Anchor-annotator 0.3.2__tar.gz → 0.4.0__tar.gz

{anchor_annotator-0.3.2 → anchor_annotator-0.4.0/Anchor_annotator.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: Anchor_annotator
-Version: 0.3.2
+Version: 0.4.0
 Summary: Anchor annotator is a program for inspecting corpora for the Montreal Forced Aligner and correcting transcriptions and pronunciations.
 Home-page: https://github.com/MontrealCorpusTools/Anchor-annotator
 Author: Montreal Corpus Tools

{anchor_annotator-0.3.2/Anchor_annotator.egg-info → anchor_annotator-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: Anchor_annotator
-Version: 0.3.2
+Version: 0.4.0
 Summary: Anchor annotator is a program for inspecting corpora for the Montreal Forced Aligner and correcting transcriptions and pronunciations.
 Home-page: https://github.com/MontrealCorpusTools/Anchor-annotator
 Author: Montreal Corpus Tools

{anchor_annotator-0.3.2 → anchor_annotator-0.4.0}/anchor/_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.3.2'
-__version_tuple__ = version_tuple = (0, 3, 2)
+__version__ = version = '0.4.0'
+__version_tuple__ = version_tuple = (0, 4, 0)

{anchor_annotator-0.3.2 → anchor_annotator-0.4.0}/anchor/main.py

@@ -540,14 +540,14 @@ class MainWindow(QtWidgets.QMainWindow):
             self.dictionary_model,
         )
         self.ui.speakerWidget.set_models(
-            self.corpus_model, self.selection_model, self.speaker_model
+            self.corpus_model, self.file_selection_model, self.speaker_model
         )
         self.ui.transcriptionWidget.set_models(self.corpus_model, self.dictionary_model)
         self.ui.alignmentWidget.set_models(self.corpus_model)
         self.ui.acousticModelWidget.set_models(self.corpus_model)
         self.ui.languageModelWidget.set_models(self.corpus_model)
         self.ui.dictionaryWidget.set_models(self.dictionary_model)
-        self.ui.diarizationWidget.set_models(self.diarization_model, self.selection_model)
+        self.ui.diarizationWidget.set_models(self.diarization_model, self.file_selection_model)
         self.ui.oovWidget.set_models(self.oov_model)
         self.file_selection_model.currentUtteranceChanged.connect(self.change_utterance)
         self.selection_model.fileViewRequested.connect(self.file_selection_model.set_current_file)
@@ -1087,8 +1087,8 @@ class MainWindow(QtWidgets.QMainWindow):
             )
             for m in (
                 session.query(anchor.db.AcousticModel)
-                .order_by(anchor.db.AcousticModel.last_used.desc())
                 .filter_by(available_locally=True)
+                .order_by(anchor.db.AcousticModel.last_used.desc())
             ):
                 a = QtGui.QAction(f"{m.path} [{m.name}]", parent=self)
                 a.setData(m.id)
@@ -1105,8 +1105,8 @@ class MainWindow(QtWidgets.QMainWindow):
 
             for m in (
                 session.query(anchor.db.Dictionary)
-                .order_by(anchor.db.Dictionary.last_used.desc())
                 .filter_by(available_locally=True)
+                .order_by(anchor.db.Dictionary.last_used.desc())
             ):
                 a = QtGui.QAction(text=f"{m.path} [{m.name}]", parent=self)
                 a.setData(m.id)
@@ -1122,8 +1122,8 @@ class MainWindow(QtWidgets.QMainWindow):
 
             for m in (
                 session.query(anchor.db.LanguageModel)
-                .order_by(anchor.db.LanguageModel.last_used.desc())
                 .filter_by(available_locally=True)
+                .order_by(anchor.db.LanguageModel.last_used.desc())
             ):
                 a = QtGui.QAction(text=f"{m.path} [{m.name}]", parent=self)
                 a.setData(m.id)
@@ -1139,8 +1139,8 @@ class MainWindow(QtWidgets.QMainWindow):
 
