Support German gender-sensitive DET, ADJ, PRON ending (from KorAP-Tokenizer)
Change-Id: I8f20ecb913c0fe514b5936ab43287ca616695f16
diff --git a/src/de/gender.xfst b/src/de/gender.xfst
index 8adc35c..dd8c7de 100644
--- a/src/de/gender.xfst
+++ b/src/de/gender.xfst
@@ -8,6 +8,12 @@
! false positives in compounds like "Nutzer/Innenarchitekt".
define genderEndingsInLower [ i n ( n e n ) ];
+! Short endings for determiners, adjectives and pronouns:
+! e, n, r, s, m, es, er, em, en
+define genderShortSuffix [ [e|n|r|s|m] | [ e [s|r|m|n] ] ];
+
+define genderSeparator [":" | Slash ( %- ) | Asterisk | "_" ];
+
! Gender-sensitive endings with frau/frauen
! (lowercase only - capitalized Frau is a standalone word)
! Note: This is now only used for words ending in "mann"
@@ -15,10 +21,21 @@
! Matches: Kaufmann, Geschäftsmann, etc. but NOT just "mann"
define genderEndingsFrau [ {frau} ( {en} ) ];
-! General gender endings (only -in/-innen forms for colon, slash, parenthetical)
-! Colon forms: Nutzer:in, Nutzer:In, Nutzer:innen
-! Slash forms for -in/-innen: Nutzer/in, Nutzer/innen, Nutzer/-in, Kosovo-Albaner/innen
-define genderIn [ ":" genderEndingsIn | Slash ( %- ) genderEndingsInLower ];
+! General gender endings for -in/-innen without continuation:
+! allow all I/i variants for :, *, _; keep slash lowercase-only to
+! avoid false positives like Nutzer/Innenarchitekt.
+define genderIn [
+ [ ":" | Asterisk | "_" ] genderEndingsIn |
+ Slash ( %- ) genderEndingsInLower
+];
+
+! Short forms for DET/ADJ/PRON with colon/slash/star/underscore:
+! gute:r, ihm/r, ein*e, ein_e, diese:r
+define genderShort [ genderSeparator genderShortSuffix ];
+
+! Parenthetical short forms:
+! eine(n), ein(e)
+define genderParenShort %( ( %- ) genderShortSuffix %);
! Slash forms for -frau: Kaufmann/frau, Kaufmann/-frau, Geschäftsmann/frau
! Only applies when word ends in "mann" (with non-empty prefix before it)
@@ -31,4 +48,12 @@
! Only applies when word ends in "mann" (with non-empty prefix before it)
define genderParenFrau {mann} %( ( %- ) genderEndingsFrau %);
-define GenderEndings [ genderIn | genderFrau | genderParenIn | genderParenFrau ];
\ No newline at end of file
+! Compound continuation for forms like Lehrer:innenfortbildung.
+! For slash, restrict to lowercase in/innen so /Innen... keeps splitting
+! (e.g. Nutzer/Innenarchitekt, Innenminister/Innenministerinnen).
+define genderInComp [
+ [ ":" | Asterisk | "_" ] genderEndingsIn Char+ |
+ Slash ( %- ) genderEndingsInLower Char+
+];
+
+define GenderEndings [ genderIn | genderInComp | genderShort | genderFrau | genderParenIn | genderParenShort | genderParenFrau ];
\ No newline at end of file