. Change the handlers SPL uses to php_pcre_*_impl(pcre_cache_entry*,....)
. All refactored funcs (match, split, replace, grep) use the above
. Change (zend|php)_error() to php_error_docref()
. Move from old to new param parsing api
. Fix memleaks in unicode mode
Draft 3 of IEEE 1003.1 200x, "2.2 The Compilation Environment"
All identifiers that begin with an underscore and either an uppercase
letter or another underscore are always reserved for any use by the
implementation.
Fix warnings compiling in maintainer mode.
Install DB/common.php and DB/mysql.php on "make install".
# We still have a warning from http_config.h:97. Rasmus? :-)
- Made it possible to specify external location of the PCRE library
- Reworked PCRE extension to use updated PCRE library API
Hopefully now everything behaves just like Perl..
-Updated preg_replace() code to handle stuff properly
-Removed ability to specify external PCRE library source to link against
(since our bundled version is patched now)
First of all, as per extensive discussion on the list, the functions are now
prefixed with "preg" instead of "pcre".
Secondly, global matching is now possible using preg_match_all. Please, give
suggestions on a better name if this one doesn't sit well with you. Possible
names are preg_global_match and preg_gmatch.
preg_match_all takes 4 arguments: a regex pattern, a subject string,
the array for capturing subpatterns, and a parameter that tells how the results
in the subpatterns array are arranged.
Basically, preg_match_all will go through the subject string and try to capture
all the matches that it finds, not just the first one like preg_match.
4th parameter can be PREG_PATTERN_ORDER (default) or PREG_SET_ORDER.
Example:
preg_match_all("|</?([^>]+)>|", "<div align=left>a test</div>", $out, PREG_PATTERN_ORDER);
This returns results so that $out[0] is an array of full pattern matches, $out[1] is an array
of first captured subpattern matches, and so on.
$out[0] -> ("<div align=left>", "</div>")
$out[1] -> ("div align=left", "div")
Example:
preg_match_all("|</?([^>]+)>|", "<div align=left>a test</div>", $out, PREG_SET_ORDER);
This returns results so that $out[0] is an array of first full pattern match and subpatterns,
$out[1] is an array of second full pattern match and subpatterns.
$out[0] -> ("<div align=left>", "div align=left")
$out[1] -> ("</div>", "div")
If anyone has a better name for these PREG_ constants and also which one should be the default,
I'd like to hear it.
PCRE library can be found at ftp://ftp.cus.cam.ac.uk/pub/software/programs/pcre/
config.m4 will be updated to be more robust later on.
perl_match() takes a regular expression, the source string, and the array
for subpattern matches.
perl_replace() takes a regular expression, the search string, and the replacement
string.
Regular expression is specified using delimiters and options. Example:
perl_match("/<[a-z]*>/i", $text, $tags);
More stuff is coming soon.
