batchalign 0.7.22.post13__tar.gz → 0.7.22.post16__tar.gz

{batchalign-0.7.22.post13/batchalign.egg-info → batchalign-0.7.22.post16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: batchalign
-Version: 0.7.22.post13
+Version: 0.7.22.post16
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalign-0.7.22.post13 → batchalign-0.7.22.post16}/batchalign/document.py

@@ -206,7 +206,7 @@ class Utterance(BaseModel):
     def __len__(self):
         return len(self.content)
 
-    def tostring(self, always_detokenize=False):
+    def tostring(self, always_detokenize=False, merge_abbrev=False):
         if self.text != None and not always_detokenize:
             t = self.text
         else:
@@ -225,10 +225,11 @@ class Utterance(BaseModel):
         t = re.sub(r" +", " ", t.strip()).strip()
         t = t.replace("+ ,", "+,").strip()
 
-        abbrevs =  [" "+" " .join(list(i))+" " for i in abbrev]
-        for i in abbrevs:
-            t = t.replace(i, " "+i.replace(" ", "")+" ")
-            t = t.replace(i.lower(), " "+i.replace(" ", "")+" ")
+        if merge_abbrev:
+            abbrevs =  [" "+" " .join(list(i))+" " for i in abbrev]
+            for i in abbrevs:
+                t = t.replace(i, " "+i.replace(" ", "")+" ")
+                t = t.replace(i.lower(), " "+i.replace(" ", "")+" ")
 
         return t
 

{batchalign-0.7.22.post13 → batchalign-0.7.22.post16}/batchalign/formats/chat/file.py

@@ -128,7 +128,8 @@ class CHATFile(BaseFormat):
                 continue
             else:
                 main.append(generate_chat_utterance(i, special and doc.langs[0] == "eng",
-                                                    write_wor=write_wor))
+                                                    write_wor=write_wor,
+                                                    merge_abbrev="eng" in doc.langs))
         main.append("@End\n")
 
         raw = "\n".join(main)

{batchalign-0.7.22.post13 → batchalign-0.7.22.post16}/batchalign/formats/chat/generator.py

@@ -11,7 +11,7 @@ import warnings
 # document[3].text = None
 # document[3].model_dump()
 
-def generate_chat_utterance(utterance: Utterance, special_mor=False, write_wor=True):
+def generate_chat_utterance(utterance: Utterance, special_mor=False, write_wor=True, merge_abbrev=False):
     """Converts at Utterance to a CHAT string.
 
     Parameters
@@ -27,7 +27,7 @@ def generate_chat_utterance(utterance: Utterance, special_mor=False, write_wor=T
         The generated string.
     """
     
-    main_line = str(utterance)
+    main_line = utterance.tostring(merge_abbrev=merge_abbrev)
     # last minut ecorrections
     # main_line = re.sub(r"<([\w ]+) \[\/", r"<\1> [/", main_line)
     main_line = re.sub(r"«", "“", main_line)
@@ -98,7 +98,7 @@ def generate_chat_utterance(utterance: Utterance, special_mor=False, write_wor=T
 
     #### WOR LINE GENERATION ####
     if has_wor and write_wor:
-        result.append("%wor:\t"+" ".join(wor_elems))
+        result.append(("%wor:\t"+" ".join(wor_elems)).replace("-\x15", "- \x15"))
     if has_coref:
         result.append("%coref:\t"+(", ".join(coref_elems)))
     if utterance.translation != None and utterance.translation.strip() not in ["", ".", "!", "?"]:

batchalign-0.7.22.post16/batchalign/version

@@ -0,0 +1,3 @@
+0.7.22-post.16
+December 27th, 2025
+fix wor lines

{batchalign-0.7.22.post13 → batchalign-0.7.22.post16/batchalign.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: batchalign
-Version: 0.7.22.post13
+Version: 0.7.22.post16
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalign-0.7.22.post13/batchalign/version

@@ -1,3 +0,0 @@
-0.7.22-post.13
-November 23rd, 2025
-Remove bullets