fix(crg): manually clean up crg-ipa -> eng-ipa

roedoejet · Aug 25, 2023 · ec105eb · ec105eb
1 parent a5f50c8
commit ec105eb
Show file tree

Hide file tree

Showing 3 changed files with 31 additions and 56 deletions.
diff --git a/g2p/mappings/langs/generated/config.yaml b/g2p/mappings/langs/generated/config.yaml
@@ -130,6 +130,7 @@ mappings:
     rule_ordering: as-written
   - authors:
       - Generated 2020-09-16 17:14:33.602655
+      - Eric Joanis manual fixes 2023-08-25
     case_sensitive: false
     display_name: crg-ipa IPA to eng-ipa IPA
     escape_special: false

diff --git a/g2p/mappings/langs/generated/crg-ipa_to_eng-ipa.json b/g2p/mappings/langs/generated/crg-ipa_to_eng-ipa.json
@@ -1,58 +1,32 @@
 [
-    {"in": "ŋ", "out": "ŋ", "context_before": "", "context_after": ""},
-    {"in": "n", "out": "n", "context_before": "", "context_after": ""},
-    {"in": "ɛ̃ː", "out": "ɛ̃", "context_before": "", "context_after": ""},
-    {"in": "oaw", "out": "oːɑw", "context_before": "", "context_after": ""},
-    {"in": "aw", "out": "ɑw", "context_before": "", "context_after": ""},
-    {"in": "aj", "out": "ɑj", "context_before": "", "context_after": ""},
-    {"in": "ɑː", "out": "ɑ", "context_before": "", "context_after": ""},
-    {"in": "waj", "out": "wɑj", "context_before": "", "context_after": ""},
-    {"in": "æ", "out": "æ", "context_before": "", "context_after": ""},
-    {"in": "iː", "out": "i", "context_before": "", "context_after": ""},
-    {"in": "eː", "out": "eː", "context_before": "", "context_after": ""},
-    {"in": "uː", "out": "u", "context_before": "", "context_after": ""},
-    {"in": "ɑ̃ː", "out": "ɑ̃", "context_before": "", "context_after": ""},
-    {"in": "ĩ", "out": "ĩ", "context_before": "", "context_after": ""},
-    {"in": "ĩː", "out": "ĩ", "context_before": "", "context_after": ""},
-    {"in": "ɔ̃ː", "out": "ɔ̃", "context_before": "", "context_after": ""},
-    {"in": "ɔ̃", "out": "ɔ̃", "context_before": "", "context_after": ""},
-    {"in": "o", "out": "ɔ", "context_before": "", "context_after": ""},
-    {"in": "ʰp", "out": "p", "context_before": "", "context_after": ""},
-    {"in": "ʰt", "out": "t", "context_before": "", "context_after": ""},
-    {"in": "ʰk", "out": "k", "context_before": "", "context_after": ""},
-    {"in": "ʃ", "out": "ʃ", "context_before": "", "context_after": ""},
-    {"in": "ʒ", "out": "ʒ", "context_before": "", "context_after": ""},
-    {"in": "tʃ", "out": "t͡ʃ", "context_before": "", "context_after": ""},
-    {"in": "ʰtʃ", "out": "t͡ʃ", "context_before": "", "context_after": ""},
-    {"in": "ɛ", "out": "ɛ", "context_before": "", "context_after": ""},
-    {"in": "ɪ", "out": "ɪ", "context_before": "", "context_after": ""},
-    {"in": "dʒ", "out": "dʒ", "context_before": "", "context_after": ""},
-    {"in": "j", "out": "j", "context_before": "", "context_after": ""},
-    {"in": "ŋ", "out": "ŋ", "context_before": "", "context_after": ""},
