batchalign 0.7.17.post12__tar.gz → 0.7.17.post14__tar.gz

{batchalign-0.7.17.post12/batchalign.egg-info → batchalign-0.7.17.post14}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: batchalign
-Version: 0.7.17.post12
+Version: 0.7.17.post14
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalign-0.7.17.post12 → batchalign-0.7.17.post14}/batchalign/formats/chat/generator.py

@@ -57,7 +57,7 @@ def generate_chat_utterance(utterance: Utterance, special_mor=False, write_wor=T
                 if j.end:
                     coref_str += ")"
                 coref_str_form.append(coref_str)
-            coref_elems.append(",".join(coref_str_form))
+            coref_elems.append(" ".join(coref_str_form))
         else:
             coref_elems.append("-")
 
@@ -94,7 +94,7 @@ def generate_chat_utterance(utterance: Utterance, special_mor=False, write_wor=T
     if has_wor and write_wor:
         result.append("%wor:\t"+" ".join(wor_elems))
     if has_coref:
-        result.append("%coref:\t"+" ".join(coref_elems))
+        result.append("%coref:\t"+(", ".join(coref_elems)))
     if utterance.translation != None:
         result.append("%xtra:\t"+utterance.translation)
 

{batchalign-0.7.17.post12 → batchalign-0.7.17.post14}/batchalign/pipelines/morphosyntax/ud.py

@@ -462,7 +462,7 @@ def parse_sentence(sentence, delimiter=".", special_forms=[], lang="$nospecial$"
             # specivl forms: recall the special form marker is xbxxx
             if "xbxxx" in word.text.strip():
                 form = special_forms.pop(0)
-                mor.append(f"x|{form.strip().replace(',', 'cm')}")
+                mor.append(f"{form[1].strip()}|{form[0].strip().replace(',', 'cm')}")
                 special_form_ids.append(word.id)
             else:
                 mor.append(mor_word)
@@ -555,7 +555,6 @@ def parse_sentence(sentence, delimiter=".", special_forms=[], lang="$nospecial$"
                           # add a deliminator
         mor_str = mor_str + " " + delimiter
 
-
     mor_str = mor_str.replace("<UNK>", "")
     gra_str = gra_str.replace("<UNK>", "")
 
@@ -843,7 +842,7 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
         special_forms_cleaned = []
         for form in special_forms:
             line_cut = line_cut.replace(form, "xbxxx")
-            special_forms_cleaned.append(re.sub(r"@[\w\:]+", "", form).strip())
+            special_forms_cleaned.append(form.split("@"))
 
         # if line cut is still nothing, we get very angry
         if line_cut == "":
@@ -942,7 +941,7 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                         if ut[i.payload].text != ",":
                             ut[i.payload].morphology = [Morphology(
                                 lemma = sents[0].tokens[i.payload].text if len(sents) > 0 and len(sents[0].tokens) > i.payload and sents[0].tokens[i.payload].text != "xbxxx" else ut[i.payload].text,
-                                pos = "x",
+                                pos = ut[i.payload].morphology[0].pos if (ut[i.payload].morphology and len(ut[i.payload].morphology) > 0) else "x",
                                 feats = ""
                             )]
                 poses = [i.morphology[0].pos.upper() for i in ut
@@ -999,6 +998,7 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                     content.dependency = form.dependency
 
         except Exception as e:
+            raise e
             pass
             # warnings.warn(f"Utterance failed parsing, skipping ud tagging... line='{line}', error='{e}'.\n")
 

batchalign-0.7.17.post14/batchalign/version

@@ -0,0 +1,3 @@
+0.7.17-post.14
+April 06th, 2025
+Japanese form fixes

{batchalign-0.7.17.post12 → batchalign-0.7.17.post14/batchalign.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: batchalign
-Version: 0.7.17.post12
+Version: 0.7.17.post14
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalign-0.7.17.post12/batchalign/version

@@ -1,3 +0,0 @@
-0.7.17-post.12
-March 26th, 2025
-better coref model