Added missing parts of refactoring json for phrase_translate

UAlbertaALTLab · Nov 19, 2024 · bb88751 · bb88751
1 parent 014a9ce
commit bb88751
Show file tree

Hide file tree

Showing 5 changed files with 78 additions and 74 deletions.
diff --git a/src/crkeng/resources/phrase_translate/noun_wordform_to_phrase.json b/src/crkeng/resources/phrase_translate/noun_wordform_to_phrase.json
@@ -0,0 +1,25 @@
+[
+    ["+N", null, 0],
+    ["+A", null, 0],
+    ["+I", null, 0],
+    ["+D", null, 0],
+    # Number
+    ["+Sg", "TagMap.COPY_TAG_NAME", 1],
+    ["+Pl", "TagMap.COPY_TAG_NAME", 1],
+    ["+Obv", "TagMap.COPY_TAG_NAME", 1],
+    ["+Loc", "TagMap.COPY_TAG_NAME", 1],
+    ["+Distr", "TagMap.COPY_TAG_NAME", 1],
+    # Diminutive
+    ["+Dim", "TagMap.COPY_TAG_NAME", 2],
+    ["+Der/Dim", "Dim+", 2],
+    # Possessives
+    ["+Px1Sg", "TagMap.COPY_TAG_NAME", 3],
+    ["+Px2Sg", "TagMap.COPY_TAG_NAME", 3],
+    ["+Px3Sg", "TagMap.COPY_TAG_NAME", 3],
+    ["+Px1Pl", "TagMap.COPY_TAG_NAME", 3],
+    ["+Px2Pl", "TagMap.COPY_TAG_NAME", 3],
+    ["+Px12Pl", "TagMap.COPY_TAG_NAME", 3],  # Maybe needs to be recoded with 12 -> 21
+    ["+Px3Pl", "TagMap.COPY_TAG_NAME", 3],
+    ["+Px4Sg/Pl", "TagMap.COPY_TAG_NAME", 3],
+    ["+PxX", "PxXPl+", 3]
+]
diff --git a/src/crkeng/resources/phrase_translate/verb_wordform_to_phrase.json b/src/crkeng/resources/phrase_translate/verb_wordform_to_phrase.json
@@ -0,0 +1,46 @@
+[ 
+    ["+V", null, 0],
+    ["+TA", null, 0],
+    ["+AI", null, 0],
+    ["+II", null, 0],
+    ["+TI", null, 0],
+    # Tense/Aspect
+    ["PV/ki+", "Prt+", 1],  # Preterite aka simple past
+    [["PV/ki+", "+Ind"], "Prt+", 1],  # Preterite aka simple past
+    [["+Fut", "+Cond"], "Cond+", 1],  # Future conditional
+    [["+Imp", "+Imm"], "Imm+", 1],  # Immediate imperative
+    [["+Imp", "+Del"], "Del+", 1],  # Delayed imperative
+    [["PV/wi+", "+Ind"], "Fut+", 1],  # Future
+    ["PV/wi+", "Fut+", 1],  # Also accept PV/wi without indicative as future
+    [["PV/e+", "+Cnj"], null, 1],  # conjunctive marker
+    # Note that these crk features as disjoint, but both are needed for the eng feature
+    [["PV/ka+", "+Ind"], "Def+", 1],
+    [["PV/ka+", "+Cnj"], "Inf+", 1],
+    [["PV/ta+", "+Cnj"], "Inf+", 1],  # future definite
+    ["+Ind", "Prs+", 1],
+    ["TagMap.DEFAULT", "Prs+", 1],  # default to present tense
+    # Person - Subject
+    ["+1Sg", "TagMap.COPY_TAG_NAME", 2],
+    ["+2Sg", "TagMap.COPY_TAG_NAME", 2],
+    ["+3Sg", "TagMap.COPY_TAG_NAME", 2],
+    ["+1Pl", "TagMap.COPY_TAG_NAME", 2],
+    ["+12Pl", "21Pl+", 2],
+    ["+2Pl", "TagMap.COPY_TAG_NAME", 2],
+    ["+3Pl", "TagMap.COPY_TAG_NAME", 2],
+    ["+4Sg/Pl", "TagMap.COPY_TAG_NAME", 2],
+    ["+5Sg/Pl", "TagMap.COPY_TAG_NAME", 2],
+    ["+X", "XPl+", 2],
+    # Person - Object
+    ["+1SgO", "TagMap.COPY_TAG_NAME", 3],
+    ["+2SgO", "TagMap.COPY_TAG_NAME", 3],
+    ["+3SgO", "TagMap.COPY_TAG_NAME", 3],
+    ["+1PlO", "TagMap.COPY_TAG_NAME", 3],
+    ["+12PlO", "21PlO+", 3],
+    ["+2PlO", "TagMap.COPY_TAG_NAME", 3],
+    ["+3PlO", "TagMap.COPY_TAG_NAME", 3],
+    ["+4Pl", "TagMap.COPY_TAG_NAME", 3],
+    ["+4Sg", "TagMap.COPY_TAG_NAME", 3],
+    ["+4Sg/PlO", "TagMap.COPY_TAG_NAME", 3],
+    ["+5Sg/PlO", "TagMap.COPY_TAG_NAME", 3],
+    ["+XO", "XPlO+", 3]
+]
diff --git a/src/morphodict/phrase_translate/source_tag_map.py b/src/morphodict/phrase_translate/source_tag_map.py
@@ -1,33 +1,7 @@
-from morphodict.analysis.tag_map import TagMap
-
-# TODO Migrate to json files
-
-COPY_TAG_NAME = TagMap.COPY_TAG_NAME
+from morphodict.phrase_translate.tag_maps import *
 
