Merge pull request #16 from bretttolbert/catalan-contd

Catalan contd

CHANGELOG.md

@@ -1,5 +1,13 @@
 # verbecc Changelog
 
+- 1.9.1 [17 December 2023]
+  - Renamed Catalan 'preterit' to 'passat-simple'
+  - Cont. added more missing conjugation templates for Catalan, decent support for most Catalan verbs now
+  - Added `localization` module with localization functions `localize_mood` and `localize_tense`
+  - Removed pre-generated model .zip files
+  - Added dummy file in models directory as workaround for installation issue
+  - fixed KeyError with Spanish verb abolir
+
 - 1.9.0 [December 2023]
   - Added limited support for Catalan language
   - Please help improve support for Catalan verb conjugation, PRs welcome
@@ -8,7 +16,7 @@
   - Updated dependencies (scikit-learn, etc.)
   - Now targetting Python 3.11
 
-- 1.8.1 [2022 December 28]
+- 1.8.1 [28 December 2022]
   - Updated from Python 3.7 to Python 3.10
   - Updated dependencies
   - Increased SGDClassifier max_iter from 4000 to 40000

EXAMPLE_CA_SER.md

@@ -32,7 +32,7 @@
                 "vosaltres éreu",
                 "ells eren"
             ],
-            "pretèrit": [
+            "passat-simple": [
                 "jo fui",
                 "tu fores",
                 "ell fou",

README.md

@@ -19,7 +19,7 @@
 - [HTTP API : /verbecc/conjugate/fr/manger](http://verbe.cc/verbecc/conjugate/fr/manger)
 
 ### Features
-* Conjugate verbs in six romance languages: French, Spanish, Portuguese, Italian, Romanian, Catalan (limited-support)
+* Conjugate verbs in six romance languages: French, Spanish, Portuguese, Italian, Romanian, Catalan
 * Uses machine learning techniques to predict conjugation of unknown verbs with 99% accurracy
 * Includes both simple and compound conjugations
 * Unit-tested

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "verbecc"
-version = "1.9.0"
+version = "1.9.1"
 dependencies = [
   "lxml",
   "lxml-stubs",

tests/test_inflector_ca.py

@@ -8,6 +8,7 @@
 
 cg = Conjugator(lang='ca')
 
+# TODO: Fix all of these missing templates
 @pytest.mark.skip("known failure")
 def test_all_verbs_have_templates():
     """Have not finished adding templates for all verbs, so this should fail"""
@@ -140,7 +141,11 @@ def test_find_verb_by_infinitive():
     ('morir', 'indicatiu', 'present',
         ['jo moro', 'tu mors', 'ell mor', 'nosaltres morim', 'vosaltres moriu', 'ells moren']),
     ('eixir', 'indicatiu', 'present',
-        ['jo ixo', 'tu ixes', 'ell ix', 'nosaltres eixim', 'vosaltres eixiu', 'ells ixen'])
+        ['jo ixo', 'tu ixes', 'ell ix', 'nosaltres eixim', 'vosaltres eixiu', 'ells ixen']),
+    ('jaure', 'indicatiu', 'present',
+        ['jo jec', 'tu jeus', 'ell jeu', 'nosaltres jaiem', 'vosaltres jaieu', 'ells jeuen']),
+    ('jeure', 'indicatiu', 'present',
+        ['jo jec', 'tu jeus', 'ell jeu', 'nosaltres jaiem', 'vosaltres jaieu', 'ells jeuen'])
 ]
 
 @pytest.mark.parametrize("infinitive,mood,tense,expected_result",

verbecc/data/conjugations-ca.xml

@@ -634,6 +634,99 @@ voler ("to want"), caldre ("must"), dir ("to say") and dur ("to take/get").
 	</Particip>
 </template>
 
