github / linguist

#5149)

* Use JSON if Yajl doesn't exist to boost perf on dependent repos.

* Also require 'json' if Yajl isn't available.

Co-authored-by: Colin Seymour <colin@github.com>

* Move issue template to new directory and rename to reflect purpose

* Update bug report template

* Add feature request template

* Add issue template config

* Move PR template to new dir structure

* Use @Alhadis's words

Co-authored-by: John Gardner <gardnerjohng@gmail.com>

* Split adding language to own template

* Add comment

* Add section for grammar URL

* More of Alhadis's words

Co-authored-by: John Gardner <gardnerjohng@gmail.com>

Co-authored-by: Colin Seymour <colin@github.com>

Not all of these cases are checking behaviors that make sense, but they
should serve as better quirk documentation for the tokenizer, and help
understanding the impact of tokenizer changes.

Co-authored-by: Colin Seymour <colin@github.com>

…bmodule (#5133)

* Remove babel-sublime grammar

* Move .jsx to JavaScript

* Move JSX sample to JavaScript

* Remove babel-sublime section

* Remove babel-sublime license

* Fix extension order

* Remove ref to old babel-sublime grammar

* Add pedantic test for heuristics exhaustiveness

Tests that there is a test in test_heuristics.rb for every extension
defined in heuristics.rb.

Add missing heuristic tests and add explicit checks for languages that
miss a rule.

Fixes #4207.

* Correct test order

I messed this up when resolving the merge conflict.

* Remove whitespace changes

These also snuck in during the conflict resolution

Co-authored-by: Colin Seymour <colin@github.com>

Move `delphi` and `objectpascal` aliases to Pascal instead of Component
Pascal. Both aliases are used for Object Pascal dialects, which is a
Pascal extension, not derived from Component Pascal.

Co-authored-by: Colin Seymour <colin@github.com>
Co-authored-by: John Gardner <gardnerjohng@gmail.com>

Added heuristic for M4Sugar with M4 fallback.

Added samples:

- M4/translit2.m4 BSD https://github.com/freebsd/freebsd/blob/d551da60d42039156f003de6644e9e147ed167a3/usr.bin/m4/tests/translit2.m4
- M4/fibo.m4 BSD https://github.com/openbsd/src/blob/d714ed3db694a238d1e0a5018a7fb156af9527f3/regress/usr.bin/m4/fibo.m4
- M4/postscript.m4 LPPL 1.3c http://mirrors.ctan.org/graphics/circuit_macros/postscript.m4- M4/translit2.m4 https://github.com/freebsd/freebsd/blob/d551da60d42039156f003de6644e9e147ed167a3/usr.bin/m4/tests/translit2.m4

Co-authored-by: Colin Seymour <colin@github.com>

Charlock_holmes is struggling to build on Ruby 3

* Drecrement number of expected errors

* Version v7.12.2

Co-authored-by: Colin Seymour <colin@github.com>

* Update ActionScript regex heuristic

* Update lib/linguist/heuristics.yml

Co-authored-by: Nathan Witmer <nathan@zerowidth.com>

Co-authored-by: Nathan Witmer <nathan@zerowidth.com>

* Switch to ruby/setup-ruby for CI

* Use versions without .x

* No need to install gems again

* Move grammar checkout to workflow

* Remove redundant script

* feat(grammar, languages): add text.browserslist

* feat(examples): add examples files

* feat(language-id): generate language-id

* fix(example): change filename

* fix(tests): delete aliases

* fix(languages): add filename: browserslist

* fix(language): add extension, relocate filename

* fix(langiage): revert filename

* fix(tests): add filenames dir in /sample )

* refactor: resolve PR comments by Alhadis

* fix(tests): add core.ignorecase to lacal git config

* feat(color): add color property for language

* Delete .browserslistrc

* Delete browserslist

Co-authored-by: Stepan Polevshchikov <spolevshchikov@ozon.ru>

* Update link to new Github documentation

* Update README.md

* Update link to new GitHub documentation

* Update Julia grammar

* Update cached Julia license

* Version v7.12.1

* Update all grammars

* Perl6 grammar scope is now source.raku

* Version v7.12.0

* Update license cache

* Add devcontainer.json to JSONC

This file is VS Code specific & allows comments: https://code.visualstudio.com/docs/remote/devcontainerjson-reference

* Add sample devcontainer.json

Co-authored-by: Colin Seymour <colin@github.com>

* Add ReScript programming language

* Remove .resi extension from ReScript language

* Add heuristics for ReScript programming language

* Add .res extension to XML and add sample

Co-authored-by: Patrick Stapfer <ryyppy@users.noreply.github.com>
Co-authored-by: Colin Seymour <colin@github.com>

* add LEEx file extension to HTML+EEX

* add LEEx sample file

Co-authored-by: Colin Seymour <colin@github.com>

Co-authored-by: Colin Seymour <colin@github.com>

* Revise some pattern for avoding Coq/Verilog misdetection

* Add a sample Verilog file used to be misclassified as Coq.

Co-authored-by: cyan <cyan@0x00a1e9.dev>

Co-authored-by: Colin Seymour <colin@github.com>

Co-authored-by: Colin Seymour <colin@github.com>

* add additional LabVIEW file types and added samples directory with multiple sample files for each type

* added LabVIEW extensions with examples

* removed binary file types and examples leaving only .lvproj and .lvlib

Co-authored-by: Colin Seymour <colin@github.com>

Co-authored-by: Colin Seymour <colin@github.com>

Some heuristic tests were not running for any sample:

- *.sol tests had incorrect paths in a glob expression, which expanded
to an empty list.

- *.h tests for C did not reference any sample.

Co-authored-by: Colin Seymour <colin@github.com>

In the tokenizer, all lexer actions created intermediate string copies.
These are unnecessary, since they are later copied again when creating
the final Ruby strings.

Removed all intermediate copies and unnecessary strlen calls (string
length is already provided by flex in the `yyleng` variable).

Now Ruby strings are created directly by macros used in lexer actions:
FEED/FEED1/FEED2/FEED_SGML/FEED_SHEBANG. Some other code such as
`strndup.c`, token types, etc, was rendered obsolete in the process and
removed.

Depending on the language, tokenization time is decreased between 20%
and 40%.

Re-generate tokenizer with flex 2.6.4 (current version on Ubuntu 20.04
and Homebrew).

* Update accuracy of Actionscript heuristic

* add new test cases

* optimize actionscript heuristic and remove angelscript fallback

* Require at least one of `intrinsics` or `extends`

* Fix YAML indentation

* Update heuristic test

Co-authored-by: Alhadis <gardnerjohng@gmail.com>
Co-authored-by: Colin Seymour <colin@github.com>