@hackage language-javascript0.4.9

Parser for JavaScript

Based (loosely) on language-python

How to build

Library:

cabal clean && cabal configure && cabal build

Tests:

cabal clean && cabal configure -fbuildtests && cabal build

Running the tests

./dist/build/runtests/runtests

To debug the grammar

happy -iparse.txt -g -a -d src/Language/JavaScript/Parser/Grammar5.y

This generates src/Language/JavaScript/Parser/Grammar.hs, delete this when done with the debug version

UTF8/Unicode version

Alex 3.0 now supports unicode natively, and has been included as a dependency in the cabal file.

Changes

0.4.9 Make alex and happy versions more explicit. Expose the AlexSpan data type, so parse error positions can be reported.

0.4.8 Close issue https://github.com/alanz/language-javascript/issues/5 by correcting the lexical analyser for DecimalLiteral

0.4.7 Continue ECMASCRIPT update, remove incorrect future reserved words for "code", "mode", "of", "one", "or" and "strict", put in by accident/stupidity.

0.4.6 Update to ECMASCRIPT 5.1 by allowing continuations in string literals Include build dependencies on alex >= 3.0 and happy.

0.4.5 Update cabal file and docs for Alex 3.0, giving out of the box unicode support

0.4.4 Allow unicode characters in comments

0.4.3 correct lexer for decimalToken, only leading digit needs to be non-zero.

0.4.2 Expose SrcSpan

0.4.1 Expose AST.Node as well.

0.4.0 Updated JSNode to include location information, and made the parse more true to the original by leaving blocks as such, not replacing with ';'.

0.3.0 ECMA-262 edition 5 compliance. Required adding JSFunctionExpression and JSPropertyAccessor to the AST.

0.2.2 Heading toward compliance with edition 5. Trailing commas allowed in array literals too.

0.2.1 Allow trailing comma in object literal

0.2.0 ECMAScript 3 allows function expressions to have names, AST.JSFunctionExpression now reflects this

0.1.0 Simplified AST by removing JSElement and JSElementList components

0.0.3 Support for unicode in source. At the moment it only supports UTF8 encoding, does not recognise byte order marks or UTF-16/UTF-32

0.0.2 Multiline comments were processed in greedy form, now end of comment recognised properly. Thanks to Tony Morris for reporting this.

0.0.1 Initial release

EOF