rpms/grep
7
0
Fork 0
Code Issues Pull Requests Releases Activity

New version

Browse Source 
Resolves: rhbz#1305035
- Dropped disable-performance-related-tests, better-encoding-errors-handling,
  Pc-consistent-results, and test-pcre-count-fix patches (all upstreamed)
- De-fuzzified man-fx-gs, and help-align patches
This commit is contained in:
Jaroslav Škarvada 2016-02-05 16:01:05 +01:00
parent 9cc608b66b
commit 08a2c56622
8 changed files with 18 additions and 893 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

81
grep-2.22-Pc-consistent-results.patch
View File

@ -1,81 +0,0 @@
diff --git a/src/grep.c b/src/grep.c
index eac540a..9fabeb8 100644
--- a/src/grep.c
+++ b/src/grep.c
@@ -1386,7 +1386,8 @@ grep (int fd, struct stat const *st)
has_nulls = true;
if (binary_files == WITHOUT_MATCH_BINARY_FILES)
return 0;
- done_on_match = out_quiet = true;
+ if (!count_matches)
+ done_on_match = out_quiet = true;
nul_zapper = eol;
skip_nuls = skip_empty_lines;
}
diff --git a/tests/Makefile.am b/tests/Makefile.am
index 2ade5be..2079ae5 100644
--- a/tests/Makefile.am
+++ b/tests/Makefile.am
@@ -106,6 +106,7 @@ TESTS = \
pcre \
pcre-abort \
pcre-context \
+ pcre-count \
pcre-infloop \
pcre-invalid-utf8-input \
pcre-jitstack \
diff --git a/tests/Makefile.in b/tests/Makefile.in
index b5bd7b5..ee516da 100644
--- a/tests/Makefile.in
+++ b/tests/Makefile.in
@@ -1442,6 +1442,7 @@ TESTS = \
pcre \
pcre-abort \
pcre-context \
+ pcre-count \
pcre-infloop \
pcre-invalid-utf8-input \
pcre-jitstack \
@@ -2250,6 +2251,13 @@ pcre-context.log: pcre-context
--log-file $$b.log --trs-file $$b.trs \
$(am__common_driver_flags) $(AM_LOG_DRIVER_FLAGS) $(LOG_DRIVER_FLAGS) -- $(LOG_COMPILE) \
"$$tst" $(AM_TESTS_FD_REDIRECT)
+pcre-count.log: pcre-count
+ @p='pcre-count'; \
+ b='pcre-count'; \
+ $(am__check_pre) $(LOG_DRIVER) --test-name "$$f" \
+ --log-file $$b.log --trs-file $$b.trs \
+ $(am__common_driver_flags) $(AM_LOG_DRIVER_FLAGS) $(LOG_DRIVER_FLAGS) -- $(LOG_COMPILE) \
+ "$$tst" $(AM_TESTS_FD_REDIRECT)
pcre-infloop.log: pcre-infloop
@p='pcre-infloop'; \
b='pcre-infloop'; \
diff --git a/tests/pcre-count b/tests/pcre-count
new file mode 100755
index 0000000..78e1c7c
--- /dev/null
+++ b/tests/pcre-count
@@ -0,0 +1,23 @@
+#! /bin/sh
+# grep -P / grep -Pc are inconsistent results
+# This bug affected grep versions 2.21 through 2.22.
+#
+# Copyright (C) 2015 Free Software Foundation, Inc.
+#
+# Copying and distribution of this file, with or without modification,
+# are permitted in any medium without royalty provided the copyright
+# notice and this notice are preserved.
+
+. "${srcdir=.}/init.sh"; path_prepend_ ../src
+require_pcre_
+
+fail=0
+
+printf 'a\n%032768d\nb\x0\n%032768d\na\n' 0 0 > in
+
+LC_ALL=C grep -P 'a' in | wc -l > exp
+
+LC_ALL=C grep -Pc 'a' in > out || fail=1
+compare exp out || fail=1
+
+Exit $fail

