This project adheres to Semantic Versioning.
- Update dependency versions (Removes security vulnerability from lodash@3)
- Multiline comment support for PHP (Sean Snyder)
- Multiline comment support for Ruby and Python
- Apply JS-Standard Coding-Style
- Change method to detect whether standardjs or Thought are installed.
- Code-Context-Functions for Handlebars
- Fix code-context detection for object properties that are functions. (i.e. "key: function(a,b) {")
- C-style multiline-comments splitted into multiple regexes (
/**, and/*)./**is marked as "used for apidocs" - Add
infoproperty to the output of.regex, which contains additional information (so far only theapidoc: trueproperty.
.singleLineCommentis no longer and array of strings (['#']but an array of objects ([ { start: '#' } ]) in order to allow them to be marked as "used for apidocs". Definitions can be mixed in the languages-source-files, but will always appear in the second form in the compiled database.
- Add function
.codeContextto return code-context parser for different languages.
- Multi-line-comments for Less
- Add
.bashas file-extension for shell-scripts.
- Remove "function()" from generated database files. This lead to an error in the test cases.
- Consecutive indented lines of single-line-comments are now recognized as a single comment by regexes
- Next-line-of-code is not matched anymode. The client has to extract the code itself. The line-number is still computed.
- Some comment delimiters are now regexes rather than strings
- Middle-prefix for C-like languages now allows a preceeding whitespace.
- Method
.regexcreates regular expressions that match comments
- Only strings are used to determine that language from the file extension
/php\d?/is replaced byphp,php3,...