             for m in (
                 session.query(anchor.db.G2PModel)
-                .order_by(anchor.db.G2PModel.last_used.desc())
                 .filter_by(available_locally=True)
+                .order_by(anchor.db.G2PModel.last_used.desc())
             ):
                 a = QtGui.QAction(text=f"{m.path} [{m.name}]", parent=self)
                 a.setData(m.id)
@@ -1385,6 +1385,8 @@ class MainWindow(QtWidgets.QMainWindow):
         self.ui.languageModelMenu.setEnabled(True)
 
     def finalize_load_g2p_model(self, generator: PyniniValidator):
+        if generator is None:
+            return
         self.dictionary_model.set_g2p_generator(generator)
         self.corpus_model.g2p_model = generator.g2p_model
         self.check_actions()
@@ -1932,7 +1934,7 @@ class MainWindow(QtWidgets.QMainWindow):
                 session.query(anchor.db.AnchorCorpus).filter_by(current=True).update(
                     {anchor.db.AnchorCorpus.acoustic_model_id: m_id}
                 )
-                session.query(anchor.db.AcousticModel).filter_by(
+                session.query(anchor.db.AcousticModel).filter(
                     anchor.db.AcousticModel.id == m_id
                 ).update(
                     {
@@ -2003,9 +2005,7 @@ class MainWindow(QtWidgets.QMainWindow):
                 session.query(anchor.db.AnchorCorpus).filter_by(current=True).update(
                     {anchor.db.AnchorCorpus.dictionary_id: m_id}
                 )
-                session.query(anchor.db.Dictionary).filter_by(
-                    anchor.db.Dictionary.id == m_id
-                ).update(
+                session.query(anchor.db.Dictionary).filter(anchor.db.Dictionary.id == m_id).update(
                     {
                         anchor.db.Dictionary.last_used: datetime.datetime.now(),
                     }
@@ -2060,7 +2060,7 @@ class MainWindow(QtWidgets.QMainWindow):
                 session.query(anchor.db.AnchorCorpus).filter_by(current=True).update(
                     {anchor.db.AnchorCorpus.language_model_id: m_id}
                 )
-                session.query(anchor.db.LanguageModel).filter_by(
+                session.query(anchor.db.LanguageModel).filter(
                     anchor.db.LanguageModel.id == m_id
                 ).update(
                     {
@@ -2139,7 +2139,7 @@ class MainWindow(QtWidgets.QMainWindow):
                         anchor.db.AnchorCorpus.g2p_model_id: m_id,
                     }
                 )
-                session.query(anchor.db.G2PModel).filter_by(anchor.db.G2PModel.id == m_id).update(
+                session.query(anchor.db.G2PModel).filter(anchor.db.G2PModel.id == m_id).update(
                     {
                         anchor.db.G2PModel.last_used: datetime.datetime.now(),
                     }
@@ -2182,7 +2182,7 @@ class MainWindow(QtWidgets.QMainWindow):
                 session.query(anchor.db.AnchorCorpus).filter_by(current=True).update(
                     {anchor.db.AnchorCorpus.ivector_extractor_id: m_id}
                 )
-                session.query(anchor.db.IvectorExtractor).filter_by(
+                session.query(anchor.db.IvectorExtractor).filter(
                     anchor.db.IvectorExtractor.id == m_id
                 ).update(
                     {

{anchor_annotator-0.3.2 → anchor_annotator-0.4.0}/anchor/models.py

@@ -24,6 +24,12 @@ from montreal_forced_aligner.corpus.acoustic_corpus import (
 )
 from montreal_forced_aligner.data import PhoneType, WordType
 from montreal_forced_aligner.db import File, Phone, Speaker, Utterance
+from montreal_forced_aligner.dictionary.mixins import (
+    DEFAULT_CLITIC_MARKERS,
+    DEFAULT_COMPOUND_MARKERS,
+    DEFAULT_PUNCTUATION,
+    DEFAULT_WORD_BREAK_MARKERS,
+)
 from montreal_forced_aligner.g2p.generator import PyniniValidator
 from montreal_forced_aligner.models import (
     AcousticModel,
@@ -45,6 +51,23 @@ if typing.TYPE_CHECKING:
 logger = logging.getLogger("anchor")
 
