This is the reference codebase for measuring the performance of Universal-ctags parsers.

We assume you may have enough storage space on your PC.

1. Get the input code for a parser for the language you are interested in with following command line:

   $ ./codebase clone <LANGUAGE>

   The following command lists available languages :

   $ ./codebase list-languages
   #           LANGUAGE    CODE
                      C    linux
                    C++    qtbase rocksdb
                     Go    buildah kubernetes
                   HTML    cockpit
             JavaScript    cockpit
               LdScript    linux
             ObjectiveC    gnustep-libs-base

2. Run Universal-ctags for the cloned code with following command line:

   $ [CTAGS=${where your ctags executable is}] ./codebase ctags <LANGUAGE> [<PROFILE>]

   codebase refers CTAGS environment variable to run ctags.

   You can run ctags with different option combination. We call such option combination PROFILE. The following command is for listing pre defined profiles:

   $ ./codebase list-profiles
   
           PROFILE         DESCRIPTION
           maximum0        Enables all extras, fields, and kinds
           minimum0        Disables all fields and extras.

   Results are displayed to your terminal. Tee'ed output goes to a file under results/ directory.

You have to write a .lcopy file and put it to lcopy.d directory. See lcopy.d/linux.lcopy as an example.

You have to write a .ctags file and put it to profile.d directory. A line started from "# @" is used as a description for the profile. You may wan to use --options-maybe to extend profile without modifying existing .ctags files.

Let's optimize ourt parsers! Masatake YAMATO <[email protected]>