github-linguist 7.30.0 → 8.0.0

data/lib/linguist/samples.rb

@@ -25,7 +25,14 @@ module Linguist
     # Hash of serialized samples object, uncached
     def self.load_samples
       serializer = defined?(Yajl) ? Yajl : JSON
-      serializer.load(File.read(PATH, encoding: 'utf-8'))
+      data = serializer.load(File.read(PATH, encoding: 'utf-8'))
+      # JSON serialization does not allow integer keys, we fix them here
+      for lang in data['centroids'].keys
+        fixed = data['centroids'][lang].to_a.map { |k,v| [k.to_i, v] }
+        data['centroids'][lang] = Hash[fixed]
+      end
+
+      data
     end
 
     # Public: Iterate over each sample.
@@ -106,6 +113,7 @@ module Linguist
         Classifier.train!(db, language_name, data)
       end
 
+      Classifier.finalize_train! db
       db['sha256'] = Linguist::SHA256.hexdigest(db)
 
       db

data/lib/linguist/sha256.rb

@@ -13,7 +13,7 @@ module Linguist
       digest = Digest::SHA256.new
 
       case obj
-      when String, Symbol, Integer
+      when String, Symbol, Integer, Float
         digest.update "#{obj.class}"
         digest.update "#{obj}"
       when TrueClass, FalseClass, NilClass

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: github-linguist
 version: !ruby/object:Gem::Version
-  version: 7.30.0
+  version: 8.0.0
 platform: ruby
 authors:
 - GitHub
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-06-07 00:00:00.000000000 Z
+date: 2024-09-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: cgi
@@ -248,6 +248,7 @@ files:
 - grammars/inline.graphql.res.json
 - grammars/inline.graphql.scala.json
 - grammars/inline.hbs.json
+- grammars/inline.peggy.json
 - grammars/inline.prisma.json
 - grammars/inline.template.json
 - grammars/liquid.injection.json
@@ -264,7 +265,6 @@ files:
 - grammars/markdown.hxml.codeblock.json
 - grammars/markdown.lean.codeblock.json
 - grammars/markdown.lean4.codeblock.json
-- grammars/markdown.mcfunction.codeblock.json
 - grammars/markdown.move.codeblock.json
 - grammars/markdown.plantuml.codeblock.json
 - grammars/markdown.prisma.codeblock.json
@@ -272,7 +272,7 @@ files:
 - grammars/markdown.rescript.codeblock.json
 - grammars/markdown.talon.codeblock.json
 - grammars/markdown.textproto.codeblock.json
-- grammars/mdx.LANGUAGE.codeblock.json
+- grammars/mdx.move.codeblock.json
 - grammars/objdump.x86asm.json
 - grammars/source.2da.json
 - grammars/source.4dm.json
@@ -348,6 +348,7 @@ files:
 - grammars/source.cache.cmake.json
 - grammars/source.cadence.json
 - grammars/source.cairo.json
+- grammars/source.cairo0.json
 - grammars/source.camlp4.ocaml.json
 - grammars/source.capnp.json
 - grammars/source.cds.json
@@ -398,6 +399,7 @@ files:
 - grammars/source.curlrc.json
 - grammars/source.curry.json
 - grammars/source.cwl.json
+- grammars/source.cylc.json
 - grammars/source.cypher.json
 - grammars/source.cython.json
 - grammars/source.d.json
@@ -428,6 +430,7 @@ files:
 - grammars/source.dosbox-conf.json
 - grammars/source.dot.json
 - grammars/source.dotenv.json
+- grammars/source.dune.json
 - grammars/source.dylan.json
 - grammars/source.earthfile.json
 - grammars/source.ebnf.json
@@ -529,6 +532,7 @@ files:
 - grammars/source.hx.type.json
 - grammars/source.hxml.json
 - grammars/source.hy.json
+- grammars/source.iCalendar.json
 - grammars/source.ice.json
 - grammars/source.icurry.json
 - grammars/source.ideal.json
@@ -580,6 +584,7 @@ files:
 - grammars/source.kusto.json
 - grammars/source.lark.json
 - grammars/source.lbnf.json
+- grammars/source.lcb.json
 - grammars/source.lcov.json
 - grammars/source.lean.json
 - grammars/source.lean.markdown.json
@@ -592,6 +597,7 @@ files:
 - grammars/source.lilypond.json
 - grammars/source.lisp.json
 - grammars/source.litcoffee.json
+- grammars/source.livecodescript.json
 - grammars/source.livescript.json
 - grammars/source.llvm.json
 - grammars/source.logos.json
@@ -670,6 +676,7 @@ files:
 - grammars/source.ninja.json
 - grammars/source.nit.json
 - grammars/source.nix.json
+- grammars/source.nr.json
 - grammars/source.nsis.json
 - grammars/source.nu.json
 - grammars/source.nunjucks.json
@@ -711,6 +718,7 @@ files:
 - grammars/source.pddl.happenings.json
 - grammars/source.pddl.json
 - grammars/source.pddl.plan.json
+- grammars/source.peggy.json
 - grammars/source.pegjs.json
 - grammars/source.pep8.json
 - grammars/source.perl.6.json
@@ -732,6 +740,7 @@ files:
 - grammars/source.postcss.json
 - grammars/source.postscript.json
 - grammars/source.pov-ray sdl.json
+- grammars/source.powerbuilder.json
 - grammars/source.powershell.json
 - grammars/source.praat.json
 - grammars/source.prisma.json
@@ -880,6 +889,7 @@ files:
 - grammars/source.varnish.vcl.json
 - grammars/source.vba.json
 - grammars/source.vbnet.json
+- grammars/source.vcard.json
 - grammars/source.velocity.html.json
 - grammars/source.velocity.json
 - grammars/source.verilog.json
@@ -963,6 +973,7 @@ files:
 - grammars/text.html.javadoc.json
 - grammars/text.html.js.json
 - grammars/text.html.jsp.json
+- grammars/text.html.jte.json
 - grammars/text.html.liquid.json
 - grammars/text.html.mako.json
 - grammars/text.html.markdown.astro.json
@@ -992,6 +1003,7 @@ files:
 - grammars/text.log.latex.json
 - grammars/text.marko.json
 - grammars/text.md.json
+- grammars/text.mdx.astro.codeblock.json
 - grammars/text.muse.json
 - grammars/text.openbsd-pkg.desc.json
 - grammars/text.plain.json
@@ -1086,7 +1098,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.5.9
+rubygems_version: 3.5.11
 signing_key:
 specification_version: 4
 summary: GitHub Language detection

