srx-languagetool 0.2.0 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d929b515d7e87e260c01373e1f5d6c35bf35fa05f8fe29fea39ba18d0078e120
-  data.tar.gz: ee821f5cfb3b4fea3d347c3ef3583459da60d044c458aa5ad907b378d5eef963
+  metadata.gz: 9f3ac5da2f53e57018c6febfd004e1a6b2ba5059fc8911aaae52d263ff52dcbd
+  data.tar.gz: 54e97225366062cc30080346dd69c49b4bef9c86db54b8ff2c08f5e5a7920c69
 SHA512:
-  metadata.gz: 2733d28e0ed99d6b5fd0c15b30311aed10059c34c8c411de051901ca545ed6a4eb299e0695190817b45d6691f74b3d78215d6a5ce68c589b0776e25ca3c3dada
-  data.tar.gz: 4344f1b2dc867dcbffb8db8b2db95da458de718f0481814e8dd23965c2b3831199702333637d56017566388de5da0d745995276e0376c9e16e63dc9f15cd8d10
+  metadata.gz: d960af93ad5200cbe3201c22094ddf09f999b05e6a89b69079c47ff1e6c07a927977bd17fc28c8ec0cd7b7da46d4b0119fde7415268ea2c51374a01ae239a62f
+  data.tar.gz: b37d5e1ba0bc5fc789005250be57de3e4aa4036416e91dd10b787472bd69775cfe0090de5a2d9bb25fe402e2cc69078319ef0f6042e4311cbb7ddfb8eab217bd

data/.github/workflows/main.yml

@@ -13,7 +13,7 @@ jobs:
         ruby-version: 2.7.3
     - name: Install
       run: |
-        gem install bundler -v 2.2.16
+        gem install bundler -v 2.2.20
         bundle install
     - name: Type check
       run: bundle exec solargraph typecheck --level typed

data/.rubocop.yml

@@ -1,7 +1,7 @@
 inherit_from: .rubocop_todo.yml
 
 AllCops:
-  TargetRubyVersion: 2.4
+  TargetRubyVersion: 2.5
   SuggestExtensions: false
   NewCops: enable
 

data/CHANGELOG.md

@@ -1,5 +1,7 @@
 ## [Unreleased]
 
+- Update rules to LanguageTool 5.4
+
 ## [0.2.0] - 2021-04-15
 
 - Update rules to LanguageTool 5.3

data/Gemfile.lock

@@ -1,14 +1,14 @@
 PATH
   remote: .
   specs:
-    srx-languagetool (0.2.0)
+    srx-languagetool (0.3.0)
       srx (< 1.0)
 
 GEM
   remote: https://rubygems.org/
   specs:
     ast (2.4.2)
-    backport (1.1.2)
+    backport (1.2.0)
     benchmark (0.1.1)
     byebug (11.1.3)
     diff-lcs (1.4.4)
@@ -19,10 +19,10 @@ GEM
     kramdown-parser-gfm (1.1.0)
       kramdown (~> 2.0)
     minitest (5.14.4)
-    nokogiri (1.11.3-x86_64-darwin)
+    nokogiri (1.11.7-x86_64-darwin)
       racc (~> 1.4)
     parallel (1.20.1)
-    parser (3.0.1.0)
+    parser (3.0.1.1)
       ast (~> 2.4.1)
     racc (1.5.2)
     rainbow (3.0.0)
@@ -35,22 +35,23 @@ GEM
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.10.0)
     rspec-support (3.10.2)
-    rubocop (1.12.1)
+    rubocop (1.17.0)
       parallel (~> 1.10)
       parser (>= 3.0.0.0)
       rainbow (>= 2.2.2, < 4.0)
       regexp_parser (>= 1.8, < 3.0)
       rexml
-      rubocop-ast (>= 1.2.0, < 2.0)
+      rubocop-ast (>= 1.7.0, < 2.0)
       ruby-progressbar (~> 1.7)
       unicode-display_width (>= 1.4.0, < 3.0)
-    rubocop-ast (1.4.1)
-      parser (>= 2.7.1.5)
+    rubocop-ast (1.7.0)
+      parser (>= 3.0.1.1)
     ruby-progressbar (1.11.0)
-    solargraph (0.40.4)
-      backport (~> 1.1)
+    solargraph (0.42.3)
+      backport (~> 1.2)
       benchmark
       bundler (>= 1.17.2)
+      diff-lcs (~> 1.4)
       e2mmap
       jaro_winkler (~> 1.5)
       kramdown (~> 2.3)
@@ -81,4 +82,4 @@ DEPENDENCIES
   srx-languagetool!
 
