' if you are
+    # using remote repo
+    path: str
+    # name of folder with a repodata folder. E.g. 'baseos', 'appstream', etc
+    folder: str
+    # Is a repo remote or local
+    is_remote: bool
+    # Is a reference repository (usually it's a RHEL repo)
+    # Layout of packages from such repository will be taken as example
+    # Only layout of specific package (which doesn't exist
+    # in a reference repository) will be taken as example
+    is_reference: bool = False
+    # The packages from 'present' repo will be added to a variant.
+    # The packages from 'absent' repo will be removed from a variant.
+    repo_type: str = 'present'
+
+
+@dataclass
+class VariantInfo:
+    # name of variant. E.g. 'BaseOS', 'AppStream', etc
+    name: AnyStr
+    # architecture of variant. E.g. 'x86_64', 'i686', etc
+    arch: AnyStr
+    # The packages which will be not added to a variant
+    excluded_packages: List[str] = field(default_factory=list)
+    # Repos of a variant
+    repos: List[RepoInfo] = field(default_factory=list)
+
+
+class PackagesGenerator:
+
+    repo_arches = defaultdict(lambda: list(('noarch',)))
+    addon_repos = {
+        'x86_64': ['i686'],
+        'ppc64le': [],
+        'aarch64': [],
+        's390x': [],
+        'i686': [],
+    }
+
+    def __init__(
+        self,
+        variants: List[VariantInfo],
+        excluded_packages: List[AnyStr],
+        included_packages: List[AnyStr],
+    ):
+        self.variants = variants
+        self.pkgs = dict()
+        self.excluded_packages = excluded_packages
+        self.included_packages = included_packages
+        self.tmp_files = []  # type: list[Path]
+        for arch, arch_list in self.addon_repos.items():
+            self.repo_arches[arch].extend(arch_list)
+            self.repo_arches[arch].append(arch)
+
+    def __del__(self):
+        for tmp_file in self.tmp_files:
+            if tmp_file.exists():
+                tmp_file.unlink()
+
+    @staticmethod
+    def _get_full_repo_path(repo_info: RepoInfo):
+        result = os.path.join(
+            repo_info.path,
+            repo_info.folder
+        )
+        if repo_info.is_remote:
+            result = urljoin(
+                repo_info.path + '/',
+                repo_info.folder,
+            )
+        return result
+
+    @staticmethod
+    def _warning_callback(warning_type, message):
+        """
+        Warning callback for createrepo_c parsing functions
+        """
+        print(f'Warning message: "{message}"; warning type: "{warning_type}"')
+        return True
+
+    def get_remote_file_content(self, file_url: AnyStr) -> AnyStr:
+        """
+        Get content from a remote file and write it to a temp file
+        :param file_url: url of a remote file
+        :return: path to a temp file
+        """
+
+        file_request = requests.get(
+            url=file_url,
+        )
+        file_request.raise_for_status()
+        with tempfile.NamedTemporaryFile(delete=False) as file_stream:
+            file_stream.write(file_request.content)
+        self.tmp_files.append(Path(file_stream.name))
+        return file_stream.name
+
+    @staticmethod
+    def _parse_repomd(repomd_file_path: AnyStr) -> Repomd:
+        """
+        Parse file repomd.xml and create object Repomd
+        :param repomd_file_path: path to local repomd.xml
+        """
+        return Repomd(repomd_file_path)
+
+    @classmethod
+    def _parse_modules_file(
+            cls,
+            modules_file_path: AnyStr,
+
+    ) -> Iterator[Any]:
+        """
+        Parse modules.yaml.gz and returns parsed data
+        :param modules_file_path: path to local modules.yaml.gz
+        :return: List of dict for each module in a repo
+        """
+
+        with open(modules_file_path, 'rb') as modules_file:
+            data = modules_file.read()
+            if is_gzip_file(data[:2]):
+                data = gzip.decompress(data)
+            elif is_xz_file(data[:2]):
+                data = lzma.decompress(data)
+            return yaml.load_all(
+                data,
+                Loader=yaml.BaseLoader,
+            )
+
+    def _get_repomd_records(
+            self,
+            repo_info: RepoInfo,
+    ) -> List[RepomdRecord]:
+        """
+        Get, parse file repomd.xml and extract from it repomd records
+        :param repo_info: structure which contains info about a current repo
+        :return: list with repomd records
+        """
+        repomd_file_path = os.path.join(
+            repo_info.path,
+            repo_info.folder,
+            'repodata',
+            'repomd.xml',
+        )
+        if repo_info.is_remote:
+            repomd_file_path = urljoin(
+                urljoin(
+                    repo_info.path + '/',
+                    repo_info.folder
+                ) + '/',
+                'repodata/repomd.xml'
+            )
+            repomd_file_path = self.get_remote_file_content(repomd_file_path)
+
+        repomd_object = self._parse_repomd(repomd_file_path)
+        if repo_info.is_remote:
+            os.remove(repomd_file_path)
+        return repomd_object.records
+
+    def _download_repomd_records(
+            self,
+            repo_info: RepoInfo,
+            repomd_records: List[RepomdRecord],
+            repomd_records_dict: Dict[str, str],
+    ):
+        """
+        Download repomd records
+        :param repo_info: structure which contains info about a current repo
+        :param repomd_records: list with repomd records
+        :param repomd_records_dict: dict with paths to repodata files
+        """
+        for repomd_record in repomd_records:
+            if repomd_record.type not in (
+                    'primary',
+                    'filelists',
+                    'other',
+            ):
+                continue
+            repomd_record_file_path = os.path.join(
+                repo_info.path,
+                repo_info.folder,
+                repomd_record.location_href,
+            )
+            if repo_info.is_remote:
+                repomd_record_file_path = self.get_remote_file_content(
+                    repomd_record_file_path)
+            repomd_records_dict[repomd_record.type] = repomd_record_file_path
+
+    def _parse_module_repomd_record(
+            self,
+            repo_info: RepoInfo,
+            repomd_records: List[RepomdRecord],
+    ) -> List[Dict]:
+        """
+        Download repomd records
+        :param repo_info: structure which contains info about a current repo
+        :param repomd_records: list with repomd records
+        """
+        for repomd_record in repomd_records:
+            if repomd_record.type != 'modules':
+                continue
+            repomd_record_file_path = os.path.join(
+                repo_info.path,
+                repo_info.folder,
+                repomd_record.location_href,
+            )
+            if repo_info.is_remote:
+                repomd_record_file_path = self.get_remote_file_content(
+                    repomd_record_file_path)
+            return list(self._parse_modules_file(
+                repomd_record_file_path,
+            ))
+        return []
+
+    @staticmethod
+    def compare_pkgs_version(package_1: Package, package_2: Package) -> int:
+        version_tuple_1 = (
+            package_1.epoch,
+            package_1.version,
+            package_1.release,
+        )
+        version_tuple_2 = (
+            package_2.epoch,
+            package_2.version,
+            package_2.release,
+        )
+        return rpm.labelCompare(version_tuple_1, version_tuple_2)
+
+    def get_packages_iterator(
+            self,
+            repo_info: RepoInfo,
+    ) -> Union[PackageIterator, Iterator]:
+        full_repo_path = self._get_full_repo_path(repo_info)
+        pkgs_iterator = self.pkgs.get(full_repo_path)
+        if pkgs_iterator is None:
+            repomd_records = self._get_repomd_records(
+                repo_info=repo_info,
+            )
+            repomd_records_dict = {}  # type: Dict[str, str]
+            self._download_repomd_records(
+                repo_info=repo_info,
+                repomd_records=repomd_records,
+                repomd_records_dict=repomd_records_dict,
+            )
+            pkgs_iterator = PackageIterator(
+                primary_path=repomd_records_dict['primary'],
+                filelists_path=repomd_records_dict['filelists'],
+                other_path=repomd_records_dict['other'],
+                warningcb=self._warning_callback,
+            )
+            pkgs_iterator, self.pkgs[full_repo_path] = tee(pkgs_iterator)
+        return pkgs_iterator
+
+    def get_package_arch(
+            self,
+            package: Package,
+            variant_arch: str,
+    ) -> str:
+        result = variant_arch
+        if package.arch in self.repo_arches[variant_arch]:
+            result = package.arch
+        return result
+
+    def is_skipped_module_package(
+            self,
+            package: Package,
+            variant_arch: str,
+    ) -> bool:
+        package_key = self.get_package_key(package, variant_arch)
+        # Even a module package will be added to packages.json if
+        # it presents in the list of included packages
+        return 'module' in package.release and not any(
+            re.search(
+                f'^{included_pkg}$',
+                package_key,
+            ) or included_pkg in (package.name, package_key)
+            for included_pkg in self.included_packages
+        )
+
+    def is_excluded_package(
+            self,
+            package: Package,
+            variant_arch: str,
+            excluded_packages: List[str],
+    ) -> bool:
+        package_key = self.get_package_key(package, variant_arch)
+        return any(
+            re.search(
+                f'^{excluded_pkg}$',
+                package_key,
+            ) or excluded_pkg in (package.name, package_key)
+            for excluded_pkg in excluded_packages
+        )
+
+    @staticmethod
+    def get_source_rpm_name(package: Package) -> str:
+        source_rpm_nvra = parse_nvra(package.rpm_sourcerpm)
+        return source_rpm_nvra['name']
+
+    def get_package_key(self, package: Package, variant_arch: str) -> str:
+        return (
+            f'{package.name}.'
+            f'{self.get_package_arch(package, variant_arch)}'
+        )
+
+    def generate_packages_json(
+            self
+    ) -> Dict[AnyStr, Dict[AnyStr, Dict[AnyStr, List[AnyStr]]]]:
+        """
+        Generate packages.json
+        """
+        packages = defaultdict(lambda: defaultdict(lambda: {
+            'variants': list(),
+        }))
+        for variant_info in self.variants:
+            for repo_info in variant_info.repos:
+                is_reference = repo_info.is_reference
+                for package in self.get_packages_iterator(repo_info=repo_info):
+                    if self.is_skipped_module_package(
+                            package=package,
+                            variant_arch=variant_info.arch,
+                    ):
+                        continue
+                    if self.is_excluded_package(
+                        package=package,
+                        variant_arch=variant_info.arch,
+                        excluded_packages=self.excluded_packages,
+                    ):
+                        continue
+                    if self.is_excluded_package(
+                        package=package,
+                        variant_arch=variant_info.arch,
+                        excluded_packages=variant_info.excluded_packages,
+                    ):
+                        continue
+                    package_key = self.get_package_key(
+                        package,
+                        variant_info.arch,
+                    )
+                    source_rpm_name = self.get_source_rpm_name(package)
+                    package_info = packages[source_rpm_name][package_key]
+                    if 'is_reference' not in package_info:
+                        package_info['variants'].append(variant_info.name)
+                        package_info['is_reference'] = is_reference
+                        package_info['package'] = package
+                    elif not package_info['is_reference'] or \
+                            package_info['is_reference'] == is_reference and \
+                            self.compare_pkgs_version(
+                        package_1=package,
+                        package_2=package_info['package'],
+                    ) > 0:
+                        package_info['variants'] = [variant_info.name]
+                        package_info['is_reference'] = is_reference
+                        package_info['package'] = package
+                    elif self.compare_pkgs_version(
+                        package_1=package,
+                        package_2=package_info['package'],
+                    ) == 0 and repo_info.repo_type != 'absent':
+                        package_info['variants'].append(variant_info.name)
+        result = defaultdict(lambda: defaultdict(
+            lambda: defaultdict(list),
+        ))
+        for variant_info in self.variants:
+            for source_rpm_name, packages_info in packages.items():
+                for package_key, package_info in packages_info.items():
+                    variant_pkgs = result[variant_info.name][variant_info.arch]
+                    if variant_info.name not in package_info['variants']:
+                        continue
+                    variant_pkgs[source_rpm_name].append(package_key)
+        return result
+
+
+def create_parser():
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        '-c',
+        '--config',
+        type=Path,
+        default=Path('config.yaml'),
+        required=False,
+        help='Path to a config',
+    )
+    parser.add_argument(
+        '-o',
+        '--json-output-path',
+        type=str,
+        help='Full path to output json file',
+        required=True,
+    )
+
+    return parser
+
+
+def read_config(config_path: Path) -> Optional[Dict]:
+    if not config_path.exists():
+        logging.error('A config by path "%s" does not exist', config_path)
+        exit(1)
+    with config_path.open('r') as config_fd:
+        return yaml.safe_load(config_fd)
+
+
+def process_config(config_data: Dict) -> Tuple[
+    List[VariantInfo],
+    List[str],
+    List[str],
+]:
+    excluded_packages = config_data.get('excluded_packages', [])
+    included_packages = config_data.get('included_packages', [])
+    variants = [VariantInfo(
+        name=variant_name,
+        arch=variant_info['arch'],
+        excluded_packages=variant_info.get('excluded_packages', []),
+        repos=[RepoInfo(
+            path=variant_repo['path'],
+            folder=variant_repo['folder'],
+            is_remote=variant_repo['remote'],
+            is_reference=variant_repo['reference'],
+            repo_type=variant_repo.get('repo_type', 'present'),
+        ) for variant_repo in variant_info['repos']]
+    ) for variant_name, variant_info in config_data['variants'].items()]
+    return variants, excluded_packages, included_packages
+
+
+def cli_main():
+    args = create_parser().parse_args()
+    variants, excluded_packages, included_packages = process_config(
+        config_data=read_config(args.config)
+    )
+    pg = PackagesGenerator(
+        variants=variants,
+        excluded_packages=excluded_packages,
+        included_packages=included_packages,
+    )
+    result = pg.generate_packages_json()
+    with open(args.json_output_path, 'w') as packages_file:
+        json.dump(
+            result,
+            packages_file,
+            indent=4,
+            sort_keys=True,
+        )
+
+
+if __name__ == '__main__':
+    cli_main()

pungi/scripts/fedmsg_notification.py

@@ -14,6 +14,9 @@ def send(cmd, data):
     topic = "compose.%s" % cmd.replace("-", ".").lower()
     try:
         msg = fedora_messaging.api.Message(topic="pungi.{}".format(topic), body=data)
+        if cmd == "ostree":
+            # https://pagure.io/fedora-infrastructure/issue/10899
+            msg.priority = 3
         fedora_messaging.api.publish(msg)
     except fedora_messaging.exceptions.PublishReturned as e:
         print("Fedora Messaging broker rejected message %s: %s" % (msg.id, e))

