Alan Lehotsky's patch for REG_STARTEND.

git-svn-id: svn://vcs.exim.org/pcre/code/trunk@332 2f5784b3-3f2a-0410-8824-cb99058d5e15

4 files changed, 39 insertions, 7 deletions

diff --git a/ChangeLog b/ChangeLog
index a97fe92..2e6e458 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -10,7 +10,7 @@ Version 7.7 05-Mar-08
     
 2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
     pre-7.6 versions, which defined a global no_arg variable instead of putting
-    it in the RE class.
+    it in the RE class. (See also #8 below.)
     
 3.  Remove a line of dead code, identified by coverity and reported by Nuno 
     Lopes. 
@@ -43,6 +43,9 @@ Version 7.7 05-Mar-08
     
 9.  Applied Craig's patch to remove the use of push_back(). 
 
+10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
+    matching function regexec().
+
 
 Version 7.6 28-Jan-08
 ---------------------
diff --git a/doc/pcreposix.3 b/doc/pcreposix.3
index 2f75c9b..e01068e 100644
--- a/doc/pcreposix.3
+++ b/doc/pcreposix.3
@@ -157,8 +157,9 @@ REG_NEWLINE action.
 .rs
 .sp
 The function \fBregexec()\fP is called to match a compiled pattern \fIpreg\fP
-against a given \fIstring\fP, which is terminated by a zero byte, subject to
-the options in \fIeflags\fP. These can be:
+against a given \fIstring\fP, which is by default terminated by a zero byte 
+(but see REG_STARTEND below), subject to the options in \fIeflags\fP. These can
+be:
 .sp
   REG_NOTBOL
 .sp
@@ -169,6 +170,17 @@ function.
 .sp
 The PCRE_NOTEOL option is set when calling the underlying PCRE matching
 function.
+.sp
+  REG_STARTEND
+.sp
+The string is considered to start at \fIstring\fP + \fIpmatch[0].rm_so\fP and
+to have a terminating NUL located at \fIstring\fP + \fIpmatch[0].rm_eo\fP
+(there need not actually be a NUL at that location), regardless of the value of
+\fInmatch\fP. This is a BSD extension, compatible with but not specified by
+IEEE Standard 1003.2 (POSIX.2), and should be used with caution in software
+intended to be portable to other systems. Note that a non-zero \fIrm_so\fP does
+not imply REG_NOTBOL; REG_STARTEND affects only the location of the string, not
+how it is matched.
 .P
 If the pattern was compiled with the REG_NOSUB flag, no data about any matched
 strings is returned. The \fInmatch\fP and \fIpmatch\fP arguments of
@@ -221,6 +233,6 @@ Cambridge CB2 3QH, England.
 .rs
 .sp
 .nf
-Last updated: 06 March 2007
-Copyright (c) 1997-2007 University of Cambridge.
+Last updated: 05 April 2008
+Copyright (c) 1997-2008 University of Cambridge.
 .fi
diff --git a/pcreposix.c b/pcreposix.c
index 6c7198b..d129c02 100644
--- a/pcreposix.c
+++ b/pcreposix.c
@@ -263,7 +263,7 @@ PCREPOSIX_EXP_DEFN int
 regexec(const regex_t *preg, const char *string, size_t nmatch,
   regmatch_t pmatch[], int eflags)
 {
-int rc;
+int rc, so, eo;
 int options = 0;
 int *ovector = NULL;
 int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
@@ -296,7 +296,23 @@ else if (nmatch > 0)
     }
   }
 
-rc = pcre_exec((const pcre *)preg->re_pcre, NULL, string, (int)strlen(string),
+/* REG_STARTEND is a BSD extension, to allow for non-NUL-terminated strings.
+The man page from OS X says "REG_STARTEND affects only the location of the
+string, not how it is matched". That is why the "so" value is used to bump the 
+start location rather than being passed as a PCRE "starting offset". */
+
+if ((eflags & REG_STARTEND) != 0) 
+  {
+  so = pmatch[0].rm_so;
+  eo = pmatch[0].rm_eo;
+  } 
+else 
+  {
+  so = 0;
+  eo = strlen(string);
+  }
+
+rc = pcre_exec((const pcre *)preg->re_pcre, NULL, string + so, (eo - so),
   0, options, ovector, nmatch * 3);
 
 if (rc == 0) rc = nmatch;    /* All captured slots were filled in */
diff --git a/pcreposix.h b/pcreposix.h
index 109376d..a73ad26 100644
--- a/pcreposix.h
+++ b/pcreposix.h
@@ -59,6 +59,7 @@ extern "C" {
 #define REG_DOTALL    0x0010   /* NOT defined by POSIX. */
 #define REG_NOSUB     0x0020
 #define REG_UTF8      0x0040   /* NOT defined by POSIX. */
+#define REG_STARTEND  0x0080   /* BSD feature: pass subject string by so,eo */
 
 /* This is not used by PCRE, but by defining it we make it easier
 to slot PCRE into existing programs that make POSIX calls. */