diocloid/LinkTitles
1
0
Fork 0
mirror of https://github.com/diocloid/LinkTitles.git synced 2025-10-22 05:42:32 +02:00
Code Issues Packages Projects Releases Wiki Activity

Add Delimiters class.

Browse Source
This commit is contained in:
Daniel Kraus
2017-08-26 22:01:52 +02:00
parent 51fe6b3910
commit 4d5554a74e
5 changed files with 238 additions and 87 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
extension.json
View File

@@ -35,6 +35,7 @@
"AutoloadClasses": {
"LinkTitles\\Extension": "includes/Extension.php",
"LinkTitles\\Targets": "includes/Targets.php",
"LinkTitles\\Delimiters": "includes/Delimiters.php",
"LinkTitles\\Config": "includes/Config.php",
"LinkTitles\\Special": "includes/Special.php",
"LinkTitles\\TestCase": "tests/phpunit/TestCase.php"
@@ -64,7 +65,6 @@
"LinkTitles\\Extension::onParserFirstCallInit"
]
},
"callback": "LinkTitles\\Extension::setup",
"ExtensionMessagesFiles": {
"LinkTitlesMagic": "includes/Magic.php"
},

45
includes/Config.php
View File

@@ -96,6 +96,43 @@ class Config {
*/
public $capitalLinks;
/**
* Whether or not to link to pages only if the page title appears at the
* start of a word on the target page (i.e., link 'MediaWiki' to a page
* 'Media', but not to a page 'Wiki').
*
* Set both $wordStartOnly and $wordEndOnly to true to enforce matching
* whole titles.
*
* @var bool $wordStartOnly;
*/
public $wordStartOnly;
/**
* Whether or not to link to pages only if the page title appears at the
* end of a word on the target page (i.e., link 'MediaWiki' to a page
* 'Wiki', but not to a page 'Media').
*
* Set both $wordStartOnly and $wordEndOnly to true to enforce matching
* whole titles.
*
* @var bool $wordEndOnly;
*/
public $wordEndOnly;
/**
* Whether or not to skip templates. If set to true, text inside transclusions
* will not be linked.
* @var bool $skipTemplates
*/
public $skipTemplates;
/**
* Whether or not to parse headings.
* @var bool $parseHeadings
*/
public $parseHeadings;
/**
* Constructs a new Config object.
*
@@ -112,6 +149,10 @@ class Config {
global $wgLinkTitlesFirstOnly;
global $wgLinkTitlesSmartMode;
global $wgCapitalLinks;
global $wgLinkTitlesWordStartOnly;
global $wgLinkTitlesWordEndOnly;
global $wgLinkTitlesSkipTemplates;
global $wgLinkTitlesParseHeadings;
$this->parseOnEdit = $wgLinkTitlesParseOnEdit;
$this->parseOnRender = $wgLinkTitlesParseOnRender;
$this->preferShortTitles = $wgLinkTitlesPreferShortTitles;
@@ -121,6 +162,10 @@ class Config {
$this->firstOnly = $wgLinkTitlesFirstOnly;
$this->smartMode = $wgLinkTitlesSmartMode;
$this->capitalLinks = $wgCapitalLinks; // MediaWiki global variable
$this->wordStartOnly = $wgLinkTitlesWordStartOnly;
$this->wordEndOnly = $wgLinkTitlesWordEndOnly;
$this->skipTemplates = $wgLinkTitlesSkipTemplates;
$this->parseHeadings = $wgLinkTitlesParseHeadings;
}
}