 BUNDLED WITH
-   2.2.16
+   2.2.20

data/lib/srx/languagetool/version.rb

@@ -2,6 +2,6 @@
 
 module Srx
   module Languagetool
-    VERSION = '0.2.0'
+    VERSION = '0.3.0'
   end
 end

data/lib/srx/segment.srx

@@ -1524,11 +1524,11 @@
 <afterbreak>[A-Za-z0-9\-]+\.(com|net|org|info|de|es|edu|co|eu|nl|io|cn|uk|gov|biz|ca|tk|ru|br|jp|pl)(\.|\b)</afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\b(Drs|Art|Afr|Am|Ar|Br|Cie|Comp|Dhr|([Pp]rof\.)?[Dd]r|Em|Fa|Kon)\.\s</beforebreak>
+<beforebreak>\b(Drs|Art|Afr|Am|Ar|Br|Cie|Comp|Dhr|([Pp]rof\.)?[Dd]r|Em|Fa|Kon|Bros)\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\b(Mej|Mevr|Mgr|Mw|Ndl|Ned|Nl|No|Prof|Secr|Chr)\.\s</beforebreak>
+<beforebreak>\b(Mej|Mevr|Mgr|Mw|Ndl|Ned|Nl|No|Prof|Secr|Chr|Jac)\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -1668,12 +1668,28 @@
 <beforebreak>\bprof\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
+<rule break="no">
+<beforebreak>[.!?…][’'"]\s</beforebreak>
+<afterbreak>[a-z]</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>[.][.]\s</beforebreak>
+<afterbreak>[a-z]</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>SP[.]</beforebreak>
+<afterbreak>A</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>Warner Bros\.</beforebreak>
+<afterbreak>[a-z]</afterbreak>
+</rule>
 <rule break="yes">
-<beforebreak>[\.!?…][\u00BB\u2019\u201D\u203A\u00AB\p{Pe}\u0002¹²³]*\s</beforebreak>
+<beforebreak>[\.!?…][’'"\u00BB\u2019\u201D\u203A\u00AB\p{Pe}\u0002¹²³]*\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="yes">
-<beforebreak>[\.!?…]['"\u00BB\u2019\u201D\u203A\u00AB\p{Pe}\u0002]*</beforebreak>
+<beforebreak>[\.!?…][’'"\u00BB\u2019\u201D\u203A\u00AB\p{Pe}\u0002]*</beforebreak>
 <afterbreak>\p{Lu}[^\p{Lu}]</afterbreak>
 </rule>
 <rule break="yes">
@@ -1713,6 +1729,7 @@
 <beforebreak>[?!.]\s</beforebreak>
 <afterbreak>['"\u00BB\u2019\u201D\u203A\u00AB\p{Pe}\u0002][A-Z][a-z]</afterbreak>
 </rule>
+<!-- ##### end of Dutch #### -->
 </languagerule>
 <languagerule languagerulename="Slovak">
 <rule break="no">
@@ -4271,7 +4288,7 @@
 <rule break="no">
 <beforebreak>\b(бульв|г|д|доп|др|е|зам|Зам|и|им|инд|исп|Исп)\.\s</beforebreak>
 <afterbreak></afterbreak>
-</rule>    
+</rule>
 <rule break="no">
 <beforebreak>\b(англ|в|вв|га|гг|гл|гос|грн|дм|долл|е|ед)\.\s</beforebreak>
 <afterbreak>\p{Ll}</afterbreak>
@@ -4557,7 +4574,7 @@
 </rule>
 <!-- Abbreviations that cannot finish sentences-->
 <rule break="no">
-<beforebreak>\b(dc|(?iu)(n|Mr|C|Dr|Dra|E|Emm|Emma|Excm|Excma|Hble|I|Il·lm|Il·lma|Il·ltre|Im|Ima|Mgfc|Mgfca|Mn|R|Rev|Sr|Sra|Sres|Srs|St|Sta|a|abr|abs|acad|add|adj|adm|admdor|admdora|admtiu|admtiva|adv|ag|agl|agr|agron|agròn|aj|ajud|al|alim|amb|ampl|ant|ap|apmt|apnt|apr|aprox|apt|arm|arq|arqueol|arquit|assign|assoc|atm|aut|aux|av|b|batx|bda|bibl|bl|bnc|butll|bxs|c|calef|cartogr|cat|catedr|catol|cf|cia|cin|cint|circul|cit|climat|col|col·l|compt|cons|constr|cont|contr|conv|corp|corr|cpl|cpt|cró|ct|cte|ctra|cts|d|dept|derog