 
+WORD_BREAK_SET = "".join(
+    sorted(
+        set(
+            DEFAULT_WORD_BREAK_MARKERS
+            + DEFAULT_PUNCTUATION
+            + DEFAULT_CLITIC_MARKERS
+            + DEFAULT_COMPOUND_MARKERS
+        )
+    )
+)
+
+if "-" in WORD_BREAK_SET:
+    WORD_BREAK_SET = "" + WORD_BREAK_SET.replace("-", "")
+
+WORD_BREAK_REGEX_SET = rf"[\s{WORD_BREAK_SET}]"
+
+
 # noinspection PyUnresolvedReferences
 @dataclass(slots=True)
 class TextFilterQuery:
@@ -76,9 +99,9 @@ class TextFilterQuery:
         if posix:
             text = text.replace(r"\b", word_break_set)
             if text.startswith(r"\b"):
-                text = r"((?<=\s)|(?<=^))" + text[2:]
+                text = rf"((?<={WORD_BREAK_REGEX_SET})|(?<=^))" + text[2:]
             if text.endswith(r"\b"):
-                text = text[:-2] + r"((?=\s)|(?=$))"
+                text = text[:-2] + rf"((?={WORD_BREAK_REGEX_SET})|(?=$))"
         if self.regex or self.word:
             if not self.case_sensitive:
                 text = "(?i)" + text
@@ -545,7 +568,6 @@ class FileSelectionModel(QtCore.QItemSelectionModel):
     resetView = QtCore.Signal()
     viewChanged = QtCore.Signal(object, object)
     selectionAudioChanged = QtCore.Signal()
-    currentTimeChanged = QtCore.Signal(object)
     currentUtteranceChanged = QtCore.Signal()
     speakerRequested = QtCore.Signal(object)
 
@@ -710,7 +732,8 @@ class FileSelectionModel(QtCore.QItemSelectionModel):
             end = None
         self.selected_min_time = begin
         self.selected_max_time = end
-        self.selectionAudioChanged.emit()
+        if self.selected_min_time != self.min_time:
+            self.selectionAudioChanged.emit()
 
     def request_start_time(self, start_time):
         if start_time >= self.max_time:
@@ -719,7 +742,8 @@ class FileSelectionModel(QtCore.QItemSelectionModel):
             return
         self.selected_min_time = start_time
         self.selected_max_time = None
-        self.selectionAudioChanged.emit()
+        if self.selected_min_time != self.min_time:
+            self.selectionAudioChanged.emit()
 
     def set_current_channel(self, channel):
         if channel == self.selected_channel:
@@ -804,7 +828,8 @@ class FileSelectionModel(QtCore.QItemSelectionModel):
             end = self.max_time
         self.selected_min_time = begin
         self.selected_max_time = end
-        self.selectionAudioChanged.emit()
+        if self.selected_min_time != self.min_time:
+            self.selectionAudioChanged.emit()
 
     def visible_utterances(self) -> typing.List[Utterance]:
         file_utts = []
@@ -843,8 +868,7 @@ class FileSelectionModel(QtCore.QItemSelectionModel):
             self.selected_max_time = None
         self.viewChanged.emit(self.min_time, self.max_time)
 
-    def set_current_file(self, info, force_update=False):
-        file_id, begin, end, utterance_id, speaker_id = info
+    def set_current_file(self, file_id, begin, end, utterance_id, speaker_id, force_update=False):
         try:
             new_file = self.model().file is None or self.model().file.id != file_id
         except sqlalchemy.orm.exc.DetachedInstanceError:
@@ -926,7 +950,7 @@ class CorpusSelectionModel(QtCore.QItemSelectionModel):
     channelChanged = QtCore.Signal()
     resetView = QtCore.Signal()
     fileAboutToChange = QtCore.Signal()
-    fileViewRequested = QtCore.Signal(object)
+    fileViewRequested = QtCore.Signal(object, object, object, object, object)
     selectionAudioChanged = QtCore.Signal()
     currentTimeChanged = QtCore.Signal(object)
     currentUtteranceChanged = QtCore.Signal()
@@ -1023,7 +1047,7 @@ class CorpusSelectionModel(QtCore.QItemSelectionModel):
         if focus:
             flags |= QtCore.QItemSelectionModel.SelectionFlag.Current
             if row == self.currentIndex().row():
-                self.fileViewRequested.emit(self.model().audio_info_for_utterance(row))
+                self.fileViewRequested.emit(*self.model().audio_info_for_utterance(row))
 
