german_corpus.py 文件源码

def _extract_positional_label_by_id(self, files: Iterable[Path]) -> Dict[str, Union[PositionalLabel, str]]:
        xml_ending = ".xml"

        microphone_endings = [
            "_Yamaha",
            "_Kinect-Beam",
            "_Kinect-RAW",
            "_Realtek",
            "_Samson",
            "_Microsoft-Kinect-Raw"
        ]

        xml_files = [file for file in files if file.name.endswith(xml_ending) if
                     self.id_filter_regex.match(name_without_extension(file))]

        return OrderedDict(
            (name_without_extension(file) + microphone_ending,
             self._extract_label_from_xml(file))
            for file in xml_files
            for microphone_ending in microphone_endings
            if (Path(file.parent) / (name_without_extension(file) + microphone_ending + ".wav")).exists())