pungi/scripts/gather_modules.py

@@ -0,0 +1,250 @@
+import gzip
+import lzma
+import os
+from argparse import ArgumentParser, FileType
+from glob import iglob
+from io import BytesIO
+from pathlib import Path
+from typing import List, AnyStr, Iterable, Union, Optional
+import logging
+from urllib.parse import urljoin
+
+import yaml
+import createrepo_c as cr
+from typing.io import BinaryIO
+
+from .create_packages_json import PackagesGenerator, is_gzip_file, is_xz_file
+
+EMPTY_FILE = '.empty'
+
+
+def read_modules_yaml(modules_yaml_path: Union[str, Path]) -> BytesIO:
+    with open(modules_yaml_path, 'rb') as fp:
+        return BytesIO(fp.read())
+
+
+def grep_list_of_modules_yaml(repos_path: AnyStr) -> Iterable[BytesIO]:
+    """
+    Find all of valid *modules.yaml.gz in repos
+    :param repos_path: path to a directory which contains repo dirs
+    :return: iterable object of content from *modules.yaml.*
+    """
+
+    return (
+        read_modules_yaml_from_specific_repo(repo_path=Path(path).parent)
+        for path in iglob(
+            str(Path(repos_path).joinpath('**/repodata')),
+            recursive=True
+        )
+    )
+
+
+def _is_remote(path: str):
+    return any(str(path).startswith(protocol)
+               for protocol in ('http', 'https'))
+
+
+def read_modules_yaml_from_specific_repo(
+        repo_path: Union[str, Path]
+) -> Optional[BytesIO]:
+    """
+    Read modules_yaml from a specific repo (remote or local)
+    :param repo_path: path/url to a specific repo
+                      (final dir should contain dir `repodata`)
+    :return: iterable object of content from *modules.yaml.*
+    """
+
+    if _is_remote(repo_path):
+        repomd_url = urljoin(
+            repo_path + '/',
+            'repodata/repomd.xml',
+        )
+        packages_generator = PackagesGenerator(
+            variants=[],
+            excluded_packages=[],
+            included_packages=[],
+        )
+        repomd_file_path = packages_generator.get_remote_file_content(
+            file_url=repomd_url
+        )
+    else:
+        repomd_file_path = os.path.join(
+            repo_path,
+            'repodata/repomd.xml',
+        )
+    repomd_obj = cr.Repomd(str(repomd_file_path))
+    for record in repomd_obj.records:
+        if record.type != 'modules':
+            continue
+        else:
+            if _is_remote(repo_path):
+                modules_yaml_url = urljoin(
+                    repo_path + '/',
+                    record.location_href,
+                )
+                modules_yaml_path = PackagesGenerator.get_remote_file_content(
+                    file_url=modules_yaml_url
+                )
+            else:
+                modules_yaml_path = os.path.join(
+                    repo_path,
+                    record.location_href,
+                )
+            return read_modules_yaml(modules_yaml_path=modules_yaml_path)
+    else:
+        return None
+
+
+def _should_grep_defaults(
+        document_type: str,
+        grep_only_modules_data: bool = False,
+        grep_only_modules_defaults_data: bool = False,
+) -> bool:
+    xor_flag = grep_only_modules_data == grep_only_modules_defaults_data
+    if document_type == 'modulemd' and (xor_flag or grep_only_modules_data):
+        return True
+    return False
+
+
+def _should_grep_modules(
+        document_type: str,
+        grep_only_modules_data: bool = False,
+        grep_only_modules_defaults_data: bool = False,
+) -> bool:
+    xor_flag = grep_only_modules_data == grep_only_modules_defaults_data
+    if document_type == 'modulemd-defaults' and \
+            (xor_flag or grep_only_modules_defaults_data):
+        return True
+    return False
+
+
+def collect_modules(
+        modules_paths: List[BinaryIO],
+        target_dir: str,
+        grep_only_modules_data: bool = False,
+        grep_only_modules_defaults_data: bool = False,
+):
+    """
+    Read given modules.yaml.gz files and export modules
+    and modulemd files from it.
+    Returns:
+        object:
+    """
+    xor_flag = grep_only_modules_defaults_data is grep_only_modules_data
+    modules_path = os.path.join(target_dir, 'modules')
+    module_defaults_path = os.path.join(target_dir, 'module_defaults')
+    if grep_only_modules_data or xor_flag:
+        os.makedirs(modules_path, exist_ok=True)
+    if grep_only_modules_defaults_data or xor_flag:
+        os.makedirs(module_defaults_path, exist_ok=True)
+    # Defaults modules can be empty, but pungi detects
+    # empty folder while copying and raises the exception in this case
+    Path(os.path.join(module_defaults_path, EMPTY_FILE)).touch()
+
+    for module_file in modules_paths:
+        data = module_file.read()
+        if is_gzip_file(data[:2]):
+            data = gzip.decompress(data)
+        elif is_xz_file(data[:2]):
+            data = lzma.decompress(data)
+        documents = yaml.load_all(data, Loader=yaml.BaseLoader)
+        for doc in documents:
+            path = None
+            if _should_grep_modules(
+                doc['document'],
+                grep_only_modules_data,
+                grep_only_modules_defaults_data,
+            ):
+                name = f"{doc['data']['module']}.yaml"
+                path = os.path.join(module_defaults_path, name)
+                logging.info('Found %s module defaults', name)
+            elif _should_grep_defaults(
+                    doc['document'],
+                    grep_only_modules_data,
+                    grep_only_modules_defaults_data,
+            ):
+                # pungi.phases.pkgset.sources.source_koji.get_koji_modules
+                stream = doc['data']['stream'].replace('-', '_')
+                doc_data = doc['data']
+                name = f"{doc_data['name']}-{stream}-" \
+                       f"{doc_data['version']}.{doc_data['context']}"
+                arch_dir = os.path.join(
+                    modules_path,
+                    doc_data['arch']
+                )
+                os.makedirs(arch_dir, exist_ok=True)
+                path = os.path.join(
+                    arch_dir,
+                    name,
+                )
+                logging.info('Found module %s', name)
+
+                if 'artifacts' not in doc['data']:
+                    logging.warning(
+                        'RPM %s does not have explicit list of artifacts',
+                        name
+                    )
+            if path is not None:
+                with open(path, 'w') as f:
+                    yaml.dump(doc, f, default_flow_style=False)
+
+
+def cli_main():
+    parser = ArgumentParser()
+    content_type_group = parser.add_mutually_exclusive_group(required=False)
+    content_type_group.add_argument(
+        '--get-only-modules-data',
+        action='store_true',
+        help='Parse and get only modules data',
+    )
+    content_type_group.add_argument(
+        '--get-only-modules-defaults-data',
+        action='store_true',
+        help='Parse and get only modules_defaults data',
+    )
+    path_group = parser.add_mutually_exclusive_group(required=True)
+    path_group.add_argument(
+        '-p', '--path',
+        type=FileType('rb'), nargs='+',
+        help='Path to modules.yaml.gz file. '
+             'You may pass multiple files by passing -p path1 path2'
+    )
+    path_group.add_argument(
+        '-rp', '--repo-path',
+        required=False,
+        type=str,
+        default=None,
+        help='Path to a directory which contains repodirs. E.g. /var/repos'
+    )
+    path_group.add_argument(
+        '-rd', '--repodata-paths',
+        required=False,
+        type=str,
+        nargs='+',
+        default=[],
+        help='Paths/urls to the directories with directory `repodata`',
+    )
+    parser.add_argument('-t', '--target', required=True)
+
+    namespace = parser.parse_args()
+    if namespace.repodata_paths:
+        modules = []
+        for repodata_path in namespace.repodata_paths:
+            modules.append(read_modules_yaml_from_specific_repo(
+                repodata_path,
+            ))
+    elif namespace.path is not None:
+        modules = namespace.path
+    else:
+        modules = grep_list_of_modules_yaml(namespace.repo_path)
+    modules = list(filter(lambda i: i is not None, modules))
+    collect_modules(
+        modules,
+        namespace.target,
+        namespace.get_only_modules_data,
+        namespace.get_only_modules_defaults_data,
+    )
+
+
+if __name__ == '__main__':
+    cli_main()

pungi/scripts/gather_rpms.py

@@ -0,0 +1,96 @@
+import re
+from argparse import ArgumentParser
+
+import os
+from glob import iglob
+from typing import List
+from pathlib import Path
+
+from dataclasses import dataclass
+from productmd.common import parse_nvra
+
+
+@dataclass
+class Package:
+    nvra: dict
+    path: Path
+
+
+def search_rpms(top_dir: Path) -> List[Package]:
+    """
+    Search for all *.rpm files recursively
+    in given top directory
+    Returns:
+        list: list of paths
+    """
+    return [Package(
+        nvra=parse_nvra(Path(path).stem),
+        path=Path(path),
+    ) for path in iglob(str(top_dir.joinpath('**/*.rpm')), recursive=True)]
+
+
+def is_excluded_package(
+        package: Package,
+        excluded_packages: List[str],
+) -> bool:
+    package_key = f'{package.nvra["name"]}.{package.nvra["arch"]}'
+    return any(
+        re.search(
+            f'^{excluded_pkg}$',
+            package_key,
+        ) or excluded_pkg in (package.nvra['name'], package_key)
+        for excluded_pkg in excluded_packages
+    )
+
+
+def copy_rpms(
+        packages: List[Package],
+        target_top_dir: Path,
+        excluded_packages: List[str],
+):
+    """
+    Search synced repos for rpms and prepare
+    koji-like structure for pungi
+
+    Instead of repos, use following structure:
+        # ls /mnt/koji/
+            i686/  noarch/  x86_64/
+    Returns:
+        Nothing:
+    """
+    for package in packages:
+        if is_excluded_package(package, excluded_packages):
+            continue
+        target_arch_dir = target_top_dir.joinpath(package.nvra['arch'])
+        target_file = target_arch_dir.joinpath(package.path.name)
+        os.makedirs(target_arch_dir, exist_ok=True)
+
+        if not target_file.exists():
+            try:
+                os.link(package.path, target_file)
+            except OSError:
+                # hardlink failed, try symlinking
+                package.path.symlink_to(target_file)
+
+
+def cli_main():
+    parser = ArgumentParser()
+    parser.add_argument('-p', '--path', required=True, type=Path)
+    parser.add_argument('-t', '--target', required=True, type=Path)
+    parser.add_argument(
+        '-e',
+        '--excluded-packages',
+        required=False,
+        nargs='+',
+        type=str,
+        default=[],
+    )
+
+    namespace = parser.parse_args()
+
+    rpms = search_rpms(namespace.path)
+    copy_rpms(rpms, namespace.target, namespace.excluded_packages)
+
+
+if __name__ == '__main__':
+    cli_main()

pungi/scripts/pungi.py

@@ -319,7 +319,6 @@ def get_arguments(config):
 
 
 def main():
-
     config = pungi.config.Config()
     opts = get_arguments(config)
 

pungi/scripts/pungi_koji.py

@@ -23,6 +23,7 @@ from pungi.phases import PHASES_NAMES
 from pungi import get_full_version, util
 from pungi.errors import UnsignedPackagesError
 from pungi.wrappers import kojiwrapper
+from pungi.util import rmtree
 
 
 # force C locales
@@ -251,9 +252,15 @@ def main():
         kobo.log.add_stderr_logger(logger)
 
     conf = util.load_config(opts.config)
-
     compose_type = opts.compose_type or conf.get("compose_type", "production")
-    if compose_type == "production" and not opts.label and not opts.no_label:
+    label = opts.label or conf.get("label")
+    if label:
+        try:
+            productmd.composeinfo.verify_label(label)
+        except ValueError as ex:
+            abort(str(ex))
+
+    if compose_type == "production" and not label and not opts.no_label:
         abort("must specify label for a production compose")
 
     if (
@@ -300,7 +307,12 @@ def main():
 
     if opts.target_dir:
         compose_dir = Compose.get_compose_dir(
-            opts.target_dir, conf, compose_type=compose_type, compose_label=opts.label
+            opts.target_dir,
+            conf,
+            compose_type=compose_type,
+            compose_label=label,
+            parent_compose_ids=opts.parent_compose_id,
+            respin_of=opts.respin_of,
         )
     else:
         compose_dir = opts.compose_dir
@@ -309,7 +321,7 @@ def main():
             ci = Compose.get_compose_info(
                 conf,
                 compose_type=compose_type,
-                compose_label=opts.label,
+                compose_label=label,
                 parent_compose_ids=opts.parent_compose_id,
                 respin_of=opts.respin_of,
             )
@@ -380,6 +392,14 @@ def run_compose(
     compose.log_info("Current timezone offset: %s" % pungi.util.get_tz_offset())
     compose.log_info("COMPOSE_ID=%s" % compose.compose_id)
 
+    installed_pkgs_log = compose.paths.log.log_file("global", "installed-pkgs")
+    compose.log_info("Logging installed packages to %s" % installed_pkgs_log)
+    try:
+        with open(installed_pkgs_log, "w") as f:
+            subprocess.Popen(["rpm", "-qa"], stdout=f)
+    except Exception as e:
+        compose.log_warning("Failed to log installed packages: %s" % str(e))
+
     compose.read_variants()
 
     # dump the config file
@@ -671,7 +691,7 @@ def cli_main():
     except (Exception, KeyboardInterrupt) as ex:
         if COMPOSE:
             COMPOSE.log_error("Compose run failed: %s" % ex)
-            COMPOSE.traceback()
+            COMPOSE.traceback(show_locals=getattr(ex, "show_locals", True))
             COMPOSE.log_critical("Compose failed: %s" % COMPOSE.topdir)
             COMPOSE.write_status("DOOMED")
         else:
@@ -680,3 +700,8 @@ def cli_main():
         sys.stdout.flush()
         sys.stderr.flush()
         sys.exit(1)
+    finally:
+        # Remove repositories cloned during ExtraFiles phase
+        process_id = os.getpid()
+        directoy_to_remove = "/tmp/pungi-temp-git-repos-" + str(process_id) + "/"
+        rmtree(directoy_to_remove)

pungi/util.py

@@ -279,7 +279,7 @@ class GitUrlResolveError(RuntimeError):
     pass
 
 
-def resolve_git_ref(repourl, ref):
+def resolve_git_ref(repourl, ref, credential_helper=None):
     """Resolve a reference in a Git repo to a commit.
 
     Raises RuntimeError if there was an error. Most likely cause is failure to
@@ -289,7 +289,7 @@ def resolve_git_ref(repourl, ref):
         # This looks like a commit ID already.
         return ref
     try:
-        _, output = git_ls_remote(repourl, ref)
+        _, output = git_ls_remote(repourl, ref, credential_helper)
     except RuntimeError as e:
         raise GitUrlResolveError(
             "ref does not exist in remote repo %s with the error %s %s"
@@ -316,7 +316,7 @@ def resolve_git_ref(repourl, ref):
     return lines[0].split()[0]
 
 
-def resolve_git_url(url):
+def resolve_git_url(url, credential_helper=None):
     """Given a url to a Git repo specifying HEAD or origin/<branch> as a ref,
     replace that specifier with actual SHA1 of the commit.
 
@@ -335,7 +335,7 @@ def resolve_git_url(url):
     scheme = r.scheme.replace("git+", "")
 
     baseurl = urllib.parse.urlunsplit((scheme, r.netloc, r.path, "", ""))
-    fragment = resolve_git_ref(baseurl, ref)
+    fragment = resolve_git_ref(baseurl, ref, credential_helper)
 
     result = urllib.parse.urlunsplit((r.scheme, r.netloc, r.path, r.query, fragment))
     if "?#" in url:
@@ -354,13 +354,18 @@ class GitUrlResolver(object):
         self.offline = offline
         self.cache = {}
 
-    def __call__(self, url, branch=None):
+    def __call__(self, url, branch=None, options=None):
+        credential_helper = options.get("credential_helper") if options else None
         if self.offline:
             return branch or url
         key = (url, branch)
         if key not in self.cache:
             try:
-                res = resolve_git_ref(url, branch) if branch else resolve_git_url(url)
+                res = (
+                    resolve_git_ref(url, branch, credential_helper)
+                    if branch
+                    else resolve_git_url(url, credential_helper)
+                )
                 self.cache[key] = res
             except GitUrlResolveError as exc:
                 self.cache[key] = exc
@@ -456,6 +461,9 @@ def get_volid(compose, arch, variant=None, disc_type=False, formats=None, **kwar
         if not variant_uid and "%(variant)s" in i:
             continue
         try:
+            # fmt: off
+            # Black wants to add a comma after kwargs, but that's not valid in
+            # Python 2.7
             args = get_format_substs(
                 compose,
                 variant=variant_uid,
@@ -467,6 +475,7 @@ def get_volid(compose, arch, variant=None, disc_type=False, formats=None, **kwar
                 base_product_version=base_product_version,
                 **kwargs
             )
+            # fmt: on
             volid = (i % args).format(**args)
         except KeyError as err:
             raise RuntimeError(
@@ -991,8 +1000,12 @@ def retry(timeout=120, interval=30, wait_on=Exception):
 
 
 @retry(wait_on=RuntimeError)
-def git_ls_remote(baseurl, ref):
-    return run(["git", "ls-remote", baseurl, ref], universal_newlines=True)
+def git_ls_remote(baseurl, ref, credential_helper=None):
+    cmd = ["git"]
+    if credential_helper:
+        cmd.extend(["-c", "credential.useHttpPath=true"])
+        cmd.extend(["-c", "credential.helper=%s" % credential_helper])
+    return run(cmd + ["ls-remote", baseurl, ref], universal_newlines=True)
 
 
 def get_tz_offset():
@@ -1137,3 +1150,16 @@ def read_json_file(file_path):
     """A helper function to read a JSON file."""
     with open(file_path) as f:
         return json.load(f)
+
+
+UNITS = ["", "Ki", "Mi", "Gi", "Ti"]
+
+
+def format_size(sz):
+    sz = float(sz)
+    unit = 0
+    while sz > 1024:
+        sz /= 1024
+        unit += 1
+
+    return "%.3g %sB" % (sz, UNITS[unit])

pungi/wrappers/comps.py

@@ -183,15 +183,16 @@ class CompsFilter(object):
         """
         all_groups = self.tree.xpath("/comps/group/id/text()") + lookaside_groups
         for environment in self.tree.xpath("/comps/environment"):
-            for group in environment.xpath("grouplist/groupid"):
-                if group.text not in all_groups:
-                    group.getparent().remove(group)
+            for parent_tag in ("grouplist", "optionlist"):
+                for group in environment.xpath("%s/groupid" % parent_tag):
+                    if group.text not in all_groups:
+                        group.getparent().remove(group)
 
-            for group in environment.xpath("grouplist/groupid[@arch]"):
-                value = group.attrib.get("arch")
-                values = [v for v in re.split(r"[, ]+", value) if v]
-                if arch not in values:
-                    group.getparent().remove(group)
+                for group in environment.xpath("%s/groupid[@arch]" % parent_tag):
+                    value = group.attrib.get("arch")
+                    values = [v for v in re.split(r"[, ]+", value) if v]
+                    if arch not in values:
+                        group.getparent().remove(group)
 
     def remove_empty_environments(self):
         """
@@ -305,6 +306,8 @@ class CompsWrapper(object):
             append_common_info(doc, group_node, group, force_description=True)
             append_bool(doc, group_node, "default", group.default)
             append_bool(doc, group_node, "uservisible", group.uservisible)
+            if group.display_order is not None:
+                append(doc, group_node, "display_order", str(group.display_order))
 
             if group.lang_only:
                 append(doc, group_node, "langonly", group.lang_only)

pungi/wrappers/fus.py

@@ -88,5 +88,12 @@ def parse_output(output):
                 packages.add((name, arch, frozenset(flags)))
             else:
                 name, arch = nevra.rsplit(".", 1)
-                modules.add(name.split(":", 1)[1])
+                # replace dash by underscore in stream of module's nerva
+                # source of name looks like
+                # module:llvm-toolset:rhel8:8040020210411062713:9f9e2e7e.x86_64
+                name = ':'.join(
+                    item.replace('-', '_') if i == 1 else item for
+                    i, item in enumerate(name.split(':')[1:])
+                )
+                modules.add(name)
     return packages, modules

pungi/wrappers/iso.py

@@ -260,20 +260,23 @@ def get_isohybrid_cmd(iso_path, arch):
     return cmd
 
 
-def get_manifest_cmd(iso_name, xorriso=False):
+def get_manifest_cmd(iso_name, xorriso=False, output_file=None):
+    if not output_file:
+        output_file = "%s.manifest" % iso_name
+
     if xorriso:
         return """xorriso -dev %s --find |
             tail -n+2 |
             tr -d "'" |
             cut -c2- |
-            sort >> %s.manifest""" % (
-            shlex_quote(iso_name),
+            sort >> %s""" % (
             shlex_quote(iso_name),
+            shlex_quote(output_file),
         )
     else:
-        return "isoinfo -R -f -i %s | grep -v '/TRANS.TBL$' | sort >> %s.manifest" % (
-            shlex_quote(iso_name),
+        return "isoinfo -R -f -i %s | grep -v '/TRANS.TBL$' | sort >> %s" % (
             shlex_quote(iso_name),
+            shlex_quote(output_file),
         )
 