703
grep-2.22-better-encoding-errors-handling.patch
View File

@ -1,703 +0,0 @@
diff --git a/doc/grep.texi b/doc/grep.texi
index e3495bb..41f4fa5 100644
--- a/doc/grep.texi
+++ b/doc/grep.texi
@@ -596,13 +596,13 @@ If a file's allocation metadata,
or if its data read before a line is selected for output,
indicate that the file contains binary data,
assume that the file is of type @var{type}.
-Non-text bytes indicate binary data; these are either data bytes
-improperly encoded for the current locale, or null bytes when the
+Non-text bytes indicate binary data; these are either output bytes that are
+improperly encoded for the current locale, or null input bytes when the
@option{-z} (@option{--null-data}) option is not given (@pxref{Other
Options}).
-By default, @var{type} is @samp{binary},
-and @command{grep} normally outputs either
+By default, @var{type} is @samp{binary}, and when @command{grep}
+discovers that a file is binary it normally outputs either
a one-line message saying that a binary file matches,
or no message if there is no match.
When processing binary data, @command{grep} may treat non-text bytes
@@ -611,7 +611,8 @@ not match a null byte, as the null byte might be treated as a line
terminator even without the @option{-z} (@option{--null-data}) option.
If @var{type} is @samp{without-match},
-@command{grep} assumes that a binary file does not match;
+when @command{grep} discovers that a file is binary
+it assumes that the rest of the file does not match;
this is equivalent to the @option{-I} option.
If @var{type} is @samp{text},
diff --git a/src/grep.c b/src/grep.c
index 2748fd3..eac540a 100644
--- a/src/grep.c
+++ b/src/grep.c
@@ -377,7 +377,6 @@ bool match_icase;
bool match_words;
bool match_lines;
char eolbyte;
-enum textbin input_textbin;
static char const *matcher;
@@ -389,6 +388,10 @@ static bool omit_dot_slash;
static bool errseen;
static bool write_error_seen;
+/* True if output from the current input file has been suppressed
+ because an output line had an encoding error. */
+static bool encoding_error_output;
+
enum directories_type
{
READ_DIRECTORIES = 2,
@@ -481,12 +484,6 @@ clean_up_stdout (void)
close_stdout ();
}
-static bool
-textbin_is_binary (enum textbin textbin)
-{
- return textbin < TEXTBIN_UNKNOWN;
-}
-
/* The high-order bit of a byte. */
enum { HIBYTE = 0x80 };
@@ -551,58 +548,60 @@ skip_easy_bytes (char const *buf)
return p;
}
-/* Return the text type of data in BUF, of size SIZE.
+/* Return true if BUF, of size SIZE, has an encoding error.
BUF must be followed by at least sizeof (uword) bytes,
- which may be arbitrarily written to or read from. */
-static enum textbin
-buffer_textbin (char *buf, size_t size)
+ the first of which may be modified. */
+static bool
+buf_has_encoding_errors (char *buf, size_t size)
{
- if (eolbyte && memchr (buf, '\0', size))
- return TEXTBIN_BINARY;
+ if (MB_CUR_MAX <= 1)
+ return false;
- if (1 < MB_CUR_MAX)
- {
- mbstate_t mbs = { 0 };
- size_t clen;
- char const *p;
+ mbstate_t mbs = { 0 };
+ size_t clen;
- buf[size] = -1;
- for (p = buf; (p = skip_easy_bytes (p)) < buf + size; p += clen)
- {
- clen = mbrlen (p, buf + size - p, &mbs);
- if ((size_t) -2 <= clen)
- return clen == (size_t) -2 ? TEXTBIN_UNKNOWN : TEXTBIN_BINARY;
- }
+ buf[size] = -1;
+ for (char const *p = buf; (p = skip_easy_bytes (p)) < buf + size; p += clen)
+ {
+ clen = mbrlen (p, buf + size - p, &mbs);
+ if ((size_t) -2 <= clen)
+ return true;
}
- return TEXTBIN_TEXT;
+ return false;
}
-/* Return the text type of a file. BUF, of size SIZE, is the initial
- buffer read from the file with descriptor FD and status ST.
- BUF must be followed by at least sizeof (uword) bytes,
+
+/* Return true if BUF, of size SIZE, has a null byte.
+ BUF must be followed by at least one byte,
which may be arbitrarily written to or read from. */
-static enum textbin
-file_textbin (char *buf, size_t size, int fd, struct stat const *st)
