srx-languagetool 0.12.0 → 0.13.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: b095dc51541d60d446b063cdff16940d7840e7d51891e1f5201117d42089b0da
-  data.tar.gz: 5376dce7bda3c17a142477a3b19f629e65527c4fb96d540bfa1a853f0ed394f8
+  metadata.gz: 037ab00a25a87ba79d5bb760cc4d292aad1289d792edea55d8b05034a2ac0f5f
+  data.tar.gz: 760147c7ba571a06943244728a709a02bedcb320244342ba38694a1148091a7b
 SHA512:
-  metadata.gz: eb9f0aac7af1e2345842bdaf21917556429d8c887d356430c4c707b18b82d0a54cc46a5b6344247de27113f3ad0730f0d3fa37012f4a7c5c7cf4987dcedcf271
-  data.tar.gz: 3370f8761982aa574c862f32d9a2702451cb992fa35dd0f842a4116224d448c7a2dd247be0ac2c61cace69f72364963c9072bd72e3fc2c46a3bc948996c8fc61
+  metadata.gz: 15184b2b378ab3c2f0667ac6168894583959833403b4181c223eb5cd51a43ab8286c2f9110c1769fd5fb9898aa386cff525e6c170e48bc6aa2247d2cf171126c
+  data.tar.gz: e30f82724c3bcb264bde6cf2a99f3ea09eb6815f43a9609b71b2b4bbb1fd2e662cbb1abee2bdfd52035130fee9ed1118fe545f48efb61b103c27acaf29773e2e

data/.github/workflows/main.yml

@@ -6,7 +6,7 @@ jobs:
   build:
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v4
     - name: Set up Ruby
       uses: ruby/setup-ruby@v1
       with:

data/CHANGELOG.md

@@ -1,5 +1,9 @@
 ## [Unreleased]
 
+## [0.13.0] - 2024-03-29
+
+- Update rules to LanguageTool 6.4
+
 ## [0.12.0] - 2023-10-07
 
 - Update rules to LanguageTool 6.3

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    srx-languagetool (0.12.0)
+    srx-languagetool (0.13.0)
       srx (< 1.0)
 
 GEM
@@ -9,69 +9,72 @@ GEM
   specs:
     ast (2.4.2)
     backport (1.2.0)
-    benchmark (0.2.1)
+    benchmark (0.3.0)
     byebug (11.1.3)
-    diff-lcs (1.5.0)
+    diff-lcs (1.5.1)
     e2mmap (0.1.0)
-    jaro_winkler (1.5.4)
-    json (2.6.3)
+    jaro_winkler (1.5.6)
+    json (2.7.1)
     kramdown (2.4.0)
       rexml
     kramdown-parser-gfm (1.1.0)
       kramdown (~> 2.0)
-    minitest (5.18.0)
-    nokogiri (1.14.2-x86_64-darwin)
+    language_server-protocol (3.17.0.3)
+    minitest (5.22.3)
+    nokogiri (1.15.6-x86_64-darwin)
       racc (~> 1.4)
-    parallel (1.22.1)
-    parser (3.2.1.1)
+    parallel (1.24.0)
+    parser (3.3.0.5)
       ast (~> 2.4.1)
-    racc (1.6.2)
+      racc
+    racc (1.7.3)
     rainbow (3.1.1)
-    rake (13.0.6)
-    regexp_parser (2.7.0)
+    rake (13.1.0)
+    rbs (2.8.4)
+    regexp_parser (2.9.0)
     reverse_markdown (2.1.1)
       nokogiri
-    rexml (3.2.5)
-    rspec-expectations (3.12.2)
+    rexml (3.2.6)
+    rspec-expectations (3.13.0)
       diff-lcs (>= 1.2.0, < 2.0)
-      rspec-support (~> 3.12.0)
-    rspec-support (3.12.0)
-    rubocop (1.48.1)
+      rspec-support (~> 3.13.0)
+    rspec-support (3.13.1)
+    rubocop (1.62.1)
       json (~> 2.3)
+      language_server-protocol (>= 3.17.0)
       parallel (~> 1.10)
-      parser (>= 3.2.0.0)
+      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 (>= 1.26.0, < 2.0)
+      rubocop-ast (>= 1.31.1, < 2.0)
       ruby-progressbar (~> 1.7)
       unicode-display_width (>= 2.4.0, < 3.0)
-    rubocop-ast (1.28.0)
-      parser (>= 3.2.1.0)
+    rubocop-ast (1.31.2)
+      parser (>= 3.3.0.4)
     ruby-progressbar (1.13.0)
-    solargraph (0.48.0)
+    solargraph (0.50.0)
       backport (~> 1.2)
       benchmark
-      bundler (>= 1.17.2)
+      bundler (~> 2.0)
       diff-lcs (~> 1.4)
       e2mmap
       jaro_winkler (~> 1.5)
       kramdown (~> 2.3)
       kramdown-parser-gfm (~> 1.1)
       parser (~> 3.0)
-      reverse_markdown (>= 1.0.5, < 3)
-      rubocop (>= 0.52)
+      rbs (~> 2.0)
+      reverse_markdown (~> 2.0)
+      rubocop (~> 1.38)
       thor (~> 1.0)
       tilt (~> 2.0)
       yard (~> 0.9, >= 0.9.24)
     srx (0.6.0)
       nokogiri (~> 1.11)
-    thor (1.2.1)
-    tilt (2.1.0)
-    unicode-display_width (2.4.2)
-    webrick (1.7.0)
-    yard (0.9.28)
-      webrick (~> 1.7.0)
+    thor (1.3.1)
+    tilt (2.3.0)
+    unicode-display_width (2.5.0)
+    yard (0.9.36)
 
