PyPI - batchalign - Versions diffs - 0.7.6a25__tar.gz → 0.7.6a27__tar.gz - Mend

@@ -890,14 +890,15 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                 collected = ""
                 # brrr
                 aligned = align(chunks_chars, ud_chars, tqdm=False)
                 for i in aligned:
                     if isinstance(i, Match):
                         if i.reference_payload not in chunks_backplate[i.payload]:
                             chunks_backplate[i.payload].append(i.reference_payload)
                     elif isinstance(i, Extra) and i.extra_type == ExtraType.PAYLOAD:
-                        if i.key == "*":
+                        if i.key == "⁎":
                             creaky = not creaky
-                            chunks_backplate[i.payload].append("*"+collected+"*")
+                            chunks_backplate[i.payload].append("⁎"+collected+"⁎")
                             collected = ""
                         elif creaky:
                             collected += i.key
@@ -938,7 +939,8 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                 retokenized_ut = re.sub(r"@ ?w ?p", "@wp", retokenized_ut)
                 retokenized_ut = retokenized_ut.replace(" @", "@")
                 retokenized_ut = re.sub(r"\*[* ]*", "*", retokenized_ut)
-                retokenized_ut = re.sub(r"\*(.*?)\*", r"*\1* ", retokenized_ut)
+                retokenized_ut = re.sub(r"⁎[⁎ ]*(.*?)[⁎ ]*⁎", r"⁎\1⁎ ", retokenized_ut)
+                retokenized_ut = re.sub(r"\[\*(.)\]", r"[* \1]", retokenized_ut)
                 # pray to everyone that it works---this will simply crash and ignore
                 # the utterance if it didn't work, so we are doing this as a sanity
                 # check rather than needing the parsed result

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.6a25
+Version: 0.7.6a27
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

@@ -0,0 +1,3 @@
+0.7.6-alpha.27
+November 10, 2024
+Japanese changes + packaging

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.6a25
+Version: 0.7.6a27
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

@@ -1,3 +0,0 @@
-0.7.6-alpha.25
-November 8, 2024
-Japanese changes + packaging

batchalign 0.7.6a25__tar.gz → 0.7.6a27__tar.gz

batchalign 0.7.6a25tar.gz → 0.7.6a27tar.gz