+static bool
+buf_has_nulls (char *buf, size_t size)
{
- enum textbin textbin = buffer_textbin (buf, size);
- if (textbin_is_binary (textbin))
- return textbin;
+ buf[size] = 0;
+ return strlen (buf) != size;
+}
+/* Return true if a file is known to contain null bytes.
+ SIZE bytes have already been read from the file
+ with descriptor FD and status ST. */
+static bool
+file_must_have_nulls (size_t size, int fd, struct stat const *st)
+{
if (usable_st_size (st))
{
if (st->st_size <= size)
- return textbin == TEXTBIN_UNKNOWN ? TEXTBIN_BINARY : textbin;
+ return false;
/* If the file has holes, it must contain a null byte somewhere. */
- if (SEEK_HOLE != SEEK_SET && eolbyte)
+ if (SEEK_HOLE != SEEK_SET)
{
off_t cur = size;
if (O_BINARY || fd == STDIN_FILENO)
{
cur = lseek (fd, 0, SEEK_CUR);
if (cur < 0)
- return TEXTBIN_UNKNOWN;
+ return false;
}
/* Look for a hole after the current location. */
@@ -612,12 +611,12 @@ file_textbin (char *buf, size_t size, int fd, struct stat const *st)
if (lseek (fd, cur, SEEK_SET) < 0)
suppressible_error (filename, errno);
if (hole_start < st->st_size)
- return TEXTBIN_BINARY;
+ return true;
}
}
}
- return TEXTBIN_UNKNOWN;
+ return false;
}
/* Convert STR to a nonnegative integer, storing the result in *OUT.
@@ -888,7 +887,7 @@ static char *label = NULL; /* Fake filename for stdin */
/* Internal variables to keep track of byte count, context, etc. */
static uintmax_t totalcc; /* Total character count before bufbeg. */
static char const *lastnl; /* Pointer after last newline counted. */
-static char const *lastout; /* Pointer after last character output;
+static char *lastout; /* Pointer after last character output;
NULL if no character has been output
or if it's conceptually before bufbeg. */
static intmax_t outleft; /* Maximum number of lines to be output. */
@@ -960,10 +959,31 @@ print_offset (uintmax_t pos, int min_width, const char *color)
pr_sgr_end_if (color);
}
-/* Print a whole line head (filename, line, byte). */
-static void
-print_line_head (char const *beg, char const *lim, char sep)
+/* Print a whole line head (filename, line, byte). The output data
+ starts at BEG and contains LEN bytes; it is followed by at least
+ sizeof (uword) bytes, the first of which may be temporarily modified.
+ The output data comes from what is perhaps a larger input line that
+ goes until LIM, where LIM[-1] is an end-of-line byte. Use SEP as
+ the separator on output.
+
+ Return true unless the line was suppressed due to an encoding error. */
+
+static bool
+print_line_head (char *beg, size_t len, char const *lim, char sep)
{
+ bool encoding_errors = false;
+ if (binary_files != TEXT_BINARY_FILES)
+ {
+ char ch = beg[len];
+ encoding_errors = buf_has_encoding_errors (beg, len);
+ beg[len] = ch;
+ }
+ if (encoding_errors)
+ {
+ encoding_error_output = done_on_match = out_quiet = true;
+ return false;
+ }
+
bool pending_sep = false;
if (out_file)
@@ -1010,22 +1030,27 @@ print_line_head (char const *beg, char const *lim, char sep)
print_sep (sep);
}
+
+ return true;
}
-static const char *
-print_line_middle (const char *beg, const char *lim,
+static char *
+print_line_middle (char *beg, char *lim,
const char *line_color, const char *match_color)
{
size_t match_size;
size_t match_offset;
- const char *cur = beg;
- const char *mid = NULL;
-
- while (cur < lim
- && ((match_offset = execute (beg, lim - beg, &match_size, cur))
- != (size_t) -1))
+ char *cur = beg;
+ char *mid = NULL;
+ char *b;
+
+ for (cur = beg;
+ (cur < lim
+ && ((match_offset = execute (beg, lim - beg, &match_size, cur))
+ != (size_t) -1));
+ cur = b + match_size)
{
- char const *b = beg + match_offset;
+ b = beg + match_offset;
/* Avoid matching the empty line at the end of the buffer. */
if (b == lim)
@@ -1045,8 +1070,11 @@ print_line_middle (const char *beg, const char *lim,
/* This function is called on a matching line only,