 

pungi/wrappers/kojimock.py

@@ -0,0 +1,299 @@
+import os
+import time
+from pathlib import Path
+
+from attr import dataclass
+from kobo.rpmlib import parse_nvra
+
+from pungi.module_util import Modulemd
+
+# just a random value which we don't
+# use in mock currently
+# originally builds are filtered by this value
+# to get consistent snapshot of tags and packages
+from pungi.scripts.gather_rpms import search_rpms
+
+LAST_EVENT_ID = 999999
+# last event time is not important but build
+# time should be less then it
+LAST_EVENT_TIME = time.time()
+BUILD_TIME = 0
+# virtual build that collects all
+# packages built for some arch
+RELEASE_BUILD_ID = 15270
+# tag that should have all packages available
+ALL_PACKAGES_TAG = 'dist-c8-compose'
+# tag that should have all modules available
+ALL_MODULES_TAG = 'dist-c8-module-compose'
+
+
+@dataclass
+class Module:
+    build_id: int
+    name: str
+    nvr: str
+    stream: str
+    version: str
+    context: str
+    arch: str
+
+
+class KojiMock:
+    """
+    Class that acts like real koji (for some needed methods)
+    but uses local storage as data source
+    """
+    def __init__(self, packages_dir, modules_dir, all_arches):
+        self._modules = self._gather_modules(modules_dir)
+        self._modules_dir = modules_dir
+        self._packages_dir = packages_dir
+        self._all_arches = all_arches
+
+    @staticmethod
+    def _gather_modules(modules_dir):
+        modules = {}
+        for index, (f, arch) in enumerate(
+            (sub_path.name, sub_path.parent.name)
+            for path in Path(modules_dir).glob('*')
+            for sub_path in path.iterdir()
+        ):
+            parsed = parse_nvra(f)
+            modules[index] = Module(
+                name=parsed['name'],
+                nvr=f,
+                version=parsed['release'],
+                context=parsed['arch'],
+                stream=parsed['version'],
+                build_id=index,
+                arch=arch,
+            )
+        return modules
+
+    @staticmethod
+    def getLastEvent(*args, **kwargs):
+        return {'id': LAST_EVENT_ID, 'ts': LAST_EVENT_TIME}
+
+    def listTagged(self, tag_name, *args, **kwargs):
+        """
+        Returns list of virtual 'builds' that contain packages by given tag
+        There are two kinds of tags: modular and distributive.
+        For now, only one kind, distributive one, is needed.
+        """
+        if tag_name != ALL_MODULES_TAG:
+            raise ValueError("I don't know what tag is %s" % tag_name)
+
+        builds = []
+        for module in self._modules.values():
+            builds.append({
+                'build_id': module.build_id,
+                'owner_name': 'centos',
+                'package_name': module.name,
+                'nvr': module.nvr,
+                'version': module.stream,
+                'release': '%s.%s' % (module.version, module.context),
+                'name': module.name,
+                'id': module.build_id,
+                'tag_name': tag_name,
+                'arch': module.arch,
+                # Following fields are currently not
+                # used but returned by real koji
+                # left them here just for reference
+                #
+                # 'task_id': None,
+                # 'state': 1,
+                # 'start_time': '2020-12-23 16:43:59',
+                # 'creation_event_id': 309485,
+                # 'creation_time': '2020-12-23 17:05:33.553748',
+                # 'epoch': None, 'tag_id': 533,
+                # 'completion_time': '2020-12-23 17:05:23',
+                # 'volume_id': 0,
+                # 'package_id': 3221,
+                # 'owner_id': 11,
+                # 'volume_name': 'DEFAULT',
+            })
+
+        return builds
+
+    @staticmethod
+    def getFullInheritance(*args, **kwargs):
+        """
+        Unneeded because we use local storage.
+        """
+        return []
+
+    def getBuild(self, build_id, *args, **kwargs):
+        """
+        Used to get information about build
+        (used in pungi only for modules currently)
+        """
+        module = self._modules[build_id]
+
+        result = {
+            'id': build_id,
+            'name': module.name,
+            'version': module.stream,
+            'release': '%s.%s' % (module.version, module.context),
+            'completion_ts': BUILD_TIME,
+            'state': 'COMPLETE',
+            'arch': module.arch,
+            'extra': {
+                'typeinfo': {
+                    'module': {
+                        'stream': module.stream,
+                        'version': module.version,
+                        'name': module.name,
+                        'context': module.context,
+                        'content_koji_tag': '-'.join([
+                            module.name,
+                            module.stream,
+                            module.version
+                        ]) + '.' + module.context
+                    }
+                }
+            }
+        }
+        return result
+
+    def listArchives(self, build_id, *args, **kwargs):
+        """
+        Originally lists artifacts for build, but in pungi used
+        only to get list of modulemd files for some module
+        """
+        module = self._modules[build_id]
+
+        return [
+            {
+                'build_id': module.build_id,
+                'filename': f'modulemd.{module.arch}.txt',
+                'btype': 'module'
+            },
+            # noone ever uses this file
+            # but it should be because pungi ignores builds
+            # with len(files) <= 1
+            {
+                'build_id': module.build_id,
+                'filename': 'modulemd.txt',
+                'btype': 'module'
+            }
+        ]
+
+    def listTaggedRPMS(self, tag_name, *args, **kwargs):
+        """
+        Get information about packages that are tagged by tag.
+        There are two kings of tags: per-module and per-distr.
+        """
+        if tag_name == ALL_PACKAGES_TAG:
+            builds, packages = self._get_release_packages()
+        else:
+            builds, packages = self._get_module_packages(tag_name)
+        return [
+            packages,
+            builds
+        ]
+
+    def _get_release_packages(self):
+        """
+        Search packages dir and keep only
+        packages that are non-modular.
+
+        This is quite the way how real koji works:
+         - modular packages are tagged by module-* tag
+         - all other packages are tagged with dist* tag
+        """
+        packages = []
+
+        # get all rpms in folder
+        rpms = search_rpms(Path(self._packages_dir))
+
+        for rpm in rpms:
+            info = parse_nvra(rpm.path.stem)
+            if 'module' in info['release']:
+                continue
+            packages.append({
+                "build_id": RELEASE_BUILD_ID,
+                "name": info['name'],
+                "extra": None,
+                "arch": info['arch'],
+                "epoch": info['epoch'] or None,
+                "version": info['version'],
+                "metadata_only": False,
+                "release": info['release'],
+                # not used currently
+                # "id": 262555,
+                # "size": 0
+            })
+        builds = []
+        return builds, packages
+
+    def _get_module_packages(self, tag_name):
+        """
+        Get list of builds for module and given module tag name.
+        """
+        builds = []
+        packages = []
+        modules = self._get_modules_by_name(tag_name)
+        for module in modules:
+            if module is None:
+                raise ValueError('Module %s is not found' % tag_name)
+            path = os.path.join(
+                self._modules_dir,
+                module.arch,
+                tag_name,
+            )
+
+            builds.append({
+                "build_id": module.build_id,
+                "package_name": module.name,
+                "nvr": module.nvr,
+                "tag_name": module.nvr,
+                "version": module.stream,
+                "release": module.version,
+                "id": module.build_id,
+                "name": module.name,
+                "volume_name": "DEFAULT",
+                # Following fields are currently not
+                # used but returned by real koji
+                # left them here just for reference
+                #
+                # "owner_name": "mbox-mbs-backend",
+                # "task_id": 195937,
+                # "state": 1,
+                # "start_time": "2020-12-22 19:20:12.504578",
+                # "creation_event_id": 306731,
+                # "creation_time": "2020-12-22 19:20:12.504578",
+                # "epoch": None,
+                # "tag_id": 1192,
+                # "completion_time": "2020-12-22 19:34:34.716615",
+                # "volume_id": 0,
+                # "package_id": 104,
+                # "owner_id": 6,
+            })
+
+            if os.path.exists(path):
+                info = Modulemd.ModuleStream.read_string(open(path).read(), strict=True)
+                for art in info.get_rpm_artifacts():
+                    data = parse_nvra(art)
+                    packages.append({
+                        "build_id": module.build_id,
+                        "name": data['name'],
+                        "extra": None,
+                        "arch": data['arch'],
+                        "epoch": data['epoch'] or None,
+                        "version": data['version'],
+                        "metadata_only": False,
+                        "release": data['release'],
+                        "id": 262555,
+                        "size": 0
+                    })
+            else:
+                raise RuntimeError('Unable to find module %s' % path)
+        return builds, packages
+
+    def _get_modules_by_name(self, tag_name):
+        modules = []
+        for arch in self._all_arches:
+            for module in self._modules.values():
+                if module.nvr != tag_name or module.arch != arch:
+                    continue
+                modules.append(module)
+        return modules

pungi/wrappers/kojiwrapper.py

@@ -14,18 +14,25 @@
 # along with this program; if not, see <https://gnu.org/licenses/>.
 
 
+import contextlib
 import os
 import re
+import socket
+import shutil
 import time
 import threading
-import contextlib
+
+import requests
 
 import koji
 from kobo.shortcuts import run, force_list
 import six
 from six.moves import configparser, shlex_quote
 import six.moves.xmlrpc_client as xmlrpclib
+from flufl.lock import Lock
+from datetime import timedelta
 
+from .kojimock import KojiMock
 from .. import util
 from ..arch_utils import getBaseArch
 
@@ -785,11 +792,10 @@ class KojiWrapper(object):
         if list_of_args is None and list_of_kwargs is None:
             raise ValueError("One of list_of_args or list_of_kwargs must be set.")
 
-        if type(list_of_args) not in [type(None), list] or type(list_of_kwargs) not in [
-            type(None),
-            list,
-        ]:
-            raise ValueError("list_of_args and list_of_kwargs must be list or None.")
+        if list_of_args is not None and not isinstance(list_of_args, list):
+            raise ValueError("list_of_args must be list or None.")
+        if list_of_kwargs is not None and not isinstance(list_of_kwargs, list):
+            raise ValueError("list_of_kwargs must be list or None.")
 
         if list_of_kwargs is None:
             list_of_kwargs = [{}] * len(list_of_args)
@@ -803,9 +809,9 @@ class KojiWrapper(object):
 
         koji_session.multicall = True
         for args, kwargs in zip(list_of_args, list_of_kwargs):
-            if type(args) != list:
+            if not isinstance(args, list):
                 args = [args]
-            if type(kwargs) != dict:
+            if not isinstance(kwargs, dict):
                 raise ValueError("Every item in list_of_kwargs must be a dict")
             koji_session_fnc(*args, **kwargs)
 
@@ -813,7 +819,7 @@ class KojiWrapper(object):
 
         if not responses:
             return None
-        if type(responses) != list:
+        if not isinstance(responses, list):
             raise ValueError(
                 "Fault element was returned for multicall of method %r: %r"
                 % (koji_session_fnc, responses)
@@ -829,7 +835,7 @@ class KojiWrapper(object):
         # a one-item array containing the result value,
         # or a struct of the form found inside the standard <fault> element.
         for response, args, kwargs in zip(responses, list_of_args, list_of_kwargs):
-            if type(response) == list:
+            if isinstance(response, list):
                 if not response:
                     raise ValueError(
                         "Empty list returned for multicall of method %r with args %r, %r"  # noqa: E501
@@ -864,6 +870,45 @@ class KojiWrapper(object):
             pass
 
 
+class KojiMockWrapper(object):
+    lock = threading.Lock()
+
+    def __init__(self, compose, all_arches):
+        self.all_arches = all_arches
+        self.compose = compose
+        try:
+            self.profile = self.compose.conf["koji_profile"]
+        except KeyError:
+            raise RuntimeError("Koji profile must be configured")
+        with self.lock:
+            self.koji_module = koji.get_profile_module(self.profile)
+            session_opts = {}
+            for key in (
+                "timeout",
+                "keepalive",
+                "max_retries",
+                "retry_interval",
+                "anon_retry",
+                "offline_retry",
+                "offline_retry_interval",
+                "debug",
+                "debug_xmlrpc",
+                "serverca",
+                "use_fast_upload",
+            ):
+                value = getattr(self.koji_module.config, key, None)
+                if value is not None:
+                    session_opts[key] = value
+            self.koji_proxy = KojiMock(
+                packages_dir=self.koji_module.config.topdir,
+                modules_dir=os.path.join(
+                    self.koji_module.config.topdir,
+                    'modules',
+                ),
+                all_arches=self.all_arches,
+            )
+
+
 def get_buildroot_rpms(compose, task_id):
     """Get build root RPMs - either from runroot or local"""
     result = []
@@ -895,3 +940,176 @@ def get_buildroot_rpms(compose, task_id):
                 continue
             result.append(i)
     return sorted(result)
+
+
+class KojiDownloadProxy:
+    def __init__(self, topdir, topurl, cache_dir, logger):
+        if not topdir:
+            # This will only happen if there is either no koji_profile
+            # configured, or the profile doesn't have a topdir. In the first
+            # case there will be no koji interaction, and the second indicates
+            # broken koji configuration.
+            # We can pretend to have local access in both cases to avoid any
+            # external requests.
+            self.has_local_access = True
+            return
+
+        self.cache_dir = cache_dir
+        self.logger = logger
+
+        self.topdir = topdir
+        self.topurl = topurl
+
+        # If cache directory is configured, we want to use it (even if we
+        # actually have local access to the storage).
+        self.has_local_access = not bool(cache_dir)
+        # This is used for temporary downloaded files. The suffix is unique
+        # per-process. To prevent threads in the same process from colliding, a
+        # thread id is added later.
+        self.unique_suffix = "%s.%s" % (socket.gethostname(), os.getpid())
+        self.session = None
+        if not self.has_local_access:
+            self.session = requests.Session()
+
+    @property
+    def path_prefix(self):
+        dir = self.topdir if self.has_local_access else self.cache_dir
+        return dir.rstrip("/") + "/"
+
+    @classmethod
+    def from_config(klass, conf, logger):
+        topdir = None
+        topurl = None
+        cache_dir = None
+        if "koji_profile" in conf:
+            koji_module = koji.get_profile_module(conf["koji_profile"])
+            topdir = koji_module.config.topdir
+            topurl = koji_module.config.topurl
+
+            cache_dir = conf.get("koji_cache")
+            if cache_dir:
+                cache_dir = cache_dir.rstrip("/") + "/"
+        return klass(topdir, topurl, cache_dir, logger)
+
+    @util.retry(wait_on=requests.exceptions.RequestException)
+    def _download(self, url, dest):
+        """Download file into given location
+
+        :param str url: URL of the file to download
+        :param str dest: file path to store the result in
+        :returns: path to the downloaded file (same as dest) or None if the URL
+        """
+        with self.session.get(url, stream=True) as r:
+            if r.status_code == 404:
+                self.logger.warning("GET %s NOT FOUND", url)
+                return None
+            if r.status_code != 200:
+                self.logger.error("GET %s %s", url, r.status_code)
+                r.raise_for_status()
+                # The exception from here will be retried by the decorator.
+
+            file_size = int(r.headers.get("Content-Length", 0))
+            self.logger.info("GET %s OK %s", url, util.format_size(file_size))
+            with open(dest, "wb") as f:
+                shutil.copyfileobj(r.raw, f)
+        return dest
+
+    def _delete(self, path):
+        """Try to delete file at given path and ignore errors."""
+        try:
+            os.remove(path)
+        except Exception:
+            self.logger.warning("Failed to delete %s", path)
+
+    def _atomic_download(self, url, dest, validator):
+        """Atomically download a file
+
+        :param str url: URL of the file to download
+        :param str dest: file path to store the result in
+        :returns: path to the downloaded file (same as dest) or None if the URL
+                  return 404.
+        """
+        temp_file = "%s.%s.%s" % (dest, self.unique_suffix, threading.get_ident())
+
+        # First download to the temporary location.
+        try:
+            if self._download(url, temp_file) is None:
+                # The file was not found.
+                return None
+        except Exception:
+            # Download failed, let's make sure to clean up potentially partial
+            # temporary file.
+            self._delete(temp_file)
+            raise
+
+        # Check if the temporary file is correct (assuming we were provided a
+        # validator function).
+        try:
+            if validator:
+                validator(temp_file)
+        except Exception:
+            # Validation failed. Let's delete the problematic file and re-raise
+            # the exception.
+            self._delete(temp_file)
+            raise
+
+        # Atomically move the temporary file into final location
+        os.rename(temp_file, dest)
+        return dest
+
+    def _download_file(self, path, validator):
+        """Ensure file on Koji volume in ``path`` is present in the local
+        cache.
+
+        :returns: path to the local file or None if file is not found
+        """
+        url = path.replace(self.topdir, self.topurl)
+        destination_file = path.replace(self.topdir, self.cache_dir)
+        util.makedirs(os.path.dirname(destination_file))
+
+        lock = Lock(destination_file + ".lock")
+        # Hold the lock for this file for 5 minutes. If another compose needs
+        # the same file but it's not downloaded yet, the process will wait.
+        #
+        # If the download finishes in time, the downloaded file will be used
+        # here.
+        #
+        # If the download takes longer, this process will steal the lock and
+        # start its own download.
+        #
+        # That should not be a problem: the same file will be downloaded and
+        # then replaced atomically on the filesystem. If the original process
+        # managed to hardlink the first file already, that hardlink will be
+        # broken, but that will only result in the same file stored twice.
+        lock.lifetime = timedelta(minutes=5)
+
+        with lock:
+            # Check if the file already exists. If yes, return the path.
+            if os.path.exists(destination_file):
+                # Update mtime of the file. This covers the case of packages in the
+                # tag that are not included in the compose. Updating mtime will
+                # exempt them from cleanup for extra time.
+                os.utime(destination_file)
+                return destination_file
+
+            return self._atomic_download(url, destination_file, validator)
+
+    def get_file(self, path, validator=None):
+        """
+        If path refers to an existing file in Koji, return a valid local path
+        to it. If no such file exists, return None.
+
+        :param validator: A callable that will be called with the path to the
+            downloaded file if and only if the file was actually downloaded.
+            Any exception raised from there will be abort the download and be
+            propagated.
+        """
+        if self.has_local_access:
+            # We have koji volume mounted locally. No transformation needed for
+            # the path, just check it exists.
+            if os.path.exists(path):
+                return path
+            return None
+        else:
+            # We need to download the file.
+            return self._download_file(path, validator)

pungi/wrappers/scm.py

@@ -20,6 +20,7 @@ import os
 import shutil
 import glob
 import six
+import threading
 from six.moves import shlex_quote
 from six.moves.urllib.request import urlretrieve
 from fnmatch import fnmatch
@@ -29,12 +30,15 @@ from kobo.shortcuts import run, force_list
 from pungi.util import explode_rpm_package, makedirs, copy_all, temp_dir, retry
 from .kojiwrapper import KojiWrapper
 
+lock = threading.Lock()
+
 
 class ScmBase(kobo.log.LoggingBase):
-    def __init__(self, logger=None, command=None, compose=None):
+    def __init__(self, logger=None, command=None, compose=None, options=None):
         kobo.log.LoggingBase.__init__(self, logger=logger)
         self.command = command
         self.compose = compose
+        self.options = options or {}
 
     @retry(interval=60, timeout=300, wait_on=RuntimeError)
     def retry_run(self, cmd, **kwargs):
@@ -156,22 +160,31 @@ class GitWrapper(ScmBase):
         if "://" not in repo:
             repo = "file://%s" % repo
 
