When parsing an incomplete string literal or regex literal at the end of a file, $end would be set to an offset higher than $length, because the code speculatively increases $end without correcting for this scenario. This is due to the assumption that the strcspn() search will end because an end character was seen, instead of simply ending because the string doesn't have any further characters. Bug: T75556 Change-Id: I2325c9aff33293c13ff414699c2d47306182aaa6
238 lines
7.4 KiB
PHP
238 lines
7.4 KiB
PHP
<?php
|
|
|
|
class JavaScriptMinifierTest extends PHPUnit_Framework_TestCase {
|
|
|
|
use MediaWikiCoversValidator;
|
|
|
|
public static function provideCases() {
|
|
return [
|
|
|
|
// Basic whitespace and comments that should be stripped entirely
|
|
[ "\r\t\f \v\n\r", "" ],
|
|
[ "/* Foo *\n*bar\n*/", "" ],
|
|
|
|
/**
|
|
* Slashes used inside block comments (T28931).
|
|
* At some point there was a bug that caused this comment to be ended at '* /',
|
|
* causing /M... to be left as the beginning of a regex.
|
|
*/
|
|
[
|
|
"/**\n * Foo\n * {\n * 'bar' : {\n * "
|
|
. "//Multiple rules with configurable operators\n * 'baz' : false\n * }\n */",
|
|
"" ],
|
|
|
|
/**
|
|
* ' Foo \' bar \
|
|
* baz \' quox ' .
|
|
*/
|
|
[
|
|
"' Foo \\' bar \\\n baz \\' quox ' .length",
|
|
"' Foo \\' bar \\\n baz \\' quox '.length"
|
|
],
|
|
[
|
|
"\" Foo \\\" bar \\\n baz \\\" quox \" .length",
|
|
"\" Foo \\\" bar \\\n baz \\\" quox \".length"
|
|
],
|
|
[ "// Foo b/ar baz", "" ],
|
|
[
|
|
"/ Foo \\/ bar [ / \\] / ] baz / .length",
|
|
"/ Foo \\/ bar [ / \\] / ] baz /.length"
|
|
],
|
|
|
|
// HTML comments
|
|
[ "<!-- Foo bar", "" ],
|
|
[ "<!-- Foo --> bar", "" ],
|
|
[ "--> Foo", "" ],
|
|
[ "x --> y", "x-->y" ],
|
|
|
|
// Semicolon insertion
|
|
[ "(function(){return\nx;})", "(function(){return\nx;})" ],
|
|
[ "throw\nx;", "throw\nx;" ],
|
|
[ "while(p){continue\nx;}", "while(p){continue\nx;}" ],
|
|
[ "while(p){break\nx;}", "while(p){break\nx;}" ],
|
|
[ "var\nx;", "var x;" ],
|
|
[ "x\ny;", "x\ny;" ],
|
|
[ "x\n++y;", "x\n++y;" ],
|
|
[ "x\n!y;", "x\n!y;" ],
|
|
[ "x\n{y}", "x\n{y}" ],
|
|
[ "x\n+y;", "x+y;" ],
|
|
[ "x\n(y);", "x(y);" ],
|
|
[ "5.\nx;", "5.\nx;" ],
|
|
[ "0xFF.\nx;", "0xFF.x;" ],
|
|
[ "5.3.\nx;", "5.3.x;" ],
|
|
|
|
// Cover failure case for incomplete hex literal
|
|
[ "0x;", false, false ],
|
|
|
|
// Cover failure case for number with no digits after E
|
|
[ "1.4E", false, false ],
|
|
|
|
// Cover failure case for number with several E
|
|
[ "1.4EE2", false, false ],
|
|
[ "1.4EE", false, false ],
|
|
|
|
// Cover failure case for number with several E (nonconsecutive)
|
|
// FIXME: This is invalid, but currently tolerated
|
|
[ "1.4E2E3", "1.4E2 E3", false ],
|
|
|
|
// Semicolon insertion between an expression having an inline
|
|
// comment after it, and a statement on the next line (T29046).