but is it selected or rejected/context? */
if (only_matching)
- print_line_head (b, lim, (out_invert ? SEP_CHAR_REJECTED
- : SEP_CHAR_SELECTED));
+ {
+ char sep = out_invert ? SEP_CHAR_REJECTED : SEP_CHAR_SELECTED;
+ if (! print_line_head (b, match_size, lim, sep))
+ return NULL;
+ }
else
{
pr_sgr_start (line_color);
@@ -1064,7 +1092,6 @@ print_line_middle (const char *beg, const char *lim,
if (only_matching)
fputs ("\n", stdout);
}
- cur = b + match_size;
}
if (only_matching)
@@ -1075,8 +1102,8 @@ print_line_middle (const char *beg, const char *lim,
return cur;
}
-static const char *
-print_line_tail (const char *beg, const char *lim, const char *line_color)
+static char *
+print_line_tail (char *beg, const char *lim, const char *line_color)
{
size_t eol_size;
size_t tail_size;
@@ -1097,14 +1124,15 @@ print_line_tail (const char *beg, const char *lim, const char *line_color)
}
static void
-prline (char const *beg, char const *lim, char sep)
+prline (char *beg, char *lim, char sep)
{
bool matching;
const char *line_color;
const char *match_color;
if (!only_matching)
- print_line_head (beg, lim, sep);
+ if (! print_line_head (beg, lim - beg - 1, lim, sep))
+ return;
matching = (sep == SEP_CHAR_SELECTED) ^ out_invert;
@@ -1124,7 +1152,11 @@ prline (char const *beg, char const *lim, char sep)
{
/* We already know that non-matching lines have no match (to colorize). */
if (matching && (only_matching || *match_color))
- beg = print_line_middle (beg, lim, line_color, match_color);
+ {
+ beg = print_line_middle (beg, lim, line_color, match_color);
+ if (! beg)
+ return;
+ }
if (!only_matching && *line_color)
{
@@ -1158,7 +1190,7 @@ prpending (char const *lim)
lastout = bufbeg;
while (pending > 0 && lastout < lim)
{
- char const *nl = memchr (lastout, eolbyte, lim - lastout);
+ char *nl = memchr (lastout, eolbyte, lim - lastout);
size_t match_size;
--pending;
if (outleft
@@ -1173,7 +1205,7 @@ prpending (char const *lim)
/* Output the lines between BEG and LIM. Deal with context. */
static void
-prtext (char const *beg, char const *lim)
+prtext (char *beg, char *lim)
{
static bool used; /* Avoid printing SEP_STR_GROUP before any output. */
char eol = eolbyte;
@@ -1181,7 +1213,7 @@ prtext (char const *beg, char const *lim)
if (!out_quiet && pending > 0)
prpending (beg);
- char const *p = beg;
+ char *p = beg;
if (!out_quiet)
{
@@ -1207,7 +1239,7 @@ prtext (char const *beg, char const *lim)
while (p < beg)
{
- char const *nl = memchr (p, eol, beg - p);
+ char *nl = memchr (p, eol, beg - p);
nl++;
prline (p, nl, SEP_CHAR_REJECTED);
p = nl;
@@ -1220,7 +1252,7 @@ prtext (char const *beg, char const *lim)
/* One or more lines are output. */
for (n = 0; p < lim && n < outleft; n++)
{
- char const *nl = memchr (p, eol, lim - p);
+ char *nl = memchr (p, eol, lim - p);
nl++;
if (!out_quiet)
prline (p, nl, SEP_CHAR_SELECTED);
@@ -1267,13 +1299,12 @@ zap_nuls (char *p, char *lim, char eol)
between matching lines if OUT_INVERT is true). Return a count of
lines printed. Replace all NUL bytes with NUL_ZAPPER as we go. */
static intmax_t
-grepbuf (char const *beg, char const *lim)
+grepbuf (char *beg, char const *lim)
{
intmax_t outleft0 = outleft;
- char const *p;
- char const *endp;
+ char *endp;
- for (p = beg; p < lim; p = endp)
+ for (char *p = beg; p < lim; p = endp)
{
size_t match_size;
size_t match_offset = execute (p, lim - p, &match_size, NULL);
@@ -1284,15 +1315,15 @@ grepbuf (char const *beg, char const *lim)
match_offset = lim - p;
match_size = 0;
}
- char const *b = p + match_offset;
+ char *b = p + match_offset;
endp = b + match_size;
/* Avoid matching the empty line at the end of the buffer. */
if (!out_invert && b == lim)
break;
if (!out_invert || p < b)
{
- char const *prbeg = out_invert ? p : b;
- char const *prend = out_invert ? b : endp;
+ char *prbeg = out_invert ? p : b;
+ char *prend = out_invert ? b : endp;
prtext (prbeg, prend);
if (!outleft || done_on_match)