 PLATFORMS
   x86_64-darwin-20

data/lib/srx/languagetool/version.rb

@@ -2,6 +2,6 @@
 
 module Srx
   module Languagetool
-    VERSION = '0.12.0'
+    VERSION = '0.13.0'
   end
 end

data/lib/srx/segment.srx

@@ -5,7 +5,7 @@
 <formathandle type="end" include="yes"></formathandle>
 <formathandle type="isolated" include="no"></formathandle>
 <okpsrx:options oneSegmentIncludesAll="no" trimLeadingWhitespaces="no" trimTrailingWhitespaces="no" useJavaRegex="yes" useIcu4JBreakRules="no" treatIsolatedCodesAsWhitespace="no"></okpsrx:options>
-<okpsrx:sample language="nl" useMappedRules="yes">Wat God buiten Christus is. 2.</okpsrx:sample>
+<okpsrx:sample language="nl" useMappedRules="yes"> ON! is een omroep.</okpsrx:sample>
 <okpsrx:rangeRule></okpsrx:rangeRule>
 </header>
 <body>
@@ -1164,7 +1164,7 @@
 <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?|[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|[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">
@@ -1273,7 +1273,7 @@
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\bLL\.[\s\u00A0]?[BM]\.[\s\u00A0]</beforebreak>
+<beforebreak>\bLL\.[\s\u00A0]?[BMD]\.[\s\u00A0]</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -1282,7 +1282,7 @@
 </rule>
 <rule break="no">
 <beforebreak>\bLL\.[\s\u00A0]?</beforebreak>
-<afterbreak>[BM]\.?</afterbreak>
+<afterbreak>[BMD]\.?</afterbreak>
 </rule>
 <rule break="no">
 <beforebreak>\b[BM]\.[\s\u00A0]?</beforebreak>
@@ -1329,7 +1329,11 @@
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\bCorp\.[\s\u00A0]</beforebreak>
+<beforebreak>\b[cC]orp\.[\s\u00A0]</beforebreak>
+<afterbreak></afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\b[Rr]eg\.[\s\u00A0]</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -1540,6 +1544,15 @@
 </rule>
 </languagerule>
 <languagerule languagerulename="Dutch">
+<rule break="no">
+<beforebreak>\sart\.\s</beforebreak>
+<afterbreak>[IVX]+[ .]</afterbreak>
+</rule>
+<!--Do not break after abbreviation of type a.b.c.-->
+<rule break="no">
+<beforebreak>\s([a-z]\.){2,10}\s</beforebreak>
+<afterbreak></afterbreak>
+</rule>
 <rule break="yes">
 <beforebreak>[ ]is[.][ ]</beforebreak>
 <afterbreak>[0-9]\.($|[ ])</afterbreak>
@@ -1582,7 +1595,7 @@
 <afterbreak>\p{Ll}</afterbreak>
 </rule>
 <rule break="yes">
-<beforebreak>\s(la|do|del)\sMar\.\s</beforebreak>
+<beforebreak>\s(la|do|del?)\sMar\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -1641,6 +1654,14 @@