         index = self.model().index(row, 0)
         if not index.isValid():
@@ -1082,7 +1106,7 @@ class CorpusSelectionModel(QtCore.QItemSelectionModel):
             return
         self.current_utterance_id = utt
         self.currentUtteranceChanged.emit()
-        self.fileViewRequested.emit(self.model().audio_info_for_utterance(row))
+        self.fileViewRequested.emit(*self.model().audio_info_for_utterance(row))
 
     def model(self) -> CorpusModel:
         return super().model()
@@ -1099,7 +1123,7 @@ class CorpusSelectionModel(QtCore.QItemSelectionModel):
             return
         self.current_utterance_id = utt_id
         self.currentUtteranceChanged.emit()
-        self.fileViewRequested.emit(self.model().audio_info_for_utterance(row))
+        self.fileViewRequested.emit(*self.model().audio_info_for_utterance(row))
 
 
 class OovModel(TableModel):
@@ -1266,7 +1290,7 @@ class DictionaryTableModel(TableModel):
             return True
         return False
 
-    def add_word(self, word, word_id):
+    def add_word(self, word, word_id=None):
         self.requestLookup.emit(word)
         self.add_pronunciation(word, word_id)
 
@@ -1697,10 +1721,10 @@ class DiarizationModel(TableModel):
         super().__init__(columns, parent=parent)
         self.settings = AnchorSettings()
         self.speaker_count = None
-        self._utterance_ids = []
-        self._file_ids = []
-        self._speaker_indices = []
-        self._suggested_indices = []
+        self.utterance_ids = []
+        self.file_ids = []
+        self.speaker_indices = []
+        self.suggested_indices = []
         self.corpus_model: Optional[CorpusModel] = None
         self.set_limit(self.settings.value(self.settings.RESULTS_PER_PAGE))
         self.speaker_filter = None
@@ -1728,11 +1752,6 @@ class DiarizationModel(TableModel):
             return self._data[index.row()][index.column()]
         return super().data(index, role)
 
-    def utterance_id_at(self, row: int):
-        if row is None:
-            return None
-        return self._utterance_ids[row]
-
     def set_threshold(self, threshold: float):
         if threshold != self.threshold:
             self.current_offset = 0
@@ -1792,32 +1811,32 @@ class DiarizationModel(TableModel):
                 self.alternate_speaker_filter = current_speaker.id
 
     def reassign_utterance(self, row: int):
-        utterance_id = self.utterance_id_at(row)
+        utterance_id = self.utterance_ids[row]
         if utterance_id is None:
             return
-        self.changeUtteranceSpeakerRequested.emit(utterance_id, self._suggested_indices[row])
+        self.changeUtteranceSpeakerRequested.emit(utterance_id, self.suggested_indices[row])
         self.layoutAboutToBeChanged.emit()
         self._data.pop(row)
-        self._utterance_ids.pop(row)
-        self._suggested_indices.pop(row)
-        self._speaker_indices.pop(row)
+        self.utterance_ids.pop(row)
+        self.suggested_indices.pop(row)
+        self.speaker_indices.pop(row)
 
         self.layoutChanged.emit()
 
     def merge_speakers(self, row: int):
-        speaker_id = self._speaker_indices[row]
+        speaker_id = self.speaker_indices[row]
         if self.inverted:
-            utterance_id = self._utterance_ids[row]
+            utterance_id = self.utterance_ids[row]
             self.corpus_model.addCommand.emit(
                 undo.ChangeSpeakerCommand([utterance_id], speaker_id, 0, self)
             )
         else:
-            self.corpus_model.merge_speakers([self._suggested_indices[row], speaker_id])
+            self.corpus_model.merge_speakers([self.suggested_indices[row], speaker_id])
         self.layoutAboutToBeChanged.emit()
         self._data.pop(row)
-        self._utterance_ids.pop(row)
-        self._suggested_indices.pop(row)
-        self._speaker_indices.pop(row)
+        self.utterance_ids.pop(row)
+        self.suggested_indices.pop(row)
+        self.speaker_indices.pop(row)
 
         self.layoutChanged.emit()
 