 noun_wordform_to_phrase = TagMap(
-    ("+N", None, 0),
-    ("+A", None, 0),
-    ("+I", None, 0),
-    ("+D", None, 0),
-    # Number
-    ("+Sg", COPY_TAG_NAME, 1),
-    ("+Pl", COPY_TAG_NAME, 1),
-    ("+Obv", COPY_TAG_NAME, 1),
-    ("+Loc", COPY_TAG_NAME, 1),
-    ("+Distr", COPY_TAG_NAME, 1),
-    # Diminutive
-    ("+Dim", COPY_TAG_NAME, 2),
-    ("+Der/Dim", "Dim+", 2),
-    # Possessives
-    ("+Px1Sg", COPY_TAG_NAME, 3),
-    ("+Px2Sg", COPY_TAG_NAME, 3),
-    ("+Px3Sg", COPY_TAG_NAME, 3),
-    ("+Px1Pl", COPY_TAG_NAME, 3),
-    ("+Px2Pl", COPY_TAG_NAME, 3),
-    ("+Px12Pl", COPY_TAG_NAME, 3),  # Maybe needs to be recoded with 12 -> 21
-    ("+Px3Pl", COPY_TAG_NAME, 3),
-    ("+Px4Sg/Pl", COPY_TAG_NAME, 3),
-    ("+PxX", "PxXPl+", 3),
+    *tagmap_json_to_tuples(read_phrase_translate_json("noun_wordform_to_phrase.json")),
 )
 
