Add provinces

[wannaphong]

I add provinces name of Thailand. (Thai and romanize)

(will resolve #467)

[pep8speaks]

Hello @wannaphong! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

• In the file pythainlp/corpus/common.py:

Line 72:80: E501 line too long (101 > 79 characters)
Line 74:80: E501 line too long (98 > 79 characters)
Line 93:1: W293 blank line contains whitespace

Comment last updated at 2020-07-23 03:04:52 UTC

[coveralls]

Coverage increased (+0.03%) to 95.038% when pulling 362eb2d on Add-provinces into d27ff7d on dev.

[p16i]

maybe we can combine this function with provinces()?
To me, because these two functions are quite similar, we can have an argument letting the user to choose which list he/she wants.

For example, the signatures could be: provinces() and provinces(details=True).

[wannaphong]

OK.

[bact]

It's tricky here.

provinces() was designed to populate a set of province names in the first run (when _THAI_THAILAND_PROVINCES is still empty, it will read the names from a file with a name _THAI_THAILAND_PROVINCES_FILENAME) and for any other subsequent call it will just return _THAI_THAILAND_PROVINCES - to save processing time and avoid IO operation).

The modification breaks this.

May be we can modify the function in this way:

• Has only one file (the CSV one) of Thai names and romanized names (province Ministry of Interior code, province abbreviation, province postal code, etc.)
• In the first call, populate as set (or sorted list?) of name pairs to _THAI_THAILAND_PROVINCES.
• In any other subsequent call, just use the existing _THAI_THAILAND_PROVINCES.
• The return type/format will be determined by value in details=

[wannaphong]

@bact OK. I updated code.

[p16i]

This is LGTM. I haven't checked the documentation yet. Should we merge it?

[bact]

Sorry for being so late, I was quite sick in the past weeks.

I made some code cleaning, fix the types, and remove the .txt file as the information there is duplicated with ones in .csv.

One question here please, what's the meaning of these field names in dict?

• provinces_th
• abridgement
• provinces_en
• HS

I propose:

• name_th
• abbr_th (abbr for abbreviation)
• name_en
• abbr_en

will this make sense?

Please continue discussion in #466