srx-languagetool 0.13.0 → 0.14.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 037ab00a25a87ba79d5bb760cc4d292aad1289d792edea55d8b05034a2ac0f5f
-  data.tar.gz: 760147c7ba571a06943244728a709a02bedcb320244342ba38694a1148091a7b
+  metadata.gz: a71886af9137758228b7fdbb92b1e0c9e8f64c1b3456597660186b43379bf098
+  data.tar.gz: 621f91e2be6bc34a564259f61953b56e02cb57aeaa7c68fdea2770f39ee2631a
 SHA512:
-  metadata.gz: 15184b2b378ab3c2f0667ac6168894583959833403b4181c223eb5cd51a43ab8286c2f9110c1769fd5fb9898aa386cff525e6c170e48bc6aa2247d2cf171126c
-  data.tar.gz: e30f82724c3bcb264bde6cf2a99f3ea09eb6815f43a9609b71b2b4bbb1fd2e662cbb1abee2bdfd52035130fee9ed1118fe545f48efb61b103c27acaf29773e2e
+  metadata.gz: 2dd8b533adbd82f274f492b1243330554a35a8e98d88af8a7790d2aaef6b2ebb85c0a3f0fc4bb87964b9a0135a5efd65554dbdae0ab8569e0838dbb2e5f504ef
+  data.tar.gz: d5270d27a3dba9622b84ebfb1a6874c5eaeb655cdc933e61001ae9827fbfd0158bd7ebe7da4491145d8acc1273501aa7aee2c687438f957026a3c16c5273c2a3

data/.github/workflows/main.yml

@@ -10,10 +10,10 @@ jobs:
     - name: Set up Ruby
       uses: ruby/setup-ruby@v1
       with:
-        ruby-version: 2.7.7
+        ruby-version: 3.2.3
     - name: Install
       run: |
-        gem install bundler -v 2.4.10
+        gem install bundler -v 2.5.7
         bundle install
     - name: Type check
       run: bundle exec solargraph typecheck --level typed

data/.ruby-version

@@ -1 +1 @@
-2.7.7
+3.2.3

data/CHANGELOG.md

@@ -1,5 +1,9 @@
 ## [Unreleased]
 
+## [0.14.0] - 2024-09-27
+
+- Update rules to LanguageTool 6.5
+
 ## [0.13.0] - 2024-03-29
 
 - Update rules to LanguageTool 6.4

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    srx-languagetool (0.13.0)
+    srx-languagetool (0.14.0)
       srx (< 1.0)
 
 GEM
@@ -13,45 +13,44 @@ GEM
     byebug (11.1.3)
     diff-lcs (1.5.1)
     e2mmap (0.1.0)
-    jaro_winkler (1.5.6)
-    json (2.7.1)
+    jaro_winkler (1.6.0)
+    json (2.7.2)
     kramdown (2.4.0)
       rexml
     kramdown-parser-gfm (1.1.0)
       kramdown (~> 2.0)
     language_server-protocol (3.17.0.3)
-    minitest (5.22.3)
-    nokogiri (1.15.6-x86_64-darwin)
+    minitest (5.25.1)
+    nokogiri (1.16.7-x86_64-darwin)
       racc (~> 1.4)
-    parallel (1.24.0)
-    parser (3.3.0.5)
+    parallel (1.26.3)
+    parser (3.3.5.0)
       ast (~> 2.4.1)
       racc
-    racc (1.7.3)
+    racc (1.8.1)
     rainbow (3.1.1)
-    rake (13.1.0)
+    rake (13.2.1)
     rbs (2.8.4)
-    regexp_parser (2.9.0)
+    regexp_parser (2.9.2)
     reverse_markdown (2.1.1)
       nokogiri
-    rexml (3.2.6)
-    rspec-expectations (3.13.0)
+    rexml (3.3.7)
+    rspec-expectations (3.13.3)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.13.0)
     rspec-support (3.13.1)
-    rubocop (1.62.1)
+    rubocop (1.66.1)
       json (~> 2.3)
       language_server-protocol (>= 3.17.0)
       parallel (~> 1.10)
       parser (>= 3.3.0.2)
       rainbow (>= 2.2.2, < 4.0)
-      regexp_parser (>= 1.8, < 3.0)
-      rexml (>= 3.2.5, < 4.0)
-      rubocop-ast (>= 1.31.1, < 2.0)
+      regexp_parser (>= 2.4, < 3.0)
+      rubocop-ast (>= 1.32.2, < 2.0)
       ruby-progressbar (~> 1.7)
       unicode-display_width (>= 2.4.0, < 3.0)
-    rubocop-ast (1.31.2)
-      parser (>= 3.3.0.4)
+    rubocop-ast (1.32.3)
+      parser (>= 3.3.1.0)
     ruby-progressbar (1.13.0)
     solargraph (0.50.0)
       backport (~> 1.2)
@@ -71,15 +70,16 @@ GEM
       yard (~> 0.9, >= 0.9.24)
     srx (0.6.0)
       nokogiri (~> 1.11)
-    thor (1.3.1)
-    tilt (2.3.0)
-    unicode-display_width (2.5.0)
-    yard (0.9.36)
+    thor (1.3.2)
+    tilt (2.4.0)
+    unicode-display_width (2.6.0)
+    yard (0.9.37)
 