@@ -1828,17 +1847,16 @@ class DiarizationModel(TableModel):
     def finish_update_data(self, result, *args, **kwargs):
         self.layoutAboutToBeChanged.emit()
         if result is None:
-            self._data, self._utterance_ids, self._suggested_indices, self._speaker_indices = (
-                [],
-                [],
-                [],
-            )
+            self._data = []
+            self.utterance_ids = []
+            self.suggested_indices = []
+            self.speaker_indices = []
         else:
             (
                 self._data,
-                self._utterance_ids,
-                self._suggested_indices,
-                self._speaker_indices,
+                self.utterance_ids,
+                self.suggested_indices,
+                self.speaker_indices,
             ) = result
         self.layoutChanged.emit()
         self.newResults.emit()

{anchor_annotator-0.3.2 → anchor_annotator-0.4.0}/anchor/plot.py

@@ -13,6 +13,11 @@ import sqlalchemy
 from Bio import pairwise2
 from montreal_forced_aligner.data import CtmInterval
 from montreal_forced_aligner.db import Speaker, Utterance
+from montreal_forced_aligner.dictionary.mixins import (
+    DEFAULT_PUNCTUATION,
+    DEFAULT_WORD_BREAK_MARKERS,
+)
+from montreal_forced_aligner.tokenization.simple import SimpleTokenizer
 from PySide6 import QtCore, QtGui, QtWidgets
 
 from anchor import workers
@@ -161,7 +166,7 @@ class UtteranceClusterView(pg.PlotWidget):
         self.setBackground(self.settings.value(self.settings.PRIMARY_VERY_DARK_COLOR))
         self.corpus_model = None
         self.speaker_model: SpeakerModel = None
-        self.selection_model: CorpusSelectionModel = None
+        self.selection_model: FileSelectionModel = None
         self.updated_indices = set()
         self.brushes = {-1: pg.mkBrush(0.5)}
         self.scatter_item = ScatterPlot()
@@ -226,7 +231,7 @@ class UtteranceClusterView(pg.PlotWidget):
     def set_models(
         self,
         corpus_model: CorpusModel,
-        selection_model: CorpusSelectionModel,
+        selection_model: FileSelectionModel,
         speaker_model: SpeakerModel,
     ):
         self.corpus_model = corpus_model
@@ -248,13 +253,12 @@ class UtteranceClusterView(pg.PlotWidget):
         if ev.button() == QtCore.Qt.MouseButton.LeftButton:
             utterance_id = int(self.speaker_model.utterance_ids[index])
             utterance = self.corpus_model.session.query(Utterance).get(utterance_id)
-            self.selection_model.set_current_utterance(utterance_id)
-            self.selection_model.current_utterance_id = utterance_id
             self.selection_model.set_current_file(
                 utterance.file_id,
                 utterance.begin,
                 utterance.end,
-                utterance.channel,
+                utterance.id,
+                utterance.speaker_id,
                 force_update=True,
             )
         else:
@@ -1395,7 +1399,7 @@ class NormalizedTextRegion(TextAttributeRegion):
 
 
 class Highlighter(QtGui.QSyntaxHighlighter):
-    WORDS = r"\S+"
+    WORDS = rf"[^\s{''.join(DEFAULT_WORD_BREAK_MARKERS)+''.join(DEFAULT_PUNCTUATION)}]+"
 
