batchalign 0.7.1b6__tar.gz → 0.7.1b7__tar.gz

{batchalign-0.7.1b6/batchalign.egg-info → batchalign-0.7.1b7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.1b6
+Version: 0.7.1b7
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalign-0.7.1b6 → batchalign-0.7.1b7}/batchalign/cli/cli.py

@@ -149,19 +149,22 @@ def transcribe(ctx, in_dir, out_dir, lang, num_speakers, **kwargs):
     def loader(file):
         return file
 
+    asr = "rev"
+    if kwargs["whisper"]:
+        asr = "whisper"
+    if kwargs["whisperx"]:
+        asr = "whisperx"
+
+
     def writer(doc, output):
+        doc.content.insert(0, CustomLine(id="Comment", type=CustomLineType.INDEPENDENT,
+                                         content=f"Batchalign {VERSION_NUMBER.strip()}, ASR Engine {asr}"))
         CHATFile(doc=doc, special_mor_=True).write(output
                                                    .replace(".wav", ".cha")
                                                    .replace(".mp4", ".cha")
                                                    .replace(".mp3", ".cha"),
                                                    write_wor=kwargs.get("wor", False))
 
-    asr = "rev"
-    if kwargs["whisper"]:
-        asr = "whisper"
-    if kwargs["whisperx"]:
-        asr = "whisperx"
-
     if kwargs.get("diarize"):
         _dispatch("transcribe_s",
                   lang, num_speakers, ["mp3", "mp4", "wav"], ctx,

{batchalign-0.7.1b6 → batchalign-0.7.1b7}/batchalign/formats/chat/parser.py

@@ -222,7 +222,7 @@ def chat_parse_doc(lines, special_mor=False):
                 continue
             # we split because there are multiple languages possible 
             elif "@Languages" in line.strip():
-                results["langs"] = [i.strip() for i in line.strip("@Languages:").strip().split(",")]
+                results["langs"] = [i.strip() for i in line.strip("@Languages:").strip().replace(" ", ",").strip().split(",") if i.strip() != ""]
                 if len(results["langs"]) > 0 and results["langs"][0] == "eng" and special_mor:
                     use_special_mor = True
             # parse participants; the number of | delinates the metedata field

{batchalign-0.7.1b6 → batchalign-0.7.1b7}/batchalign/pipelines/morphosyntax/ud.py

@@ -808,7 +808,27 @@ def morphoanalyze(doc: Document, retokenize:bool, status_hook:callable = None, *
                 ut, end = chat_parse_utterance(" ".join([i.text for i in sents[0].tokens])+" "+ending,
                                           mor, gra,
                                           None, None)
+                # JANK add a space after every form being analyzed
+                text_fixed = []
+                text_orig = i.text
+                # we do this to force one replacement of the token
+                # for every input token
+                for i in sents[0].tokens:
+                    try:
+                        before, after = text_orig.split(i.text, 1)
+                        text_fixed.append(before.strip())
+                        text_fixed.append(i.text.strip())
+                        text_orig = after
+                    except ValueError:
+                        # we give up on that token; likely not found
+                        # because there we tokenization issues (i.e.
+                        # existing tokenization)
+                        continue
+                text_fixed.append(text_orig.strip())
+                text_fixed = " ".join(text_fixed).strip()
+                text_fixed = re.sub(r" +", " ", text_fixed)
                 doc.content[indx] = Utterance(content=ut,
+                                              text=text_fixed,
                                               tier=doc.content[indx].tier,
                                               time=doc.content[indx].time,
                                               custom_dependencies=doc.content[indx].custom_dependencies)

batchalign-0.7.1b7/batchalign/version

@@ -0,0 +1,3 @@
+0.7.1-beta.7
+May 21st, 2024
+insert debug info to transcribe file

{batchalign-0.7.1b6 → batchalign-0.7.1b7/batchalign.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: batchalign
-Version: 0.7.1b6
+Version: 0.7.1b7
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalign-0.7.1b6/batchalign/version

@@ -1,3 +0,0 @@
-0.7.1-beta.6
-May 17th, 2024
-remove Hebrew for UD