[feature] Generate a cache to store all pinyin of characters

[Freed-Wu]

Comes from petronny/pinyin-completion#18:

ngram is not fast. Can this project provide a tool to generate a cache (format can be json or other formats) to store all pinyin of characters? Such as:

python -m pypinyin.tools.generate_cache --format=json --style=FIRST_LETTER > /the/path/of/pinyin_FIRST_LETTER.json
jq -S . /the/path/of/pinyin_FIRST_LETTER.json
{
...
"你": "n"
"您": "n"
...
}
# a very large json

[mozillazg]

Thanks for your suggestion. But I'm too busy to do it. Feel free to send a pull request.

IMHO, some names like generate_all_pinyins may be a better name than generate_cache.

[Freed-Wu]

Currently I use cpp-pinyin to replace pypinyin to speed up, looks like better than cache.