{
@@ -1313,7 +1344,6 @@ static intmax_t
grep (int fd, struct stat const *st)
{
intmax_t nlines, i;
- enum textbin textbin;
size_t residue, save;
char oldc;
char *beg;
@@ -1322,6 +1352,7 @@ grep (int fd, struct stat const *st)
char nul_zapper = '\0';
bool done_on_match_0 = done_on_match;
bool out_quiet_0 = out_quiet;
+ bool has_nulls = false;
if (! reset (fd, st))
return 0;
@@ -1333,6 +1364,7 @@ grep (int fd, struct stat const *st)
after_last_match = 0;
pending = 0;
skip_nuls = skip_empty_lines && !eol;
+ encoding_error_output = false;
seek_data_failed = false;
nlines = 0;
@@ -1345,26 +1377,20 @@ grep (int fd, struct stat const *st)
return 0;
}
- if (binary_files == TEXT_BINARY_FILES)
- textbin = TEXTBIN_TEXT;
- else
+ for (bool firsttime = true; ; firsttime = false)
{
- textbin = file_textbin (bufbeg, buflim - bufbeg, fd, st);
- if (textbin_is_binary (textbin))
+ if (!has_nulls && eol && binary_files != TEXT_BINARY_FILES
+ && (buf_has_nulls (bufbeg, buflim - bufbeg)
+ || (firsttime && file_must_have_nulls (buflim - bufbeg, fd, st))))
{
+ has_nulls = true;
if (binary_files == WITHOUT_MATCH_BINARY_FILES)
return 0;
done_on_match = out_quiet = true;
nul_zapper = eol;
skip_nuls = skip_empty_lines;
}
- else if (execute != Pexecute)
- textbin = TEXTBIN_TEXT;
- }
- for (;;)
- {
- input_textbin = textbin;
lastnl = bufbeg;
if (lastout)
lastout = bufbeg;
@@ -1415,13 +1441,8 @@ grep (int fd, struct stat const *st)
}
/* Detect whether leading context is adjacent to previous output. */
- if (lastout)
- {
- if (textbin == TEXTBIN_UNKNOWN)
- textbin = TEXTBIN_TEXT;
- if (beg != lastout)
- lastout = 0;
- }
+ if (beg != lastout)
+ lastout = 0;
/* Handle some details and read more data to scan. */
save = residue + lim - beg;
@@ -1434,22 +1455,6 @@ grep (int fd, struct stat const *st)
suppressible_error (filename, errno);
goto finish_grep;
}
-
- /* If the file's textbin has not been determined yet, assume
- it's binary if the next input buffer suggests so. */
- if (textbin == TEXTBIN_UNKNOWN)
- {
- enum textbin tb = buffer_textbin (bufbeg, buflim - bufbeg);
- if (textbin_is_binary (tb))
- {
- if (binary_files == WITHOUT_MATCH_BINARY_FILES)
- return 0;
- textbin = tb;
- done_on_match = out_quiet = true;
- nul_zapper = eol;
- skip_nuls = skip_empty_lines;
- }
- }
}
if (residue)
{
@@ -1463,7 +1468,7 @@ grep (int fd, struct stat const *st)
finish_grep:
done_on_match = done_on_match_0;
out_quiet = out_quiet_0;
- if (textbin_is_binary (textbin) && !out_quiet && nlines != 0)
+ if ((has_nulls || encoding_error_output) && !out_quiet && nlines != 0)
printf (_("Binary file %s matches\n"), filename);
return nlines;
}
diff --git a/src/grep.h b/src/grep.h
index 580eb11..2e4527c 100644
--- a/src/grep.h
+++ b/src/grep.h
@@ -29,22 +29,4 @@ extern bool match_words; /* -w */
extern bool match_lines; /* -x */
extern char eolbyte; /* -z */
-/* An enum textbin describes the file's type, inferred from data read
- before the first line is selected for output. */
-enum textbin
- {
- /* Binary, as it contains null bytes and the -z option is not in effect,
- or it contains encoding errors. */
- TEXTBIN_BINARY = -1,
-
- /* Not known yet. Only text has been seen so far. */
- TEXTBIN_UNKNOWN = 0,
-
- /* Text. */
- TEXTBIN_TEXT = 1
- };
-
-/* Input file type. */
-extern enum textbin input_textbin;
-
#endif
diff --git a/src/pcresearch.c b/src/pcresearch.c
index b1f8310..a446b2c 100644
--- a/src/pcresearch.c
+++ b/src/pcresearch.c
@@ -194,32 +194,13 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
error. */
char const *subject = buf;
- /* If the input type is unknown, the caller is still testing the
- input, which means the current buffer cannot contain encoding
- errors and a multiline search is typically more efficient.
- Otherwise, a single-line search is typically faster, so that
- pcre_exec doesn't waste time validating the entire input
- buffer. */
- bool multiline = input_textbin == TEXTBIN_UNKNOWN;