 <beforebreak>\b(geb|[Gg]em|get|gld|id|[Ii]ncl|ind|inf|ing|intern|[Ss]ec|inz|ir|jhr|jkvr)\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
+<rule break="yes">
+<beforebreak>\s(tel|red|min)\.\s</beforebreak>
+<afterbreak>[A-Z]</afterbreak>
+</rule>
+<rule break="yes">
+<beforebreak>\.(nl|be|com)\.\s</beforebreak>
+<afterbreak></afterbreak>
+</rule>
 <rule break="no">
 <beforebreak>\b(jl|jr|kr|kt|lic|ll|lt|lw|max|[Mm]evr|mi|[Mm]in|mld)\.\s</beforebreak>
 <afterbreak></afterbreak>
@@ -1662,6 +1683,10 @@
 <afterbreak>[A-Z]</afterbreak>
 </rule>
 <rule break="yes">
+<beforebreak>\svitamine [A-Z]\.\s</beforebreak>
+<afterbreak>[A-Z]</afterbreak>
+</rule>
+<rule break="yes">
 <beforebreak>°C\.\s</beforebreak>
 <afterbreak>[A-Z][a-z]</afterbreak>
 </rule>
@@ -1714,6 +1739,34 @@
 <afterbreak>\p{L}\.\s</afterbreak>
 </rule>
 <rule break="no">
+<beforebreak>\set al\.\s</beforebreak>
+<afterbreak></afterbreak>
+</rule>
+<!--pa. as (wrong) abbrev for pag.-->
+<rule break="no">
+<beforebreak>\spa\.\s</beforebreak>
+<afterbreak>[0-9]</afterbreak>
+</rule>
+<!--op. as abbrev for opus-->
+<rule break="no">
+<beforebreak>\sop\.\s</beforebreak>
+<afterbreak>[0-9]|cit\.</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\soa\.\s</beforebreak>
+<afterbreak>[a-z]</afterbreak>
+</rule>
+<!--al. as abbrev for alinea-->
+<rule break="no">
+<beforebreak>\sal\.\s</beforebreak>
+<afterbreak>[0-9]</afterbreak>
+</rule>
+<!--Break also when the next sentence has no capital-->
+<rule break="yes">
+<beforebreak>\s((is|op|in|af|ik|ze|om|me|je|na|nu|al|ja|VS|EU|er|we|tv|he|ga|hè|hé|TV|as|ei|SP|pc|wc|PC|IS|NS|ok|AD|OK|at|OM|cd|VN|it|EK|In|pa|AZ|up|IT|FM|VI|ui|la|CD|CV|pr|ie|cv|WW|GB|Jo|Aa|UK|HD|oa|VU))\.\s</beforebreak>
+<afterbreak></afterbreak>
+</rule>
+<rule break="no">
 <beforebreak>\b\p{L}\.</beforebreak>
 <afterbreak>\p{L}\.</afterbreak>
 </rule>
@@ -1737,6 +1790,10 @@
 <beforebreak>\b\p{Lu}\.\p{Lu}\.\s</beforebreak>
 <afterbreak></afterbreak>
 </rule>
+<rule break="yes">
+<beforebreak>\s(op)\sX\.\s</beforebreak>
+<afterbreak></afterbreak>
+</rule>
 <rule break="no">
 <beforebreak>[^\.]\s[A-Z]\.\s</beforebreak>
 <afterbreak></afterbreak>
@@ -1771,10 +1828,18 @@
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\s[A-Z].+!\s</beforebreak>
+<beforebreak>(^|\s)[A-Z].+!\s</beforebreak>
