# -*- coding: utf-8 -*-
import os
import re
import json
import pprint

TOKEN_REGEX = r"""(\d+)(-)(\d+)(\.)(.+)"""
TOKEN_REGEX2 = r"""(?:\|\s*)([^\s]*)(?:\s*)(?:\|\s*)([^\s]*)(?:\s*)(?:\|\s*)([^\s]*)(?:\s*)(?:\|\s*)([^\s]*)(?:\s*)"""

def md2csv(mdfile):
    state = 0
    log_dict = {}
    log_dict['__keys__'] = []
    curr_node = None
    for line in open(mdfile).readlines():
        match = re.match(TOKEN_REGEX, line)
        if match:
            logclass1, _, logclass2, _, title = match.groups()
            state = 1
            assert((logclass1 + '-' + logclass2) not in log_dict)
            curr_node = {}
            curr_node['__keys__'] = []
            log_key = logclass1 + '-' + logclass2
            log_dict[log_key] = curr_node
            log_dict['__keys__'].append(log_key)
            continue
        match2 = re.match(TOKEN_REGEX2, line)
        if match2 and state == 1:
            field1, field2, field3, field4 = match2.groups()
            if field1[0] != '-' and field1 != '字段':
                field_name = field1.replace('\\', '')
                en_name = field2.replace('\\', '')
                curr_node[field_name] = {
                    'short_name': field_name,
                    'en_name': en_name,
                    'desc': field4
                }
                assert (field_name != 'ext') or (field_name == 'ext' and en_name == 'ext')
                curr_node['__keys__'].append(field_name)
                if not (en_name in curr_node):
                    curr_node['__keys__'].append(en_name)
                    curr_node[en_name] = curr_node[field_name]
    #end for
    print(json.dumps(log_dict, sort_keys=True, indent=4, separators=(',', ': ')))

md2csv('/root/opensource/doc/api说明/log_normal.md')