|des|desp|dg|dip|disp|distr|div|dj|dl|doc|drec|ds|dt|dta|dte|dupl|dv|e|econ|ed|ef|entl|esc|esp|espf|esq|ex|exc|exp|exped|ext|f|fac|fca|febr|fig|figs|fra|gen|gov|gral|i|imp|impr|impt|inc|insp|inst|int|inv|j|jul|jur|jurispr|leg|llic|loc|ltda|làm|merc|mil·l|màx|mín|neg|nov|nre|núm|o|oct|op|p|pàg|pàgs|paq|par|pda|pg|pl|pobl|pol|ppda|ppt|pral|prev|prof|progr|prov|pta|ptes|ptge|pvt|pàg|quadr|quint|r|rbla|ref|reg|rev|secr|serv|sgt|sotsp|subsp|supl|supt|t|tel|telegr|tit|trad|trans|transcr|transf|trav|tripl|trv|tt|tèc|univ|urb|v|var|veg|venc|vid|vig|vocab|vs|x|àt|íd))\.\s</beforebreak>
+<beforebreak>\b(dc|(?iu)(n|Mr|C|Dr|Dra|Dra\. Ma|Sta\. Ma|E|Emm|Emma|Excm|Excma|Hble|I|Il·lm|Il·lma|Il·ltre|Im|Ima|Mgfc|Mgfca|Mn|R|Rev|Sr|Sra|Sres|Srs|St|Sta|a|abr|abs|acad|add|adj|adm|admdor|admdora|admtiu|admtiva|adv|ag|agl|agr|agron|agròn|aj|ajud|al|alim|amb|ampl|ant|ap|apmt|apnt|apr|aprox|apt|arm|arq|arqueol|arquit|assign|assoc|atm|aut|aux|av|b|batx|bda|bibl|bl|bnc|butll|bxs|c|calef|cartogr|cat|catedr|catol|cf|cia|cin|cint|circul|cit|climat|col|col·l|compt|cons|constr|cont|contr|conv|corp|corr|cpl|cpt|cró|ct|cte|ctra|cts|d|dept|derog|des|desp|dg|dip|disp|distr|div|dj|dl|doc|drec|ds|dt|dta|dte|dupl|dv|e|econ|ed|ef|entl|esc|esp|espf|esq|ex|exc|exp|exped|ext|f|fac|fca|febr|fig|figs|fra|gen|gov|gral|i|imp|impr|impt|inc|insp|inst|int|inv|j|jul|jur|jurispr|leg|llic|loc|ltda|làm|merc|mil·l|màx|mín|neg|nov|nre|núm|o|oct|op|p|pàg|pàgs|paq|par|pda|pg|pl|pobl|pol|ppda|ppt|pral|prev|prof|progr|prov|pta|ptes|ptge|pvt|pàg|quadr|quint|r|rbla|ref|reg|rev|secr|serv|sgt|sotsp|subsp|supl|supt|t|tel|telegr|tit|trad|trans|transcr|transf|trav|tripl|trv|tt|tèc|univ|urb|v|var|veg|venc|vid|vig|vocab|vs|x|àt|íd))\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <!-- Abbreviations that can finish sentences -->
@@ -4693,7 +4710,7 @@
 </rule>
 <!-- Abbreviations that cannot finish sentences-->
 <rule break="no">
-<beforebreak>\b(dc|(?iu)(n|Mr|C|Dr|Dra|E|Emm|Emma|Excm|Excma|Hble|I|Il·lm|Il·lma|Il·ltre|Im|Ima|Mgfc|Mgfca|Mn|R|Rev|Sr|Sra|Sres|Sras|Srs|St|Sta|a|abr|abs|acad|add|adj|adm|admdor|admdora|admtiu|admtiva|adv|ag|agl|agr|agron|agròn|aj|ajud|al|alim|amb|ampl|ant|ap|apmt|apnt|apr|aprox|apt|arm|arq|arqueol|arquit|assign|assoc|atm|aut|aux|av|b|batx|bda|bibl|bl|bnc|butll|bxs|c|calef|cartogr|cat|catedr|catol|cf|cia|cin|cint|circul|cit|climat|col|col·l|compt|cons|constr|cont|contr|conv|corp|corr|cpl|cpt|cró|ct|cte|ctra|cts|d|dept|derog|des|desp|dg|dip|disp|distr|div|dj|dl|doc|drec|ds|dt|dta|dte|dupl|dv|e|econ|ed|ef|entl|esc|esp|espf|esq|ex|exc|exp|exped|ext|f|fac|fca|febr|fig|figs|fra|gen|gov|gral|i|imp|impr|impt|inc|insp|inst|int|inv|j|jul|jur|jurispr|leg|llic|loc|ltda|làm|merc|mil·l|màx|mín|neg|nov|nre|núm|o|oct|op|p|pàg|pàgs|paq|par|pda|pg|pl|pobl|pol|ppda|ppt|pral|prev|prof|progr|prov|pta|ptes|ptge|pvt|pàg|quadr|quint|r|rbla|ref|reg|rev|secr|serv|sgt|sotsp|subsp|supl|supt|t|tel|telegr|tit|trad|trans|transcr|transf|trav|tripl|trv|tt|tèc|univ|urb|v|var|veg|venc|vid|vig|vocab|vs|x|àt|íd))\.\s</beforebreak>