+<afterbreak>[a-z]</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\s[A-Z].+z\.\s</beforebreak>
 <afterbreak>[a-z]</afterbreak>
 </rule>
 <rule break="no">
+<beforebreak>\sart\.\s</beforebreak>
+<afterbreak>[0-9]</afterbreak>
+</rule>
+<rule break="no">
 <beforebreak>\b(jan|mrt|mar|jun|jul|aug|sept|okt|sep|spt|nov|dec|.*opp)\.\s</beforebreak>
 <afterbreak>[a-z]</afterbreak>
 </rule>
@@ -5066,11 +5131,11 @@
 </rule>
 <!-- German abbreviations -->
 <rule break="no">
-<beforebreak>\b(betr|Geb|Stk|ggü|Mag|mtl|[Pp]arl|Bsp|versch|[Dd]iesbzgl|[Dd]bzgl[Ss]tellv|d|Übers|usw|[Bb]zw|Ab[hkst]|[Aa]bzü?gl|\d+-tlg|tlg|ggfls|[Ll]tda|[Ee]inschl|[Vv]mtl|Ev|bezgl|lit|Abzw|[Vv]sl|ahd|Akk|aktual|[Öö]ffentl|prof|allg|alltagsspr|altdt|alttest|amerikan|Anh|Ank|Anm|Art|[Aa]utom|Auftragsnr|Az|Bat|bayr|Bde?|bearb|Bed|Bem|bes|bez|Bez|Bhf|Blvd|[Bb]spw|btto|bw|Dtl|[Gg]esetzl|Dez|[Jj]gdfr|[Ee]ff)\.[\u00A0\s]{1,2}</beforebreak>
+<beforebreak>\b(betr|Geb|Stk|ggü|Mag|mtl|Flgh?|[Pp]arl|Bsp|versch|[Dd]iesbzgl|[Zz]ykl|[Dd]bzgl[Ss]tellv|d|Übers|usw|[Bb]zw|Ab[hkst]|[Ee]ig|[Aa]bzü?gl|\d+-tlg|tlg|[Gg]gfls|[Ff]achspr|[Ll]tda|[Ee]inschl|[Vv]mtl|[Ss]tellv|Ev|[Bb]ezgl|lit|Abzw|[Vv]sl|ahd|Akk|aktual|[Öö]ffentl|prof|allg|alltagsspr|altdt|alttest|amerikan|Anh|Ank|Anm|Art|[Aa]utom|Auftragsnr|Az|Bat|bayr|Bde?|bearb|Bed|Bem|bes|bez|wsl|vsl|Bez|Bhf|Blvd|[Bb]spw|btto|bw|Dtl|[Gg]esetzl|Dez|[Jj]gdfr|[Ee]ff)\.[\u00A0\s]{1,2}</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\b(cts?|[Cc]a|chem|chin|Chr|cresc|[Dd]at|desgl|ders|dgl|Dipl|Dir?|Doz?|durchg|durchges|Dr|dt|ebd|Ed|[Ee]igt?l|akt|[Ee]ngl|Erg|al|et[cw]|Etw|ev|[Ee]vtl?|[Ee]xkl|Expl|Exz)\.[\u00A0\s]{1,2}</beforebreak>
+<beforebreak>\b(cts?|[Cc]a|chem|chin|Chr|cresc|[Dd]at|desgl|ders|dgl|Dipl|Dir?|Doz?|durchg|durchges|Dr|[Dd]t|ebd|Ed|[Ee]igt?l|akt|[Ee]ngl|Erg|al|et[cw]|Etw|ev|[Ee]vtl?|[Ee]xkl|Expl|Exz)\.[\u00A0\s]{1,2}</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -5094,7 +5159,7 @@
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\b(Part|Per[fs]|Pfd|Pl(ur)?|pl|Plusq|Pos|pp|Prä[ps]|Prät|Pro[vf]|rd|reg|resp|Rhld|rit|Sa|südl|Br|se[ln]|Sept|Sing|sign|So|sog|Sp|Std?|stacc|Str|stud|Subst|sva|svw|sZ)\.[\u00A0\s]{1,2}</beforebreak>
+<beforebreak>\b(Part|Per[fs]|Pfd|Pl(ur)?|pl|Plusq|Pos|pp|Prä[ps]|Prät|Pro[vf]|rd|reg|resp|Rhld|rit|Sa|südl|Br|se[ln]|Sept|Sing|sign|So|sog|Sp|[Ss]td?|stacc|Str|stud|Subst|sva|svw|sZ)\.[\u00A0\s]{1,2}</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -5227,7 +5292,7 @@