+        git_cmd = ["git"]
+        if "credential_helper" in self.options:
+            git_cmd.extend(["-c", "credential.useHttpPath=true"])
+            git_cmd.extend(
+                ["-c", "credential.helper=%s" % self.options["credential_helper"]]
+            )
+
         run(["git", "init"], workdir=destdir)
         try:
-            run(["git", "fetch", "--depth=1", repo, branch], workdir=destdir)
+            run(git_cmd + ["fetch", "--depth=1", repo, branch], workdir=destdir)
             run(["git", "checkout", "FETCH_HEAD"], workdir=destdir)
         except RuntimeError as e:
             # Fetch failed, to do a full clone we add a remote to our empty
             # repo, get its content and check out the reference we want.
             self.log_debug(
                 "Trying to do a full clone because shallow clone failed: %s %s"
-                % (e, e.output)
+                % (e, getattr(e, "output", ""))
             )
             try:
                 # Re-run git init in case of previous failure breaking .git dir
                 run(["git", "init"], workdir=destdir)
                 run(["git", "remote", "add", "origin", repo], workdir=destdir)
-                self.retry_run(["git", "remote", "update", "origin"], workdir=destdir)
+                self.retry_run(
+                    git_cmd + ["remote", "update", "origin"], workdir=destdir
+                )
                 run(["git", "checkout", branch], workdir=destdir)
             except RuntimeError:
                 if self.compose:
@@ -185,19 +198,38 @@ class GitWrapper(ScmBase):
                     copy_all(destdir, debugdir)
                 raise
 
-        self.run_process_command(destdir)
+    def get_temp_repo_path(self, scm_root, scm_branch):
+        scm_repo = scm_root.split("/")[-1]
+        process_id = os.getpid()
+        tmp_dir = (
+            "/tmp/pungi-temp-git-repos-"
+            + str(process_id)
+            + "/"
+            + scm_repo
+            + "-"
+            + scm_branch
+        )
+        return tmp_dir
+
+    def setup_repo(self, scm_root, scm_branch):
+        tmp_dir = self.get_temp_repo_path(scm_root, scm_branch)
+        if not os.path.isdir(tmp_dir):
+            makedirs(tmp_dir)
+            self._clone(scm_root, scm_branch, tmp_dir)
+        self.run_process_command(tmp_dir)
+        return tmp_dir
 
     def export_dir(self, scm_root, scm_dir, target_dir, scm_branch=None):
         scm_dir = scm_dir.lstrip("/")
         scm_branch = scm_branch or "master"
 
-        with temp_dir() as tmp_dir:
-            self.log_debug(
-                "Exporting directory %s from git %s (branch %s)..."
-                % (scm_dir, scm_root, scm_branch)
-            )
+        self.log_debug(
+            "Exporting directory %s from git %s (branch %s)..."
+            % (scm_dir, scm_root, scm_branch)
+        )
 
-            self._clone(scm_root, scm_branch, tmp_dir)
+        with lock:
+            tmp_dir = self.setup_repo(scm_root, scm_branch)
 
             copy_all(os.path.join(tmp_dir, scm_dir), target_dir)
 
@@ -205,15 +237,15 @@ class GitWrapper(ScmBase):
         scm_file = scm_file.lstrip("/")
         scm_branch = scm_branch or "master"
 
-        with temp_dir() as tmp_dir:
-            target_path = os.path.join(target_dir, os.path.basename(scm_file))
+        target_path = os.path.join(target_dir, os.path.basename(scm_file))
 
-            self.log_debug(
-                "Exporting file %s from git %s (branch %s)..."
-                % (scm_file, scm_root, scm_branch)
-            )
+        self.log_debug(
+            "Exporting file %s from git %s (branch %s)..."
+            % (scm_file, scm_root, scm_branch)
+        )
 
-            self._clone(scm_root, scm_branch, tmp_dir)
+        with lock:
+            tmp_dir = self.setup_repo(scm_root, scm_branch)
 
             makedirs(target_dir)
             shutil.copy2(os.path.join(tmp_dir, scm_file), target_path)
@@ -361,15 +393,19 @@ def get_file_from_scm(scm_dict, target_path, compose=None):
         scm_file = os.path.abspath(scm_dict)
         scm_branch = None
         command = None
+        options = {}
     else:
         scm_type = scm_dict["scm"]
         scm_repo = scm_dict["repo"]
         scm_file = scm_dict["file"]
         scm_branch = scm_dict.get("branch", None)
         command = scm_dict.get("command")
+        options = scm_dict.get("options", {})
 
     logger = compose._logger if compose else None
-    scm = _get_wrapper(scm_type, logger=logger, command=command, compose=compose)
+    scm = _get_wrapper(
+        scm_type, logger=logger, command=command, compose=compose, options=options
+    )
 
     files_copied = []
     for i in force_list(scm_file):
@@ -450,15 +486,19 @@ def get_dir_from_scm(scm_dict, target_path, compose=None):
         scm_dir = os.path.abspath(scm_dict)
         scm_branch = None
         command = None
+        options = {}
     else:
         scm_type = scm_dict["scm"]
         scm_repo = scm_dict.get("repo", None)
         scm_dir = scm_dict["dir"]
         scm_branch = scm_dict.get("branch", None)
         command = scm_dict.get("command")
+        options = scm_dict.get("options", {})
 
     logger = compose._logger if compose else None
-    scm = _get_wrapper(scm_type, logger=logger, command=command, compose=compose)
+    scm = _get_wrapper(
+        scm_type, logger=logger, command=command, compose=compose, options=options
+    )
 
     with temp_dir(prefix="scm_checkout_") as tmp_dir:
         scm.export_dir(scm_repo, scm_dir, scm_branch=scm_branch, target_dir=tmp_dir)

pungi/wrappers/variants.py

@@ -276,7 +276,6 @@ class Variant(object):
         modules=None,
         modular_koji_tags=None,
     ):
-
         environments = environments or []
         buildinstallpackages = buildinstallpackages or []
 

pungi_utils/orchestrator.py

@@ -1,705 +0,0 @@
-# -*- coding: utf-8 -*-
-
-from __future__ import print_function
-
-import argparse
-import atexit
-import errno
-import json
-import logging
-import os
-import re
-import shutil
-import subprocess
-import sys
-import tempfile
-import time
-import threading
-from collections import namedtuple
-
-import kobo.conf
-import kobo.log
-import productmd
-from kobo import shortcuts
-from six.moves import configparser, shlex_quote
-
-import pungi.util
-from pungi.compose import get_compose_dir
-from pungi.linker import linker_pool
-from pungi.phases.pkgset.sources.source_koji import get_koji_event_raw
-from pungi.util import find_old_compose, parse_koji_event, temp_dir
-from pungi.wrappers.kojiwrapper import KojiWrapper
-
-
-Config = namedtuple(
-    "Config",
-    [
-        # Path to directory with the compose
-        "target",
-        "compose_type",
-        "label",
-        # Path to the selected old compose that will be reused
-        "old_compose",
-        # Path to directory with config file copies
-        "config_dir",
-        # Which koji event to use (if any)
-        "event",
-        # Additional arguments to pungi-koji executable
-        "extra_args",
-    ],
-)
-
-log = logging.getLogger(__name__)
-
-
-class Status(object):
-    # Ready to start
-    READY = "READY"
-    # Waiting for dependencies to finish.
-    WAITING = "WAITING"
-    # Part is currently running
-    STARTED = "STARTED"
-    # A dependency failed, this one will never start.
-    BLOCKED = "BLOCKED"
-
-
-class ComposePart(object):
-    def __init__(self, name, config, just_phase=[], skip_phase=[], dependencies=[]):
-        self.name = name
-        self.config = config
-        self.status = Status.WAITING if dependencies else Status.READY
-        self.just_phase = just_phase
-        self.skip_phase = skip_phase
-        self.blocked_on = set(dependencies)
-        self.depends_on = set(dependencies)
-        self.path = None
-        self.log_file = None
-        self.failable = False
-
-    def __str__(self):
-        return self.name
-
-    def __repr__(self):
-        return (
-            "ComposePart({0.name!r},"
-            " {0.config!r},"
-            " {0.status!r},"
-            " just_phase={0.just_phase!r},"
-            " skip_phase={0.skip_phase!r},"
-            " dependencies={0.depends_on!r})"
-        ).format(self)
-
-    def refresh_status(self):
-        """Refresh status of this part with the result of the compose. This
-        should only be called once the compose finished.
-        """
-        try:
-            with open(os.path.join(self.path, "STATUS")) as fh:
-                self.status = fh.read().strip()
-        except IOError as exc:
-            log.error("Failed to update status of %s: %s", self.name, exc)
-            log.error("Assuming %s is DOOMED", self.name)
-            self.status = "DOOMED"
-
-    def is_finished(self):
-        return "FINISHED" in self.status
-
-    def unblock_on(self, finished_part):
-        """Update set of blockers for this part. If it's empty, mark us as ready."""
-        self.blocked_on.discard(finished_part)
-        if self.status == Status.WAITING and not self.blocked_on:
-            log.debug("%s is ready to start", self)
-            self.status = Status.READY
-
-    def setup_start(self, global_config, parts):
-        substitutions = dict(
-            ("part-%s" % name, p.path) for name, p in parts.items() if p.is_finished()
-        )
-        substitutions["configdir"] = global_config.config_dir
-
-        config = pungi.util.load_config(self.config)
-
-        for f in config.opened_files:
-            # apply substitutions
-            fill_in_config_file(f, substitutions)
-
-        self.status = Status.STARTED
-        self.path = get_compose_dir(
-            os.path.join(global_config.target, "parts"),
-            config,
-            compose_type=global_config.compose_type,
-            compose_label=global_config.label,
-        )
-        self.log_file = os.path.join(global_config.target, "logs", "%s.log" % self.name)
-        log.info("Starting %s in %s", self.name, self.path)
-
-    def get_cmd(self, global_config):
-        cmd = ["pungi-koji", "--config", self.config, "--compose-dir", self.path]
-        cmd.append("--%s" % global_config.compose_type)
-        if global_config.label:
-            cmd.extend(["--label", global_config.label])
-        for phase in self.just_phase:
-            cmd.extend(["--just-phase", phase])
-        for phase in self.skip_phase:
-            cmd.extend(["--skip-phase", phase])
-        if global_config.old_compose:
-            cmd.extend(
-                ["--old-compose", os.path.join(global_config.old_compose, "parts")]
-            )
-        if global_config.event:
-            cmd.extend(["--koji-event", str(global_config.event)])
-        if global_config.extra_args:
-            cmd.extend(global_config.extra_args)
-        cmd.extend(["--no-latest-link"])
-        return cmd
-
-    @classmethod
-    def from_config(cls, config, section, config_dir):
-        part = cls(
-            name=section,
-            config=os.path.join(config_dir, config.get(section, "config")),
-            just_phase=_safe_get_list(config, section, "just_phase", []),
-            skip_phase=_safe_get_list(config, section, "skip_phase", []),
-            dependencies=_safe_get_list(config, section, "depends_on", []),
-        )
-        if config.has_option(section, "failable"):
-            part.failable = config.getboolean(section, "failable")
-        return part
-
-
-def _safe_get_list(config, section, option, default=None):
-    """Get a value from config parser. The result is split into a list on
-    commas or spaces, and `default` is returned if the key does not exist.
-    """
-    if config.has_option(section, option):
-        value = config.get(section, option)
-        return [x.strip() for x in re.split(r"[, ]+", value) if x]
-    return default
-
-
-def fill_in_config_file(fp, substs):
-    """Templating function. It works with Jinja2 style placeholders such as
-    {{foo}}. Whitespace around the key name is fine. The file is modified in place.
-
-    :param fp string: path to the file to process
-    :param substs dict: a mapping for values to put into the file
-    """
-
-    def repl(match):
-        try:
-            return substs[match.group(1)]
-        except KeyError as exc:
-            raise RuntimeError(
-                "Unknown placeholder %s in %s" % (exc, os.path.basename(fp))
-            )
-
-    with open(fp, "r") as f:
-        contents = re.sub(r"{{ *([a-zA-Z-_]+) *}}", repl, f.read())
-    with open(fp, "w") as f:
-        f.write(contents)
-
-
-def start_part(global_config, parts, part):
-    part.setup_start(global_config, parts)
-    fh = open(part.log_file, "w")
-    cmd = part.get_cmd(global_config)
-    log.debug("Running command %r", " ".join(shlex_quote(x) for x in cmd))
-    return subprocess.Popen(cmd, stdout=fh, stderr=subprocess.STDOUT)
-
-
-def handle_finished(global_config, linker, parts, proc, finished_part):
-    finished_part.refresh_status()
-    log.info("%s finished with status %s", finished_part, finished_part.status)
-    if proc.returncode == 0:
-        # Success, unblock other parts...
-        for part in parts.values():
-            part.unblock_on(finished_part.name)
-        # ...and link the results into final destination.
-        copy_part(global_config, linker, finished_part)
-        update_metadata(global_config, finished_part)
-    else:
-        # Failure, other stuff may be blocked.
-        log.info("See details in %s", finished_part.log_file)
-        block_on(parts, finished_part.name)
-
-
-def copy_part(global_config, linker, part):
-    c = productmd.Compose(part.path)
-    for variant in c.info.variants:
-        data_path = os.path.join(part.path, "compose", variant)
-        link = os.path.join(global_config.target, "compose", variant)
-        log.info("Hardlinking content %s -> %s", data_path, link)
-        hardlink_dir(linker, data_path, link)
-
-
-def hardlink_dir(linker, srcdir, dstdir):
-    for root, dirs, files in os.walk(srcdir):
-        root = os.path.relpath(root, srcdir)
-        for f in files:
-            src = os.path.normpath(os.path.join(srcdir, root, f))
-            dst = os.path.normpath(os.path.join(dstdir, root, f))
-            linker.queue_put((src, dst))
-
-
-def update_metadata(global_config, part):
-    part_metadata_dir = os.path.join(part.path, "compose", "metadata")
-    final_metadata_dir = os.path.join(global_config.target, "compose", "metadata")
-    for f in os.listdir(part_metadata_dir):
-        # Load the metadata
-        with open(os.path.join(part_metadata_dir, f)) as fh:
-            part_metadata = json.load(fh)
-        final_metadata = os.path.join(final_metadata_dir, f)
-        if os.path.exists(final_metadata):
-            # We already have this file, will need to merge.
-            merge_metadata(final_metadata, part_metadata)
-        else:
-            # A new file, just copy it.
-            copy_metadata(global_config, final_metadata, part_metadata)
-
-
-def copy_metadata(global_config, final_metadata, source):
-    """Copy file to final location, but update compose information."""
-    with open(
-        os.path.join(global_config.target, "compose/metadata/composeinfo.json")
-    ) as f:
-        composeinfo = json.load(f)
-    try:
-        source["payload"]["compose"].update(composeinfo["payload"]["compose"])
-    except KeyError:
-        # No [payload][compose], probably OSBS metadata
-        pass
-    with open(final_metadata, "w") as f:
-        json.dump(source, f, indent=2, sort_keys=True)
-
-
-def merge_metadata(final_metadata, source):
-    with open(final_metadata) as f:
-        metadata = json.load(f)
-
-    try:
-        key = {
-            "productmd.composeinfo": "variants",
-            "productmd.modules": "modules",
-            "productmd.images": "images",
-            "productmd.rpms": "rpms",
-        }[source["header"]["type"]]
-        # TODO what if multiple parts create images for the same variant
-        metadata["payload"][key].update(source["payload"][key])
-    except KeyError:
-        # OSBS metadata, merge whole file
-        metadata.update(source)
-    with open(final_metadata, "w") as f:
-        json.dump(metadata, f, indent=2, sort_keys=True)
-
-
-def block_on(parts, name):
-    """Part ``name`` failed, mark everything depending on it as blocked."""
-    for part in parts.values():
-        if name in part.blocked_on:
-            log.warning("%s is blocked now and will not run", part)
-            part.status = Status.BLOCKED
-            block_on(parts, part.name)
-
-
-def check_finished_processes(processes):
-    """Walk through all active processes and check if something finished."""
-    for proc in processes.keys():
-        proc.poll()
-        if proc.returncode is not None:
-            yield proc, processes[proc]
-
-
-def run_all(global_config, parts):
-    # Mapping subprocess.Popen -> ComposePart
-    processes = dict()
-    remaining = set(p.name for p in parts.values() if not p.is_finished())
-
-    with linker_pool("hardlink") as linker:
-        while remaining or processes:
-            update_status(global_config, parts)
-
-            for proc, part in check_finished_processes(processes):
-                del processes[proc]
-                handle_finished(global_config, linker, parts, proc, part)
-
-            # Start new available processes.
-            for name in list(remaining):
-                part = parts[name]
-                # Start all ready parts
-                if part.status == Status.READY:
-                    remaining.remove(name)
-                    processes[start_part(global_config, parts, part)] = part
-                # Remove blocked parts from todo list
-                elif part.status == Status.BLOCKED:
-                    remaining.remove(part.name)
-
-            # Wait for any child process to finish if there is any.
-            if processes:
-                pid, reason = os.wait()
-                for proc in processes.keys():
-                    # Set the return code for process that we caught by os.wait().
-                    # Calling poll() on it would not set the return code properly
-                    # since the value was already consumed by os.wait().
-                    if proc.pid == pid:
-                        proc.returncode = (reason >> 8) & 0xFF
-
-        log.info("Waiting for linking to finish...")
-    return update_status(global_config, parts)
-
-
-def get_target_dir(config, compose_info, label, reldir=""):
-    """Find directory where this compose will be.
-
-    @param reldir: if target path in config is relative, it will be resolved
-                   against this directory
-    """
-    dir = os.path.realpath(os.path.join(reldir, config.get("general", "target")))
-    target_dir = get_compose_dir(
-        dir,
-        compose_info,
-        compose_type=config.get("general", "compose_type"),
-        compose_label=label,
-    )
-    return target_dir
-
-
-def setup_logging(debug=False):
-    FORMAT = "%(asctime)s: %(levelname)s: %(message)s"
-    level = logging.DEBUG if debug else logging.INFO
-    kobo.log.add_stderr_logger(log, log_level=level, format=FORMAT)
-    log.setLevel(level)
-
-
-def compute_status(statuses):
-    if any(map(lambda x: x[0] in ("STARTED", "WAITING"), statuses)):
-        # If there is anything still running or waiting to start, the whole is
-        # still running.
-        return "STARTED"
-    elif any(map(lambda x: x[0] in ("DOOMED", "BLOCKED") and not x[1], statuses)):
-        # If any required part is doomed or blocked, the whole is doomed
-        return "DOOMED"
-    elif all(map(lambda x: x[0] == "FINISHED", statuses)):
-        # If all parts are complete, the whole is complete
-        return "FINISHED"
-    else:
-        return "FINISHED_INCOMPLETE"
-
-
-def update_status(global_config, parts):
-    log.debug("Updating status metadata")
-    metadata = {}
-    statuses = set()
-    for part in parts.values():
-        metadata[part.name] = {"status": part.status, "path": part.path}
-        statuses.add((part.status, part.failable))
-    metadata_path = os.path.join(
-        global_config.target, "compose", "metadata", "parts.json"
-    )
-    with open(metadata_path, "w") as fh:
-        json.dump(metadata, fh, indent=2, sort_keys=True, separators=(",", ": "))
-
-    status = compute_status(statuses)
-    log.info("Overall status is %s", status)
-    with open(os.path.join(global_config.target, "STATUS"), "w") as fh:
-        fh.write(status)
-
-    return status != "DOOMED"
-
-
-def prepare_compose_dir(config, args, main_config_file, compose_info):
-    if not hasattr(args, "compose_path"):
-        # Creating a brand new compose
-        target_dir = get_target_dir(
-            config, compose_info, args.label, reldir=os.path.dirname(main_config_file)
-        )
-        for dir in ("logs", "parts", "compose/metadata", "work/global"):
-            try:
-                os.makedirs(os.path.join(target_dir, dir))
-            except OSError as exc:
-                if exc.errno != errno.EEXIST:
-                    raise
-        with open(os.path.join(target_dir, "STATUS"), "w") as fh:
-            fh.write("STARTED")
-        # Copy initial composeinfo for new compose
-        shutil.copy(
-            os.path.join(target_dir, "work/global/composeinfo-base.json"),
-            os.path.join(target_dir, "compose/metadata/composeinfo.json"),
-        )
-    else:
-        # Restarting a particular compose
-        target_dir = args.compose_path
-
-    return target_dir
-
-
-def load_parts_metadata(global_config):
-    parts_metadata = os.path.join(global_config.target, "compose/metadata/parts.json")
-    with open(parts_metadata) as f:
-        return json.load(f)
-
-
-def setup_for_restart(global_config, parts, to_restart):
-    has_stuff_to_do = False
-    metadata = load_parts_metadata(global_config)
-    for key in metadata:
-        # Update state to match what is on disk
-        log.debug(
-            "Reusing %s (%s) from %s",
-            key,
-            metadata[key]["status"],
-            metadata[key]["path"],
-        )
-        parts[key].status = metadata[key]["status"]
-        parts[key].path = metadata[key]["path"]
-    for key in to_restart:
-        # Set restarted parts to run again
-        parts[key].status = Status.WAITING
-        parts[key].path = None
-
-    for key in to_restart:
-        # Remove blockers that are already finished
-        for blocker in list(parts[key].blocked_on):
-            if parts[blocker].is_finished():
-                parts[key].blocked_on.discard(blocker)
-        if not parts[key].blocked_on:
-            log.debug("Part %s in not blocked", key)
-            # Nothing blocks it; let's go
-            parts[key].status = Status.READY
-            has_stuff_to_do = True
-
-    if not has_stuff_to_do:
-        raise RuntimeError("All restarted parts are blocked. Nothing to do.")
-
-
-def run_kinit(config):
-    if not config.getboolean("general", "kerberos"):
-        return
-
-    keytab = config.get("general", "kerberos_keytab")
-    principal = config.get("general", "kerberos_principal")
-
-    fd, fname = tempfile.mkstemp(prefix="krb5cc_pungi-orchestrate_")
-    os.close(fd)
-    os.environ["KRB5CCNAME"] = fname
-    shortcuts.run(["kinit", "-k", "-t", keytab, principal])
-    log.debug("Created a kerberos ticket for %s", principal)
-
-    atexit.register(os.remove, fname)
-
-
-def get_compose_data(compose_path):
-    try:
-        compose = productmd.compose.Compose(compose_path)
-        data = {
-            "compose_id": compose.info.compose.id,
-            "compose_date": compose.info.compose.date,
-            "compose_type": compose.info.compose.type,
-            "compose_respin": str(compose.info.compose.respin),
-            "compose_label": compose.info.compose.label,
-            "release_id": compose.info.release_id,
-            "release_name": compose.info.release.name,
-            "release_short": compose.info.release.short,
-            "release_version": compose.info.release.version,
-            "release_type": compose.info.release.type,
-            "release_is_layered": compose.info.release.is_layered,
-        }
-        if compose.info.release.is_layered:
-            data.update(
-                {
-                    "base_product_name": compose.info.base_product.name,
-                    "base_product_short": compose.info.base_product.short,
-                    "base_product_version": compose.info.base_product.version,
-                    "base_product_type": compose.info.base_product.type,
-                }
-            )
-        return data
-    except Exception:
-        return {}
-
-
-def get_script_env(compose_path):
-    env = os.environ.copy()
-    env["COMPOSE_PATH"] = compose_path
-    for key, value in get_compose_data(compose_path).items():
-        if isinstance(value, bool):
-            env[key.upper()] = "YES" if value else ""
-        else:
-            env[key.upper()] = str(value) if value else ""
-    return env
-
-
-def run_scripts(prefix, compose_dir, scripts):
-    env = get_script_env(compose_dir)
-    for idx, script in enumerate(scripts.strip().splitlines()):
-        command = script.strip()
-        logfile = os.path.join(compose_dir, "logs", "%s%s.log" % (prefix, idx))
-        log.debug("Running command: %r", command)
-        log.debug("See output in %s", logfile)
-        shortcuts.run(command, env=env, logfile=logfile)
-
-
-def try_translate_path(parts, path):
-    translation = []
-    for part in parts.values():
-        conf = pungi.util.load_config(part.config)
-        translation.extend(conf.get("translate_paths", []))
-    return pungi.util.translate_path_raw(translation, path)
-
-
-def send_notification(compose_dir, command, parts):
-    if not command:
-        return
-    from pungi.notifier import PungiNotifier
-
-    data = get_compose_data(compose_dir)
-    data["location"] = try_translate_path(parts, compose_dir)
-    notifier = PungiNotifier([command])
-    with open(os.path.join(compose_dir, "STATUS")) as f:
-        status = f.read().strip()
-    notifier.send("status-change", workdir=compose_dir, status=status, **data)
-
-
-def setup_progress_monitor(global_config, parts):
-    """Update configuration so that each part send notifications about its
-    progress to the orchestrator.
-
-    There is a file to which the notification is written. The orchestrator is
-    reading it and mapping the entries to particular parts. The path to this
-    file is stored in an environment variable.
-    """
-    tmp_file = tempfile.NamedTemporaryFile(prefix="pungi-progress-monitor_")
-    os.environ["_PUNGI_ORCHESTRATOR_PROGRESS_MONITOR"] = tmp_file.name
-    atexit.register(os.remove, tmp_file.name)
-
-    global_config.extra_args.append(
-        "--notification-script=pungi-notification-report-progress"
-    )
-
-    def reader():
-        while True:
-            line = tmp_file.readline()
-            if not line:
-                time.sleep(0.1)
-                continue
-            path, msg = line.split(":", 1)
-            for part in parts:
-                if parts[part].path == os.path.dirname(path):
-                    log.debug("%s: %s", part, msg.strip())
-                    break
-
-    monitor = threading.Thread(target=reader)
-    monitor.daemon = True
-    monitor.start()
-
-
-def run(work_dir, main_config_file, args):
-    config_dir = os.path.join(work_dir, "config")
-    shutil.copytree(os.path.dirname(main_config_file), config_dir)
-
-    # Read main config
-    parser = configparser.RawConfigParser(
-        defaults={
-            "kerberos": "false",
-            "pre_compose_script": "",
-            "post_compose_script": "",
-            "notification_script": "",
-        }
-    )
-    parser.read(main_config_file)
-
-    # Create kerberos ticket
-    run_kinit(parser)
-
-    compose_info = dict(parser.items("general"))
-    compose_type = parser.get("general", "compose_type")
-
-    target_dir = prepare_compose_dir(parser, args, main_config_file, compose_info)
-    kobo.log.add_file_logger(log, os.path.join(target_dir, "logs", "orchestrator.log"))
-    log.info("Composing %s", target_dir)
-
-    run_scripts("pre_compose_", target_dir, parser.get("general", "pre_compose_script"))
-
-    old_compose = find_old_compose(
-        os.path.dirname(target_dir),
-        compose_info["release_short"],
-        compose_info["release_version"],
-        "",
-    )
-    if old_compose:
-        log.info("Reusing old compose %s", old_compose)
-
-    global_config = Config(
-        target=target_dir,
-        compose_type=compose_type,
-        label=args.label,
-        old_compose=old_compose,
-        config_dir=os.path.dirname(main_config_file),
-        event=args.koji_event,
-        extra_args=_safe_get_list(parser, "general", "extra_args"),
-    )
-
-    if not global_config.event and parser.has_option("general", "koji_profile"):
-        koji_wrapper = KojiWrapper(parser.get("general", "koji_profile"))
-        event_file = os.path.join(global_config.target, "work/global/koji-event")
-        result = get_koji_event_raw(koji_wrapper, None, event_file)
-        global_config = global_config._replace(event=result["id"])
-
-    parts = {}
-    for section in parser.sections():
-        if section == "general":
-            continue
-        parts[section] = ComposePart.from_config(parser, section, config_dir)
-
-    if hasattr(args, "part"):
-        setup_for_restart(global_config, parts, args.part)
-
-    setup_progress_monitor(global_config, parts)
-
-    send_notification(target_dir, parser.get("general", "notification_script"), parts)
-
-    retcode = run_all(global_config, parts)
-
-    if retcode:
-        # Only run the script if we are not doomed.
-        run_scripts(
-            "post_compose_", target_dir, parser.get("general", "post_compose_script")
-        )
-
-    send_notification(target_dir, parser.get("general", "notification_script"), parts)
-
-    return retcode
-
-
-def parse_args(argv):
-    parser = argparse.ArgumentParser()
-    parser.add_argument("--debug", action="store_true")
-    parser.add_argument("--koji-event", metavar="ID", type=parse_koji_event)
-    subparsers = parser.add_subparsers()
-    start = subparsers.add_parser("start")
-    start.add_argument("config", metavar="CONFIG")
-    start.add_argument("--label")
-
-    restart = subparsers.add_parser("restart")
-    restart.add_argument("config", metavar="CONFIG")
-    restart.add_argument("compose_path", metavar="COMPOSE_PATH")
-    restart.add_argument(
-        "part", metavar="PART", nargs="*", help="which parts to restart"
-    )
-    restart.add_argument("--label")
-
-    return parser.parse_args(argv)
-
-
-def main(argv=None):
-    args = parse_args(argv)
-    setup_logging(args.debug)
-
-    main_config_file = os.path.abspath(args.config)
-
-    with temp_dir() as work_dir:
-        try:
-            if not run(work_dir, main_config_file, args):
-                sys.exit(1)
-        except Exception:
-            log.exception("Unhandled exception!")
-            sys.exit(1)

