Spaces:
Runtime error
Runtime error
| import streamlit as st | |
| import epitran | |
| import langcodes | |
| from langcodes import LanguageTagError | |
| from pathlib import Path | |
| from operator import itemgetter | |
| from collections import defaultdict | |
| # TODO: reverse transliterate? | |
| def get_lang_description_from_mapping_name(string_to_check): | |
| if "generic-Latn" == string_to_check: | |
| return "Generic Latin Script" | |
| if len(string_to_check)<2: | |
| return None | |
| substrings = string_to_check.split("-") | |
| substrings = substrings[:2] # first two are ISO 639-3 language, and ISO 15924 script | |
| string_to_check = "-".join(substrings) | |
| description = None | |
| lang = langcodes.get(string_to_check) | |
| if lang: | |
| items = [] | |
| for key, value in lang.describe().items(): | |
| if key == "language": | |
| iso_code = lang.to_alpha3() | |
| value = f"[{value}](https://iso639-3.sil.org/code/{iso_code})" | |
| items.append(f"{key}: {value}") | |
| description = ", ".join(items) | |
| if substrings[-1] == "red": | |
| description = description + " (reduced)" | |
| return description | |
| def get_valid_epitran_mappings_list(): | |
| map_path = Path(epitran.__path__[0]) / "data" / "map" | |
| map_files = map_path.glob("*.*") | |
| valid_mappings = [map_file.stem for map_file in map_files] | |
| valid_mappings.append("cmn-Hans") # special case | |
| problem_mappings = ['generic-Latn', | |
| 'tur-Latn-bab', | |
| 'ood-Latn-sax', | |
| 'vie-Latn-so', | |
| 'vie-Latn-ce', | |
| 'vie-Latn-no', | |
| 'kaz-Cyrl-bab'] # https://github.com/dmort27/epitran/issues/98 | |
| filtered_mappings = [mapping for mapping in valid_mappings if mapping not in problem_mappings] | |
| return filtered_mappings | |
| if __name__ == "__main__": | |
| st.write("# Phonemize your text with [Epitran](https://github.com/dmort27/epitran)!") | |
| st.write("Epitran is a library and tool for transliterating orthographic text as IPA (International Phonetic Alphabet), by Mortensen, David R. and Dalmia, Siddharth and Littell, Patrick.") | |
| valid_epitran_mappings = get_valid_epitran_mappings_list() | |
| index_of_swa_latn = valid_epitran_mappings.index("swa-Latn") | |
| st.write(f"It supports converting many writing sytems to IPA symbols, including approximately {len(valid_epitran_mappings)} languages/scripts, listed below:") | |
| #st.write(valid_epitran_mappings) | |
| selected_mapping = st.selectbox("Select input language/script:", valid_epitran_mappings, index=index_of_swa_latn) | |
| description = get_lang_description_from_mapping_name(selected_mapping) | |
| st.write(f"Selected input language/script: {description}") | |
| if selected_mapping == "cmn-Hans": | |
| st.info("Chinese requires a special dictionary. Downloading now") | |
| epitran.download.cedict() | |
| st.info("attempting to instantiate epitran transliterator for your language/script") | |
| epi = epitran.Epitran(selected_mapping) | |
| examples = defaultdict(lambda: 'Try typing some words in the language you chose, and they will be transliterated.') | |
| examples['cmn-Hans'] = 'ε€ͺεζιοΌιδΈη₯εε¨οΌιε°±ζ―η₯γ' | |
| examples['swa-Latn'] = 'Mwanzoni Kabla ulimwengu haujaumbwa alikuwepo Neno Huyo Neno alikuwa pamoja na Mungu, na Neno alikuwa Mungu.' | |
| input_text = st.text_area(label="Whatever you type here will be transliterated!", value=examples[selected_mapping]) | |
| # combined_code = "-".join([iso_lang_code, iso_script_code]) | |
| # st.write(f"Combined code: {combined_code}") | |
| st.info(f"transliterating `{input_text}`\n\tusing {selected_mapping}...") | |
| transliteration = epi.transliterate(input_text) | |
| output = { | |
| "original": input_text, | |
| "transliteration":transliteration, | |
| } | |
| st.write(output) | |