batchalign 0.7.7.post1__tar.gz → 0.7.9__tar.gz

{batchalign-0.7.7.post1/batchalign.egg-info → batchalign-0.7.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.7.post1
+Version: 0.7.9
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalign-0.7.7.post1 → batchalign-0.7.9}/batchalign/constants.py

@@ -1,5 +1,5 @@
 # CHAT punctuation specifications
-ENDING_PUNCT = [".", "?", "!", "+//.", "+/.", "+...", "+\"/.", "+..?", "+\".", "+//?", "+.", "+!?", "+/?", "..."]
+ENDING_PUNCT = [".", "?", "!", "+//.", "+/.", "+...", "+\"/.", "+..?", "+\".", "+//?", "+.", "+!?", "+/?", "...", "(.)"]
 MOR_PUNCT = ["‡", "„", ","]
 CHAT_IGNORE = ["xxx", "yyy", "www"]
 

{batchalign-0.7.7.post1 → batchalign-0.7.9}/batchalign/formats/chat/generator.py

@@ -41,9 +41,9 @@ def generate_chat_utterance(utterance: Utterance, special_mor=False, write_wor=T
         gras.append(i.dependency)
         if i.time:
             has_wor = True
-            wor_elems.append(f"{i.text} \x15{str(i.time[0])}_{str(i.time[1])}\x15")
+            wor_elems.append(re.sub(r"@\w+", "", f"{i.text} \x15{str(i.time[0])}_{str(i.time[1])}\x15"))
         else:
-            wor_elems.append(i.text)
+            wor_elems.append(re.sub(r"@\w+", "", i.text))
 
         if i.coreference:
             has_coref = True
@@ -124,7 +124,7 @@ def generate_chat_preamble(doc, birthdays=[]):
     header = []
     header.append("@Languages:\t"+", ".join(doc.langs))
     header.append("@Participants:\t"+", ".join([f"{i.id} {i.name}" for i in doc.tiers]))
-    header.append("@Options:\tmulti")
+    # header.append("@Options:\tmulti")
     header.append("\n".join([f"@ID:\t{i.lang}|{i.corpus}|{i.id}|{i.birthday}|{i.additional[0]}|{i.additional[1]}|{i.additional[2]}|{i.name}|{i.additional[3]}|{i.additional[4]}|" for i in doc.tiers]))
     for i in birthdays:
         header.append(f"@{i.id}:\t{i.content}")

{batchalign-0.7.7.post1 → batchalign-0.7.9}/batchalign/formats/chat/lexer.py

@@ -197,7 +197,7 @@ class UtteranceLexer:
         while True:
             res = self.__pull()
             try:
-                if res == "" or res == False or res in ENDING_PUNCT or (res[-1] in ENDING_PUNCT and re.findall("\w", res)):
+                if res == "" or res == False or res in ENDING_PUNCT or (res[-1] in ENDING_PUNCT and re.findall(r"\w", res)):
                     break
             except IndexError:
                 raise CHATValidationException(f"Lexer failed! Utterance ended without ending punct. Utterance: {self.raw}")

{batchalign-0.7.7.post1 → batchalign-0.7.9}/batchalign/formats/chat/parser.py

@@ -39,13 +39,13 @@ def chat_parse_utterance(text, mor, gra, wor, additional):
 
     # scan the timing
     # lex the utterance
-    to_lex = re.compile("\x15\d+_\d+\x15").sub("", text).strip()
+    to_lex = re.compile(r"\x15\d+_\d+\x15").sub("", text).strip()
 
     # if the first form has a < in it and has no words,
     # its probably a beginning delimiter which we do not lex
     if (len(to_lex) > 0 and
         ("<" in to_lex.split(" ")[0]  or "+" in to_lex.split(" ")[0] )
-        and not re.findall("\w", to_lex.split(" ")[0])):
+        and not re.findall(r"\w", to_lex.split(" ")[0])):
         beg = to_lex.split(" ")[0]
         to_lex = to_lex.replace(beg, "", 1)
 
@@ -64,9 +64,11 @@ def chat_parse_utterance(text, mor, gra, wor, additional):
     # fix commas for people that don't annotate commas with a space
     to_lex = to_lex.replace(",", " ,")
 
-    to_lex = re.sub(r"\([\d.:]+\)", "", to_lex)
+    to_lex = re.sub(r"\([\d.:]+\)(?!$)", "", to_lex)
     to_lex = re.sub(r"↫.*?↫", "", to_lex)
 
+    to_lex = re.sub(r"\(.\)$", r"$END_SPC$", to_lex)
+
     # if there is a punct, move it
     for end in sorted(ENDING_PUNCT, key=len, reverse=True):
         if end in to_lex:
@@ -76,6 +78,8 @@ def chat_parse_utterance(text, mor, gra, wor, additional):
     to_lex = to_lex.replace("  ", " ")
 
     tokens = lex(to_lex)
+    if tokens[-1][0] == "END_SPC":
+        tokens = tokens[:-1] + [("(.)", TokenType.PUNCT)]
 
     # correct 0 forms
     res = []

{batchalign-0.7.7.post1 → batchalign-0.7.9}/batchalign/pipelines/morphosyntax/ud.py

@@ -793,7 +793,7 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
         except AttributeError:
             breakpoint()
 
-        if re.findall("\w", ending):
+        if re.findall(r"\w", ending):
             ending = "."
             line_cut = i.strip(join_with_spaces=True)
         else:

batchalign-0.7.9/batchalign/version

@@ -0,0 +1,3 @@
+0.7.9
+Janurary 8st, 2025
+align fixes

{batchalign-0.7.7.post1 → batchalign-0.7.9/batchalign.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.7.post1
+Version: 0.7.9
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalign-0.7.7.post1/batchalign/version

@@ -1,3 +0,0 @@
-0.7.7-post.1
-Janurary 3st, 2025
-releasing new full version