batchalign 0.7.1b11__tar.gz → 0.7.1b13__tar.gz

{batchalign-0.7.1b11/batchalign.egg-info → batchalign-0.7.1b13}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.1b11
+Version: 0.7.1b13
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalign-0.7.1b11 → batchalign-0.7.1b13}/batchalign/document.py

@@ -83,6 +83,7 @@ class TokenType(IntEnum):
     ANNOT = 4 # &~ject &~head
     PUNCT = 5 # ‡„,
     CORRECTION = 6 # test [= test]
+    VOCAL = 7 # test@o
 
 class CustomLineType(IntEnum):
     DEPENDENT = 0 # %com

{batchalign-0.7.1b11 → batchalign-0.7.1b13}/batchalign/formats/chat/lexer.py

@@ -93,6 +93,8 @@ class UtteranceLexer:
             self.__clauses.append((form, TokenType.FEAT))
         elif annotation_clean(form).strip() in CHAT_IGNORE:
             self.__clauses.append((annotation_clean(form).strip(), TokenType.ANNOT))
+        elif "@" in form:
+            self.__clauses.append((annotation_clean(form).strip(), TokenType.VOCAL))
         else:
             self.__clauses.append((annotation_clean(form).strip(), TokenType.REGULAR))
 

{batchalign-0.7.1b11 → batchalign-0.7.1b13}/batchalign/formats/chat/parser.py

@@ -90,9 +90,10 @@ def chat_parse_utterance(text, mor, gra, wor, additional):
     # seperate out main words by whether it should have phonation/morphology and add ending punct
     words = list(enumerate(tokens))
     lexed_words = [tok for tok in words if tok[1][1] in [TokenType.REGULAR,
-                                                            TokenType.PUNCT]]
+                                                         TokenType.PUNCT]]
     phonated_words = [tok for tok in words if tok[1][1] in [TokenType.REGULAR,
                                                             TokenType.RETRACE,
+                                                            TokenType.VOCAL,
                                                             TokenType.PUNCT,
                                                             TokenType.FP]]
     # create base forms

{batchalign-0.7.1b11 → batchalign-0.7.1b13}/batchalign/pipelines/dispatch.py

@@ -73,7 +73,7 @@ def dispatch_pipeline(pkg_str, lang, num_speakers=None, **arg_overrides):
             packages.append("disfluency")
         if "retracing" not in packages:
             packages.append("retracing")
-        if "utterance" not in packages and resolve("utterance", lang) == None and lang not in ["heb"]:
+        if "utterance" not in packages and resolve("utterance", lang) == None and lang not in ["heb", "fra"]:
             packages.append("utterance")
     if "fa" in packages:
         if "utr" not in packages:

{batchalign-0.7.1b11 → batchalign-0.7.1b13}/batchalign/pipelines/morphosyntax/ud.py

@@ -130,6 +130,8 @@ def handler(word, lang=None):
 
     # fix dash
     target = target.replace("-", "–")
+    if target == "“":
+        target = word.text
 
     return f"{'' if not unknown else '0'}{word.upos.lower()}|{target}"
 
@@ -739,6 +741,7 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
             line_cut = i.strip(join_with_spaces=True)
             ending = '.'
 
+
         # clean the sentence
         line_cut = clean_sentence(line_cut)
 
@@ -813,12 +816,12 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                 chunks = list(enumerate(doc.content[indx].text.split(" ")))
                 # filter out everything that could not possibly align
                 chunks_align = [(i,j) for i,j in chunks
-                                if len(j) != 0 and (j[0] not in ["<", "[", "&", "\x15"]) and (j[-1] not in ["]"])
+                                if len(j) != 0 and (j[0] not in ["<", "[", "&", "\x15", "(", ")"]) and (j[-1] not in ["]"])
                                                    and ("@" not in j)
                                 and j.strip() not in ENDING_PUNCT + MOR_PUNCT + CHAT_IGNORE + ["++"]]
                 # hollow out anything we are trying to align, and leave everything else
                 chunks_backplate = [[j] 
-                                    if not (len(j) != 0 and (j[0] not in ["<", "[", "&", "\x15"]) and (j[-1] not in ["]"])
+                                    if not (len(j) != 0 and (j[0] not in ["<", "[", "&", "\x15", "(", ")"]) and (j[-1] not in ["]"])
                                     and ("@" not in j)
                                             and j.strip() not in ENDING_PUNCT + MOR_PUNCT + CHAT_IGNORE + ["++"])
                                     else
@@ -857,6 +860,7 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                 retokenized_ut = retokenized_ut.replace(": <", ": <")
                 retokenized_ut = re.sub(r"@ ?w ?p", "@wp", retokenized_ut)
                 retokenized_ut = retokenized_ut.replace(" @", "@")
+                retokenized_ut = re.sub(r" ↑", "↑", 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

batchalign-0.7.1b13/batchalign/version

@@ -0,0 +1,3 @@
+0.7.1-beta.13
+June 2nd, 2024
+more ud fixes

{batchalign-0.7.1b11 → batchalign-0.7.1b13/batchalign.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.1b11
+Version: 0.7.1b13
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalign-0.7.1b11/batchalign/version

@@ -1,3 +0,0 @@
-0.7.1-beta.11
-May 27th, 2024
-patch minor ud bugs