requirements.txt

@@ -1,7 +1,8 @@
 # Some packages must be installed via dnf/yum first, see doc/contributing.rst
 dict.sorted
 dogpile.cache
-fedmsg
+flufl.lock ; python_version >= '3.0'
+flufl.lock < 3.0 ; python_version <= '2.7'
 funcsigs
 jsonschema
 kobo

setup.cfg

@@ -0,0 +1,2 @@
+[sdist]
+formats=bztar

setup.py

@@ -5,14 +5,9 @@
 import os
 import glob
 
-import distutils.command.sdist
 from setuptools import setup
 
 
-# override default tarball format with bzip2
-distutils.command.sdist.sdist.default_format = {"posix": "bztar"}
-
-
 # recursively scan for python modules to be included
 package_root_dirs = ["pungi", "pungi_utils"]
 packages = set()
@@ -25,7 +20,7 @@ packages = sorted(packages)
 
 setup(
     name="pungi",
-    version="4.3.6",
+    version="4.5.0",
     description="Distribution compose tool",
     url="https://pagure.io/pungi",
     author="Dennis Gilmore",
@@ -41,16 +36,21 @@ setup(
             "pungi-patch-iso = pungi.scripts.patch_iso:cli_main",
             "pungi-make-ostree = pungi.ostree:main",
             "pungi-notification-report-progress = pungi.scripts.report_progress:main",
-            "pungi-orchestrate = pungi_utils.orchestrator:main",
             "pungi-wait-for-signed-ostree-handler = pungi.scripts.wait_for_signed_ostree_handler:main",  # noqa: E501
             "pungi-koji = pungi.scripts.pungi_koji:cli_main",
             "pungi-gather = pungi.scripts.pungi_gather:cli_main",
             "pungi-config-dump = pungi.scripts.config_dump:cli_main",
             "pungi-config-validate = pungi.scripts.config_validate:cli_main",
+            "pungi-cache-cleanup = pungi.scripts.cache_cleanup:main",
+            "pungi-gather-modules = pungi.scripts.gather_modules:cli_main",
+            "pungi-gather-rpms = pungi.scripts.gather_rpms:cli_main",
+            "pungi-generate-packages-json = pungi.scripts.create_packages_json:cli_main",  # noqa: E501
+            "pungi-create-extra-repo = pungi.scripts.create_extra_repo:cli_main"
         ]
     },
     scripts=["contrib/yum-dnf-compare/pungi-compare-depsolving"],
     data_files=[
+        ("/usr/lib/tmpfiles.d", glob.glob("contrib/tmpfiles.d/*.conf")),
         ("/usr/share/pungi", glob.glob("share/*.xsl")),
         ("/usr/share/pungi", glob.glob("share/*.ks")),
         ("/usr/share/pungi", glob.glob("share/*.dtd")),
@@ -66,5 +66,5 @@ setup(
         "dogpile.cache",
     ],
     extras_require={':python_version=="2.7"': ["enum34", "lockfile"]},
-    tests_require=["mock", "pytest", "pytest-cov"],
+    tests_require=["mock", "pytest", "pytest-cov", "pyfakefs"],
 )

tests/data/dummy-comps.xml

@@ -108,6 +108,7 @@
       <groupid>core</groupid>
     </grouplist>
     <optionlist>
+      <groupid arch="x86_64">standard</groupid>
     </optionlist>
   </environment>
 

tests/data/test_create_extra_repo/test_repo/repodata/repomd.xml

@@ -0,0 +1,36 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<repomd xmlns="http://linux.duke.edu/metadata/repo" xmlns:rpm="http://linux.duke.edu/metadata/rpm">
+  <revision>1612479076</revision>
+  <data type="primary">
+    <checksum type="sha256">08941fae6bdb14f3b22bfad38b9d7dcb685a9df58fe8f515a3a0b2fe1af903bb</checksum>
+    <open-checksum type="sha256">2a15e618f049a883d360ccbf3e764b30640255f47dc526c633b1722fe23cbcbc</open-checksum>
+    <location href="repodata/08941fae6bdb14f3b22bfad38b9d7dcb685a9df58fe8f515a3a0b2fe1af903bb-primary.xml.gz"/>
+    <timestamp>1612479075</timestamp>
+    <size>1240</size>
+    <open-size>3888</open-size>
+  </data>
+  <data type="filelists">
+    <checksum type="sha256">e37a0b4a63b2b245dca1727195300cd3961f80aebc82ae7b9849dbf7482f5d0f</checksum>
+    <open-checksum type="sha256">b1782bc4207a5b7c3e64115d5a1d001802e8d363f022ea165df7cdab6f14651c</open-checksum>
+    <location href="repodata/e37a0b4a63b2b245dca1727195300cd3961f80aebc82ae7b9849dbf7482f5d0f-filelists.xml.gz"/>
+    <timestamp>1612479075</timestamp>
+    <size>439</size>
+    <open-size>1295</open-size>
+  </data>
+  <data type="other">
+    <checksum type="sha256">92992176bce71dcde9e4b6ad1442e7b5c7f3de9b7f019a2cd27d042ab38ea2b1</checksum>
+    <open-checksum type="sha256">3b847919691ad32279b13463de6c08f1f8b32f51e87b7d8d7e95a3ec2f46ef51</open-checksum>
+    <location href="repodata/92992176bce71dcde9e4b6ad1442e7b5c7f3de9b7f019a2cd27d042ab38ea2b1-other.xml.gz"/>
+    <timestamp>1612479075</timestamp>
+    <size>630</size>
+    <open-size>1911</open-size>
+  </data>
+  <data type="modules">
+    <checksum type="sha256">e7a671401f8e207e4cd3b90b4ac92d621f84a34dc9026f57c3f427fbed444c57</checksum>
+    <open-checksum type="sha256">d59fee86c18018cc18bb7325aa74aa0abf923c64d29a4ec45e08dcd01a0c3966</open-checksum>
+    <location href="repodata/e7a671401f8e207e4cd3b90b4ac92d621f84a34dc9026f57c3f427fbed444c57-modules.yaml.gz"/>
+    <timestamp>1612479075</timestamp>
+    <size>920</size>
+    <open-size>3308</open-size>
+  </data>
+</repomd>