-
for (; p < buf + size; p = line_start = line_end + 1)
{
- bool too_big;
-
- if (multiline)
- {
- size_t pcre_size_max = MIN (INT_MAX, SIZE_MAX - 1);
- size_t scan_size = MIN (pcre_size_max + 1, buf + size - p);
- line_end = memrchr (p, eolbyte, scan_size);
- too_big = ! line_end;
- }
- else
- {
- line_end = memchr (p, eolbyte, buf + size - p);
- too_big = INT_MAX < line_end - p;
- }
-
- if (too_big)
+ /* A single-line search is typically faster, so that
+ pcre_exec doesn't waste time validating the entire input
+ buffer. */
+ line_end = memchr (p, eolbyte, buf + size - p);
+ if (INT_MAX < line_end - p)
error (EXIT_TROUBLE, 0, _("exceeded PCRE's line length limit"));
for (;;)
@@ -247,27 +228,11 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
int options = 0;
if (!bol)
options |= PCRE_NOTBOL;
- if (multiline)
- options |= PCRE_NO_UTF8_CHECK;
e = jit_exec (subject, line_end - subject, search_offset,
options, sub);
if (e != PCRE_ERROR_BADUTF8)
- {
- if (0 < e && multiline && sub[1] - sub[0] != 0)
- {
- char const *nl = memchr (subject + sub[0], eolbyte,
- sub[1] - sub[0]);
- if (nl)
- {
- /* This match crosses a line boundary; reject it. */
- p = subject + sub[0];
- line_end = nl;
- continue;
- }
- }
- break;
- }
+ break;
int valid_bytes = sub[0];
/* Try to match the string before the encoding error. */
@@ -337,15 +302,6 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
beg = matchbeg;
end = matchend;
}
- else if (multiline)
- {
- char const *prev_nl = memrchr (line_start - 1, eolbyte,
- matchbeg - (line_start - 1));
- char const *next_nl = memchr (matchend, eolbyte,
- line_end + 1 - matchend);
- beg = prev_nl + 1;
- end = next_nl + 1;
- }
else
{
beg = line_start;
diff --git a/tests/Makefile.am b/tests/Makefile.am
index d379821..2ade5be 100644
--- a/tests/Makefile.am
+++ b/tests/Makefile.am
@@ -70,6 +70,7 @@ TESTS = \
empty \
empty-line \
empty-line-mb \
+ encoding-error \
epipe \
equiv-classes \
ere \
diff --git a/tests/Makefile.in b/tests/Makefile.in
index 6de6f49..b5bd7b5 100644
--- a/tests/Makefile.in
+++ b/tests/Makefile.in
@@ -1406,6 +1406,7 @@ TESTS = \
empty \
empty-line \
empty-line-mb \
+ encoding-error \
epipe \
equiv-classes \
ere \
@@ -1997,6 +1998,13 @@ empty-line-mb.log: empty-line-mb
--log-file $$b.log --trs-file $$b.trs \
$(am__common_driver_flags) $(AM_LOG_DRIVER_FLAGS) $(LOG_DRIVER_FLAGS) -- $(LOG_COMPILE) \
"$$tst" $(AM_TESTS_FD_REDIRECT)
+encoding-error.log: encoding-error
+ @p='encoding-error'; \
+ b='encoding-error'; \
+ $(am__check_pre) $(LOG_DRIVER) --test-name "$$f" \
+ --log-file $$b.log --trs-file $$b.trs \
+ $(am__common_driver_flags) $(AM_LOG_DRIVER_FLAGS) $(LOG_DRIVER_FLAGS) -- $(LOG_COMPILE) \
+ "$$tst" $(AM_TESTS_FD_REDIRECT)
epipe.log: epipe
@p='epipe'; \
b='epipe'; \
diff --git a/tests/encoding-error b/tests/encoding-error
new file mode 100755
index 0000000..fe52de2
--- a/dev/null
+++ b/tests/encoding-error
@@ -0,0 +1,41 @@
+#! /bin/sh
+# Test grep's behavior on encoding errors.
+#
+# Copyright 2015 Free Software Foundation, Inc.
+#
+# Copying and distribution of this file, with or without modification,
+# are permitted in any medium without royalty provided the copyright
+# notice and this notice are preserved.
+
+. "${srcdir=.}/init.sh"; path_prepend_ ../src
+
+require_en_utf8_locale_
+
+LC_ALL=en_US.UTF-8
+export LC_ALL
+
+printf 'Alfred Jones\n' > a || framework_failure_
+printf 'John Smith\n' >j || framework_failure_
+printf 'Pedro P\xe9rez\n' >p || framework_failure_
+cat a p j >in || framework_failure_
+
+fail=0
+
+grep '^A' in >out || fail=1
+compare a out || fail=1
+
+grep '^P' in >out || fail=1
+printf 'Binary file in matches\n' >exp || framework_failure_
+compare exp out || fail=1
+
+grep '^J' in >out || fail=1
+compare j out || fail=1
+
+grep '^X' in >out
+test $? = 1 || fail=1
+compare /dev/null out || fail=1
+
+grep -a . in >out || fail=1
+compare in out
+
+Exit $fail

