#! /usr/bin/env python3
# Time-stamp: <2019-07-18 15:12:18 christophe@pallier.org>


""" take a list of tables (.tsv, .csv, .rds, ... ) as arguments and generates
the embryon of the json file for openlexicon.
"""

import os
import sys
import hashlib

def md5(fname):
    hash_md5 = hashlib.md5()
    with open(fname, "rb") as f:
        for chunk in iter(lambda: f.read(4096), b""):
            hash_md5.update(chunk)
    return hash_md5.hexdigest()



HEADER="""
{
      "_comment": "generated by create_json.py",
      "name": "XXXX",
      "description": "XXXXXX",
      "website": "XXXXXX",
      "readme": "XXXXXX",
      "urls" : [
"""


FOOTER="""      ], 
       "type": "tsv",
       "mandatory_columns": [],
       "tags": ["french", "frequencies"]
}
"""

if __name__ == '__main__':

   print(HEADER, end='')

   first = True
   for fname in sys.argv[1:]:
        md5sum = md5(fname)
        nbytes = os.path.getsize(fname)
        url =  "http://www.lexique.org" + os.path.abspath(fname).replace('/var/www','')

        if not first:
            print(',')
        first = False

        print('         {')
        print(f'            "url": "{url}",')
        print(f'            "bytes": {nbytes},')
        print(f'            "md5sum": "{md5sum}"')
        print('         }')

        
        

   print(FOOTER)