batchalign 0.7.3b0__tar.gz → 0.7.3b2__tar.gz

{batchalign-0.7.3b0/batchalign.egg-info → batchalign-0.7.3b2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.3b0
+Version: 0.7.3b2
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalign-0.7.3b0 → batchalign-0.7.3b2}/batchalign/document.py

@@ -256,6 +256,7 @@ class Utterance(BaseModel):
         detokenized = " ".join(result)
         # replace suprious spaces caused by edge fudging
         detokenized = re.sub(r"< +", "<", detokenized)
+        detokenized = re.sub(r" >", ">", detokenized)
 
         # check and seperate punct 
         last_tok = result[-1]

{batchalign-0.7.3b0 → batchalign-0.7.3b2}/batchalign/pipelines/asr/whisper.py

@@ -36,6 +36,8 @@ class WhisperEngine(BatchalignEngine):
         language = pycountry.languages.get(alpha_3=lang).name
         if language == "Yue Chinese":
             language = "Cantonese"
+        if "greek" in language.lower():
+            language = "Greek"
             
         self.__whisper = WhisperASRModel(model, base=base, language=language)
         self.__lang = lang

{batchalign-0.7.3b0 → batchalign-0.7.3b2}/batchalign/pipelines/cleanup/retrace.py

@@ -43,6 +43,8 @@ class NgramRetraceEngine(BatchalignEngine):
                         for j in content[begin:begin+n]:
                             if j.type != TokenType.FP:
                                 j.type = TokenType.RETRACE
+                                for p in ENDING_PUNCT + MOR_PUNCT:
+                                    j.text = j.text.replace(p, "").strip()
                         root = root+n
                     # we scan grams forward one by one
                     begin += 1

{batchalign-0.7.3b0 → batchalign-0.7.3b2}/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", "fra", "ara"]:
+        if "utterance" not in packages and resolve("utterance", lang) == None and lang not in ["heb", "fra", "ara", "nld", "hun", "eus", "grc", "ell"]:
             packages.append("utterance")
     if "fa" in packages:
         if "utr" not in packages:

batchalign-0.7.3b2/batchalign/version

@@ -0,0 +1,3 @@
+0.7.3-beta.2
+June 7, 2024
+more languages

{batchalign-0.7.3b0 → batchalign-0.7.3b2/batchalign.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.3b0
+Version: 0.7.3b2
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalign-0.7.3b0/batchalign/version

@@ -1,3 +0,0 @@
-0.7.3-beta.0
-June 6, 2024
-Skip everything that's just POS of X and PUNCT