jenkins-bot has submitted this change and it was merged.
Change subject: Improve MultiStringReplacer performance
......................................................................
Improve MultiStringReplacer performance
Don't call MultiStringMatcher::searchIn, because that way we end up having to
walk the results array again to construct the replacement matches array.
Instead, perform the search in MultiStringReplacer itself.
Change-Id: I2101f9764c4ddb8f4f7d2b2d90368a9ca04c6862
---
M .gitignore
M bench/bench.php
M src/MultiStringReplacer.php
3 files changed, 14 insertions(+), 9 deletions(-)
Approvals:
Ori.livneh: Looks good to me, approved
jenkins-bot: Verified
diff --git a/.gitignore b/.gitignore
index f6f384e..4f51d1a 100644
--- a/.gitignore
+++ b/.gitignore
@@ -2,3 +2,5 @@
/doc/html
vendor/
composer.lock
+bench/23835-0.txt
+bench/ZhConversion.php
diff --git a/bench/bench.php b/bench/bench.php
index 7562be1..b537025 100644
--- a/bench/bench.php
+++ b/bench/bench.php
@@ -6,11 +6,12 @@
use AhoCorasick\MultiStringMatcher;
if ( !file_exists( __DIR__ . '/23835-0.txt' ) ) {
- die( 'Please download http://www.gutenberg.org/files/23835/23835-0.txt'
);
+ die( "Please download
http://www.gutenberg.org/files/23835/23835-0.txt\n"; );
}
if ( !file_exists( __DIR__ . '/ZhConversion.php' ) ) {
- die( 'You need ZhConversion.php, from http://git.io/vIMst' );
+ die( "You need ZhConversion.php, from " .
+
"https://github.com/wikimedia/mediawiki/blob/master/includes/ZhConversion.php\n";
);
}
require_once __DIR__ . '/ZhConversion.php';
diff --git a/src/MultiStringReplacer.php b/src/MultiStringReplacer.php
index ee1ace0..ab3ebc4 100644
--- a/src/MultiStringReplacer.php
+++ b/src/MultiStringReplacer.php
@@ -72,14 +72,16 @@
* @endcode
*/
public function searchAndReplace( $text ) {
- if ( !$this->searchKeywords || $text === '' ) {
- return $text;
- }
-
+ $state = 0;
+ $length = strlen( $text );
$matches = array();
- foreach ( $this->searchIn( $text ) as $result ) {
- list( $offset, $match ) = $result;
- $matches[$offset] = $match;
+ for ( $i = 0; $i < $length; $i++ ) {
+ $ch = $text[$i];
+ $state = $this->nextState( $state, $ch );
+ foreach ( $this->outputs[$state] as $match ) {
+ $offset = $i - $this->searchKeywords[$match] +
1;
+ $matches[$offset] = $match;
+ }
}
ksort( $matches );
--
To view, visit https://gerrit.wikimedia.org/r/218121
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: merged
Gerrit-Change-Id: I2101f9764c4ddb8f4f7d2b2d90368a9ca04c6862
Gerrit-PatchSet: 2
Gerrit-Project: AhoCorasick
Gerrit-Branch: master
Gerrit-Owner: Ori.livneh <o...@wikimedia.org>
Gerrit-Reviewer: Ori.livneh <o...@wikimedia.org>
Gerrit-Reviewer: jenkins-bot <>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