 # Normally having a preverb tag excludes a wordform from auto-translation; this
@@ -38,48 +12,5 @@
 
 # Cree tense/aspects:
 verb_wordform_to_phrase = TagMap(
-    ("+V", None, 0),
-    ("+TA", None, 0),
-    ("+AI", None, 0),
-    ("+II", None, 0),
-    ("+TI", None, 0),
-    # Tense/Aspect
-    ("PV/ki+", "Prt+", 1),  # Preterite aka simple past
-    (("PV/ki+", "+Ind"), "Prt+", 1),  # Preterite aka simple past
-    (("+Fut", "+Cond"), "Cond+", 1),  # Future conditional
-    (("+Imp", "+Imm"), "Imm+", 1),  # Immediate imperative
-    (("+Imp", "+Del"), "Del+", 1),  # Delayed imperative
-    (("PV/wi+", "+Ind"), "Fut+", 1),  # Future
-    ("PV/wi+", "Fut+", 1),  # Also accept PV/wi without indicative as future
-    (("PV/e+", "+Cnj"), None, 1),  # conjunctive marker
-    # Note that these crk features as disjoint, but both are needed for the eng feature
-    (("PV/ka+", "+Ind"), "Def+", 1),
-    (("PV/ka+", "+Cnj"), "Inf+", 1),
-    (("PV/ta+", "+Cnj"), "Inf+", 1),  # future definite
-    ("+Ind", "Prs+", 1),
-    (TagMap.DEFAULT, "Prs+", 1),  # default to present tense
-    # Person - Subject
-    ("+1Sg", COPY_TAG_NAME, 2),
-    ("+2Sg", COPY_TAG_NAME, 2),
-    ("+3Sg", COPY_TAG_NAME, 2),
-    ("+1Pl", COPY_TAG_NAME, 2),
-    ("+12Pl", "21Pl+", 2),
-    ("+2Pl", COPY_TAG_NAME, 2),
-    ("+3Pl", COPY_TAG_NAME, 2),
-    ("+4Sg/Pl", COPY_TAG_NAME, 2),
-    ("+5Sg/Pl", COPY_TAG_NAME, 2),
-    ("+X", "XPl+", 2),
-    # Person - Object
-    ("+1SgO", COPY_TAG_NAME, 3),
-    ("+2SgO", COPY_TAG_NAME, 3),
-    ("+3SgO", COPY_TAG_NAME, 3),
-    ("+1PlO", COPY_TAG_NAME, 3),
-    ("+12PlO", "21PlO+", 3),
-    ("+2PlO", COPY_TAG_NAME, 3),
-    ("+3PlO", COPY_TAG_NAME, 3),
-    ("+4Pl", COPY_TAG_NAME, 3),
-    ("+4Sg", COPY_TAG_NAME, 3),
-    ("+4Sg/PlO", COPY_TAG_NAME, 3),
-    ("+5Sg/PlO", COPY_TAG_NAME, 3),
-    ("+XO", "XPlO+", 3),
+    *tagmap_json_to_tuples(read_phrase_translate_json("verb_wordform_to_phrase.json")),
 )
diff --git a/src/morphodict/phrase_translate/tag_maps.py b/src/morphodict/phrase_translate/tag_maps.py
@@ -52,6 +52,8 @@ def element_process(x):
             return tuple(x)
         if x == "TagMap.DEFAULT":
             return TagMap.DEFAULT
+        if x == "TagMap.COPY_TAG_NAME":
+            return TagMap.COPY_TAG_NAME
         return x
 
     return tuple(element_process(x) for x in l)

diff --git a/src/morphodict/phrase_translate/to_target/__init__.py b/src/morphodict/phrase_translate/to_target/__init__.py
@@ -1,5 +1,5 @@
 import logging
-
+from django.conf import settings
 from morphodict.analysis import RichAnalysis
 from morphodict.phrase_translate.source_tag_map import (
     noun_wordform_to_phrase,
@@ -17,7 +17,7 @@
 def inflect_target_language_phrase(analysis, lemma_definition):
     if isinstance(analysis, tuple):
         analysis = RichAnalysis(analysis)
-    cree_wordform_tag_list = analysis.prefix_tags + analysis.suffix_tags
+    cree_wordform_tag_list = analysis.prefix_tags + analysis.suffix_tags + settings.DEFAULT_TARGET_LANGUAGE_PHRASE_TAGS
 
     if "+N" in cree_wordform_tag_list:
         tags_for_phrase = noun_wordform_to_phrase.map_tags(cree_wordform_tag_list)