[MediaWiki-commits] [Gerrit] mediawiki/core[master]: Protect -{...}- variant constructs in definition lists.
jenkins-bot has submitted this change and it was merged. (
https://gerrit.wikimedia.org/r/327112 )
Change subject: Protect -{...}- variant constructs in definition lists.
..
Protect -{...}- variant constructs in definition lists.
Given the wikitext:
;-{zh-cn:AAA;zh-tw:BBB}-
Prevent `doBlockLevels` from trying to split the definition list at the
embedded colon and using `AAA;zh-tw:BBB}-` as the `` portion.
Bug: T153135
Change-Id: I3a4d02f1fbd0d0fe8278d6b7c66005f0dd3dd36b
---
M includes/parser/BlockLevelPass.php
M tests/parser/parserTests.txt
2 files changed, 89 insertions(+), 34 deletions(-)
Approvals:
Tim Starling: Looks good to me, approved
C. Scott Ananian: Looks good to me, approved
jenkins-bot: Verified
diff --git a/includes/parser/BlockLevelPass.php
b/includes/parser/BlockLevelPass.php
index cbacd34..e16cfd4 100644
--- a/includes/parser/BlockLevelPass.php
+++ b/includes/parser/BlockLevelPass.php
@@ -38,6 +38,7 @@
const COLON_STATE_COMMENT = 5;
const COLON_STATE_COMMENTDASH = 6;
const COLON_STATE_COMMENTDASHDASH = 7;
+ const COLON_STATE_LC = 8;
/**
* Make lists from lines starting with ':', '*', '#', etc.
@@ -389,15 +390,14 @@
* @return string The position of the ':', or false if none found
*/
private function findColonNoLinks( $str, &$before, &$after ) {
- $colonPos = strpos( $str, ':' );
- if ( $colonPos === false ) {
+ if ( !preg_match( '/:|<|-\{/', $str, $m, PREG_OFFSET_CAPTURE )
) {
# Nothing to find!
return false;
}
- $ltPos = strpos( $str, '<' );
- if ( $ltPos === false || $ltPos > $colonPos ) {
+ if ( $m[0][0] === ':' ) {
# Easy; no tag nesting to worry about
+ $colonPos = $m[0][1];
$before = substr( $str, 0, $colonPos );
$after = substr( $str, $colonPos + 1 );
return $colonPos;
@@ -405,9 +405,10 @@
# Ugly state machine to walk through avoiding tags.
$state = self::COLON_STATE_TEXT;
- $level = 0;
+ $ltLevel = 0;
+ $lcLevel = 0;
$len = strlen( $str );
- for ( $i = 0; $i < $len; $i++ ) {
+ for ( $i = $m[0][1]; $i < $len; $i++ ) {
$c = $str[$i];
switch ( $state ) {
@@ -418,7 +419,7 @@
$state = self::COLON_STATE_TAGSTART;
break;
case ":":
- if ( $level === 0 ) {
+ if ( $ltLevel === 0 ) {
# We found it!
$before = substr( $str, 0, $i );
$after = substr( $str, $i + 1 );
@@ -428,35 +429,44 @@
break;
default:
# Skip ahead looking for something
interesting
- $colonPos = strpos( $str, ':', $i );
- if ( $colonPos === false ) {
+ if ( !preg_match( '/:|<|-\{/', $str,
$m, PREG_OFFSET_CAPTURE, $i ) ) {
# Nothing else interesting
return false;
}
- $ltPos = strpos( $str, '<', $i );
- if ( $level === 0 ) {
- if ( $ltPos === false ||
$colonPos < $ltPos ) {
- # We found it!
- $before = substr( $str,
0, $colonPos );
- $after = substr( $str,
$colonPos + 1 );
- return $i;
- }
+ if ( $m[0][0] === '-{' ) {
+ $state = self::COLON_STATE_LC;
+ $lcLevel++;
+ $i = $m[0][1] + 1;
+ } else {
+ # Skip ahead to next
interesting character.
