From 0efedaf8864d1caa8ed0e7f8fb0b50d5231cacfa Mon Sep 17 00:00:00 2001 From: ph10 Date: Fri, 22 Jun 2018 16:29:56 +0000 Subject: [PATCH] Fix bug when \K is used in a lookbehind in a substitute pattern. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@948 6239d852-aaf2-0410-a92c-79f79f948069 Petr Písař : Ported to 10.31. Signed-off-by: Petr Písař --- doc/html/pcre2api.html | 14 ++++++++++++-- doc/pcre2.txt | 14 ++++++++++++-- doc/pcre2api.3 | 3 ++- src/pcre2_error.c | 2 +- src/pcre2_substitute.c | 6 +++--- testdata/testinput2 | 3 +++ testdata/testoutput2 | 6 +++++- 7 files changed, 38 insertions(+), 10 deletions(-) diff --git a/doc/html/pcre2api.html b/doc/html/pcre2api.html index ba3b2ca..af904e6 100644 --- a/doc/html/pcre2api.html +++ b/doc/html/pcre2api.html @@ -2549,7 +2549,7 @@ calls to pcre2_match() if you are making repeated calls to find other matches in the same subject string.

-WARNING: When PCRE2_NO_UTF_CHECK is set, the effect of passing an invalid +Warning: When PCRE2_NO_UTF_CHECK is set, the effect of passing an invalid string as a subject, or an invalid value of startoffset, is undefined. Your program may crash or loop indefinitely. 

@@ -2756,6 +2756,15 @@ branch of the group, but it is not on the matching path. On the other hand,
 when this pattern fails to match "bx", the returned name is B.
 

 

+Warning: By default, certain start-of-match optimizations are used to
+give a fast "no match" result in some situations. For example, if the anchoring
+is removed from the pattern above, there is an initial check for the presence
+of "c" in the subject before running the matching engine. This check fails for
+"bx", causing a match failure without seeing any marks. You can disable the
+start-of-match optimizations by setting the PCRE2_NO_START_OPTIMIZE option for
+pcre2_compile() or starting the pattern with (*NO_START_OPT).
+

+

 After a successful match, a partial match, or one of the invalid UTF errors
 (for example, PCRE2_ERROR_UTF8_ERR5), pcre2_get_startchar() can be
 called. After a successful or partial match it returns the code unit offset of
@@ -3310,7 +3319,8 @@ replacement string, with more particular errors being PCRE2_ERROR_BADREPESCAPE
 (invalid escape sequence), PCRE2_ERROR_REPMISSINGBRACE (closing curly bracket
 not found), PCRE2_ERROR_BADSUBSTITUTION (syntax error in extended group
 substitution), and PCRE2_ERROR_BADSUBSPATTERN (the pattern match ended before
-it started, which can happen if \K is used in an assertion).
+it started or the match started earlier than the current position in the 
+subject, which can happen if \K is used in an assertion).
 

 

 As for all PCRE2 errors, a text message that describes the error can be
diff --git a/doc/pcre2.txt b/doc/pcre2.txt
index 79d94e3..e5b941f 100644
--- a/doc/pcre2.txt
+++ b/doc/pcre2.txt
@@ -2498,7 +2498,7 @@ MATCHING A PATTERN: THE TRADITIONAL FUNCTION
        second and subsequent calls to pcre2_match() if you are making repeated
        calls to find other matches in the same subject string.
 
-       WARNING: When PCRE2_NO_UTF_CHECK is  set,  the  effect  of  passing  an
+       Warning: When PCRE2_NO_UTF_CHECK is  set,  the  effect  of  passing  an
        invalid  string  as  a  subject, or an invalid value of startoffset, is
        undefined.  Your program may crash or loop indefinitely.
 
@@ -2683,6 +2683,15 @@ OTHER INFORMATION ABOUT A MATCH
        the other hand, when this pattern fails to  match  "bx",  the  returned
        name is B.
 
