New dataset on LCOM components

aravij commented 4 years ago

Issue #469

For each file in dataset we extract all classes and for each class extract all LCOM components. Information about class and component is stored in dataset in fields class_name and component_index.
Almost all patterns are now accept AST instead of file path. Exception:
- var_diff (P20) - not refactored
- "P28", "P9" - text based patterns
Only NCSS metric present in new dataset. Cognitive complexity needs refactoring.
Dataset calculation procedure changed. It now has a timout in 60 seconds for single file processing (can be changed with -t CLI parameter). All exceptions are printed right in the console. Failure of file processing does not stop the rest of computation. Progress bar is added.

Most of the files you will find here are patterns and their tests.

aravij commented 4 years ago

@acheshkov please merge

acheshkov commented 4 years ago

@rultor merge

acheshkov commented 4 years ago

@rultor merge

rultor commented 4 years ago

@rultor merge

@acheshkov OK, I'll try to merge now. You can check the progress of the merge here

rultor commented 4 years ago

@rultor merge

@acheshkov Done! FYI, the full log is here (took me 41min)

rultor commented 4 years ago

@rultor merge

@acheshkov @aravij The pull request is closed already, so I can't merge it

rultor commented 4 years ago

@rultor merge

@acheshkov I'm sorry, I don't understand you :( Check this page and try again please

cqfn / aibolit