+ $i = $m[0][1] - 1;
}
- if ( $ltPos === false ) {
-
[MediaWiki-commits] [Gerrit] mediawiki/core[master]: Protect -{...}- variant constructs in definition lists.
C. Scott Ananian has uploaded a new change for review. (
https://gerrit.wikimedia.org/r/327112 )
Change subject: Protect -{...}- variant constructs in definition lists.
..
Protect -{...}- variant constructs in definition lists.
Given the wikitext:
;-{zh-cn:AAA;zh-tw:BBB}-
Prevent `doBlockLevels` from trying to split the definition list at the
embedded colon and using `AAA;zh-tw:BBB}-` as the `` portion.
Bug: T153135
Change-Id: I3a4d02f1fbd0d0fe8278d6b7c66005f0dd3dd36b
---
M includes/parser/BlockLevelPass.php
M tests/parser/parserTests.txt
2 files changed, 40 insertions(+), 31 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/mediawiki/core
refs/changes/12/327112/1
diff --git a/includes/parser/BlockLevelPass.php
b/includes/parser/BlockLevelPass.php
index cbacd34..1bb3c49 100644
--- a/includes/parser/BlockLevelPass.php
+++ b/includes/parser/BlockLevelPass.php
@@ -38,6 +38,7 @@
const COLON_STATE_COMMENT = 5;
const COLON_STATE_COMMENTDASH = 6;
const COLON_STATE_COMMENTDASHDASH = 7;
+ const COLON_STATE_LC = 8;
/**
* Make lists from lines starting with ':', '*', '#', etc.
@@ -389,15 +390,14 @@
* @return string The position of the ':', or false if none found
*/
private function findColonNoLinks( $str, &$before, &$after ) {
- $colonPos = strpos( $str, ':' );
- if ( $colonPos === false ) {
+ if ( !preg_match( '/:|<|-\{/', $str, $m, PREG_OFFSET_CAPTURE )
) {
# Nothing to find!
return false;
}
- $ltPos = strpos( $str, '<' );
- if ( $ltPos === false || $ltPos > $colonPos ) {
+ if ( $m[0][0] === ':' ) {
# Easy; no tag nesting to worry about
+ $colonPos = $m[0][1];
$before = substr( $str, 0, $colonPos );
$after = substr( $str, $colonPos + 1 );
return $colonPos;
@@ -405,9 +405,10 @@
# Ugly state machine to walk through avoiding tags.
$state = self::COLON_STATE_TEXT;
- $level = 0;
+ $ltLevel = 0;
+ $lcLevel = 0;
$len = strlen( $str );
- for ( $i = 0; $i < $len; $i++ ) {
+ for ( $i = $m[0][1]; $i < $len; $i++ ) {
$c = $str[$i];
switch ( $state ) {
@@ -418,7 +419,7 @@
$state = self::COLON_STATE_TAGSTART;
break;
case ":":
- if ( $level === 0 ) {
+ if ( $ltLevel === 0 ) {
# We found it!
$before = substr( $str, 0, $i );
$after = substr( $str, $i + 1 );
@@ -428,35 +429,44 @@
break;
default:
# Skip ahead looking for something
interesting
- $colonPos = strpos( $str, ':', $i );
- if ( $colonPos === false ) {
+ if ( !preg_match( '/:|<|-\{/', $str,
$m, PREG_OFFSET_CAPTURE, $i ) ) {
# Nothing else interesting
return false;
}
- $ltPos = strpos( $str, '<', $i );
- if ( $level === 0 ) {
- if ( $ltPos === false ||
$colonPos < $ltPos ) {
- # We found it!
- $before = substr( $str,
0, $colonPos );
- $after = substr( $str,
$colonPos + 1 );
- return $i;
- }
+ if ( $m[0][0] === '-{' ) {
+ $state = self::COLON_STATE_LC;
+ $lcLevel++;
+ $i = $m[0][1] + 1;
+ } else {
+ # Skip ahead to next
interesting character.
+ $i = $m[0][1] - 1;
}
- if ( $ltPos === false ) {
-