-    {"in": "n", "out": "n", "context_before": "", "context_after": ""},
-    {"in": "ɛ̃ː", "out": "ɛ̃", "context_before": "", "context_after": ""},
-    {"in": "oaw", "out": "oːɑw", "context_before": "", "context_after": ""},
-    {"in": "aw", "out": "ɑw", "context_before": "", "context_after": ""},
-    {"in": "aj", "out": "ɑj", "context_before": "", "context_after": ""},
-    {"in": "ɑː", "out": "ɑ", "context_before": "", "context_after": ""},
-    {"in": "waj", "out": "wɑj", "context_before": "", "context_after": ""},
-    {"in": "æ", "out": "æ", "context_before": "", "context_after": ""},
-    {"in": "iː", "out": "i", "context_before": "", "context_after": ""},
-    {"in": "eː", "out": "eː", "context_before": "", "context_after": ""},
-    {"in": "uː", "out": "u", "context_before": "", "context_after": ""},
-    {"in": "ɑ̃ː", "out": "ɑ̃", "context_before": "", "context_after": ""},
-    {"in": "ĩ", "out": "ĩ", "context_before": "", "context_after": ""},
-    {"in": "ĩː", "out": "ĩ", "context_before": "", "context_after": ""},
-    {"in": "ɔ̃", "out": "ɔ̃", "context_before": "", "context_after": ""},
-    {"in": "ʰp", "out": "p", "context_before": "", "context_after": ""},
-    {"in": "ʰt", "out": "t", "context_before": "", "context_after": ""},
-    {"in": "ʰk", "out": "k", "context_before": "", "context_after": ""},
-    {"in": "ʃ", "out": "ʃ", "context_before": "", "context_after": ""},
-    {"in": "ʒ", "out": "ʒ", "context_before": "", "context_after": ""},
-    {"in": "tʃ", "out": "t͡ʃ", "context_before": "", "context_after": ""},
-    {"in": "ʰtʃ", "out": "t͡ʃ", "context_before": "", "context_after": ""},
-    {"in": "ɛ", "out": "ɛ", "context_before": "", "context_after": ""},
-    {"in": "ɪ", "out": "ɪ", "context_before": "", "context_after": ""},
-    {"in": "dʒ", "out": "dʒ", "context_before": "", "context_after": ""},
-    {"in": "j", "out": "j", "context_before": "", "context_after": ""}
+    {"in": "ŋ", "out": "ŋ"},
+    {"in": "n", "out": "n"},
+    {"in": "ɛ̃ː", "out": "ɛ̃"},
+    {"in": "oaw", "out": "oːɑw"},
+    {"in": "aw", "out": "ɑw"},
+    {"in": "aj", "out": "ɑj"},
+    {"in": "ɑː", "out": "ɑ"},
+    {"in": "waj", "out": "wɑj"},
+    {"in": "æ", "out": "æ"},
+    {"in": "iː", "out": "i"},
+    {"in": "eː", "out": "eː"},
+    {"in": "uː", "out": "u"},
+    {"in": "ɑ̃ː", "out": "ɑ̃"},
+    {"in": "ĩ", "out": "ĩ"},
+    {"in": "ĩː", "out": "ĩ"},
+    {"in": "ɔ̃ː", "out": "ɔ̃"},
+    {"in": "ɔ̃", "out": "ɔ̃"},
+    {"in": "o", "out": "ɔ"},
+    {"in": "ʊ", "out": "ʊ"},
+    {"in": "ʰp", "out": "p"},
+    {"in": "ʰt", "out": "t"},
+    {"in": "ʰk", "out": "k"},
+    {"in": "ʃ", "out": "ʃ"},
+    {"in": "ʒ", "out": "ʒ"},
+    {"in": "tʃ", "out": "t͡ʃ"},
+    {"in": "ʰtʃ", "out": "t͡ʃ"},
+    {"in": "ɛ", "out": "ɛ"},
+    {"in": "ɪ", "out": "ɪ"},
+    {"in": "dʒ", "out": "dʒ"},
+    {"in": "j", "out": "j"}
 ]
diff --git a/g2p/mappings/langs/langs.pkl b/g2p/mappings/langs/langs.pkl