 </rule>
 <rule break="no">
 <beforebreak>\bLL\.\s?</beforebreak>
-<afterbreak>[BM]\.?</afterbreak>
+<afterbreak>[BMD]\.?</afterbreak>
 </rule>
 <rule break="no">
 <beforebreak>\b[BM]\.\s?</beforebreak>
@@ -5526,7 +5591,7 @@
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
-<beforebreak>\bLL\.[\s\u00A0]?[BM]\.[\s\u00A0]</beforebreak>
+<beforebreak>\bLL\.[\s\u00A0]?[BMD]\.[\s\u00A0]</beforebreak>
 <afterbreak></afterbreak>
 </rule>
 <rule break="no">
@@ -5535,7 +5600,7 @@
 </rule>
 <rule break="no">
 <beforebreak>\bLL\.[\s\u00A0]?</beforebreak>
-<afterbreak>[BM]\.?</afterbreak>
+<afterbreak>[BMD]\.?</afterbreak>
 </rule>
 <rule break="no">
 <beforebreak>\b[BM]\.[\s\u00A0]?</beforebreak>
@@ -5591,6 +5656,14 @@
 <afterbreak>\p{Lu}\p{Ll}</afterbreak>
 </rule>
 </languagerule>
+
+<languagerule languagerulename="Crimean Tatar">
+<rule break="no">
+<beforebreak>\b[0-9]+(\.|:)[0-9][0-9][\s\u00A0\u202F]</beforebreak>
+<afterbreak></afterbreak>
+</rule>
+</languagerule>
+
 <languagerule languagerulename="Ukrainian">
 <!-- when sentence starts with ellipsis: ...Мазій і Юхим теж. -->
 <rule break="no">
@@ -5791,7 +5864,11 @@
 <!-- статус правових держав. — Авт.). -->
 <rule break="no">
 <beforebreak></beforebreak>
-<afterbreak>[\h\v]*[‐-―-][\h\v]*([Рр]ед|[Аа]вт)[\h\v]*\.[\)\]]</afterbreak>
+<afterbreak>[\h\v]*[‐-―-][\h\v]*([Рр]ед|[Аа]вт)\.[\h\v]*[\)\]]</afterbreak>
+</rule>
+<rule break="no">
+<beforebreak>\b([Рр]ед)\.[\h\v]*</beforebreak>
+<afterbreak>[А-ЯІЇЄҐ]</afterbreak>
 </rule>
 <!-- Цензор.НЕТ -->
 <rule break="no">
@@ -6837,6 +6914,7 @@
 <languagemap languagepattern="(ML|ml).*" languagerulename="Generic"></languagemap>
 <languagemap languagepattern="(TL|tl).*" languagerulename="Generic"></languagemap>
 <languagemap languagepattern="(AST|ast).*" languagerulename="Generic"></languagemap>
+<languagemap languagepattern="(CRH|crh).*" languagerulename="Generic"></languagemap>
 <languagemap languagepattern=".*" languagerulename="Default"></languagemap>
 </maprules>
 </body>

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: srx-languagetool
 version: !ruby/object:Gem::Version
-  version: 0.12.0
+  version: 0.13.0
 platform: ruby
 authors:
 - Aaron Madlon-Kay
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2023-10-07 00:00:00.000000000 Z
+date: 2024-03-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: srx