138
includes/Delimiters.php Normal file
View File

@@ -0,0 +1,138 @@
<?php
/**
* The Delimiters class caches a regular expression that delimits text to be parsed.
*
* Copyright 2012-2017 Daniel Kraus <bovender@bovender.de> ('bovender')
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
* MA 02110-1301, USA.
*
* @author Daniel Kraus <bovender@bovender.de>
*/
namespace LinkTitles;
/**
* Caches a regular expression that delimits text to be parsed.
*/
class Delimiters {
private static $instance;
/**
* Singleton factory.
*
* @param Config $config LinkTitles configuration.
*/
public static function default( Config $config ) {
if ( self::$instance === null ) {
self::$instance = new Delimiters( $config );
}
return self::$instance;
}
/**
* Invalidates the singleton instance.
*
* Used for unit testing.
*/
public static function invalidate() {
self::$instance = null;
}
protected function __construct( Config $config) {
$this->config = $config;
$this->buildDelimiters();
}
/**
* The splitting expression that separates text to be parsed from text that
* must not be parsed.
* @var String $splitter
*/
public $splitter;
/**
* Regex that matches the start of a word; this expression depends on the
* setting of LinkTitles\Config->wordStartOnly;
* @var String $wordStart
*/
public $wordStart;
/**
* Regex that matches the end of a word; this expression depends on the
* setting of LinkTitles\Config->wordEndOnly;
* @var String $wordEnd
*/
public $wordEnd;
private $config;
/*
* Builds the delimiter that is used in a regexp to separate
* text that should be parsed from text that should not be
* parsed (e.g. inside existing links etc.)
*/
private function buildDelimiters() {
// Use unicode character properties rather than \b escape sequences
// to detect whole words containing non-ASCII characters as well.
// Note that this requires a PCRE library that was compiled with
// --enable-unicode-properties
( $this->config->wordStartOnly ) ? $this->wordStart = '(?<!\pL)' : $this->wordStart = '';
( $this->config->wordEndOnly ) ? $this->wordEnd = '(?!\pL)' : $this->wordEnd = '';
if ( $this->config->skipTemplates )
{
// Use recursive regex to balance curly braces;
// see http://www.regular-expressions.info/recurse.html
$templatesDelimiter = '{{(?>[^{}]|(?R))*}}|';
} else {
// Match template names (ignoring any piped [[]] links in them)
// along with the trailing pipe and parameter name or closing
// braces; also match sequences of '|wordcharacters=' (without
// spaces in them) that usually only occur as parameter names in
// transclusions (but could also occur as wiki table cell contents).
// TODO: Find a way to match parameter names in transclusions, but
// not in table cells or other sequences involving a pipe character
// and equal sign.
$templatesDelimiter = '{{[^|]*?(?:(?:\[\[[^]]+]])?)[^|]*?(?:\|(?:\w+=)?|(?:}}))|\|\w+=|';
}
// Build a regular expression that will capture existing wiki links ("[[...]]"),
// wiki headings ("= ... =", "== ... ==" etc.),
// urls ("http://example.com", "[http://example.com]", "[http://example.com Description]",
// and email addresses ("mail@example.com").
// Since there is a user option to skip headings, we make this part of the expression
// optional. Note that in order to use preg_split(), it is important to have only one
// capturing subpattern (which precludes the use of conditional subpatterns).
( $this->config->parseHeadings ) ? $delimiter = '' : $delimiter = '=+.+?=+|';
$urlPattern = '[a-z]+?\:\/\/(?:\S+\.)+\S+(?:\/.*)?';
$this->splitter = '/(' . // exclude from linking:
'\[\[.*?\]\]|' . // links
$delimiter . // titles (if requested)
$templatesDelimiter . // templates (if requested)
'^ .+?\n|\n .+?\n|\n .+?$|^ .+?$|' . // preformatted text
'<nowiki>.*?<.nowiki>|<code>.*?<\/code>|' . // nowiki/code
'<pre>.*?<\/pre>|<html>.*?<\/html>|' . // pre/html
'<script>.*?<\/script>|' . // script
'<gallery>.*?<\/gallery>|' . // gallery
'<div.+?>|<\/div>|' . // attributes of div elements
'<span.+?>|<\/span>|' . // attributes of span elements
'<file>[^<]*<\/file>|' . // stuff inside file elements
'style=".+?"|class=".+?"|' . // styles and classes (e.g. of wikitables)
'<noautolinks>.*?<\/noautolinks>|' . // custom tag 'noautolinks'
'\[' . $urlPattern . '\s.+?\]|'. $urlPattern . '(?=\s|$)|' . // urls
'(?<=\b)\S+\@(?:\S+\.)+\S+(?=\b)' . // email addresses
')/ismS';
}
}