+<!-- TODO: This should inherit from 'perd:re' as 'jau:re' -->
+<template name="j:aure">
+	<Indicatiu>
+		<present>
+			<p><i>ec</i></p>
+			<p><i>eus</i></p>
+			<p><i>eu</i></p>
+			<p><i>aiem</i></p>
+			<p><i>aieu</i></p>
+			<p><i>euen</i></p>
+		</present>
+		<imperfet>
+			<p><i>eia</i></p>
+			<p><i>eies</i></p>
+			<p><i>eia</i></p>
+			<p><i>èiem</i></p>
+			<p><i>èieu</i></p>
+			<p><i>eien</i></p>
+		</imperfet>
+		<passat-simple>
+			<p><i>aguí</i></p>
+			<p><i>agueres</i></p>
+			<p><i>agué</i></p>
+			<p><i>aguérem</i></p>
+			<p><i>aguéreu</i></p>
+			<p><i>aguéren</i></p>
+		</passat-simple>
+		<futur>
+			<p><i>auré</i></p>
+			<p><i>auràs</i></p>
+			<p><i>aurà</i></p>
+			<p><i>aurem</i></p>
+			<p><i>aureu</i></p>
+			<p><i>auran</i></p>
+		</futur>
+		</Indicatiu>
+	<Subjuntiu>
+		<present>
+			<p><i>egui</i></p>
+			<p><i>eguis</i></p>
+			<p><i>egui</i></p>
+			<p><i>eguem</i></p>
+			<p><i>egueu</i></p>
+			<p><i>eguin</i></p>
+		</present>
+		<imperfet>
+			<p><i>agués</i></p>
+			<p><i>aguessis</i></p>
+			<p><i>agués</i></p>
+			<p><i>aguéssim</i></p>
+			<p><i>aguéssiu</i></p>
+			<p><i>aguessin</i></p>
+		</imperfet>
+	</Subjuntiu>
+	<Imperatiu>
+		<imperatiu-present>
+			<p><i>ei</i></p>
+			<p><i>egui</i></p>
+			<p><i>eguem</i></p>
+			<p><i>aieu</i></p>
+			<p><i>eguin</i></p>
+		</imperatiu-present>
+	</Imperatiu>
+	<Condicional>
+		<present>
+			<p><i>auria</i></p>
+			<p><i>auries</i></p>
+			<p><i>auria</i></p>
+			<p><i>auríem</i></p>
+			<p><i>auríeu</i></p>
+			<p><i>aurien</i></p>
+		</present>
+		</Condicional>
+	<Infinitiu>
+		<infinitiu-present>
+			<p><i>eure</i></p>
+		</infinitiu-present>
+	</Infinitiu>
+	<Gerundi>
+		<gerundi>
+			<p><i>aient</i></p>
+		</gerundi>
+	</Gerundi>
+	<Particip>
+		<particip>
+			<p><i>agut</i></p>
+			<p><i>aguda</i></p>
+			<p><i>aguts</i></p>
+			<p><i>agudes</i></p>
+		</particip>
+	</Particip>
+</template>
+
 <!-- TODO: This should inherit from 'perd:re' as 'pod:er' -->
 <template name=":poder">
 	<Indicatiu>

verbecc/data/verbs-ca.xml

@@ -5941,9 +5941,9 @@
 <v><i>jaspiar</i><t>canvi:ar</t><en></en></v>
 <v><i>jaumetar</i><t>cant:ar</t><en></en></v>
 <v><i>jaupar</i><t>cant:ar</t><en></en></v>
-<v><i>jaure</i><t>:perdre</t><en></en></v>
+<v><i>jaure</i><t>j:aure</t><en>lie down</en></v>
 <v><i>jerarquitzar</i><t>cant:ar</t><en></en></v>
-<v><i>jeure</i><t>:jaure</t><en></en></v>
+<v><i>jeure</i><t>j:aure</t><en>lie</en></v>
 <v><i>joguinejar</i><t>enve:jar</t><en></en></v>
 <v><i>jornalejar</i><t>enve:jar</t><en></en></v>
 <v><i>jovenejar</i><t>enve:jar</t><en></en></v>

verbecc/inflector.py

@@ -6,7 +6,6 @@
 
 from verbecc import grammar_defines
 from verbecc import exceptions
-from verbecc import inflector
 from verbecc import conjugations_parser
 from verbecc import conjugation_template
 from verbecc import verb
@@ -62,30 +61,34 @@ def find_verb_by_infinitive(self, infinitive) -> verb.Verb:
     def find_template(self, name: str) -> conjugation_template.ConjugationTemplate:
         return self._conj_parser.find_template(name)
 
