Convert Hong Kong Chinese sentences to the International Phonetic Alphabet (IPA).

The zhhk2ipa project provides tools to convert text from Hong Kong Chinese into the International Phonetic Alphabet (IPA). It supports phrase-level and word-level transcription based on a comprehensive vocabulary table.

Install the required dependencies by running:

pip install -r requirements.txt

To use zhhk2ipa, follow these steps:

import zhhk2ipa

# Load the vocabulary table
zhhk2ipa.init("data/zhhk_ipa.txt")

# Convert text to IPA
output = zhhk2ipa.convert("同朕check吓")
print(output)

output = zhhk2ipa.convert("返屋企,樂而忘返。")
print(output)

[('同', ['tʰʊŋ˨˩']), ('朕', ['tsɐm˨']), ('check吓', ['tsʰɛ:k˥', 'ha:˩˧'])]
[('返屋企', ['fa:n˥', 'ŋʊk˥', 'kʰei˧˥']), (',', ','), ('樂而忘返', ['lɔ:k˨', 'ji:˨˩', 'mɔ:ŋ˨˩', 'fa:n˧˥']), ('。', '。')]

If a token is not in the vocabulary table, it will be marked as None in the output:

output = zhhk2ipa.convert("我ABCD你")
print(output)

[('我', ['ŋɔ:˩˧']), ('A', None), ('B', ['pi:˥']), ('C', None), ('D', None), ('你', ['nei˩˧'])]

Please note that the functionality of this software has been tested with a limited number of test cases. While preliminary results are promising, the reliability across diverse and extensive datasets has not been fully established. Users are encouraged to conduct their own tests and report any issues or discrepancies they encounter. This will assist in improving the tool and extending its reliability for broader usage.

The vocabulary for zhhk2ipa is adapted from data provided by lotusfa/IPA-Translator with additional entries included for broader coverage.