98
includes/Extension.php
View File

@@ -45,22 +45,9 @@ class Extension {
/// as a string.
private static $targetTitleText;
/// Delimiter used in a regexp split operation to seperate those parts
/// of the page that should be parsed from those that should not be
/// parsed (e.g. inside pre-existing links etc.).
private static $delimiter;
private static $wordStartDelim;
private static $wordEndDelim;
public static $ltConsoleOutput;
public static $ltConsoleOutputDebug;
/// Setup method
public static function setup() {
self::BuildDelimiters();
}
/// Event handler that is hooked to the PageContentSave event.
public static function onPageContentSave( &$wikiPage, &$user, &$content, &$summary,
$isMinor, $isWatch, $section, &$flags, &$status ) {
@@ -120,7 +107,9 @@ class Extension {
self::$currentTitle = $title;
$newText = $text;
$targets = Targets::default( $title, new Config() );
$config = new Config();
$delimiters = Delimiters::default( $config );
$targets = Targets::default( $title, $config );
// Iterate through the page titles
foreach( $targets->queryResult as $row ) {
@@ -132,7 +121,7 @@ class Extension {
// split the page content by [[...]] groups
// credits to inhan @ StackOverflow for suggesting preg_split
// see http://stackoverflow.com/questions/10672286
$arr = preg_split( self::$delimiter, $newText, -1, PREG_SPLIT_DELIM_CAPTURE );
$arr = preg_split( $delimiters->splitter, $newText, -1, PREG_SPLIT_DELIM_CAPTURE );
// Escape certain special characters in the page title to prevent
// regexp compilation errors
@@ -146,20 +135,19 @@ class Extension {
// the title has to be searched for either in a strictly case-sensitive
// way, or in a 'fuzzy' way where the first letter of the title may
// be either case.
if ( $wgCapitalLinks && ( $quotedTitle[0] != '\\' )) {
if ( $config->capitalLinks && ( $quotedTitle[0] != '\\' )) {
$searchTerm = '((?i)' . $quotedTitle[0] . '(?-i)' .
substr($quotedTitle, 1) . ')';
} else {
$searchTerm = '(' . $quotedTitle . ')';
}
$regex = '/(?<![\:\.\@\/\?\&])' . self::$wordStartDelim .
$searchTerm . self::$wordEndDelim . '/S';
$regex = '/(?<![\:\.\@\/\?\&])' . $delimiters->wordStart . $searchTerm . $delimiters->wordEnd . '/S';
for ( $i = 0; $i < count( $arr ); $i+=2 ) {
// even indexes will point to text that is not enclosed by brackets
$arr[$i] = preg_replace_callback( $regex,
'LinkTitles\Extension::simpleModeCallback', $arr[$i], $limit, $count );
if ( $wgLinkTitlesFirstOnly && ( $count > 0 ) ) {
if ( $config->firstOnly && ( $count > 0 ) ) {
$limitReached = true;
break;
};
@@ -169,16 +157,16 @@ class Extension {
// If smart mode is turned on, the extension will perform a second
// pass on the page and add links with aliases where the case does
// not match.
if ( $wgLinkTitlesSmartMode && !$limitReached ) {
$arr = preg_split( self::$delimiter, $newText, -1, PREG_SPLIT_DELIM_CAPTURE );
if ( $config->smartMode && !$limitReached ) {
$arr = preg_split( $delimiters->splitter, $newText, -1, PREG_SPLIT_DELIM_CAPTURE );
for ( $i = 0; $i < count( $arr ); $i+=2 ) {
// even indexes will point to text that is not enclosed by brackets
$arr[$i] = preg_replace_callback( '/(?<![\:\.\@\/\?\&])' .
self::$wordStartDelim . '(' . $quotedTitle . ')' .
self::$wordEndDelim . '/iS', 'LinkTitles\Extension::smartModeCallback',
$delimiters->wordStart . '(' . $quotedTitle . ')' .
$delimiters->wordEnd . '/iS', 'LinkTitles\Extension::smartModeCallback',
$arr[$i], $limit, $count );
if ( $wgLinkTitlesFirstOnly && ( $count > 0 )) {
if ( $config->firstOnly && ( $count > 0 )) {
break;
};
};
@@ -368,68 +356,6 @@ class Extension {
return true;
}
/// Builds the delimiter that is used in a regexp to separate
/// text that should be parsed from text that should not be
/// parsed (e.g. inside existing links etc.)
private static function BuildDelimiters() {
// Configuration variables need to be defined here as globals.
global $wgLinkTitlesParseHeadings;