-    def get_verbs_that_start_with(self, query, max_results):
+    def get_verbs_that_start_with(self, query: str, max_results: int):
         query = query.lower()
         matches = self._verb_parser.get_verbs_that_start_with(query, max_results)
         return matches
 
-    def _get_verb_stem(self, infinitive, template_name):
-        template_beg, template_ending = template_name.split(u':')
+    def _get_verb_stem(self, infinitive: str, template_name: str):
+        """Get the verb stem given an ininitive and a colon-delimited template name.
+        E.g. infinitive='parler' template_name='aim:er' -> 'parl'
+        Note: Catalan overrides this base class implementation to allow looser matching
+        (only requires the last n-1 chars of template ending to match infinitive ending)"""
+        _, template_ending = template_name.split(u':')
         if not infinitive.endswith(template_ending):
             raise exceptions.ConjugatorError(
                 "Template {} ending doesn't "
                 "match infinitive {}"
                 .format(template_name, infinitive))
         return infinitive[:len(infinitive) - len(template_ending)]
 
-    def _is_impersonal_verb(self, infinitive):
+    def _is_impersonal_verb(self, infinitive: str):
         return False
 
-    def _verb_can_be_reflexive(self, infinitive):
+    def _verb_can_be_reflexive(self, infinitive: str):
         return not self._is_impersonal_verb(infinitive)
 
-    def _split_reflexive(self, infinitive):
+    def _split_reflexive(self, infinitive: str):
         return (False, infinitive)
 
-    def _add_reflexive_pronoun(self, s):
+    def _add_reflexive_pronoun(self, s: str):
         pass
 
     def _add_subjunctive_relative_pronoun(self, s: str, tense_name: str):

verbecc/inflector_ca.py

@@ -2,6 +2,7 @@
 
 from typing import Dict, List, Tuple
 
+from verbecc import exceptions
 from verbecc import inflector
 
 class InflectorCa(inflector.Inflector):
@@ -67,6 +68,7 @@ def _get_alternate_hv_inflection(self, s: str) -> str:
 
     def _get_compound_conjugations_aux_verb_map(self) -> Dict[str, Dict[str, Tuple[str, ...]]]:
         """
+        TODO: Implement all these compound tenses (Spanish compound tenses in this commment, for reference)
         return {
             'indicatiu': {
                 'pretèrit-perfet-compuest': ('indicatiu', 'present'),
@@ -85,3 +87,26 @@ def _get_compound_conjugations_aux_verb_map(self) -> Dict[str, Dict[str, Tuple[s
         }
         """
         return {}
+
+    def _get_verb_stem(self, infinitive: str, template_name: str):
+        """Get the verb stem given an ininitive and a colon-delimited template name.
+        E.g. infinitive='parlar' template_name='cant:ar' -> 'parl'
+
+        Note: Base class _get_verb_stem raises exception if template ending doesn't 
+        match infinitive ending exactly but for Catalan, some verbs
+        have endings where at least the first letter doesn't match.
+
+        E.g. both 'jaure' and and 'jeure' are apparently conjugated
+        identically, so we want either one to use the 'j:aure' template.
+        So since this is Catalan, let it pass if the last n-1 letters of the 
+        template ending match the infinitive ending
+        """
+        _, template_ending = template_name.split(u':')
+        if not infinitive.endswith(template_ending) \
+        and not infinitive.endswith(template_ending[1:]):
+                raise exceptions.ConjugatorError(
+                    "Template {} ending doesn't "
+                    "match infinitive {},"
+                    "not even a little bit"
+                    .format(template_name, infinitive))
+        return infinitive[:len(infinitive) - len(template_ending)]