Tim Starling has uploaded a new change for review. (
https://gerrit.wikimedia.org/r/394905 )
Change subject: Add a tool to modify and standardize file headers
......................................................................
Add a tool to modify and standardize file headers
See the core commit Ie0cea0ef5 for the details of the resulting changes.
Change-Id: I4689786bb87ac8fe86564e148d0997cb578dbb47
---
A bin/fixFileHeader.php
M composer.json
A src/FileHeaderFixer.php
3 files changed, 386 insertions(+), 0 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/mediawiki/tools/namespaceizer
refs/changes/05/394905/1
diff --git a/bin/fixFileHeader.php b/bin/fixFileHeader.php
new file mode 100644
index 0000000..0334ac9
--- /dev/null
+++ b/bin/fixFileHeader.php
@@ -0,0 +1,29 @@
+<?php
+
+use MediaWiki\Tool\Namespaceizer\FileHeaderFixer;
+
+if ( PHP_SAPI !== 'cli' ) {
+ exit( 1 );
+}
+
+require __DIR__ .'/../vendor/autoload.php';
+
+function fixFileHeader( $argv ) {
+ if ( count( $argv ) < 2 ) {
+ echo "Usage: {$argv[0]} <file> ...\n";
+ exit( 1 );
+ }
+ $files = array_slice( $argv, 1 );
+
+ foreach ( $files as $fileName ) {
+ $input = file_get_contents( $fileName );
+ $result = FileHeaderFixer::fix( $input, function ( $line, $msg
) use ( $fileName ) {
+ echo "$fileName:$line: $msg\n";
+ } );
+ if ( $result !== false && $input !== $result ) {
+ file_put_contents( $fileName, $result );
+ }
+ }
+}
+
+fixFileHeader( $argv );
diff --git a/composer.json b/composer.json
index 1a2d50d..6fb2e15 100644
--- a/composer.json
+++ b/composer.json
@@ -5,5 +5,10 @@
"psr-4": {
"MediaWiki\\Tool\\Namespaceizer\\": "src/"
}
+ },
+ "minimum-stability": "alpha",
+ "prefer-stable": true,
+ "require": {
+ "nikic/php-parser": "~4.0"
}
}
diff --git a/src/FileHeaderFixer.php b/src/FileHeaderFixer.php
new file mode 100644
index 0000000..a7affaa
--- /dev/null
+++ b/src/FileHeaderFixer.php
@@ -0,0 +1,352 @@
+<?php
+
+namespace MediaWiki\Tool\Namespaceizer;
+
+use PhpParser\Lexer;
+use PhpParser\NodeTraverser;
+use PhpParser\NodeVisitor;
+use PhpParser\Parser;
+use PhpParser\PrettyPrinter;
+use PhpParser\Comment;
+
+class FileHeaderFixer {
+ private $text;
+
+ public static function fix( $text, $errorCallback ) {
+ $fixer = new self( $text, $errorCallback );
+ try {
+ return $fixer->execute();
+ } catch ( Error $error ) {
+ return false;
+ }
+ }
+
+ private function __construct( $text, $errorCallback ) {
+ $this->text = $text;
+ $this->errorCallback = $errorCallback;
+ }
+
+ private function execute() {
+ $lexer = new Lexer\Emulative( [
+ 'usedAttributes' => [
+ 'comments',
+ 'startLine', 'endLine',
+ 'startTokenPos', 'endTokenPos',
+ ],
+ ] );
+ $parser = new Parser\Php7( $lexer );
+ $oldAst = $parser->parse( $this->text );
+
+ $traverser = new NodeTraverser;
+ $traverser->addVisitor( new NodeVisitor\CloningVisitor );
+ $newAst = $traverser->traverse( $oldAst );
+
+ if ( count( $newAst ) < 1 ) {
+ $this->fatal( 0, "empty file" );
+ }
+
+ $nonClass = $this->findNonClassCode( $newAst );
+ if ( $nonClass ) {
+ $this->fatal( $nonClass->getLine(), "non-class code
found" );
+ }
+
+ $firstNode = $newAst[0];
+ $comments = $firstNode->getAttribute( 'comments' );
+ if ( !count( $comments ) ) {
+ $this->error( $firstNode->getLine(), "First node has no
comments" );
+ return $this->text;
+ }
+
+ $commentText = $comments[0]->getText();
+ $parts = $this->splitComment( $firstNode->getLine(),
$commentText );
+
+ if ( isset( $parts['start'] ) ) {
+ $parts['start'] = $this->removeCreationDate(
$parts['start'] );
+ }
+
+ $ingroup = false;
+ if ( isset( $parts['end'] ) ) {
+ $ingroup = $this->getAnnotation( $parts['end'],
'ingroup' );
+ $parts['end'] = $this->removeAnnotation( $parts['end'],
'ingroup' );
+ }
