-
Features:
- Support for regular expressions in list of pages to exclude.
-
Other:
- Update dependencies
-
Bugs:
- #37: Fix bug where relative protocol URIs caused crawler to escape to other domains.
- #40: Fix length check of array for load errors.
- Fix bug where load error warnings where not shown for local files.
-
Other:
- Update licence information.
- Bugs:
- Fix bug where crawler reached maximum call stack size on large sites.
- Fix so that links pointing to URLs with other protocols than http(s) are not followed.
-
Features:
- #30: Option for setting User agent header. Any header can now be set in configuration.
-
Bugs:
- #32: Stops unexpectedly without any error message.
- #29: Follows links to binaries. Now checks content type of response.
-
Other:
- Better error handling and reporting (#9), including:
- Warning at the end if any resources did not load.
- Timeouts.
- Increased default timeout from 4 seconds to 10 seconds.
- "nosummary" option renamed to "silent".
- Better error handling and reporting (#9), including:
-
Refactoring that has implications for customization:
- Added error message parameter to logger.
- Renamed output.standard to output.report, and changed/renamed the "summary" parameter to "silent".
- Add load_error property to the result object, with a list of all errors on loading CSS and HTML.
- Bugs:
- #26: Skip mailto:url
- Bugs:
- Whitelisted rules not output correctly (#25).
- Bugs:
- Updated help text, which had wrong name for the config file.
-
Features:
- Added option --full, to show full details for all rules.
- Removed option --used (functionality is covered by --full).
- New output report, with more information.
-
Refactoring:
- Result object has new format. This has implications for custom output functions, as well as library use.
- Added CSS selector position(s) in source file to result object.
- Features:
- Added possibility to use custom logger and custom result handler.
- Bugs/issues fixed:
- Parameters are now kept for links in pages.include (but removed from links found by the crawler).
- Request pooling disabled, to avoid hangs (see request/request#465).
- Added (configurable) timeout for requests.
- Improved logging and error handling.
- Handles links without href attributes.
-
Features:
- Follow links/crawl (#3). When running from command line, crawling is the new default.
-
Other:
- Replace jsdom and jQuery with Cheerio.
- Use request module instead of http/https (except in helpers/login.js).
- Major refactoring, which includes introducing the promises pattern for managing callbacks (using Q).
- Bug fixes:
- Make sure @keyframe and @font-face does not crash uCSS (issues #15, #16)
- Bug fixes:
- Fix loading of config from relative paths.
- Changes:
- Change config file format from JSON to Node module NOTE: This breaks backwards compability.
- Bug fixes:
- Fix custom login functions to work.
- Features:
- Adds support for fetching CSS over HTTP.
- Bug/stability fixes:
- Now handles grouped selectors (issue #11)
- Now tests selectors inside e.g. a media query (issue #10)
- Now properly reads selectors after e.g. a media query (issue #10)
- Dependency changes:
- Adds cssom CSS parser, to properly extract selectors
- Stability fixes:
- Ignore pseudo part of a selector.
- Ignore at-rules (like @media and @font-face).
- First release.