def create_or_update_weights_map_main(base_dir: str,
merge_name: str,
weights_merge_name: str,
template_map: Optional[str] = None):
merge_dir = get_merged_dir(base_dir, merge_name)
assert os.path.isdir(merge_dir)
orig_prep_dir = get_merged_dir(base_dir, weights_merge_name)
assert os.path.isdir(orig_prep_dir)
logger = getLogger(__name__)
logger.info(f"Creating/updating weights map for {weights_merge_name}...")
if template_map is not None:
_template_map = SymbolsMap.load(template_map)
else:
_template_map = None
if weights_map_exists(merge_dir, weights_merge_name):
existing_map = load_weights_map(merge_dir, weights_merge_name)
else:
existing_map = None
weights_map, symbols = create_or_update_weights_map(
orig=load_merged_symbol_converter(orig_prep_dir).get_all_symbols(),
dest=load_merged_symbol_converter(merge_dir).get_all_symbols(),
existing_map=existing_map,
template_map=_template_map,
)
save_weights_map(merge_dir, weights_merge_name, weights_map)
save_weights_symbols(merge_dir, weights_merge_name, symbols)
def add_text(base_dir: str,
merge_name: str,
text_name: str,
filepath: Optional[str],
lang: Language,
text: Optional[str] = None):
assert text_name is not None and text_name != ""
logger = getLogger(__name__)
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
if not os.path.isdir(merge_dir):
logger.error("Please prepare data first.")
else:
logger.info("Adding text...")
text_input = ""
if filepath is None:
assert text is not None
text_input = text
else:
text_input = read_text(filepath)
symbol_ids, data = infer_add(
text=text_input,
lang=lang,
logger=logger,
)
print("\n" + data.get_formatted(symbol_id_dict=symbol_ids,
accent_id_dict=load_merged_accents_ids(
merge_dir)))
text_dir = get_text_dir(merge_dir, text_name, create=True)
_save_text_csv(text_dir, data)
save_text_symbol_converter(text_dir, symbol_ids)
_accent_template(base_dir, merge_name, text_name)
_check_for_unknown_symbols(base_dir, merge_name, text_name)
def map_text(base_dir: str,
merge_name: str,
text_name: str,
symbols_map_path: str,
ignore_arcs: bool = True):
logger = getLogger(__name__)
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
text_dir = get_text_dir(merge_dir, text_name, create=False)
if not os.path.isdir(text_dir):
logger.error("Please add text first.")
else:
symbol_ids, updated_sentences = sents_map(
sentences=load_text_csv(text_dir),
text_symbols=load_text_symbol_converter(text_dir),
symbols_map=SymbolsMap.load(symbols_map_path),
ignore_arcs=ignore_arcs,
logger=logger,
)
print("\n" + updated_sentences.get_formatted(
symbol_id_dict=symbol_ids,
accent_id_dict=load_merged_accents_ids(merge_dir)))
_save_text_csv(text_dir, updated_sentences)
save_text_symbol_converter(text_dir, symbol_ids)
_accent_template(base_dir, merge_name, text_name)
_check_for_unknown_symbols(base_dir, merge_name, text_name)
def ipa_convert_text(base_dir: str,
merge_name: str,
text_name: str,
ignore_tones: bool = False,
ignore_arcs: bool = True,
consider_ipa_annotations: bool = False,
mode: Optional[EngToIpaMode] = None):
logger = getLogger(__name__)
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
text_dir = get_text_dir(merge_dir, text_name, create=False)
if not os.path.isdir(text_dir):
logger.error("Please add text first.")
else:
logger.info("Converting text to IPA...")
symbol_ids, updated_sentences = sents_convert_to_ipa(
sentences=load_text_csv(text_dir),
text_symbols=load_text_symbol_converter(text_dir),
ignore_tones=ignore_tones,
ignore_arcs=ignore_arcs,
mode=mode,
consider_ipa_annotations=consider_ipa_annotations,
logger=logger,
)
print("\n" + updated_sentences.get_formatted(
symbol_id_dict=symbol_ids,
accent_id_dict=load_merged_accents_ids(merge_dir)))
_save_text_csv(text_dir, updated_sentences)
save_text_symbol_converter(text_dir, symbol_ids)
_accent_template(base_dir, merge_name, text_name)
_check_for_unknown_symbols(base_dir, merge_name, text_name)
def create_or_update_inference_map_main(base_dir: str,
merge_name: str,
template_map: Optional[str] = None):
logger = getLogger(__name__)
logger.info("Creating/updating inference map...")