+ if ( isset( $parts['license'] ) ) {
+ $newLicenseComment = "/*\n" . $parts['license'] . "
*/\n";
+ } else {
+ $newLicenseComment = null;
+ }
+
+ if ( isset( $parts['license'] ) && ( isset( $parts['start'] )
|| isset( $parts['end'] ) ) ) {
+ $newFileComment = "/**\n";
+ if ( isset( $parts['start'] ) ) {
+ $newFileComment .= $parts['start'];
+ }
+ if ( isset( $parts['end'] ) ) {
+ $newFileComment .= $parts['end'];
+ }
+ $newFileComment .= " */";
+
+ $newComments = [];
+ if ( !$this->isPracticallyEmpty( $newFileComment ) ) {
+ $newComments[] = new Comment( $newFileComment );
+ }
+ $newComments[] = new Comment( $newLicenseComment );
+ array_splice( $comments, 0, 1, $newComments );
+ } elseif ( isset( $parts['license'] ) ) {
+ $comments[0] = new Comment( $newLicenseComment );
+ }
+
+ $firstNode->setAttribute( 'comments', $comments );
+
+ if ( $ingroup !== false ) {
+ $this->forEachClass( $newAst, function ( $node ) use (
$ingroup ) {
+ $this->addAnnotation( $node, 'ingroup',
$ingroup );
+ } );
+ }
+
+ $printer = new PrettyPrinter\Standard();
+ return $printer->printFormatPreserving( $newAst, $oldAst,
$lexer->getTokens() );
+ }
+
+ /**
+ * Split a string into an array of lines, assuming that the final line
is
+ * terminated by an LF character
+ */
+ private function explodeLines( $text ) {
+ $lines = explode( "\n", $text );
+ if ( end( $lines ) === '' ) {
+ array_pop( $lines );
+ }
+ return $lines;
+ }
+
+ /**
+ * Reassemble a string from an array of lines, adding a trailing line
break
+ */
+ private function implodeLines( $lines ) {
+ return $lines ? implode( "\n", $lines ) . "\n" : '';
+ }
+
+ private function splitComment( $startLine, $text ) {
+ $lines = $this->explodeLines( $text );
+ $n = count( $lines );
+ if ( $lines[0] !== '/**' ) {
+ $this->fatal( $startLine, "File starts with non-doc
comment" );
+ }
+ if ( $lines[$n - 1] !== ' */' ) {
+ $this->fatal( $startLine + $n - 1, "Unexpected comment
end" );
+ }
+
+ $licenseStart = $licenseEnd = null;
+ for ( $i = 1; $i < $n - 1; $i++ ) {
+ if ( strpos( $lines[$i], 'This program is free
software' ) !== false ) {
+ // Include copyright notices immediately before
the license in the license section
+ if ( $i >= 3 && $lines[$i - 1] === ' *'
+ && $this->isCopyright( $lines[$i - 2] )
+ ) {
+ $i -= 2;
+ }
+ while ( $i >= 2 && $this->isCopyright(
$lines[$i - 1] ) ) {
+ $i--;
+ }
+ $licenseStart = $i;
+ break;
+ }
+ }
+ if ( $licenseStart !== null ) {
+ for ( ; $i < $n - 1; $i++ ) {
+ if ( strpos( $lines[$i], '51 Franklin Street' )
!== false ) {
+ if ( strpos( $lines[$i + 1],
'www.gnu.org' ) !== false ) {
+ $i++;
+ }
+ $licenseEnd = $i;
+ if ( $lines[$i + 1] === ' *' ) {
+ $i++;
+ }
+ $footerStart = $i + 1;
+ break;
+ }
+ }
+ }
+
+ $parts = [];
+ if ( $licenseStart === null ) {
+ $parts['start'] = $this->implodeLines(
+ array_slice( $lines, 1, $n - 2 ) );
+ } else {
+ if ( $licenseStart > 1 ) {
+ $parts['start'] = $this->implodeLines(
+ array_slice( $lines, 1, $licenseStart -
1 ) );
+ }
+ if ( $licenseEnd === null ) {
+ $parts['license'] = $this->implodeLines(
+ array_slice( $lines, $licenseStart, $n
- $licenseStart - 1 ) );
+ } else {
+ $parts['license'] = $this->implodeLines(
+ array_slice( $lines, $licenseStart,
$licenseEnd - $licenseStart + 1 ) );
+ $parts['end'] = $this->implodeLines(
+ array_slice( $lines, $footerStart, $n -
$footerStart - 1 ) );
+ }
+ }
+
+ return $parts;
+ }
+
+ private function isCopyright( $line ) {
+ return !!preg_match( '!' .
+ 'copyright|' . // Regular
+ 'https://www.mediawiki.org/\s*$|' . // Brionic
+ 'You may copy this code freely' . // Dairikite
+ '!i', $line );
+ }
+
+ private function getAnnotation( $text, $name ) {
+ foreach ( explode( "\n", $text ) as $line ) {
+ if ( preg_match( "/@" . preg_quote( $name, '/' ) .