tests/data/test_create_packages_json/test_repo/repodata/repomd.xml

@@ -0,0 +1,55 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<repomd xmlns="http://linux.duke.edu/metadata/repo" xmlns:rpm="http://linux.duke.edu/metadata/rpm">
+  <revision>1666177486</revision>
+  <data type="primary">
+    <checksum type="sha256">89cb9cc1181635c9147864a7076d91fb81072641d481cd202832a2d257453576</checksum>
+    <open-checksum type="sha256">07255d9856f7531b52a6459f6fc7701c6d93c6d6c29d1382d83afcc53f13494a</open-checksum>
+    <location href="repodata/89cb9cc1181635c9147864a7076d91fb81072641d481cd202832a2d257453576-primary.xml.gz"/>
+    <timestamp>1666177486</timestamp>
+    <size>1387</size>
+    <open-size>6528</open-size>
+  </data>
+  <data type="filelists">
+    <checksum type="sha256">f69ca03957574729fd5150335b0d87afddcfb37a97aed5b06272212854f1773d</checksum>
+    <open-checksum type="sha256">c2e1e674d7d48bccaa16cae0a5f70cb55ef4cd7352b4d9d4fdaa619075d07dbc</open-checksum>
+    <location href="repodata/f69ca03957574729fd5150335b0d87afddcfb37a97aed5b06272212854f1773d-filelists.xml.gz"/>
+    <timestamp>1666177486</timestamp>
+    <size>1252</size>
+    <open-size>5594</open-size>
+  </data>
+  <data type="other">
+    <checksum type="sha256">b3827bd6c9ea67ffa3912002515c64e4d9fe5c4dacbf7c46b0d8768b7abbb84f</checksum>
+    <open-checksum type="sha256">9ce24c526239e349d023c577b2ae3872c8b0f1888aed1fb24b9b9aa12063fdf3</open-checksum>
+    <location href="repodata/b3827bd6c9ea67ffa3912002515c64e4d9fe5c4dacbf7c46b0d8768b7abbb84f-other.xml.gz"/>
+    <timestamp>1666177486</timestamp>
+    <size>999</size>
+    <open-size>6320</open-size>
+  </data>
+  <data type="primary_db">
+    <checksum type="sha256">ab8df35061dfa0285069b843f24a7076e31266d9a8abe8282340bcb936aa61d7</checksum>
+    <open-checksum type="sha256">2bce9554ce4496cef34b5cd69f186f7f3143c7cabae8fa384fc5c9eeab326f7f</open-checksum>
+    <location href="repodata/ab8df35061dfa0285069b843f24a7076e31266d9a8abe8282340bcb936aa61d7-primary.sqlite.bz2"/>
+    <timestamp>1666177486</timestamp>
+    <size>3558</size>
+    <open-size>106496</open-size>
+    <database_version>10</database_version>
+  </data>
+  <data type="filelists_db">
+    <checksum type="sha256">8bcf6d40db4e922934ac47e8ac7fb8d15bdacf579af8c819d2134ed54d30550b</checksum>
+    <open-checksum type="sha256">f7001d1df7f5f7e4898919b15710bea8ed9711ce42faf68e22b757e63169b1fb</open-checksum>
+    <location href="repodata/8bcf6d40db4e922934ac47e8ac7fb8d15bdacf579af8c819d2134ed54d30550b-filelists.sqlite.bz2"/>
+    <timestamp>1666177486</timestamp>
+    <size>2360</size>
+    <open-size>28672</open-size>
+    <database_version>10</database_version>
+  </data>
+  <data type="other_db">
+    <checksum type="sha256">01b82e9eb7ee9151f283c6e761ae450de18ed2d64b5e32de88689eaf95216a80</checksum>
+    <open-checksum type="sha256">07f5b9750af1e440d37ca216e719dd288149e79e9132f2fdccb6f73b2e5dd541</open-checksum>
+    <location href="repodata/01b82e9eb7ee9151f283c6e761ae450de18ed2d64b5e32de88689eaf95216a80-other.sqlite.bz2"/>
+    <timestamp>1666177486</timestamp>
+    <size>2196</size>
+    <open-size>32768</open-size>
+    <database_version>10</database_version>
+  </data>
+</repomd>

tests/data/test_create_packages_json/test_repo_2/repodata/repomd.xml

@@ -0,0 +1,55 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<repomd xmlns="http://linux.duke.edu/metadata/repo" xmlns:rpm="http://linux.duke.edu/metadata/rpm">
+  <revision>1666177500</revision>
+  <data type="primary">
+    <checksum type="sha256">a1d342aa7cef3a2034fc3f9d6ee02d63572780bc76e61749a57e50b6b3ca9869</checksum>
+    <open-checksum type="sha256">a9e3eae447dd44282d7d96db5f15f049b757925397adb752f4df982176bab7e0</open-checksum>
+    <location href="repodata/a1d342aa7cef3a2034fc3f9d6ee02d63572780bc76e61749a57e50b6b3ca9869-primary.xml.gz"/>
+    <timestamp>1666177500</timestamp>
+    <size>3501</size>
+    <open-size>37296</open-size>
+  </data>
+  <data type="filelists">
+    <checksum type="sha256">6778922d5853d20f213ae7702699a76f1e87e55d6bfb5e4ac6a117d904d47b3c</checksum>
+    <open-checksum type="sha256">e30b666d9d88a70de69a08f45e6696bcd600c45485d856bd0213395d7da7bd49</open-checksum>
+    <location href="repodata/6778922d5853d20f213ae7702699a76f1e87e55d6bfb5e4ac6a117d904d47b3c-filelists.xml.gz"/>
+    <timestamp>1666177500</timestamp>
+    <size>27624</size>
+    <open-size>318187</open-size>
+  </data>
+  <data type="other">
+    <checksum type="sha256">5a60d79d8bce6a805f4fdb22fd891524359dce8ccc665c0b54e7299e79debe84</checksum>
+    <open-checksum type="sha256">b18138f4a3de45714e578fb1f30b7ec54fdcdaf1a22585891625b6af0894388e</open-checksum>
+    <location href="repodata/5a60d79d8bce6a805f4fdb22fd891524359dce8ccc665c0b54e7299e79debe84-other.xml.gz"/>
+    <timestamp>1666177500</timestamp>
+    <size>1876</size>
+    <open-size>28701</open-size>
+  </data>
+  <data type="primary_db">
+    <checksum type="sha256">c27bc2ce947173aba305041552c3c6d8db71442c1a2e5dcaf35ff750fe0469fc</checksum>
+    <open-checksum type="sha256">586e1af8934229925adb9e746ae5ced119859dfd97f4e3237399bb36a7d7f071</open-checksum>
+    <location href="repodata/c27bc2ce947173aba305041552c3c6d8db71442c1a2e5dcaf35ff750fe0469fc-primary.sqlite.bz2"/>
+    <timestamp>1666177500</timestamp>
+    <size>11528</size>
+    <open-size>126976</open-size>
+    <database_version>10</database_version>
+  </data>
+  <data type="filelists_db">
+    <checksum type="sha256">ed350865982e7a1e45b144839b56eac888e5d8f680571dd2cd06b37dc83e0fd8</checksum>
+    <open-checksum type="sha256">697903989d0f77de2d44a2b603e75c9b4ca23b3795eb136d175caf5666ce6459</open-checksum>
+    <location href="repodata/ed350865982e7a1e45b144839b56eac888e5d8f680571dd2cd06b37dc83e0fd8-filelists.sqlite.bz2"/>
+    <timestamp>1666177500</timestamp>
+    <size>20440</size>
+    <open-size>163840</open-size>
+    <database_version>10</database_version>
+  </data>
+  <data type="other_db">
+    <checksum type="sha256">35eff699131e0976429144c6f4514d21568177dc64bb4091c3ff62f76b293725</checksum>
+    <open-checksum type="sha256">3bd999a1bdf300df836a4607b7b75f845d8e1432e3e4e1ab6f0c7cc8a853db39</open-checksum>
+    <location href="repodata/35eff699131e0976429144c6f4514d21568177dc64bb4091c3ff62f76b293725-other.sqlite.bz2"/>
+    <timestamp>1666177500</timestamp>
+    <size>4471</size>
+    <open-size>49152</open-size>
+    <database_version>10</database_version>
+  </data>
+</repomd>

tests/fixtures/extraiso-tweaked-expected.treeinfo

@@ -0,0 +1,58 @@
+[checksums]
+images/boot.iso = sha256:fc8a4be604b6425746f12fa706116eb940f93358f036b8fbbe518b516cb6870c
+
+[general]
+; WARNING.0 = This section provides compatibility with pre-productmd treeinfos.
+; WARNING.1 = Read productmd documentation for details about new format.
+arch = x86_64
+family = Test
+name = Test 1.0
+packagedir = Packages
+platforms = x86_64,xen
+repository = .
+timestamp = 1531881582
+variant = Server
+variants = Client,Server
+version = 1.0
+
+[header]
+type = productmd.treeinfo
+version = 1.2
+
+[images-x86_64]
+boot.iso = images/boot.iso
+
+[images-xen]
+initrd = images/pxeboot/initrd.img
+kernel = images/pxeboot/vmlinuz
+
+[release]
+name = Test
+short = T
+version = 1.0
+
+[stage2]
+mainimage = images/install.img
+
+[tree]
+arch = x86_64
+build_timestamp = 1531881582
+platforms = x86_64,xen
+variants = Client,Server
+
+[variant-Client]
+id = Client
+name = Client
+packages = ../../../Client/x86_64/os/Packages
+repository = ../../../Client/x86_64/os
+type = variant
+uid = Client
+
+[variant-Server]
+id = Server
+name = Server
+packages = Packages
+repository = .
+type = variant
+uid = Server
+

tests/fixtures/mmds/modules/x86_64/module:master-20190318-abcdef

@@ -0,0 +1,20 @@
+---
+document: modulemd
+version: 2
+data:
+  name: module
+  stream: master
+  version: 20190318
+  context: abcdef
+  arch: x86_64
+  summary: Dummy module
+  description: Dummy module
+  license:
+    module:
+    - Beerware
+    content:
+    - Beerware
+  artifacts:
+    rpms:
+      - foobar-0:1.0-1.noarch
+...

tests/fixtures/mmds/modules/x86_64/modulemd.x86_64.txt

@@ -0,0 +1,20 @@
+---
+document: modulemd
+version: 2
+data:
+  name: module
+  stream: master
+  version: 20190318
+  context: abcdef
+  arch: x86_64
+  summary: Dummy module
+  description: Dummy module
+  license:
+    module:
+    - Beerware
+    content:
+    - Beerware
+  artifacts:
+    rpms:
+      - foobar-0:1.0-1.noarch
+...

tests/fixtures/mmds/modules/x86_64/scratch-module.x86_64.txt

@@ -0,0 +1,20 @@
+---
+document: modulemd
+version: 2
+data:
+  name: scratch-module
+  stream: master
+  version: 20200710
+  context: abcdef
+  arch: x86_64
+  summary: Dummy module
+  description: Dummy module
+  license:
+    module:
+    - Beerware
+    content:
+    - Beerware
+  artifacts:
+    rpms:
+      - foobar-0:1.0-1.noarch
+...

tests/fixtures/mmds/modules/x86_64/scratch-module:master-20190318-abcdef

@@ -0,0 +1,20 @@
+---
+document: modulemd
+version: 2
+data:
+  name: scratch-module
+  stream: master
+  version: 20200710
+  context: abcdef
+  arch: x86_64
+  summary: Dummy module
+  description: Dummy module
+  license:
+    module:
+    - Beerware
+    content:
+    - Beerware
+  artifacts:
+    rpms:
+      - foobar-0:1.0-1.noarch
+...

tests/helpers.py

@@ -7,7 +7,7 @@ import shutil
 import tempfile
 from collections import defaultdict
 
-import mock
+from unittest import mock
 import six
 from kobo.rpmlib import parse_nvr
 
@@ -21,6 +21,15 @@ from pungi import paths, checks
 from pungi.module_util import Modulemd
 
 
+GIT_WITH_CREDS = [
+    "git",
+    "-c",
+    "credential.useHttpPath=true",
+    "-c",
+    "credential.helper=!ch",
+]
+
+
 class BaseTestCase(unittest.TestCase):
     def assertFilesEqual(self, fn1, fn2):
         with open(fn1, "rb") as f1:
@@ -158,6 +167,20 @@ class IterableMock(mock.Mock):
         return iter([])
 
 
+class FSKojiDownloader(object):
+    """Mock for KojiDownloadProxy that checks provided path."""
+
+    def get_file(self, path, validator=None):
+        return path if os.path.isfile(path) else None
+
+
+class DummyKojiDownloader(object):
+    """Mock for KojiDownloadProxy that always finds the file in original location."""
+
+    def get_file(self, path, validator=None):
+        return path
+
+
 class DummyCompose(object):
     def __init__(self, topdir, config):
         self.supported = True
@@ -232,6 +255,8 @@ class DummyCompose(object):
         self.cache_region = None
         self.containers_metadata = {}
         self.load_old_compose_config = mock.Mock(return_value=None)
+        self.koji_downloader = DummyKojiDownloader()
+        self.koji_downloader.path_prefix = "/prefix"
 
     def setup_optional(self):
         self.all_variants["Server-optional"] = MockVariant(
@@ -272,7 +297,7 @@ class DummyCompose(object):
         return tempfile.mkdtemp(suffix=suffix, prefix=prefix, dir=self.topdir)
 
 
-def touch(path, content=None):
+def touch(path, content=None, mode=None):
     """Helper utility that creates an dummy file in given location. Directories
     will be created."""
     content = content or (path + "\n")
@@ -284,6 +309,8 @@ def touch(path, content=None):
         content = content.encode()
     with open(path, "wb") as f:
         f.write(content)
+    if mode:
+        os.chmod(path, mode)
     return path
 
 

tests/test_arch.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-import mock
+from unittest import mock
 import unittest
 
 from pungi.arch import (

tests/test_arguments.py

@@ -1,4 +1,4 @@
-import mock
+from unittest import mock
 
 try:
     import unittest2 as unittest

tests/test_buildinstall.py

@@ -1,15 +1,15 @@
 # -*- coding: utf-8 -*-
 
-
 try:
     import unittest2 as unittest
 except ImportError:
     import unittest
 
-import mock
+from unittest import mock
 import six
 from copy import copy
 from six.moves import StringIO
+from ddt import ddt, data
 
 import os
 
@@ -1209,6 +1209,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "lorax",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "runroot_weights": {"buildinstall": 123},
             },
         )
@@ -1308,6 +1309,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "lorax_use_koji_plugin": True,
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "runroot_weights": {"buildinstall": 123},
             },
         )
@@ -1412,6 +1414,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "buildinstall",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
             },
         )
 
@@ -1500,6 +1503,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "buildinstall",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "failable_deliverables": [("^.+$", {"*": ["buildinstall"]})],
             },
         )
@@ -1542,6 +1546,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "lorax",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "failable_deliverables": [("^.+$", {"*": ["buildinstall"]})],
             },
         )
@@ -1591,6 +1596,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "lorax",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "failable_deliverables": [("^.+$", {"*": ["buildinstall"]})],
             },
         )
@@ -1663,6 +1669,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "lorax",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "failable_deliverables": [("^.+$", {"*": ["buildinstall"]})],
             },
         )
@@ -1701,6 +1708,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "lorax",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "runroot_weights": {"buildinstall": 123},
                 "buildinstall_topdir": "/buildinstall_topdir",
             },
@@ -1810,6 +1818,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
                 "buildinstall_method": "lorax",
                 "runroot_tag": "rrt",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
             },
         )
 
@@ -2005,6 +2014,7 @@ class BuildinstallThreadTestCase(PungiTestCase):
         self.assertEqual(ret, None)
 
 
+@ddt
 class TestSymlinkIso(PungiTestCase):
     def setUp(self):
         super(TestSymlinkIso, self).setUp()
@@ -2020,8 +2030,13 @@ class TestSymlinkIso(PungiTestCase):
     @mock.patch("pungi.phases.buildinstall.get_file_size")
     @mock.patch("pungi.phases.buildinstall.iso")
     @mock.patch("pungi.phases.buildinstall.run")
-    def test_hardlink(self, run, iso, get_file_size, get_mtime, ImageCls):
-        self.compose.conf = {"buildinstall_symlink": False, "disc_types": {}}
+    @data(['Server'], ['BaseOS'])
+    def test_hardlink(self, netinstall_variants, run, iso, get_file_size, get_mtime, ImageCls):
+        self.compose.conf = {
+            "buildinstall_symlink": False,
+            "disc_types": {},
+            "netinstall_variants": netinstall_variants,
+        }
         get_file_size.return_value = 1024
         get_mtime.return_value = 13579
 
@@ -2071,9 +2086,14 @@ class TestSymlinkIso(PungiTestCase):
         self.assertEqual(image.bootable, True)
         self.assertEqual(image.implant_md5, iso.get_implanted_md5.return_value)
         self.assertEqual(image.can_fail, False)
-        self.assertEqual(
-            self.compose.im.add.mock_calls, [mock.call("Server", "x86_64", image)]
-        )
+        if 'Server' in netinstall_variants:
+            self.assertEqual(
+                self.compose.im.add.mock_calls, [mock.call("Server", "x86_64", image)]
+            )
+        else:
+            self.assertEqual(
+                self.compose.im.add.mock_calls, []
+            )
 
     @mock.patch("pungi.phases.buildinstall.Image")
     @mock.patch("pungi.phases.buildinstall.get_mtime")
@@ -2086,6 +2106,7 @@ class TestSymlinkIso(PungiTestCase):
         self.compose.conf = {
             "buildinstall_symlink": False,
             "disc_types": {"boot": "netinst"},
+            "netinstall_variants": ['Server'],
         }
         get_file_size.return_value = 1024
         get_mtime.return_value = 13579

tests/test_checks.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-import mock
+from unittest import mock
 
 try:
     import unittest2 as unittest

tests/test_compose.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 
 import logging
-import mock
+from unittest import mock
 
 try:
     import unittest2 as unittest
@@ -628,6 +628,7 @@ class ComposeTestCase(unittest.TestCase):
         ci_copy = dict(self.ci_json)
         ci_copy["header"]["version"] = "1.2"
         mocked_response = mock.MagicMock()
+        mocked_response.status_code = 200
         mocked_response.text = json.dumps(self.ci_json)
         mocked_requests.post.return_value = mocked_response
 
@@ -655,6 +656,7 @@ class ComposeTestCase(unittest.TestCase):
             mocked_requests.post.assert_called_once_with(
                 "https://cts.localhost.tld/api/1/composes/",
                 auth=mock.ANY,
+                data=None,
                 json=expected_json,
             )
 
@@ -793,12 +795,16 @@ class TracebackTest(unittest.TestCase):
         shutil.rmtree(self.tmp_dir)
         self.patcher.stop()
 