54
grep-2.22-disable-performance-related-tests.patch
View File

@ -1,54 +0,0 @@
From e938d22e2131972a6e9eaddbf850c6a04a4f750c Mon Sep 17 00:00:00 2001
From: Jim Meyering <meyering@fb.com>
Date: Wed, 4 Nov 2015 10:57:07 -0800
Subject: [PATCH] tests: mark performance-related tests as expensive
These performance-related tests are slightly failure prone due to
varying system load during the two runs.
Marking these tests as "expensive" makes it so they are no longer run
via "make check". You can still run them via make "check-expensive".
This makes them less likely to be run by regular users.
* tests/long-pattern-perf: Use expensive_.
* tests/mb-non-UTF8-performance: Likewise.
Reported by Jaroslav Skarvada in http://debbugs.gnu.org/21826
and by Andreas Schwab in http://debbugs.gnu.org/21812.
---
tests/long-pattern-perf | 5 +++++
tests/mb-non-UTF8-performance | 5 +++++
2 files changed, 10 insertions(+)
diff --git a/tests/long-pattern-perf b/tests/long-pattern-perf
index c222c02..2c9d080 100755
--- a/tests/long-pattern-perf
+++ b/tests/long-pattern-perf
@@ -20,6 +20,11 @@
fail=0
+# This test is susceptible to failure due to differences in
+# system load during the two test runs, so we'll mark it as
+# "expensive", making it less likely to be run by regular users.
+expensive_
+
echo x > in || framework_failure_
# We could use seq -s '' (avoiding the tr filter), but I
# suspect some version of seq does not honor that option.
diff --git a/tests/mb-non-UTF8-performance b/tests/mb-non-UTF8-performance
index 228361d..9bd5d39 100755
--- a/tests/mb-non-UTF8-performance
+++ b/tests/mb-non-UTF8-performance
@@ -22,6 +22,11 @@
fail=0
+# This test is susceptible to failure due to differences in
+# system load during the two test runs, so we'll mark it as
+# "expensive", making it less likely to be run by regular users.
+expensive_
+
# Make this large enough so that even on high-end systems
# it incurs at least 5-10ms of user time.
yes $(printf '%078d' 0) | head -400000 > in || framework_failure_
--
2.6.0

4
grep-2.22-help-align.patch
View File

@ -1,8 +1,8 @@
diff --git a/src/grep.c b/src/grep.c
index 7315bd0..2748fd3 100644
index 7ffd122..6ef8a7c 100644
--- a/src/grep.c
+++ b/src/grep.c
@@ -1813,17 +1813,20 @@ Output control:\n\
@@ -1854,17 +1854,20 @@ Output control:\n\
-D, --devices=ACTION how to handle devices, FIFOs and sockets;\n\
ACTION is 'read' or 'skip'\n\
-r, --recursive like --directories=recurse\n\

12
grep-2.22-man-fix-gs.patch
View File

@ -1,8 +1,8 @@
diff --git a/doc/grep.in.1 b/doc/grep.in.1
index 2b513a8..40ac40d 100644
index ade11b4..6b91700 100644
--- a/doc/grep.in.1
+++ b/doc/grep.in.1
@@ -314,7 +314,7 @@ Print
@@ -320,7 +320,7 @@ Print
.I NUM
lines of trailing context after matching lines.
Places a line containing a group separator
@ -11,7 +11,7 @@ index 2b513a8..40ac40d 100644
between contiguous groups of matches.
With the
.B \-o
@@ -327,7 +327,7 @@ Print
@@ -333,7 +333,7 @@ Print
.I NUM
lines of leading context before matching lines.
Places a line containing a group separator
@ -20,7 +20,7 @@ index 2b513a8..40ac40d 100644
between contiguous groups of matches.
With the
.B \-o
@@ -340,13 +340,24 @@ Print
@@ -346,13 +346,24 @@ Print
.I NUM
lines of output context.
Places a line containing a group separator
@ -47,10 +47,10 @@ index 2b513a8..40ac40d 100644
.TP
.BR \-a ", " \-\^\-text
diff --git a/src/grep.c b/src/grep.c
index 2c5e09a..7315bd0 100644
index 73c3651..7ffd122 100644
--- a/src/grep.c
+++ b/src/grep.c
@@ -1837,6 +1837,8 @@ Context control:\n\
@@ -1878,6 +1878,8 @@ Context control:\n\
"));
printf (_("\
-NUM same as --context=NUM\n\

