batchalign 0.7.6a11__tar.gz → 0.7.6a13__tar.gz

{batchalign-0.7.6a11/batchalign.egg-info → batchalign-0.7.6a13}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.6a11
+Version: 0.7.6a13
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalign-0.7.6a11 → batchalign-0.7.6a13}/batchalign/document.py

@@ -324,8 +324,7 @@ class Utterance(BaseModel):
             to_include.append(TokenType.RETRACE)
         if include_fp:
             to_include.append(TokenType.FP)
-        filtered = filter(lambda x:x.type in to_include,
-                          self.content)
+        filtered = filter(lambda x:x.type in to_include, self.content)
         # chain them together
         if join_with_spaces:
             return " ".join([i.text for i in filtered])

{batchalign-0.7.6a11 → batchalign-0.7.6a13}/batchalign/formats/chat/utils.py

@@ -146,7 +146,7 @@ def annotation_clean(content, special=False):
     cleaned_word = cleaned_word.replace("~","").replace("&~","")
     cleaned_word = cleaned_word.replace(">","").replace("<","")
     cleaned_word = cleaned_word.replace("〕","").replace("//","").replace(";","")
-    cleaned_word = re.sub(r"@[^abcefpoqs]", '', cleaned_word)
+    cleaned_word = re.sub(r"@[^abcefpoqsw]", '', cleaned_word)
     cleaned_word = re.sub(r"&.", '', cleaned_word)
 
     return cleaned_word

{batchalign-0.7.6a11 → batchalign-0.7.6a13}/batchalign/pipelines/morphosyntax/ud.py

@@ -280,6 +280,8 @@ def handler__PUNCT(word, lang=None):
                                         # instead of the lemma, which maybe entirely weird
         if word.text == "もん":
             return f"part|{word.text}"
+        if word.text == ",":
+            return f"cm|cm"
         else:
             return f"x|{word.text}"
 
@@ -835,11 +837,17 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                 L.debug(f"Encountered an utterance that's likely devoid of morphological information; skipping... utterance='{doc.content[indx]}'")
                 continue
 
+
             if retokenize:
                 # rewrite the sentence with our desired tokenizations
                 ut, end = chat_parse_utterance(" ".join([i.text for i in sents[0].tokens])+" "+ending,
                                           mor, gra,
                                           None, None)
+                # fix xbxxx
+                for i in ut:
+                    if i.text == "xbxxx" and len(i.morphology) > 0:
+                        i.text = i.morphology[0].lemma
+
                 # split the text up into previous chunks
                 chunks = list(enumerate(doc.content[indx].text.split(" ")))
                 # filter out everything that could not possibly align
@@ -876,11 +884,12 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                     # we want to replace the morphology of forms that are not actually
                     # supposed to be analyzed
                     elif isinstance(i, Extra) and i.extra_type == ExtraType.REFERENCE:
-                        ut[i.payload].morphology = [Morphology(
-                            lemma = sents[0].tokens[i.payload].text,
-                            pos = "x",
-                            feats = ""
-                        )]
+                        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",
+                                feats = ""
+                            )]
 
                 poses = [i.morphology[0].pos.upper() for i in ut
                          if i.morphology

batchalign-0.7.6a13/batchalign/version

@@ -0,0 +1,3 @@
+0.7.6-alpha.13
+October 13, 2024
+patch bug regarding comma structure

{batchalign-0.7.6a11 → batchalign-0.7.6a13/batchalign.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.6a11
+Version: 0.7.6a13
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalign-0.7.6a11/batchalign/version

@@ -1,3 +0,0 @@
-0.7.6-alpha.11
-October 10, 2024
-patch urgent UD bug