-    def assertTraceback(self, filename):
+    def assertTraceback(self, filename, show_locals=True):
         self.assertTrue(
             os.path.isfile("%s/logs/global/%s.global.log" % (self.tmp_dir, filename))
         )
         self.assertEqual(
-            self.Traceback.mock_calls, [mock.call(), mock.call().get_traceback()]
+            self.Traceback.mock_calls,
+            [
+                mock.call(show_locals=show_locals),
+                mock.call(show_locals=show_locals).get_traceback(),
+            ],
         )
 
     def test_traceback_default(self):
@@ -811,6 +817,7 @@ class TracebackTest(unittest.TestCase):
 
 
 class RetryRequestTest(unittest.TestCase):
+    @mock.patch("time.sleep", new=lambda x: x)
     @mock.patch("pungi.compose.requests")
     def test_retry_timeout(self, mocked_requests):
         mocked_requests.post.side_effect = [
@@ -822,8 +829,22 @@ class RetryRequestTest(unittest.TestCase):
         self.assertEqual(
             mocked_requests.mock_calls,
             [
-                mock.call.post(url, json=None, auth=None),
-                mock.call.post(url, json=None, auth=None),
+                mock.call.post(url, data=None, json=None, auth=None),
+                mock.call.post(url, data=None, json=None, auth=None),
             ],
         )
         self.assertEqual(rv.status_code, 200)
+
+    @mock.patch("pungi.compose.requests")
+    def test_no_retry_on_client_error(self, mocked_requests):
+        mocked_requests.post.side_effect = [
+            mock.Mock(status_code=400, json=lambda: {"message": "You made a mistake"}),
+        ]
+        url = "http://locahost/api/1/composes/"
+        with self.assertRaises(RuntimeError):
+            retry_request("post", url)
+
+        self.assertEqual(
+            mocked_requests.mock_calls,
+            [mock.call.post(url, data=None, json=None, auth=None)],
+        )

tests/test_config.py

@@ -7,7 +7,7 @@ except ImportError:
     import unittest
 
 import six
-import mock
+from unittest import mock
 
 from pungi import checks
 from tests.helpers import load_config, PKGSET_REPOS
@@ -440,7 +440,7 @@ class LiveMediaConfigTestCase(ConfigTestCase):
             live_media_version="Rawhide",
         )
 
-        resolve_git_url.side_effect = lambda x: x.replace("HEAD", "CAFE")
+        resolve_git_url.side_effect = lambda x, _helper: x.replace("HEAD", "CAFE")
 
         self.assertValidation(cfg)
         self.assertEqual(cfg["live_media_ksurl"], "git://example.com/repo.git#CAFE")

tests/test_config_validate_script.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 
 
-import mock
+from unittest import mock
 import os
 import six
 

tests/test_create_extra_repo.py

@@ -0,0 +1,228 @@
+# coding=utf-8
+
+import os
+from unittest import TestCase, mock, main
+
+import yaml
+
+from pungi.scripts.create_extra_repo import CreateExtraRepo, ExtraVariantInfo, RepoInfo
+
+FOLDER_WITH_TEST_DATA = os.path.join(
+    os.path.dirname(
+        os.path.abspath(__file__)
+    ),
+    'data/test_create_extra_repo/',
+)
+
+TEST_MODULE_INFO = yaml.load("""
+---
+document: modulemd
+version: 2
+data:
+  name: perl-App-cpanminus
+  stream: 1.7044
+  version: 8030020210126085450
+  context: 3a33b840
+  arch: x86_64
+  summary: Get, unpack, build and install CPAN modules
+  description: >
+    This is a CPAN client that requires zero configuration, and stands alone but it's
+    maintainable and extensible with plug-ins and friendly to shell scripting.
+  license:
+    module:
+    - MIT
+    content:
+    - (GPL+ or Artistic) and GPLv2+
+    - ASL 2.0
+    - GPL+ or Artistic
+  dependencies:
+  - buildrequires:
+      perl: [5.30]
+      platform: [el8.3.0]
+    requires:
+      perl: [5.30]
+      perl-YAML: []
+      platform: [el8]
+  references:
+    community: https://metacpan.org/release/App-cpanminus
+  profiles:
+    common:
+      description: App-cpanminus distribution
+      rpms:
+      - perl-App-cpanminus
+  api:
+    rpms:
+    - perl-App-cpanminus
+  filter:
+    rpms:
+    - perl-CPAN-DistnameInfo-dummy
+    - perl-Test-Deep
+  buildopts:
+    rpms:
+      macros: >
+        %_without_perl_CPAN_Meta_Check_enables_extra_test 1
+  components:
+    rpms:
+      perl-App-cpanminus:
+        rationale: The API.
+        ref: perl-App-cpanminus-1.7044-5.module+el8.2.0+4278+abcfa81a.src.rpm
+        buildorder: 1
+        arches: [i686, x86_64]
+      perl-CPAN-DistnameInfo:
+        rationale: Run-time dependency.
+        ref: stream-0.12-rhel-8.3.0
+        arches: [i686, x86_64]
+      perl-CPAN-Meta-Check:
+        rationale: Run-time dependency.
+        ref: perl-CPAN-Meta-Check-0.014-6.module+el8.2.0+4278+abcfa81a.src.rpm
+        buildorder: 1
+        arches: [i686, x86_64]
+      perl-File-pushd:
+        rationale: Run-time dependency.
+        ref: perl-File-pushd-1.014-6.module+el8.2.0+4278+abcfa81a.src.rpm
+        arches: [i686, x86_64]
+      perl-Module-CPANfile:
+        rationale: Run-time dependency.
+        ref: perl-Module-CPANfile-1.1002-7.module+el8.2.0+4278+abcfa81a.src.rpm
+        arches: [i686, x86_64]
+      perl-Parse-PMFile:
+        rationale: Run-time dependency.
+        ref: perl-Parse-PMFile-0.41-7.module+el8.2.0+4278+abcfa81a.src.rpm
+        arches: [i686, x86_64]
+      perl-String-ShellQuote:
+        rationale: Run-time dependency.
+        ref: perl-String-ShellQuote-1.04-24.module+el8.2.0+4278+abcfa81a.src.rpm
+        arches: [i686, x86_64]
+      perl-Test-Deep:
+        rationale: Build-time dependency.
+        ref: stream-1.127-rhel-8.3.0
+        arches: [i686, x86_64]
+  artifacts:
+    rpms:
+    - perl-App-cpanminus-0:1.7044-5.module_el8.3.0+2027+c8990d1d.noarch
+    - perl-App-cpanminus-0:1.7044-5.module_el8.3.0+2027+c8990d1d.src
+    - perl-CPAN-Meta-Check-0:0.014-6.module_el8.3.0+2027+c8990d1d.noarch
+    - perl-CPAN-Meta-Check-0:0.014-6.module_el8.3.0+2027+c8990d1d.src
+    - perl-File-pushd-0:1.014-6.module_el8.3.0+2027+c8990d1d.noarch
+    - perl-File-pushd-0:1.014-6.module_el8.3.0+2027+c8990d1d.src
+    - perl-Module-CPANfile-0:1.1002-7.module_el8.3.0+2027+c8990d1d.noarch
+    - perl-Module-CPANfile-0:1.1002-7.module_el8.3.0+2027+c8990d1d.src
+    - perl-Parse-PMFile-0:0.41-7.module_el8.3.0+2027+c8990d1d.noarch
+    - perl-Parse-PMFile-0:0.41-7.module_el8.3.0+2027+c8990d1d.src
+    - perl-String-ShellQuote-0:1.04-24.module_el8.3.0+2027+c8990d1d.noarch
+    - perl-String-ShellQuote-0:1.04-24.module_el8.3.0+2027+c8990d1d.src
+...
+""", Loader=yaml.BaseLoader)
+
+TEST_REPO_INFO = RepoInfo(
+    path=FOLDER_WITH_TEST_DATA,
+    folder='test_repo',
+    is_remote=False,
+)
+TEST_VARIANT_INFO = ExtraVariantInfo(
+    name='TestRepo',
+    arch='x86_64',
+    packages=[],
+    modules=[],
+    repos=[TEST_REPO_INFO]
+)
+
+BS_BUILD_INFO = {
+    'build_platforms': [
+        {
+            'architectures': ['non_fake_arch', 'fake_arch'],
+            'name': 'fake_platform'
+        }
+    ]
+}
+
+
+class TestCreteExtraRepo(TestCase):
+
+    maxDiff = None
+
+    def test_01_get_repo_info_from_bs_repo(self):
+        auth_token = 'fake_auth_token'
+        build_id = 'fake_build_id'
+        arch = 'fake_arch'
+        packages = ['fake_package1', 'fake_package2']
+        modules = ['fake_module1', 'fake_module2']
+
+        request_object = mock.Mock()
+        request_object.raise_for_status = lambda: True
+        request_object.json = lambda: BS_BUILD_INFO
+        with mock.patch(
+            'pungi.scripts.create_extra_repo.requests.get',
+            return_value=request_object,
+        ) as mock_request_get:
+            repos_info = CreateExtraRepo.get_repo_info_from_bs_repo(
+                auth_token=auth_token,
+                build_id=build_id,
+                arch=arch,
+                packages=packages,
+                modules=modules,
+            )
+            self.assertEqual(
+                [
+                    ExtraVariantInfo(
+                        name=f'{build_id}-fake_platform-{arch}',
+                        arch=arch,
+                        packages=packages,
+                        modules=modules,
+                        repos=[
+                            RepoInfo(
+                                path='https://build.cloudlinux.com/'
+                                     f'build_repos/{build_id}/fake_platform',
+                                folder=arch,
+                                is_remote=True,
+                            )
+                        ]
+                    )
+                ],
+                repos_info,
+            )
+            mock_request_get.assert_called_once_with(
+                url=f'https://build.cloudlinux.com/api/v1/builds/{build_id}',
+                headers={
+                    'Authorization': f'Bearer {auth_token}',
+                }
+            )
+
+    def test_02_create_extra_repo(self):
+        with mock.patch(
+            'pungi.scripts.create_extra_repo.'
+            'CreateExtraRepo._read_local_modules_yaml',
+            return_value=[],
+        ) as mock__read_local_modules_yaml, mock.patch(
+            'pungi.scripts.create_extra_repo.'
+            'CreateExtraRepo._download_rpm_to_local_repo',
+        ) as mock__download_rpm_to_local_repo, mock.patch(
+            'pungi.scripts.create_extra_repo.'
+            'CreateExtraRepo._dump_local_modules_yaml'
+        ) as mock__dump_local_modules_yaml, mock.patch(
+            'pungi.scripts.create_extra_repo.'
+            'CreateExtraRepo._create_local_extra_repo'
+        ) as mock__create_local_extra_repo:
+            cer = CreateExtraRepo(
+                variants=[TEST_VARIANT_INFO],
+                bs_auth_token='fake_auth_token',
+                local_repository_path='/path/to/local/repo',
+                clear_target_repo=False,
+            )
+            mock__read_local_modules_yaml.assert_called_once_with()
+            cer.create_extra_repo()
+            mock__download_rpm_to_local_repo.assert_called_once_with(
+                package_location='perl-App-cpanminus-1.7044-5.'
+                                 'module_el8.3.0+2027+c8990d1d.noarch.rpm',
+                repo_info=TEST_REPO_INFO,
+            )
+            mock__dump_local_modules_yaml.assert_called_once_with()
+            mock__create_local_extra_repo.assert_called_once_with()
+            self.assertEqual(
+                [TEST_MODULE_INFO],
+                cer.local_modules_data,
+            )
+
+
+if __name__ == '__main__':
+    main()

tests/test_create_packages_json.py

@@ -0,0 +1,112 @@
+# coding=utf-8
+
+import os
+from collections import defaultdict
+from unittest import TestCase, mock, main
+
+from pungi.scripts.create_packages_json import (
+    PackagesGenerator,
+    RepoInfo,
+    VariantInfo,
+)
+
+FOLDER_WITH_TEST_DATA = os.path.join(
+    os.path.dirname(
+        os.path.abspath(__file__)
+    ),
+    'data/test_create_packages_json/',
+)
+
+test_repo_info = RepoInfo(
+    path=FOLDER_WITH_TEST_DATA,
+    folder='test_repo',
+    is_remote=False,
+    is_reference=True,
+)
+
+test_repo_info_2 = RepoInfo(
+    path=FOLDER_WITH_TEST_DATA,
+    folder='test_repo_2',
+    is_remote=False,
+    is_reference=True,
+)
+variant_info_1 = VariantInfo(
+    name='TestRepo',
+    arch='x86_64',
+    repos=[test_repo_info]
+)
+variant_info_2 = VariantInfo(
+    name='TestRepo2',
+    arch='x86_64',
+    repos=[test_repo_info_2]
+)
+
+
+class TestPackagesJson(TestCase):
+    def test_01_get_remote_file_content(self):
+        """
+        Test the getting of content from a remote file
+        """
+        request_object = mock.Mock()
+        request_object.raise_for_status = lambda: True
+        request_object.content = b'TestContent'
+        with mock.patch(
+            'pungi.scripts.create_packages_json.requests.get',
+            return_value=request_object,
+        ) as mock_requests_get, mock.patch(
+            'pungi.scripts.create_packages_json.tempfile.NamedTemporaryFile',
+        ) as mock_tempfile:
+            mock_tempfile.return_value.__enter__.return_value.name = 'tmpfile'
+            packages_generator = PackagesGenerator(
+                variants=[],
+                excluded_packages=[],
+                included_packages=[],
+            )
+            file_name = packages_generator.get_remote_file_content(
+                file_url='fakeurl')
+            mock_requests_get.assert_called_once_with(url='fakeurl')
+            mock_tempfile.assert_called_once_with(delete=False)
+            mock_tempfile.return_value.__enter__().\
+                write.assert_called_once_with(b'TestContent')
+            self.assertEqual(
+                file_name,
+                'tmpfile',
+            )
+
+    def test_02_generate_additional_packages(self):
+        pg = PackagesGenerator(
+            variants=[
+                variant_info_1,
+                variant_info_2,
+            ],
+            excluded_packages=['zziplib-utils'],
+            included_packages=['vim-file*'],
+        )
+        test_packages = defaultdict(
+            lambda: defaultdict(
+                lambda: defaultdict(
+                    list,
+                )
+            )
+        )
+        test_packages['TestRepo']['x86_64']['zziplib'] = \
+            [
+                'zziplib.i686',
+                'zziplib.x86_64',
+            ]
+        test_packages['TestRepo2']['x86_64']['vim'] = \
+            [
+                'vim-X11.i686',
+                'vim-common.i686',
+                'vim-enhanced.i686',
+                'vim-filesystem.noarch',
+            ]
+        result = pg.generate_packages_json()
+        self.assertEqual(
+            test_packages,
+            result,
+        )
+
+
+if __name__ == '__main__':
+    main()

tests/test_createiso_phase.py

@@ -2,7 +2,7 @@
 
 
 import logging
-import mock
+from unittest import mock
 import six
 
 import os
@@ -552,6 +552,7 @@ class CreateisoThreadTest(helpers.PungiTestCase):
                 "release_version": "1.0",
                 "runroot_tag": "f25-build",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
             },
         )
         cmd = {
@@ -633,6 +634,7 @@ class CreateisoThreadTest(helpers.PungiTestCase):
                 "release_version": "1.0",
                 "runroot_tag": "f25-build",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "create_jigdo": False,
                 "runroot_weights": {"createiso": 123},
             },
@@ -717,6 +719,7 @@ class CreateisoThreadTest(helpers.PungiTestCase):
                 "buildinstall_method": "lorax",
                 "runroot_tag": "f25-build",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
             },
         )
         cmd = {
@@ -807,6 +810,7 @@ class CreateisoThreadTest(helpers.PungiTestCase):
                 "release_version": "1.0",
                 "runroot_tag": "f25-build",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
             },
         )
         cmd = {
@@ -839,6 +843,7 @@ class CreateisoThreadTest(helpers.PungiTestCase):
                 "release_version": "1.0",
                 "runroot_tag": "f25-build",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "failable_deliverables": [("^.*$", {"*": "iso"})],
             },
         )
@@ -881,6 +886,7 @@ class CreateisoThreadTest(helpers.PungiTestCase):
                 "release_version": "1.0",
                 "runroot_tag": "f25-build",
                 "koji_profile": "koji",
+                "koji_cache": "/tmp",
                 "failable_deliverables": [("^.*$", {"*": "iso"})],
             },
         )

tests/test_createiso_script.py

@@ -1,6 +1,7 @@
 # -*- coding: utf-8 -*-
 
-import mock
+from unittest import mock
+from parameterized import parameterized
 
 import os
 from six.moves import StringIO
@@ -391,3 +392,27 @@ class CreateIsoScriptTest(helpers.PungiTestCase):
                 ),
             ]
         )
+
+    @parameterized.expand(
+        [("644", 0o644), ("664", 0o664), ("666", 0o666), ("2644", 0o2644)]
+    )
+    def test_get_perms_non_executable(self, test_name, mode):
+        path = helpers.touch(os.path.join(self.topdir, "f"), mode=mode)
+        self.assertEqual(createiso._get_perms(path), 0o444)
+
+    @parameterized.expand(
+        [
+            ("544", 0o544),
+            ("554", 0o554),
+            ("555", 0o555),
+            ("744", 0o744),
+            ("755", 0o755),
+            ("774", 0o774),
+            ("775", 0o775),
+            ("777", 0o777),
+            ("2775", 0o2775),
+        ]
+    )
+    def test_get_perms_executable(self, test_name, mode):
+        path = helpers.touch(os.path.join(self.topdir, "f"), mode=mode)
+        self.assertEqual(createiso._get_perms(path), 0o555)

tests/test_createrepophase.py

@@ -8,7 +8,7 @@ except ImportError:
 import glob
 import os
 
-import mock
+from unittest import mock
 import six
 
 from pungi.module_util import Modulemd

tests/test_extra_files_phase.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-import mock
+from unittest import mock
 import os
 
 from productmd.extra_files import ExtraFiles

tests/test_extra_isos_phase.py

@@ -1,8 +1,8 @@
 # -*- coding: utf-8 -*-
-
-import logging
-import mock
+from typing import AnyStr, List
+from unittest import mock
 import six
+import logging
 
 import os
 
@@ -614,6 +614,7 @@ class GetExtraFilesTest(helpers.PungiTestCase):
         )
 