global $wgLinkTitlesSkipTemplates;
global $wgLinkTitlesWordStartOnly;
global $wgLinkTitlesWordEndOnly;
// Use unicode character properties rather than \b escape sequences
// to detect whole words containing non-ASCII characters as well.
// Note that this requires a PCRE library that was compiled with
// --enable-unicode-properties
( $wgLinkTitlesWordStartOnly ) ? self::$wordStartDelim = '(?<!\pL)' : self::$wordStartDelim = '';
( $wgLinkTitlesWordEndOnly ) ? self::$wordEndDelim = '(?!\pL)' : self::$wordEndDelim = '';
if ( $wgLinkTitlesSkipTemplates )
{
// Use recursive regex to balance curly braces;
// see http://www.regular-expressions.info/recurse.html
$templatesDelimiter = '{{(?>[^{}]|(?R))*}}|';
} else {
// Match template names (ignoring any piped [[]] links in them)
// along with the trailing pipe and parameter name or closing
// braces; also match sequences of '|wordcharacters=' (without
// spaces in them) that usually only occur as parameter names in
// transclusions (but could also occur as wiki table cell contents).
// TODO: Find a way to match parameter names in transclusions, but
// not in table cells or other sequences involving a pipe character
// and equal sign.
$templatesDelimiter = '{{[^|]*?(?:(?:\[\[[^]]+]])?)[^|]*?(?:\|(?:\w+=)?|(?:}}))|\|\w+=|';
}
// Build a regular expression that will capture existing wiki links ("[[...]]"),
// wiki headings ("= ... =", "== ... ==" etc.),
// urls ("http://example.com", "[http://example.com]", "[http://example.com Description]",
// and email addresses ("mail@example.com").
// Since there is a user option to skip headings, we make this part of the expression
// optional. Note that in order to use preg_split(), it is important to have only one
// capturing subpattern (which precludes the use of conditional subpatterns).
( $wgLinkTitlesParseHeadings ) ? $delimiter = '' : $delimiter = '=+.+?=+|';
$urlPattern = '[a-z]+?\:\/\/(?:\S+\.)+\S+(?:\/.*)?';
self::$delimiter = '/(' . // exclude from linking:
'\[\[.*?\]\]|' . // links
$delimiter . // titles (if requested)
$templatesDelimiter . // templates (if requested)
'^ .+?\n|\n .+?\n|\n .+?$|^ .+?$|' . // preformatted text
'<nowiki>.*?<.nowiki>|<code>.*?<\/code>|' . // nowiki/code
'<pre>.*?<\/pre>|<html>.*?<\/html>|' . // pre/html
'<script>.*?<\/script>|' . // script
'<gallery>.*?<\/gallery>|' . // gallery
'<div.+?>|<\/div>|' . // attributes of div elements
'<span.+?>|<\/span>|' . // attributes of span elements
'<file>[^<]*<\/file>|' . // stuff inside file elements
'style=".+?"|class=".+?"|' . // styles and classes (e.g. of wikitables)
'<noautolinks>.*?<\/noautolinks>|' . // custom tag 'noautolinks'
'\[' . $urlPattern . '\s.+?\]|'. $urlPattern . '(?=\s|$)|' . // urls
'(?<=\b)\S+\@(?:\S+\.)+\S+(?=\b)' . // email addresses
')/ismS';
}
/// Local Debugging output function which can send output to console as well
public static function ltDebugLog($text) {
if ( self::$ltConsoleOutputDebug ) {

42
tests/phpunit/DelimitersTest.php Normal file
View File

@@ -0,0 +1,42 @@
<?php
/**
* @group bovender
*/
class DelimitersTest extends MediaWikiTestCase {
/**
* @dataProvider provideStartOnly
*/
public function testDelimitersWordStartOnly( $enabled, $delimiter ) {
$config = new LinkTitles\Config();
$config->wordStartOnly = $enabled;
LinkTitles\Delimiters::invalidate();
$d = LinkTitles\Delimiters::default( $config );
$this->assertSame( $delimiter, $d->wordStart );
}
public static function provideStartOnly() {
return [
[ true, '(?<!\pL)' ],
[ false, '' ]
];
}
/**
* @dataProvider provideEndOnly
*/
public function testDelimitersWordEndOnly( $enabled, $delimiter ) {
$config = new LinkTitles\Config();
$config->wordEndOnly = $enabled;
LinkTitles\Delimiters::invalidate();
$d = LinkTitles\Delimiters::default( $config );
$this->assertSame( $delimiter, $d->wordEnd );
}
public static function provideEndOnly() {
return [
[ true, '(?!\pL)' ],
[ false, '' ]
];
}
}