     def __init__(self, *args):
         super(Highlighter, self).__init__(*args)
@@ -1424,8 +1428,20 @@ class Highlighter(QtGui.QSyntaxHighlighter):
     def highlightBlock(self, text):
         self.settings.sync()
         self.spellcheck_format.setUnderlineColor(self.settings.error_color)
+        tokenizers = self.dictionary_model.corpus_model.corpus.get_tokenizers()
+        dictionary_id = self.dictionary_model.corpus_model.corpus.get_dict_id_for_speaker(
+            self.speaker_id
+        )
+        words = self.WORDS
+        if isinstance(tokenizers, dict) and dictionary_id is not None:
+            tokenizer = self.dictionary_model.corpus_model.corpus.get_tokenizer(dictionary_id)
+        else:
+            tokenizer = tokenizers
+        if isinstance(tokenizer, SimpleTokenizer):
+            extra_symbols = "".join(tokenizer.punctuation) + "".join(tokenizer.word_break_markers)
+            words = rf"[^\s{extra_symbols}]+"
         if self.dictionary_model is not None and self.dictionary_model.word_sets:
-            for word_object in re.finditer(self.WORDS, text):
+            for word_object in re.finditer(words, text):
                 if not self.dictionary_model.check_word(word_object.group(), self.speaker_id):
                     self.setFormat(
                         word_object.start(),

{anchor_annotator-0.3.2 → anchor_annotator-0.4.0}/anchor/settings.py

@@ -855,12 +855,27 @@ class AnchorSettings(QtCore.QSettings):
             margin: 0 -2px; /* expand outside the groove */
         }}
         QTableWidget, QTableView, QTreeView, QTreeWidget {{
-            alternate-background-color: {table_even_color};
+            border: 4px solid {enabled_color};
+        }}
+        QTableView::item, QTreeWidget::item {{
+            background-color: {table_odd_color};
+            color: {table_text_color};
             selection-background-color: {selection_color};
             selection-color: {text_edit_color};
-            background-color: {table_odd_color};
+        }}
+        QTableView::item:selected, QTreeWidget::item:selected {{
+            background-color: {selection_color};
+            color: {text_edit_color};
+        }}
+        QTableView::item:alternate, QTreeWidget::item:alternate {{
+            background-color: {table_even_color};
             color: {table_text_color};
-            border: 4px solid {enabled_color};
+            selection-background-color: {selection_color};
+            selection-color: {text_edit_color};
+        }}
+        QTableView::item:alternate:selected, QTreeWidget::item:alternate:selected {{
+            background-color: {selection_color};
+            color: {text_edit_color};
         }}
        QTreeView QLabel, QTreeWidget QLabel{{
             color: {table_text_color};

{anchor_annotator-0.3.2 → anchor_annotator-0.4.0}/anchor/undo.py

@@ -39,8 +39,11 @@ class CorpusCommand(QtGui.QUndoCommand):
 
     def redo(self) -> None:
         with self.corpus_model.edit_lock:
-            self._redo(self.corpus_model.session)
-            self.corpus_model.session.commit()
+            try:
+                self._redo(self.corpus_model.session)
+                self.corpus_model.session.commit()
+            except Exception:
+                self.corpus_model.session.rollback()
             # while True:
             #    try:
             #        with self.corpus_model.session.begin_nested():
@@ -53,8 +56,11 @@ class CorpusCommand(QtGui.QUndoCommand):
 
     def undo(self) -> None:
         with self.corpus_model.edit_lock:
-            self._undo(self.corpus_model.session)
-            self.corpus_model.session.commit()
+            try:
+                self._undo(self.corpus_model.session)
+                self.corpus_model.session.commit()
+            except Exception:
+                self.corpus_model.session.rollback()
             # while True:
             #    try:
             #        with self.corpus_model.session.begin_nested():
@@ -453,6 +459,11 @@ class UpdateUtteranceTextCommand(FileCommand):
         self.new_text = other.new_text
         return True
 
+    def update_data(self):
+        super().update_data()
+        self.corpus_model.changeCommandFired.emit()
+        self.corpus_model.update_utterance_table_row(self.utterance)
+
 
 class ReplaceAllCommand(CorpusCommand):
     def __init__(