+<beforebreak>\b(dc|(?iu)(n|[Aa]yto|Mr|C|Dr|Dra|E|Emm|Emma|Excm|Excma|Hble|I|Il·lm|Il·lma|Il·ltre|Im|Ima|Mgfc|Mgfca|Mn|R|Rev|Sr|Sra|Sres|Sras|Srs|St|Sta|a|abr|abs|acad|add|adj|adm|admdor|admdora|admtiu|admtiva|adv|ag|agl|agr|agron|agròn|aj|ajud|al|alim|amb|ampl|ant|ap|apmt|apnt|apr|aprox|apt|arm|arq|arqueol|arquit|assign|assoc|atm|aut|aux|av|b|batx|bda|bibl|bl|bnc|butll|bxs|c|calef|cartogr|cat|catedr|catol|cf|cia|cin|cint|circul|cit|climat|col|col·l|compt|cons|constr|cont|contr|conv|corp|corr|cpl|cpt|cró|ct|cte|ctra|cts|d|dept|derog|des|desp|dg|dip|disp|distr|div|dj|dl|doc|drec|ds|dt|dta|dte|dupl|dv|e|econ|ed|ef|entl|esc|esp|espf|esq|ex|exc|exp|exped|ext|f|fac|fca|febr|fig|figs|fra|gen|gov|gral|i|imp|impr|impt|inc|insp|inst|int|inv|j|jul|jur|jurispr|leg|llic|loc|ltda|làm|merc|mil·l|màx|mín|neg|nov|nre|núm|o|oct|op|p|pàg|pàgs|paq|par|pda|pg|pl|pobl|pol|ppda|ppt|pral|prev|prof|progr|prov|pta|ptes|ptge|pvt|pàg|quadr|quint|r|rbla|ref|reg|rev|secr|serv|sgt|sotsp|subsp|supl|supt|t|tel|telegr|tit|trad|trans|transcr|transf|trav|tripl|trv|tt|tèc|univ|urb|v|var|veg|venc|vid|vig|vocab|vs|x|àt|íd))\.\s</beforebreak>
 <afterbreak/>
 </rule>
 <rule break="no">
@@ -4879,7 +4896,7 @@
 <afterbreak>(und|oder|bis)\s</afterbreak>
 </rule>
 <!-- einige deutsche Monate, vor denen eine Zahl erscheinen kann,
-         ohne dass eine Satzgrenze erkannt wird 
+         ohne dass eine Satzgrenze erkannt wird
          (z.B. "am 13. Dezember" -> keine Satzgrenze) -->
 <rule break="no">
 <beforebreak>\d+\.\s</beforebreak>
@@ -4910,7 +4927,7 @@
 </rule>
 <!-- German abbreviations -->
 <rule break="no">
-<beforebreak>\b(Mag|mtl|versch|d|Übers|usw|Bzw|bzw|Ab[hkst]|abzgl|Abzw|ahd|Akk|aktual|allg|alltagsspr|altdt|alttest|amerikan|Anh|Ank|Anm|Art|autom|Auftragsnr|Az|Bat|bayr|Bde?|bearb|Bed|Bem|bes|bez|Bez|Bhf|bspw|btto|bw)\.\s</beforebreak>
+<beforebreak>\b(ggü|Mag|mtl|versch|d|Übers|usw|Bzw|bzw|Ab[hkst]|abzgl|Abzw|ahd|Akk|aktual|allg|alltagsspr|altdt|alttest|amerikan|Anh|Ank|Anm|Art|autom|Auftragsnr|Az|Bat|bayr|Bde?|bearb|Bed|Bem|bes|bez|Bez|Bhf|bspw|btto|bw)\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -5180,7 +5197,12 @@
 </rule>
 <rule break="no"><!-- Subdomains without "www." (e.g. foo.MyDomain.com)-->
 <beforebreak>\b[A-Za-z0-9\-]+\.</beforebreak>
-<afterbreak>[A-Za-z0-9\-]+\.(com|net|org|info|de|es|edu|co|eu|nl|io|cn|uk|gov|biz|ca|tk|ru|br|jp|pl)(\.|\b)</afterbreak>
+<afterbreak>[A-Za-z0-9\-]+\.(fr|com|net|org|info|de|es|edu|co|eu|nl|io|cn|uk|gov|biz|ca|tk|ru|br|jp|pl)(\.|\b)</afterbreak>
+</rule>
+<rule break="no">
+<!-- gaffa.org -->
+<beforebreak>\b[A-Za-z0-9\-]+\.</beforebreak>
+<afterbreak>[A-Za-z]{2,5}(\.|\b)</afterbreak>
 </rule>
 <!-- French abbreviations -->
 <rule break="no">
@@ -5192,7 +5214,7 @@
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\b(apr|ave?|boul|Mr?|Mrs|MM?)\.\s</beforebreak>
+<beforebreak>\b(apr|ave?|boul|Mr?|Mrs|MM?|Mlle)\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -5810,7 +5832,7 @@