'\s+(.*)$/', $line, $m ) ) {
+ return $m[1];
+ }
+ }
+ return false;
+ }
+
+ private function removeAnnotation( $text, $name ) {
+ $lines = $this->explodeLines( $text );
+ $removeStart = null;
+ $removeLength = 1;
+ foreach ( $lines as $i => $line ) {
+ $pos = strpos( $line, "@$name" );
+ if ( $pos !== false ) {
+ $removeStart = $i;
+ if ( $i > 0 && $i < count( $lines ) - 2
+ && $line[$i - 1] === ' *'
+ && $line[$i + 1] === ' *'
+ ) {
+ $removeLength = 2;
+ }
+ break;
+ }
+ }
+
+ if ( $removeStart !== null ) {
+ array_splice( $lines, $removeStart, $removeLength );
+ }
+ return $this->implodeLines( $lines );
+ }
+
+ private function addAnnotation( $node, $name, $value ) {
+ $comments = $node->getComments();
+ if ( $comments ) {
+ $lastIndex = count( $comments ) - 1;
+ $lastComment = (string)$comments[$lastIndex];
+
+ if ( strpos( $lastComment, "@$name $value" ) !== false
) {
+ // Don't add duplicate
+ return;
+ }
+
+ if ( strpos( $lastComment, '@file' ) !== false
+ || strpos( $lastComment, 'This program is free
software' ) !== false
+ ) {
+ $comments[] = new Comment( "/**\n * @$name
$value\n */" );
+ } else {
+ $lastComment = str_replace( "\n */", "\n *
@$name $value\n */", $lastComment );
+ array_splice( $comments, $lastIndex, 1, [
+ new Comment( $lastComment )
+ ] );
+ }
+ } else {
+ $comments = [
+ new Comment( "/**\n * @$name $value\n */" )
+ ];
+ }
+ $node->setAttribute( 'comments', $comments );
+ }
+
+ /**
+ * Determine if a full comment string has nothing useful in it
+ */
+ private function isPracticallyEmpty( $text ) {
+ foreach ( $this->explodeLines( $text ) as $line ) {
+ if ( $line !== ''
+ && $line !== ' *'
+ && $line !== ' * @file'
+ && $line !== '/**'
+ && $line !== ' */'
+ ) {
+ return false;
+ }
+ }
+ return true;
+ }
+
+ private function removeCreationDate( $text ) {
+ $lines = $this->explodeLines( $text );
+ $n = count( $lines );
+ foreach ( $lines as $i => $line ) {
+ if ( preg_match( '/^ \* Created on .*20\d\d$/', $line )
) {
+ // Comments of this kind usually have an excess
of empty lines
+ // preceding them. Remove those lines.
+ $startBlock = $i;
+ while ( $startBlock > 0 && $lines[$startBlock -
1] === ' *' ) {
+ $startBlock--;
+ }
+
+ array_splice( $lines, $startBlock, $i -
$startBlock + 1, [] );
+ return $this->implodeLines( $lines );
+ }
+ }
+ return $text;
+ }
+
+ private function error( $line, $msg ) {
+ if ( $this->errorCallback ) {
+ call_user_func( $this->errorCallback, $line, $msg );
+ }
+ }
+
+ private function fatal( $line, $msg ) {
+ $this->error( $line, $msg );
+ throw new Error( $msg );
+ }
+
+ private function info( $msg ) {
+ fwrite( STDERR, "$msg\n" );
+ }
+
+ private function findNonClassCode( $ast ) {
+ foreach ( $ast as $node ) {
+ $type = $node->getType();
+ if ( in_array( $type, [
+ 'Stmt_Class',
+ 'Stmt_Interface',
+ 'Stmt_Trait',
+ 'Stmt_Namespace',
+ 'Stmt_Use'] )
+ ) {
+ continue;
+ }
+ if ( $type === 'Stmt_Expression' ) {
+ if ( $node->expr->getType() === 'Expr_FuncCall'
+ && !empty( $node->args )
+ && $node->args[0]->getType() ===
'Scalar_String'
+ && $node->args[0]->value ===
'class_alias'
+ ) {
+ continue;
+ }
+ }
+ return $node;
+ }
+ return false;
+ }
+
+ private function forEachClass( $nodeList, $callback ) {
+ foreach ( $nodeList as $node ) {
+ if ( in_array( $node->getType(), [
+ 'Stmt_Class',
+ 'Stmt_Interface',
+ 'Stmt_Trait' ] )
+ ) {
+ $callback( $node );
+ }
+ }
+ }
+}
--
To view, visit https://gerrit.wikimedia.org/r/394905
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I4689786bb87ac8fe86564e148d0997cb578dbb47
Gerrit-PatchSet: 1
Gerrit-Project: mediawiki/tools/namespaceizer
Gerrit-Branch: master
Gerrit-Owner: Tim Starling <tstarl...@wikimedia.org>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