 PLATFORMS
   x86_64-darwin-20
   x86_64-darwin-21
   x86_64-darwin-22
+  x86_64-darwin-23
 
 DEPENDENCIES
   byebug
@@ -91,4 +91,4 @@ DEPENDENCIES
   srx-languagetool!
 
 BUNDLED WITH
-   2.4.10
+   2.5.7

data/lib/srx/languagetool/version.rb

@@ -2,6 +2,6 @@
 
 module Srx
   module Languagetool
-    VERSION = '0.13.0'
+    VERSION = '0.14.0'
   end
 end

data/lib/srx/segment.srx

@@ -1107,6 +1107,14 @@
 </rule>
 </languagerule>
 <languagerule languagerulename="English">
+<rule break="no"><!-- https://www.seven.one/ -->
+<beforebreak>\b[Se]even\.</beforebreak>
+<afterbreak>[Oo]ne\b</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\b[1-9]\.[\s\u00A0]</beforebreak>
+<afterbreak>[a-z]</afterbreak>
+</rule>
 <rule break="no">
 <beforebreak>[\u00A0\s]</beforebreak>
 <afterbreak>\n</afterbreak>
@@ -1164,7 +1172,7 @@
 <afterbreak>D\.?</afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\b([Aa]vg|[Ee]d|pp|[Vv]iz|i\.?[\s\u00A0]*e|[Vvol]|[Rr]col|maj|Lt|[Ff]ig|[Ff]igs|[Vv]iz|[Vv]ols|[Aa]pprox|[Aa]cq|[Ii]ncl?|[Ee]xcl|[Aa]cc|Pres|[Cc]orp|[Ee]x|[Cc]onn|[Dd]ept|[Ll]tda|[Mm]in|[Mm]ax|[Gg]ovt|[Rr]etd|Ing|lb|lbf|ft|c\.?[\s\u00A0]*f|vs|dia|lbs|\d+-(:?oz|kc|in|h[rp]|ml)|M?sec)\.[\s\u00A0]</beforebreak>
+<beforebreak>\b([Aa]vg|[Ee]d|pp|[Vv]iz|i\.?[\s\u00A0]*e|[Vvol]|[Rr]col|maj|Lt|[Ff]ig|[Ee]xt|[Ff]igs|[Vv]iz|[Vv]ols|[Aa]pprox|[Aa]cq|[Ii]ncl?|[Ee]xcl|[Aa]cc|Pres|[Cc]orp|[Ee]x|[Cc]onn|[Dd]ept|[Ll]tda|[Mm]in|[Mm]ax|[Gg]ovt|[Rr]etd|Ing|lb|lbf|ft|c\.?[\s\u00A0]*f|vs|dia|lbs|\d+-(:?oz|kc|in|h[rp]|ml)|M?sec)\.[\s\u00A0]</beforebreak>
 <afterbreak>[^\p{Lu}]|I</afterbreak>
 </rule>
 <rule break="no">
@@ -4753,6 +4761,19 @@
 <beforebreak>\.\[\d+\][\s\u00A0]</beforebreak>
 <afterbreak></afterbreak>
 </rule>
+<!-- unknown abbreviations inside parentheses -->
+<rule break="no">
+<beforebreak>\([^\)]*\.[\s\u00A0]</beforebreak>
+<afterbreak>[^\)\r\n]*\)</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\[[^\]]*\.[\s\u00A0]</beforebreak>
+<afterbreak>[^\]\r\n]*\]</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\{[^\}]*\.[\s\u00A0]</beforebreak>
+<afterbreak>[^\}\r\n]*\}</afterbreak>
+</rule>
 <!-- initials: A. C. Jones. Problem: [...] d'Alfons I. Ell era [...] -->
 <rule break="no">
 <beforebreak>\b[A-ZÀÉÈÍÓÒÚ]\.[\s\u00A0]</beforebreak>
@@ -4760,7 +4781,7 @@
 </rule>
 <!-- Abbreviations that cannot finish sentences-->
 <rule break="no">
-<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\u00A0]</beforebreak>
+<beforebreak>\b(dc|inst|(?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\u00A0]</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <!-- Abbreviations that can finish sentences -->
@@ -4815,12 +4836,12 @@