|
|
[
|
|
"var a = this //foo bar \n for ( b = 0; c < d; b++ ) {}",
|
|
"var a=this\nfor(b=0;c<d;b++){}"
|
|
],
|
|
|
|
// Cover failure case of incomplete regexp at end of file (T75556)
|
|
// FIXME: This is invalid, but currently tolerated
|
|
[ "*/", "*/", false ],
|
|
|
|
// Cover failure case of incomplete string at end of file (T75556)
|
|
// FIXME: This is invalid, but currently tolerated
|
|
[ "'a", "'a", false ],
|
|
|
|
// Token separation
|
|
[ "x in y", "x in y" ],
|
|
[ "/x/g in y", "/x/g in y" ],
|
|
[ "x in 30", "x in 30" ],
|
|
[ "x + ++ y", "x+ ++y" ],
|
|
[ "x ++ + y", "x++ +y" ],
|
|
[ "x / /y/.exec(z)", "x/ /y/.exec(z)" ],
|
|
|
|
// State machine
|
|
[ "/ x/g", "/ x/g" ],
|
|
[ "(function(){return/ x/g})", "(function(){return/ x/g})" ],
|
|
[ "+/ x/g", "+/ x/g" ],
|
|
[ "++/ x/g", "++/ x/g" ],
|
|
[ "x/ x/g", "x/x/g" ],
|
|
[ "(/ x/g)", "(/ x/g)" ],
|
|
[ "if(/ x/g);", "if(/ x/g);" ],
|
|
[ "(x/ x/g)", "(x/x/g)" ],
|
|
[ "([/ x/g])", "([/ x/g])" ],
|
|
[ "+x/ x/g", "+x/x/g" ],
|
|
[ "{}/ x/g", "{}/ x/g" ],
|
|
[ "+{}/ x/g", "+{}/x/g" ],
|
|
[ "(x)/ x/g", "(x)/x/g" ],
|
|
[ "if(x)/ x/g", "if(x)/ x/g" ],
|
|
[ "for(x;x;{}/ x/g);", "for(x;x;{}/x/g);" ],
|
|
[ "x;x;{}/ x/g", "x;x;{}/ x/g" ],
|
|
[ "x:{}/ x/g", "x:{}/ x/g" ],
|
|
[ "switch(x){case y?z:{}/ x/g:{}/ x/g;}", "switch(x){case y?z:{}/x/g:{}/ x/g;}" ],
|
|
[ "function x(){}/ x/g", "function x(){}/ x/g" ],
|
|
[ "+function x(){}/ x/g", "+function x(){}/x/g" ],
|
|
|
|
// Multiline quoted string
|
|
[ "var foo=\"\\\nblah\\\n\";", "var foo=\"\\\nblah\\\n\";" ],
|
|
|
|
// Multiline quoted string followed by string with spaces
|
|
[
|
|
"var foo=\"\\\nblah\\\n\";\nvar baz = \" foo \";\n",
|
|
"var foo=\"\\\nblah\\\n\";var baz=\" foo \";"
|
|
],
|
|
|
|
// URL in quoted string ( // is not a comment)
|
|
[
|
|
"aNode.setAttribute('href','http://foo.bar.org/baz');",
|
|
"aNode.setAttribute('href','http://foo.bar.org/baz');"
|
|
],
|
|
|
|
// URL in quoted string after multiline quoted string
|
|
[
|
|
"var foo=\"\\\nblah\\\n\";\naNode.setAttribute('href','http://foo.bar.org/baz');",
|
|
"var foo=\"\\\nblah\\\n\";aNode.setAttribute('href','http://foo.bar.org/baz');"
|
|
],
|
|
|
|
// Division vs. regex nastiness
|
|
[
|
|
"alert( (10+10) / '/'.charCodeAt( 0 ) + '//' );",
|
|
"alert((10+10)/'/'.charCodeAt(0)+'//');"
|
|
],
|
|
[ "if(1)/a /g.exec('Pa ss');", "if(1)/a /g.exec('Pa ss');" ],
|
|
|
|
// newline insertion after 1000 chars: break after the "++", not before
|
|
[ str_repeat( ';', 996 ) . "if(x++);", str_repeat( ';', 996 ) . "if(x++\n);" ],
|
|
|
|
// Unicode letter characters should pass through ok in identifiers (T33187)
|
|
[ "var KaŝSkatolVal = {}", 'var KaŝSkatolVal={}' ],
|
|
|
|
// Per spec unicode char escape values should work in identifiers,
|
|
// as long as it's a valid char. In future it might get normalized.
