Merge remote-tracking branch 'origin/master' into strategies

* origin/master: (165 commits)
  Add F# and GLSL samples.  Add Forth and GLSL extension .fs. Add heuristic to disambiguate between F#, Forth, and GLSL.
  byebug requires ruby 2.0
  Remove test for removed extension
  Fix typo in test
  add rake interpreter
  add python3 interpreter
  Remove old wrong_shebang.rb sample
  Add byebug
  Link to Lightshow in CONTRIBUTING.md
  Switch to a better F# grammar
  Bump Rugged again
  Checkout the master for testing
  Rugged 0.22.0b3
  Reordering
  Bump version to 4.0.3
  Add some docs for tm_scope
  Change NONE to none
  Checking other case for Chart.jS
  Test that all languages have grammars
  Fix RHTML's tm_scope
  ...

Conflicts:
	lib/linguist/language.rb
This commit is contained in:
Brandon Keepers
2014-11-27 10:52:44 -05:00
129 changed files with 7316 additions and 622 deletions

View File

@@ -4,7 +4,7 @@ module Linguist
ACTIVE = true
def self.call(blob, languages)
find_by_heuristics(blob.data, langauges)
find_by_heuristics(blob.data, langauges.map(&:name))
end
# Public: Given an array of String language names,
@@ -17,23 +17,34 @@ module Linguist
# Returns an array of Languages or []
def self.find_by_heuristics(data, languages)
if active?
result = []
if languages.all? { |l| ["Perl", "Prolog"].include?(l) }
result = disambiguate_pl(data, languages)
result = disambiguate_pl(data)
end
if languages.all? { |l| ["ECL", "Prolog"].include?(l) }
result = disambiguate_ecl(data, languages)
result = disambiguate_ecl(data)
end
if languages.all? { |l| ["IDL", "Prolog"].include?(l) }
result = disambiguate_pro(data, languages)
result = disambiguate_pro(data)
end
if languages.all? { |l| ["Common Lisp", "OpenCL"].include?(l) }
result = disambiguate_cl(data, languages)
result = disambiguate_cl(data)
end
if languages.all? { |l| ["Hack", "PHP"].include?(l) }
result = disambiguate_hack(data)
end
if languages.all? { |l| ["Scala", "SuperCollider"].include?(l) }
result = disambiguate_sc(data, languages)
result = disambiguate_sc(data)
end
if languages.all? { |l| ["AsciiDoc", "AGS Script"].include?(l) }
result = disambiguate_asc(data, languages)
result = disambiguate_asc(data)
end
if languages.all? { |l| ["FORTRAN", "Forth"].include?(l) }
result = disambiguate_f(data)
end
if languages.all? { |l| ["F#", "Forth", "GLSL"].include?(l) }
result = disambiguate_fs(data)
end
return result
end
@@ -43,28 +54,37 @@ module Linguist
# We want to shortcut look for Objective-C _and_ now C++ too!
#
# Returns an array of Languages or []
def self.disambiguate_c(data, languages)
def self.disambiguate_c(data)
matches = []
matches << Language["Objective-C"] if data.include?("@interface")
matches << Language["C++"] if data.include?("#include <cstdint>")
if data.include?("@interface")
matches << Language["Objective-C"]
elsif data.include?("#include <cstdint>")
matches << Language["C++"]
end
matches
end
def self.disambiguate_pl(data, languages)
def self.disambiguate_pl(data)
matches = []
matches << Language["Prolog"] if data.include?(":-")
matches << Language["Perl"] if data.include?("use strict")
if data.include?("use strict")
matches << Language["Perl"]
elsif data.include?(":-")
matches << Language["Prolog"]
end
matches
end
def self.disambiguate_ecl(data, languages)
def self.disambiguate_ecl(data)
matches = []
matches << Language["Prolog"] if data.include?(":-")
matches << Language["ECL"] if data.include?(":=")
if data.include?(":-")
matches << Language["Prolog"]
elsif data.include?(":=")
matches << Language["ECL"]
end
matches
end
def self.disambiguate_pro(data, languages)
def self.disambiguate_pro(data)
matches = []
if (data.include?(":-"))
matches << Language["Prolog"]
@@ -74,7 +94,7 @@ module Linguist
matches
end
def self.disambiguate_ts(data, languages)
def self.disambiguate_ts(data)
matches = []
if (data.include?("</translation>"))
matches << Language["XML"]
@@ -84,21 +104,34 @@ module Linguist
matches
end
def self.disambiguate_cl(data, languages)
def self.disambiguate_cl(data)
matches = []
matches << Language["Common Lisp"] if data.include?("(defun ")
matches << Language["OpenCL"] if /\/\* |\/\/ |^\}/.match(data)
if data.include?("(defun ")
matches << Language["Common Lisp"]
elsif /\/\* |\/\/ |^\}/.match(data)
matches << Language["OpenCL"]
end
matches
end
def self.disambiguate_r(data, languages)
def self.disambiguate_r(data)
matches = []
matches << Language["Rebol"] if /\bRebol\b/i.match(data)
matches << Language["R"] if data.include?("<-")
matches
end
def self.disambiguate_sc(data, languages)
def self.disambiguate_hack(data)
matches = []
if data.include?("<?hh")
matches << Language["Hack"]
elsif /<?[^h]/.match(data)
matches << Language["PHP"]
end
matches
end
def self.disambiguate_sc(data)
matches = []
if (/\^(this|super)\./.match(data) || /^\s*(\+|\*)\s*\w+\s*{/.match(data) || /^\s*~\w+\s*=\./.match(data))
matches << Language["SuperCollider"]
@@ -109,12 +142,34 @@ module Linguist
matches
end
def self.disambiguate_asc(data, languages)
def self.disambiguate_asc(data)
matches = []
matches << Language["AsciiDoc"] if /^=+(\s|\n)/.match(data)
matches
end
def self.disambiguate_f(data)
matches = []
if /^: /.match(data)
matches << Language["Forth"]
elsif /^([c*][^a-z]| subroutine\s)/i.match(data)
matches << Language["FORTRAN"]
end
matches
end
def self.disambiguate_fs(data)
matches = []
if /^(: |new-device)/.match(data)
matches << Language["Forth"]
elsif /^(#light|import|let|module|namespace|open|type)/.match(data)
matches << Language["F#"]
elsif /^(#include|#pragma|precision|uniform|varying|void)/.match(data)
matches << Language["GLSL"]
end
matches
end
def self.active?
!!ACTIVE
end