 </rule>
 <!-- Ellipsis: ... lowercase -->
 <rule break="no">
-<beforebreak>[^\s\u00A0](\Q...\E|…)[\s\u00A0]</beforebreak>
+<beforebreak>[^\s\u00A0](\.\.\.|…)[\s\u00A0]</beforebreak>
 <afterbreak>\p{Ll}</afterbreak>
 </rule>
 <!-- (enum...) -->
 <rule break="no">
-<beforebreak>\b(\Q...\E|…)[\p{Pe}»"’”][\s\u00A0]</beforebreak>
+<beforebreak>\b(\.\.\.|…)[\p{Pe}»"’”][\s\u00A0]</beforebreak>
 <afterbreak>\p{Ll}</afterbreak>
 </rule>
 <!-- pero ¡ah! no estaba
@@ -4844,6 +4865,19 @@
 </rule>
 </languagerule>
 <languagerule languagerulename="Spanish">
+<!-- unknown abbreviations inside parentheses -->
+<rule break="no">
+<beforebreak>\([^\)]*\.[\s\u00A0]</beforebreak>
+<afterbreak>[^\)\r\n]*\)</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\[[^\]]*\.[\s\u00A0]</beforebreak>
+<afterbreak>[^\]\r\n]*\]</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\{[^\}]*\.[\s\u00A0]</beforebreak>
+<afterbreak>[^\}\r\n]*\}</afterbreak>
+</rule>
 <rule break="no">
 <beforebreak>¿[^?]+:[\s\u00A0]</beforebreak>
 <afterbreak>.</afterbreak>
@@ -4867,12 +4901,12 @@
 </rule>
 <!-- Ellipsis: ... lowercase -->
 <rule break="no">
-<beforebreak>[^\s\u00A0](\Q...\E|…)[\s\u00A0]</beforebreak>
+<beforebreak>[^\s\u00A0](\.\.\.|…)[\s\u00A0]</beforebreak>
 <afterbreak>\p{Ll}</afterbreak>
 </rule>
 <!-- (enum...) -->
 <rule break="no">
-<beforebreak>\b(\Q...\E|…)[\p{Pe}»"’”][\s\u00A0]</beforebreak>
+<beforebreak>\b(\.\.\.|…)[\p{Pe}»"’”][\s\u00A0]</beforebreak>
 <afterbreak>\p{Ll}</afterbreak>
 </rule>
 <!-- Abbreviations that can finish sentences -->
@@ -4982,6 +5016,10 @@
 <beforebreak>\b(https?|ftp|file|chrome|chromium|android|(chrome|moz)\-extension):///?[A-Za-z0-9\-]+\.</beforebreak>
 <afterbreak>[A-Za-z0-9\-]+(\.|\b)</afterbreak>
 </rule>
+<rule break="no"><!-- https://www.seven.one/ -->
+<beforebreak>\b[Se]even\.</beforebreak>
+<afterbreak>[Oo]nes?\b</afterbreak>
+</rule>
 <rule break="no">
 <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>
@@ -5091,7 +5129,7 @@
 </rule>
 <!-- don't split at cases like "Friedrich II. wird auch..." -->
 <rule break="no">
-<beforebreak>[\u00A0\s ][IVX]+\.[\u00A0\s]{1,2}</beforebreak>
+<beforebreak>[\u00A0\s ][IVX]+\.[\u00A0\s]{1,2}</beforebreak>
 <afterbreak>[^\p{Lu}]+</afterbreak>
 </rule>
 <!-- don't split at cases like "im 13. oder 14. Jahrhundert" -->
@@ -5724,7 +5762,7 @@
 </rule>
 <!-- Ів. Франко (але Ів Бутільє) -->
 <rule break="no">
-<beforebreak>(^|[\h\v])(Ів|Дж)\.[\h\v]+</beforebreak>
+<beforebreak>(^|[\h\v])(Ів|Дж|Ол)\.[\h\v]+</beforebreak>
 <afterbreak>[А-ЯІЇЄҐA-Z]</afterbreak>
 </rule>
 <!-- Year: 2000 р.:
@@ -5815,7 +5853,7 @@
 </rule>
 <!-- abbreviation with proper noun: проф. Грицько, о. Лісове -->