|
|
[ "var Ka\\u015dSkatolVal = {}", 'var Ka\\u015dSkatolVal={}' ],
|
|
|
|
// Some structures that might look invalid at first sight
|
|
[ "var a = 5.;", "var a=5.;" ],
|
|
[ "5.0.toString();", "5.0.toString();" ],
|
|
[ "5..toString();", "5..toString();" ],
|
|
// Cover failure case for too many decimal points
|
|
[ "5...toString();", false ],
|
|
[ "5.\n.toString();", '5..toString();' ],
|
|
|
|
// Boolean minification (!0 / !1)
|
|
[ "var a = { b: true };", "var a={b:!0};" ],
|
|
[ "var a = { true: 12 };", "var a={true:12};", false ],
|
|
[ "a.true = 12;", "a.true=12;", false ],
|
|
[ "a.foo = true;", "a.foo=!0;" ],
|
|
[ "a.foo = false;", "a.foo=!1;" ],
|
|
];
|
|
}
|
|
|
|
/**
|
|
* @dataProvider provideCases
|
|
* @covers JavaScriptMinifier::minify
|
|
* @covers JavaScriptMinifier::parseError
|
|
*/
|
|
public function testMinifyOutput( $code, $expectedOutput, $expectedValid = true ) {
|
|
$minified = JavaScriptMinifier::minify( $code );
|
|
|
|
// JSMin+'s parser will throw an exception if output is not valid JS.
|
|
// suppression of warnings needed for stupid crap
|
|
if ( $expectedValid ) {
|
|
MediaWiki\suppressWarnings();
|
|
$parser = new JSParser();
|
|
MediaWiki\restoreWarnings();
|
|
$parser->parse( $minified, 'minify-test.js', 1 );
|
|
}
|
|
|
|
$this->assertEquals(
|
|
$expectedOutput,
|
|
$minified,
|
|
"Minified output should be in the form expected."
|
|
);
|
|
}
|
|
|
|
public static function provideExponentLineBreaking() {
|
|
return [
|
|
[
|
|
// This one gets interpreted all together by the prior code;
|
|
// no break at the 'E' happens.
|
|
'1.23456789E55',
|
|
],
|
|
[
|
|
// This one breaks under the bad code; splits between 'E' and '+'
|
|
'1.23456789E+5',
|
|
],
|
|
[
|
|
// This one breaks under the bad code; splits between 'E' and '-'
|
|
'1.23456789E-5',
|
|
],
|
|
];
|
|
}
|
|
|
|
/**
|
|
* @dataProvider provideExponentLineBreaking
|
|
* @covers JavaScriptMinifier::minify
|
|
*/
|
|
public function testExponentLineBreaking( $num ) {
|
|
// Long line breaking was being incorrectly done between the base and
|
|
// exponent part of a number, causing a syntax error. The line should
|
|
// instead break at the start of the number. (T34548)
|
|
$prefix = 'var longVarName' . str_repeat( '_', 973 ) . '=';
|
|
$suffix = ',shortVarName=0;';
|
|
|
|
$input = $prefix . $num . $suffix;
|
|
$expected = $prefix . "\n" . $num . $suffix;
|
|
|
|
$minified = JavaScriptMinifier::minify( $input );
|
|
|
|
$this->assertEquals( $expected, $minified, "Line breaks must not occur in middle of exponent" );
|
|
}
|
|
}
|