merge_dir = get_merged_dir(base_dir, merge_name)
assert os.path.isdir(merge_dir)
all_symbols = get_all_symbols(merge_dir)
if template_map is not None:
_template_map = SymbolsMap.load(template_map)
else:
_template_map = None
if infer_map_exists(merge_dir):
existing_map = load_infer_map(merge_dir)
else:
existing_map = None
infer_map, symbols = create_or_update_inference_map(
orig=load_merged_symbol_converter(merge_dir).get_all_symbols(),
dest=all_symbols,
existing_map=existing_map,
template_map=_template_map,
)
save_infer_map(merge_dir, infer_map)
save_infer_symbols(merge_dir, symbols)
def map_to_prep_symbols(base_dir: str,
merge_name: str,
text_name: str,
ignore_arcs: bool = True):
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
assert os.path.isdir(merge_dir)
assert infer_map_exists(merge_dir)
symb_map_path = get_infer_map_path(merge_dir)
map_text(base_dir, merge_name, text_name, symb_map_path, ignore_arcs)
def get_infer_sentences(base_dir: str, merge_name: str,
text_name: str) -> InferSentenceList:
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
text_dir = get_text_dir(merge_dir, text_name, create=False)
if not os.path.isdir(text_dir):
print(f"The text '{text_name}' doesn't exist.")
assert False
result = InferSentenceList.from_sentences(
sentences=load_text_csv(text_dir),
accents=load_merged_accents_ids(merge_dir),
symbols=load_text_symbol_converter(text_dir))
return result
def _accent_template(base_dir: str, merge_name: str, text_name: str):
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
text_dir = get_text_dir(merge_dir, text_name, create=False)
if not os.path.isdir(text_dir):
print("Please add text first.")
else:
print("Updating accent template...")
accented_symbol_list = sents_accent_template(
sentences=load_text_csv(text_dir),
text_symbols=load_text_symbol_converter(text_dir),
accent_ids=load_merged_accents_ids(merge_dir),
)
_save_accents_csv(text_dir, accented_symbol_list)
def _check_for_unknown_symbols(base_dir: str, merge_name: str, text_name: str):
infer_sents = get_infer_sentences(base_dir, merge_name, text_name)
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
logger = getLogger(__name__)
unknown_symbols_exist = infer_sents.replace_unknown_symbols(
model_symbols=load_merged_symbol_converter(merge_dir), logger=logger)
if unknown_symbols_exist:
logger.info(
"Some symbols are not in the prepared dataset symbolset. You need to create an inference map and then apply it to the symbols."
)
else:
logger.info(
"All symbols are in the prepared dataset symbolset. You can now synthesize this text."
)
def accent_apply(base_dir: str, merge_name: str, text_name: str):
logger = getLogger(__name__)
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
text_dir = get_text_dir(merge_dir, text_name, create=False)
if not os.path.isdir(text_dir):
logger.error("Please add text first.")
else:
logger.info("Applying accents...")
updated_sentences = sents_accent_apply(
sentences=load_text_csv(text_dir),
accented_symbols=_load_accents_csv(text_dir),
accent_ids=load_merged_accents_ids(merge_dir),
)
print("\n" + updated_sentences.get_formatted(
symbol_id_dict=load_text_symbol_converter(text_dir),
accent_id_dict=load_merged_accents_ids(merge_dir)))
_save_text_csv(text_dir, updated_sentences)
_check_for_unknown_symbols(base_dir, merge_name, text_name)
def normalize_text(base_dir: str, merge_name: str, text_name: str):
logger = getLogger(__name__)
merge_dir = get_merged_dir(base_dir, merge_name, create=False)
text_dir = get_text_dir(merge_dir, text_name, create=False)
if not os.path.isdir(text_dir):
logger.error("Please add text first.")
else:
logger.info("Normalizing text...")
symbol_ids, updated_sentences = sents_normalize(
sentences=load_text_csv(text_dir),
text_symbols=load_text_symbol_converter(text_dir),
logger=logger,
)
print("\n" + updated_sentences.get_formatted(
symbol_id_dict=symbol_ids,
accent_id_dict=load_merged_accents_ids(merge_dir)))
_save_text_csv(text_dir, updated_sentences)
save_text_symbol_converter(text_dir, symbol_ids)
_accent_template(base_dir, merge_name, text_name)
_check_for_unknown_symbols(base_dir, merge_name, text_name)