+       Warning: By default, certain start-of-match optimizations are  used  to
+       give  a  fast "no match" result in some situations. For example, if the
+       anchoring is removed from the pattern above, there is an initial  check
+       for  the  presence  of  "c"  in the subject before running the matching
+       engine. This check fails for "bx", causing a match failure without see-
+       ing any marks. You can disable the start-of-match optimizations by set-
+       ting the PCRE2_NO_START_OPTIMIZE option for pcre2_compile() or starting
+       the pattern with (*NO_START_OPT).
+
        After  a  successful  match, a partial match, or one of the invalid UTF
        errors (for example, PCRE2_ERROR_UTF8_ERR5), pcre2_get_startchar()  can
        be called. After a successful or partial match it returns the code unit
@@ -3209,7 +3218,8 @@ CREATING A NEW STRING WITH SUBSTITUTIONS
        PCRE2_ERROR_BADREPESCAPE  (invalid  escape  sequence), PCRE2_ERROR_REP-
        MISSINGBRACE (closing curly bracket not found),  PCRE2_ERROR_BADSUBSTI-
        TUTION   (syntax   error   in   extended   group   substitution),   and
-       PCRE2_ERROR_BADSUBSPATTERN (the pattern match ended before it  started,
+       PCRE2_ERROR_BADSUBSPATTERN (the pattern match ended before  it  started
+       or  the match started earlier than the current position in the subject,
        which can happen if \K is used in an assertion).
 
        As for all PCRE2 errors, a text message that describes the error can be
diff --git a/doc/pcre2api.3 b/doc/pcre2api.3
index 786b314..ac6e246 100644
--- a/doc/pcre2api.3
+++ b/doc/pcre2api.3
@@ -3302,7 +3302,8 @@ replacement string, with more particular errors being PCRE2_ERROR_BADREPESCAPE
 (invalid escape sequence), PCRE2_ERROR_REPMISSINGBRACE (closing curly bracket
 not found), PCRE2_ERROR_BADSUBSTITUTION (syntax error in extended group
 substitution), and PCRE2_ERROR_BADSUBSPATTERN (the pattern match ended before
-it started, which can happen if \eK is used in an assertion).
+it started or the match started earlier than the current position in the 
+subject, which can happen if \eK is used in an assertion).
 .P
 As for all PCRE2 errors, a text message that describes the error can be
 obtained by calling the \fBpcre2_get_error_message()\fP function (see
diff --git a/src/pcre2_error.c b/src/pcre2_error.c
index d98cae9..a1f98d4 100644
--- a/src/pcre2_error.c
+++ b/src/pcre2_error.c
@@ -255,7 +255,7 @@ static const unsigned char match_error_texts[] =
   "expected closing curly bracket in replacement string\0"
   "bad substitution in replacement string\0"
   /* 60 */
-  "match with end before start is not supported\0"
+  "match with end before start or start moved backwards is not supported\0"
   "too many replacements (more than INT_MAX)\0"
   "bad serialized data\0"
   "heap limit exceeded\0"
diff --git a/src/pcre2_substitute.c b/src/pcre2_substitute.c
index 8da951f..955370a 100644
--- a/src/pcre2_substitute.c
+++ b/src/pcre2_substitute.c
@@ -361,9 +361,9 @@ do
     }
 
   /* Handle a successful match. Matches that use \K to end before they start
-  are not supported. */
-
-  if (ovector[1] < ovector[0])
+  or start before the current point in the subject are not supported. */
+  
+  if (ovector[1] < ovector[0] || ovector[0] < start_offset)
     {
     rc = PCRE2_ERROR_BADSUBSPATTERN;
     goto EXIT;
diff --git a/testdata/testinput2 b/testdata/testinput2
index 5d3a80e..3499042 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -4643,6 +4643,9 @@ B)x/alt_verbnames,mark
 
 /(?=a\K)/replace=z
     BaCaD
+    
+/(?<=\K.)/g,replace=-
+    ab
 
 /(?'abcdefghijklmnopqrstuvwxyzABCDEFG'toolong)/
 
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index fcaac8f..f9e128d 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -14899,7 +14899,11 @@ Subject length lower bound = 1
 
 /(?=a\K)/replace=z
     BaCaD
-Failed: error -60: match with end before start is not supported
+Failed: error -60: match with end before start or start moved backwards is not supported
+    
+/(?<=\K.)/g,replace=-
+    ab
+Failed: error -60: match with end before start or start moved backwards is not supported
 
 /(?'abcdefghijklmnopqrstuvwxyzABCDEFG'toolong)/
 Failed: error 148 at offset 36: subpattern name is too long (maximum 32 characters)
-- 
2.14.4