 <rule break="no">
-<beforebreak>\b(ап|[Аа]кад|[Пп]роф|[Дд]оц|[Аа]сист|[Рр]еж|[Аа]рх|[Сс]вв?|о|оз|ім|інж|дир|тов|упоряд|тт|чл\.-кор|[Пп]реп)\.[\h\v]*</beforebreak>
+<beforebreak>\b(ап|[Аа]кад|[Пп]роф|[Дд]оц|[Аа]сист|[Рр]еж|[Аа]рх|[Сс]вв?|о|оз|ім|інж|дир|тов|упоряд|тт|чл\.-кор|[Пп]реп|[сС]вт)\.[\h\v]*</beforebreak>
 <afterbreak>[\h\v]*[А-ЯІЇЄҐA-Z]</afterbreak>
 </rule>
 <rule break="no">
@@ -5833,8 +5871,8 @@
 </rule>
 <!-- TODO: арт. - артист -->
 <rule break="no">
-<beforebreak>\b([Аа]рт|[Мм]ал|[Рр]ис)\.[\h\v]*</beforebreak>
-<afterbreak>[\h\v]*[0-9]</afterbreak>
+<beforebreak>\b([Аа]рт|[Мм]ал|[Рр]ис|[Сс]пр)\.[\h\v]*</beforebreak>
+<afterbreak>[\h\v]*(№[\h\v]*)?[0-9]</afterbreak>
 </rule>
 <!-- ХІІ р., 3-6 арт., 2-3 тт. -->
 <rule break="no">
@@ -6359,7 +6397,7 @@
 </rule>
 <!-- Not break for ellipses (...) -->
 <rule break="no">
-<beforebreak>[^\s](\Q...\E|…)\s</beforebreak>
+<beforebreak>[^\s](\.\.\.|…)\s</beforebreak>
 <afterbreak>\p{Ll}</afterbreak>
 </rule>
 <!-- z.B. "bla (...) blubb" -> without ending sentence -->
@@ -6580,7 +6618,7 @@
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>[^\.]\s[ضصثقفغعهخحجچشسیبلاتنمکگ\ظطزرذدپوًٌٍَُِّْA-Z]\.\s</beforebreak>
+<beforebreak>[^\.]\s[ضصثقفغعهخحجچشسیبلاتنمکگ\ظطزرذدپوًٌٍَُِّْA-Z]\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -6660,7 +6698,7 @@
 </rule>
 <!--Не раздвајај у случају као на пр.: "Петар I дошао је ..."-->
 <rule break="no">
-<beforebreak>[\s ][IVX]+\s</beforebreak>
+<beforebreak>[\s ][IVX]+\s</beforebreak>
 <afterbreak>[^\p{Lu}]+</afterbreak>
 </rule>
 <!--Не раздвајај у случају као "од 13. до 14. века"-->

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: srx-languagetool
 version: !ruby/object:Gem::Version
-  version: 0.13.0
+  version: 0.14.0
 platform: ruby
 authors:
 - Aaron Madlon-Kay
-autorequire: 
+autorequire:
 bindir: exe
 cert_chain: []
-date: 2024-03-28 00:00:00.000000000 Z
+date: 2024-09-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: srx
@@ -24,7 +24,7 @@ dependencies:
     - - "<"
       - !ruby/object:Gem::Version
         version: '1.0'
-description: 
+description:
 email:
 - aaron@madlon-kay.com
 executables: []
@@ -58,7 +58,7 @@ metadata:
   source_code_uri: https://github.com/amake/srx-languagetool-ruby.git
   changelog_uri: https://github.com/amake/srx-languagetool-ruby/blob/master/CHANGELOG.md
   rubygems_mfa_required: 'true'
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -73,8 +73,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.6
-signing_key: 
+rubygems_version: 3.5.7
+signing_key:
 specification_version: 4
 summary: SRX segmentation rules from LanguageTool
 test_files: []