26
grep-2.22-test-pcre-count-fix.patch
View File

@ -1,26 +0,0 @@
diff --git a/tests/pcre-count b/tests/pcre-count
--- a/tests/pcre-count
+++ b/tests/pcre-count
@@ -13,11 +13,17 @@ require_pcre_
fail=0
-printf 'a\n%032768d\nb\x0\n%032768d\na\n' 0 0 > in
+printf 'a\n%032768d\nb\0\n%032768d\na\n' 0 0 > in || framework_failure_
-LC_ALL=C grep -P 'a' in | wc -l > exp
+# grep will discover that the input is a binary file sooner if the
+# page size is larger, so allow for either possible output.
+printf 'a\nBinary file in matches\n' >exp1a || framework_failure_
+printf 'Binary file in matches\n' >exp1b || framework_failure_
+LC_ALL=C grep -P 'a' in >out || fail=1
+compare exp1a out || compare exp1b out || fail=1
-LC_ALL=C grep -Pc 'a' in > out || fail=1
-compare exp out || fail=1
+printf '2\n' >exp2 || framework_failure_
+LC_ALL=C grep -Pc 'a' in >out || fail=1
+compare exp2 out || fail=1
Exit $fail

29
grep.spec
View File

@ -2,8 +2,8 @@
Summary: Pattern matching utilities
Name: grep
Version: 2.22
Release: 7%{?dist}
Version: 2.23
Release: 1%{?dist}
License: GPLv3+
URL: http://www.gnu.org/software/grep/
Group: Applications/Text
@ -17,17 +17,6 @@ Source4: grepconf.sh
Patch0: grep-2.22-man-fix-gs.patch
# upstream ticket 39445
Patch1: grep-2.22-help-align.patch
# backported from upstream, mb-non-UTF8-performance and long-pattern-perf tests
# marked as expensive and not run by default. The result is decided according
# to measured runtime, which doesn't work reliable on the builders with
# variable load.
Patch2: grep-2.22-disable-performance-related-tests.patch
# backported from upstream
Patch3: grep-2.22-better-encoding-errors-handling.patch
# backported from upstream, upstream bug#22028
Patch4: grep-2.22-Pc-consistent-results.patch
# backported from upstream, upstream bug#22350
Patch5: grep-2.22-test-pcre-count-fix.patch
Requires(post): /sbin/install-info
Requires(preun): /sbin/install-info
@ -47,13 +36,6 @@ GNU grep is needed by many scripts, so it shall be installed on every system.
%setup -q
%patch0 -p1 -b .man-fix-gs
%patch1 -p1 -b .help-align
%patch2 -p1 -b .disable-performance-related-tests
%patch3 -p1 -b .better-encoding-errors-handling
%patch4 -p1 -b .Pc-consistent-results
%patch5 -p1 -b .test-pcre-count-fix
chmod 755 tests/encoding-error
chmod 755 tests/pcre-count
%build
%global BUILD_FLAGS $RPM_OPT_FLAGS
@ -104,6 +86,13 @@ fi
%{_libexecdir}/grepconf.sh
%changelog
* Fri Feb 5 2016 Jaroslav Škarvada <jskarvad@redhat.com> - 2.23-1
- New version
Resolves: rhbz#1305035
- Dropped disable-performance-related-tests, better-encoding-errors-handling,
Pc-consistent-results, and test-pcre-count-fix patches (all upstreamed)
- De-fuzzified man-fx-gs, and help-align patches
* Wed Feb 03 2016 Fedora Release Engineering <releng@fedoraproject.org> - 2.22-7
- Rebuilt for https://fedoraproject.org/wiki/Fedora_24_Mass_Rebuild

2
sources
View File

@ -1 +1 @@
e1015e951a49a82b02e38891026ef5df grep-2.22.tar.xz
f46aa9d0d2577b9212a104348a286787 grep-2.23.tar.xz