data/grammars/markdown.mcfunction.codeblock.json

@@ -1 +0,0 @@
-{"scopeName":"markdown.mcfunction.codeblock","patterns":[{"include":"#codeblock.outer"}],"repository":{"codeblock.inner":{"contentName":"meta.embedded.block.mcfunction","begin":"(^|\\G)","while":"(^|\\G)(?!\\s*([`~]{3,})\\s*$)","patterns":[{"include":"source.mcfunction"}]},"codeblock.outer":{"name":"markup.fenced_code.block.markdown","begin":"(^|\\G)(\\s*)(`{3,}|~{3,})\\s*(?i:(mcfunction)((\\s+|:|\\{|\\?)[^`~]*)?$)","end":"(^|\\G)(\\2|\\s{0,3})(\\3)\\s*$","patterns":[{"include":"#codeblock.inner"}],"beginCaptures":{"3":{"name":"punctuation.definition.markdown"},"4":{"name":"fenced_code.block.language.markdown"},"5":{"name":"fenced_code.block.language.attributes.markdown"}},"endCaptures":{"3":{"name":"punctuation.definition.markdown"}}}}}

data/grammars/mdx.LANGUAGE.codeblock.json

@@ -1 +0,0 @@
-{"scopeName":"mdx.LANGUAGE.codeblock","patterns":[{"include":"#LANGUAGE-code-block"}],"repository":{"LANGUAGE-code-block":{"name":"markup.code.other.mdx","end":"(^|\\G)(\\2|\\s{0,3})(\\3)\\s*$","patterns":[{"contentName":"meta.embedded.block.LANGUAGE","begin":"(^|\\G)(\\s*)(.*)","while":"(^|\\G)(?!\\s*([`~]{3,})\\s*$)","patterns":[{}]}],"beginCaptures":{"3":{"name":"string.other.begin.code.fenced.mdx"},"4":{"name":"entity.name.function.mdx"}},"endCaptures":{"3":{"name":"string.other.end.code.fenced.mdx"}}}}}