 
+@mock.patch("pungi.phases.extra_isos.tweak_repo_treeinfo")
 @mock.patch("pungi.phases.extra_isos.tweak_treeinfo")
 @mock.patch("pungi.wrappers.iso.write_graft_points")
 @mock.patch("pungi.wrappers.iso.get_graft_points")
@@ -623,7 +624,7 @@ class GetIsoContentsTest(helpers.PungiTestCase):
         self.compose = helpers.DummyCompose(self.topdir, {})
         self.variant = self.compose.variants["Server"]
 
-    def test_non_bootable_binary(self, ggp, wgp, tt):
+    def test_non_bootable_binary(self, ggp, wgp, tt, trt):
         gp = {
             "compose/Client/x86_64/os/Packages": {"f/foo.rpm": "/mnt/f/foo.rpm"},
             "compose/Client/x86_64/os/repodata": {
@@ -693,7 +694,15 @@ class GetIsoContentsTest(helpers.PungiTestCase):
             ],
         )
 
-    def test_inherit_extra_files(self, ggp, wgp, tt):
+        # Check correct call to tweak_repo_treeinfo
+        self._tweak_repo_treeinfo_call_list_checker(
+            trt_mock=trt,
+            main_variant='Server',
+            addon_variants=['Client'],
+            sub_path='x86_64/os',
+        )
+
+    def test_inherit_extra_files(self, ggp, wgp, tt, trt):
         gp = {
             "compose/Client/x86_64/os/Packages": {"f/foo.rpm": "/mnt/f/foo.rpm"},
             "compose/Client/x86_64/os/repodata": {
@@ -767,7 +776,15 @@ class GetIsoContentsTest(helpers.PungiTestCase):
             ],
         )
 
-    def test_source(self, ggp, wgp, tt):
+        # Check correct call to tweak_repo_treeinfo
+        self._tweak_repo_treeinfo_call_list_checker(
+            trt_mock=trt,
+            main_variant='Server',
+            addon_variants=['Client'],
+            sub_path='x86_64/os',
+        )
+
+    def test_source(self, ggp, wgp, tt, trt):
         gp = {
             "compose/Client/source/tree/Packages": {"f/foo.rpm": "/mnt/f/foo.rpm"},
             "compose/Client/source/tree/repodata": {
@@ -837,7 +854,15 @@ class GetIsoContentsTest(helpers.PungiTestCase):
             ],
         )
 
-    def test_bootable(self, ggp, wgp, tt):
+        # Check correct call to tweak_repo_treeinfo
+        self._tweak_repo_treeinfo_call_list_checker(
+            trt_mock=trt,
+            main_variant='Server',
+            addon_variants=['Client'],
+            sub_path='source/tree',
+        )
+
+    def test_bootable(self, ggp, wgp, tt, trt):
         self.compose.conf["buildinstall_method"] = "lorax"
 
         bi_dir = os.path.join(self.topdir, "work/x86_64/buildinstall/Server")
@@ -939,6 +964,42 @@ class GetIsoContentsTest(helpers.PungiTestCase):
             ],
         )
 
+        # Check correct call to tweak_repo_treeinfo
+        self._tweak_repo_treeinfo_call_list_checker(
+            trt_mock=trt,
+            main_variant='Server',
+            addon_variants=['Client'],
+            sub_path='x86_64/os',
+        )
+
+    def _tweak_repo_treeinfo_call_list_checker(
+            self,
+            trt_mock: mock.Mock,
+            main_variant: AnyStr,
+            addon_variants: List[AnyStr],
+            sub_path: AnyStr) -> None:
+        """
+        Check correct call to tweak_repo_treeinfo
+        """
+        path_to_treeinfo = os.path.join(
+            self.topdir,
+            'compose',
+            main_variant,
+            sub_path,
+            '.treeinfo',
+        )
+        self.assertEqual(
+            trt_mock.call_args_list,
+            [
+                mock.call(
+                    self.compose,
+                    addon_variants,
+                    path_to_treeinfo,
+                    path_to_treeinfo,
+                )
+            ]
+        )
+
 
 class GetFilenameTest(helpers.PungiTestCase):
     def test_use_original_name(self):
@@ -1016,6 +1077,15 @@ class TweakTreeinfoTest(helpers.PungiTestCase):
 
         self.assertFilesEqual(output, expected)
 
+    def test_repo_tweak(self):
+        compose = helpers.DummyCompose(self.topdir, {})
+        input = os.path.join(helpers.FIXTURE_DIR, "extraiso.treeinfo")
+        output = os.path.join(self.topdir, "actual-treeinfo")
+        expected = os.path.join(helpers.FIXTURE_DIR, "extraiso-tweaked-expected.treeinfo")
+        extra_isos.tweak_repo_treeinfo(compose, ["Client"], input, output)
+
+        self.assertFilesEqual(output, expected)
+
 
 class PrepareMetadataTest(helpers.PungiTestCase):
     @mock.patch("pungi.metadata.create_media_repo")

tests/test_fus_wrapper.py

@@ -153,7 +153,10 @@ class TestParseOutput(unittest.TestCase):
         self.assertEqual(modules, set())
 
     def test_extracts_modules(self):
-        touch(self.file, "module:mod:master:20181003:cafebeef.x86_64@repo-0\n")
+        touch(
+            self.file,
+            "module:mod:master-1:20181003:cafebeef.x86_64@repo-0\n"
+        )
         packages, modules = fus.parse_output(self.file)
         self.assertEqual(packages, set())
-        self.assertEqual(modules, set(["mod:master:20181003:cafebeef"]))
+        self.assertEqual(modules, set(["mod:master_1:20181003:cafebeef"]))

tests/test_gather_method_deps.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-import mock
+from unittest import mock
 
 from pungi.phases.gather.methods import method_deps as deps
 from tests import helpers

tests/test_gather_method_hybrid.py

@@ -2,7 +2,7 @@
 
 from collections import namedtuple
 import copy
-import mock
+from unittest import mock
 import os
 
 import six

tests/test_gather_method_nodeps.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-import mock
+from unittest import mock
 import os
 
 import six

tests/test_gather_modules.py

@@ -0,0 +1,124 @@
+# -*- coding: utf-8 -*-
+import gzip
+import os
+from io import StringIO
+import yaml
+from pungi.scripts.gather_modules import collect_modules, EMPTY_FILE
+import unittest
+from pyfakefs.fake_filesystem_unittest import TestCase
+
+MARIADB_MODULE = yaml.load("""
+---
+document: modulemd
+version: 2
+data:
+  name: mariadb-devel
+  stream: 10.3-1
+  version: 8010020200108182321
+  context: cdc1202b
+  arch: x86_64
+  summary: MariaDB Module
+  description: >-
+    MariaDB is a community developed branch of MySQL. 
+  components:
+    rpms:
+      Judy:
+        rationale: MariaDB dependency for OQgraph computation engine
+        ref: a3583b33f939e74a530f2a1dff0552dff2c8ea73
+        buildorder: 4
+        arches: [aarch64, i686, ppc64le, x86_64]
+  artifacts:
+    rpms:
+    - Judy-0:1.0.5-18.module_el8.1.0+217+4d875839.i686
+    - Judy-debuginfo-0:1.0.5-18.module_el8.1.0+217+4d875839.i686
+""", Loader=yaml.BaseLoader)
+
+JAVAPACKAGES_TOOLS_MODULE = yaml.load("""
+---
+document: modulemd
+version: 2
+data:
+  name: javapackages-tools
+  stream: 201801
+  version: 8000020190628172923
+  context: b07bea58
+  arch: x86_64
+  summary: Tools and macros for Java packaging support
+  description: >-
+    Java Packages Tools is a collection of tools that make it easier to build RPM
+    packages containing software running on Java platform.
+  components:
+    rpms:
+      ant:
+        rationale: "Runtime dependency of ant-contrib"
+        ref: 2eaf095676540e2805ee7e8c7f6f78285c428fdc
+        arches: [aarch64, i686, ppc64le, x86_64]
+  artifacts:
+    rpms:
+    - ant-0:1.10.5-1.module_el8.0.0+30+832da3a1.noarch
+    - ant-0:1.10.5-1.module_el8.0.0+30+832da3a1.src
+""", Loader=yaml.BaseLoader)
+
+ANT_DEFAULTS = yaml.load("""
+data:
+  module: ant
+  profiles:
+    '1.10':
+    - common
+  stream: '1.10'
+document: modulemd-defaults
+version: '1'
+""", Loader=yaml.BaseLoader)
+
+
+PATH_TO_KOJI = '/path/to/koji'
+MODULES_YAML_GZ = 'modules.yaml.gz'
+
+
+class TestModulesYamlParser(TestCase):
+
+    maxDiff = None
+
+    def setUp(self):
+        self.setUpPyfakefs()
+
+    def _prepare_test_data(self):
+        """
+        Create modules.yaml.gz with some test data
+        """
+        os.makedirs(PATH_TO_KOJI)
+        modules_gz_path = os.path.join(PATH_TO_KOJI, MODULES_YAML_GZ)
+        # dump modules into compressed file as in generic repos for rpm
+        io = StringIO()
+        yaml.dump_all([MARIADB_MODULE, JAVAPACKAGES_TOOLS_MODULE, ANT_DEFAULTS], io)
+        with open(os.path.join(PATH_TO_KOJI, MODULES_YAML_GZ), 'wb') as f:
+            f.write(gzip.compress(io.getvalue().encode()))
+        return modules_gz_path
+
+    def test_export_modules(self):
+        modules_gz_path = self._prepare_test_data()
+
+        paths = [open(modules_gz_path, 'rb')]
+        collect_modules(paths, PATH_TO_KOJI)
+
+        # check directory structure matches expected
+        self.assertEqual([MODULES_YAML_GZ, 'modules', 'module_defaults'], os.listdir(PATH_TO_KOJI))
+        self.assertEqual(['mariadb-devel-10.3_1-8010020200108182321.cdc1202b',
+                          'javapackages-tools-201801-8000020190628172923.b07bea58'],
+                         os.listdir(os.path.join(PATH_TO_KOJI, 'modules/x86_64')))
+        self.assertEqual([EMPTY_FILE, 'ant.yaml'],
+                         os.listdir(os.path.join(PATH_TO_KOJI, 'module_defaults')))
+
+        # check that modules were exported
+        self.assertEqual(MARIADB_MODULE, yaml.safe_load(
+            open(os.path.join(PATH_TO_KOJI, 'modules/x86_64', 'mariadb-devel-10.3_1-8010020200108182321.cdc1202b'))))
+        self.assertEqual(JAVAPACKAGES_TOOLS_MODULE, yaml.safe_load(
+            open(os.path.join(PATH_TO_KOJI, 'modules/x86_64', 'javapackages-tools-201801-8000020190628172923.b07bea58'))))
+
+        # check that defaults were copied
+        self.assertEqual(ANT_DEFAULTS, yaml.safe_load(
+            open(os.path.join(PATH_TO_KOJI, 'module_defaults', 'ant.yaml'))))
+
+
+if __name__ == '__main__':
+    unittest.main()

tests/test_gather_phase.py

@@ -4,7 +4,7 @@ import copy
 import json
 import os
 
-import mock
+from unittest import mock
 
 try:
     import unittest2 as unittest

tests/test_gather_rpms.py

@@ -0,0 +1,151 @@
+# -*- coding: utf-8 -*-
+import os
+import unittest
+from pathlib import Path
+
+from pyfakefs.fake_filesystem_unittest import TestCase
+
+from pungi.scripts.gather_rpms import search_rpms, copy_rpms, Package
+from productmd.common import parse_nvra
+
+PATH_TO_REPOS = '/path/to/repos'
+MODULES_YAML_GZ = 'modules.yaml.gz'
+
+
+class TestGatherRpms(TestCase):
+    maxDiff = None
+
+    FILES_TO_CREATE = [
+        'powertools/Packages/libvirt-6.0.0-28.module_el'
+        '8.3.0+555+a55c8938.i686.rpm',
+        'powertools/Packages/libgit2-devel-0.26.8-2.el8.x86_64.rpm',
+        'powertools/Packages/xalan-j2-2.7.1-38.module_el'
+        '8.0.0+30+832da3a1.noarch.rpm',
+        'appstream/Packages/bnd-maven-plugin-3.5.0-4.module_el'
+        '8.0.0+30+832da3a1.noarch.rpm',
+        'appstream/Packages/OpenEXR-devel-2.2.0-11.el8.i686.rpm',
+        'appstream/Packages/mingw-binutils-generic-2.30-1.el8.x86_64.rpm',
+        'appstream/Packages/somenonrpm',
+    ]
+
+    def setUp(self):
+        self.setUpPyfakefs()
+
+        os.makedirs(PATH_TO_REPOS)
+
+        for filepath in self.FILES_TO_CREATE:
+            os.makedirs(
+                os.path.join(PATH_TO_REPOS, os.path.dirname(filepath)),
+                exist_ok=True,
+            )
+            open(os.path.join(PATH_TO_REPOS, filepath), 'w').close()
+
+    def test_gather_rpms(self):
+        self.assertEqual(
+            [Package(nvra=parse_nvra('libvirt-6.0.0-28.module_'
+                                     'el8.3.0+555+a55c8938.i686'),
+                     path=Path(
+                         f'{PATH_TO_REPOS}/powertools/Packages/'
+                         f'libvirt-6.0.0-28.module_el'
+                         f'8.3.0+555+a55c8938.i686.rpm'
+                     )),
+             Package(nvra=parse_nvra('libgit2-devel-0.26.8-2.el8.x86_64'),
+                     path=Path(
+                         f'{PATH_TO_REPOS}/powertools/Packages/'
+                         f'libgit2-devel-0.26.8-2.el8.x86_64.rpm'
+                     )),
+             Package(nvra=parse_nvra('xalan-j2-2.7.1-38.module_el'
+                          '8.0.0+30+832da3a1.noarch'),
+                     path=Path(
+                         f'{PATH_TO_REPOS}/powertools/Packages/'
+                         f'xalan-j2-2.7.1-38.module_el'
+                         f'8.0.0+30+832da3a1.noarch.rpm'
+                     )),
+             Package(nvra=parse_nvra('bnd-maven-plugin-3.5.0-4.module_el'
+                          '8.0.0+30+832da3a1.noarch'),
+                     path=Path(
+                         '/path/to/repos/appstream/Packages/'
+                         'bnd-maven-plugin-3.5.0-4.module_el'
+                         '8.0.0+30+832da3a1.noarch.rpm'
+                     )),
+             Package(nvra=parse_nvra('OpenEXR-devel-2.2.0-11.el8.i686'),
+                     path=Path(
+                         f'{PATH_TO_REPOS}/appstream/Packages/'
+                         f'OpenEXR-devel-2.2.0-11.el8.i686.rpm'
+                     )),
+             Package(nvra=parse_nvra('mingw-binutils-generic-'
+                                     '2.30-1.el8.x86_64'),
+                     path=Path(
+                         f'{PATH_TO_REPOS}/appstream/Packages/'
+                         f'mingw-binutils-generic-2.30-1.el8.x86_64.rpm'
+                     ))
+             ],
+            search_rpms(Path(PATH_TO_REPOS))
+        )
+
+    def test_copy_rpms(self):
+        target_path = Path('/mnt/koji')
+        packages = [
+
+            Package(nvra=parse_nvra('libvirt-6.0.0-28.module_'
+                                    'el8.3.0+555+a55c8938.i686'),
+                    path=Path(
+                        f'{PATH_TO_REPOS}/powertools/Packages/'
+                        f'libvirt-6.0.0-28.module_el'
+                        f'8.3.0+555+a55c8938.i686.rpm'
+                    )),
+            Package(nvra=parse_nvra('libgit2-devel-0.26.8-2.el8.x86_64'),
+                    path=Path(
+                        f'{PATH_TO_REPOS}/powertools/Packages/'
+                        f'libgit2-devel-0.26.8-2.el8.x86_64.rpm'
+                    )),
+            Package(nvra=parse_nvra('xalan-j2-2.7.1-38.module_'
+                                    'el8.0.0+30+832da3a1.noarch'),
+                    path=Path(
+                        f'{PATH_TO_REPOS}/powertools/Packages/'
+                        f'xalan-j2-2.7.1-38.module_el'
+                        f'8.0.0+30+832da3a1.noarch.rpm'
+                    )),
+            Package(nvra=parse_nvra('bnd-maven-plugin-3.5.0-4.module_el'
+                         '8.0.0+30+832da3a1.noarch'),
+                    path=Path(
+                        '/path/to/repos/appstream/Packages/'
+                         'bnd-maven-plugin-3.5.0-4.module_el'
+                        '8.0.0+30+832da3a1.noarch.rpm'
+                    )),
+            Package(nvra=parse_nvra('OpenEXR-devel-2.2.0-11.el8.i686'),
+                    path=Path(
+                        f'{PATH_TO_REPOS}/appstream/Packages/'
+                        f'OpenEXR-devel-2.2.0-11.el8.i686.rpm'
+                    )),
+            Package(nvra=parse_nvra('mingw-binutils-generic-'
+                                    '2.30-1.el8.x86_64'),
+                    path=Path(
+                        f'{PATH_TO_REPOS}/appstream/Packages/'
+                        f'mingw-binutils-generic-2.30-1.el8.x86_64.rpm'
+                    ))
+        ]
+        copy_rpms(packages, target_path, [])
+
+        self.assertCountEqual([
+            'xalan-j2-2.7.1-38.module_el8.0.0+30+832da3a1.noarch.rpm',
+            'bnd-maven-plugin-3.5.0-4.module_el8.0.0+30+832da3a1.noarch.rpm'
+        ], os.listdir(target_path / 'noarch'))
+
+        self.assertCountEqual([
+            'libgit2-devel-0.26.8-2.el8.x86_64.rpm',
+            'mingw-binutils-generic-2.30-1.el8.x86_64.rpm'
+        ], os.listdir(target_path / 'x86_64'))
+
+        self.assertCountEqual([
+            'libvirt-6.0.0-28.module_el8.3.0+555+a55c8938.i686.rpm',
+            'OpenEXR-devel-2.2.0-11.el8.i686.rpm'
+        ], os.listdir(target_path / 'i686'))
+
+        self.assertCountEqual([
+            'i686', 'x86_64', 'noarch'
+        ], os.listdir(target_path))
+
+
+if __name__ == '__main__':
+    unittest.main()

tests/test_gather_source_module.py

@@ -5,7 +5,7 @@ try:
 except ImportError:
     import unittest
 
-import mock
+from unittest import mock
 import six
 
 from pungi.phases.gather.sources.source_module import GatherSourceModule