All comments stored in the `comments` property of the `grammar` node.
Comments extracted only if the `extractComments` options set to `true` when you generate parser.
This property is object with mapping start offset of comment to comment object, that looks like:
```js
{
text: 'text in the comment, just after // or /* and before */',
multiline: true|false,// true for /**/ comments, false for // comments
location: location()
}
```
This commit adds support for '.js' files to be passed to the '-c', '--config' or '--extra-options-file' options on the CLI, allowing the developer to do some extra work before the parser is generated (if they wish), or dynamically set options based on the enviroment.
* Split 'consts' collection by content types into:
- literals: for literal expressions, like `"a"`
- classes: for character class expressions, like `[a]`
- expectations: for constants describing expected values when parse failed
- functions: for constants with function code
* Move any JavaScript code generation from 'generateBytecode' to 'generateJs'.
* Rename opcode 'MATCH_REGEXP' to 'MATCH_CLASS' (name reflects purpose, not implementation).
* Replace 'PUSH' opcode with 'PUSH_EMPTY_STRING' opcode because it is only used with empty strings
For compatibility with pre-0.11 plugins, this class lives on the same namespace as the orignal visitor helper and also exports a static method called 'build'.
Before this commit, the PEG.js parser always created the AST using a plain JavaSctript object, but allthough simple and effective for the job, this method sacrificies performance slightly.
From now on the parser shall call a Node creator. This should help with performance, as well as in the future move some AST helpers into the new AST functions.
Before there used to be some internal utility methods for arrays and objects, but as the code base moved to ES5+ use case only, these were removed in favour of native alternatives, but most of these were only beneficial for arrays.
This commit add's common utility methods for objects, and also exposes these as they can be used by plugin developer's on the PEG.js AST.
There seem's to be an edgecase where a spawned process (in this case node.exe it self) is opened and gulp straight away thinks it closed. At first I thought it was a problem with Gulp, but after fiddleing it seem's to be a Window's problem, so I just updated the `node()` spawner to work on all platforms by using Gulp's callback to create an async task. Problem solved :)
Before this commit error looks like (for input `start = break:'a'`)
> Expected "!", "$", "&", "(", "*", "+", ".", "/", "/*", "//", ";", "?", character class, code block, comment, end of line, identifier, literal, or whitespace but ":" found.
After this error looks like
> Label can't be a reserved word "break".
* Optimization: do not generate unreachable calls |peg$fail| and constants for it (local to the rule).
* Optimization: do not generate unreachable calls |peg$fail| and constants for it (non-local to rule).
* Remove stack manipulations from FAIL opcode and rename FAIL to EXPECT
* Always save the expected values regardless of the match result of expression
* Remove unnecessary check for match result in the `named` node
* Collect and process expectations from predicates
This commit ensures that all modules outside 'lib' are importing 'lib/peg.js' so that VS Code automatically gets 'lib/peg.d.ts'.
An alias module for 'lib/peg.js' called 'pegjs-dev' was made for test files at 'test/node_modules/pegjs-dev.js'
The css files used in the benchmark contained 2 files for IE that contained syntax errors. These files weren't actually used by the benchmark, but were still in the repository.