 <afterbreak>['"«¡¿\p{Ps}\p{Pi}]?\p{Lu}\p{Ll}*</afterbreak>
 </rule>
 </languagerule>
-<languagerule languagerulename="Japanese">
+<languagerule languagerulename="Ideographic">
 <rule break="no">
 <beforebreak>[:]+[\p{Pe}\p{Pf}\p{Po}"-[\u002C\u003A\u003B\u055D\u060C\u061B\u0703\u0704\u0705\u0706\u0707\u0708\u0709\u07F8\u1363\u1364\u1365\u1366\u1802\u1804\u1808\u204F\u205D\u3001\uA60D\uFE10\uFE11\uFE13\uFE14\uFE50\uFE51\uFE54\uFE55\uFF0C\uFF1A\uFF1B\uFF64]]*</beforebreak>
 <afterbreak>\s+\P{Lu}</afterbreak>
@@ -5878,7 +5900,7 @@
 </rule>
 <!-- Abbreviations that cannot finish sentences-->
 <rule break="no">
-<beforebreak>\b(a|Ab|abrev|absol|acad|Açor|A\. ?D|add|adj|adv|advers|Aeron|afér|Agric|Álg|aprox|art|Artilh|auxil|av|Av)\.\s?</beforebreak>
+<beforebreak>\b(a|Ab|abrev|absol|acad|Açor|A\. ?D|add|adj|adv|advers|Aeron|afér|Agric|Álg|aprox|[Aa]rts?|Artilh|auxil|av|Av)\.\s?</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -6606,10 +6628,8 @@
 <languagemap languagepattern="(UK|uk).*" languagerulename="Ukrainian"></languagemap>
 <languagemap languagepattern="(BE|be).*" languagerulename="Belarusian"></languagemap>
 <languagemap languagepattern="(GL|gl).*" languagerulename="Galician"></languagemap>
-<languagemap languagepattern="(JA|ja).*" languagerulename="Japanese"></languagemap>
-<!-- Modification: Japanese rules work well for Chinese; keep original name to
-     minimize diff -->
-<languagemap languagepattern="(ZH|zh).*" languagerulename="Japanese"></languagemap>
+<languagemap languagepattern="(JA|ja).*" languagerulename="Ideographic"></languagemap>
+<languagemap languagepattern="(ZH|zh).*" languagerulename="Ideographic"></languagemap>
 <languagemap languagepattern="(BR|br).*" languagerulename="Breton"></languagemap>
 <languagemap languagepattern="(PT|pt).*" languagerulename="Portuguese"></languagemap>
 <languagemap languagepattern="(IT|it).*" languagerulename="Italian"></languagemap>

data/srx-languagetool.gemspec

@@ -11,7 +11,7 @@ Gem::Specification.new do |spec|
   spec.summary       = 'SRX segmentation rules from LanguageTool'
   spec.homepage      = 'https://github.com/amake/srx-languagetool-ruby'
   spec.license       = 'LGPL-2.1'
-  spec.required_ruby_version = Gem::Requirement.new('>= 2.4.0')
+  spec.required_ruby_version = Gem::Requirement.new('>= 2.5.0')
 
   spec.metadata['homepage_uri'] = spec.homepage
   spec.metadata['source_code_uri'] = 'https://github.com/amake/srx-languagetool-ruby.git'

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: srx-languagetool
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors:
 - Aaron Madlon-Kay
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2021-04-15 00:00:00.000000000 Z
+date: 2021-06-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: srx
@@ -149,7 +149,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: 2.4.0
+      version: 2.5.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="