langchao_fayuan/dump_tables.py

import re
import codecs


def find_table_names(_data):
    s = []
    bi = []

    def find(p, offset=1):
        results = re.findall(p, _data, re.IGNORECASE)
        for l in results:
            # print('l', l)
            text = l
            if type(l) is tuple:
                if l[0] == 'lawe':
                    text = l[1]
                elif len(l[1]) == 1 or l[1] == 'thd' or l[1] == 'AIiKi' or l[1] == 'XVuSA':
                    text = l[0]
                else:
                    text = l[1]
            # print('>', text)
            if not text.startswith('bi_') and not text.startswith('v_bi_') and text != 'lawe' and text != 'icm' and text != 'icm_bpm':
                s.append(text)
            if text.startswith('bi_'):
                bi.append(text)

    find(r'CREATE TABLE `(\w+)`')

    find(r'from lawe\.(\w+)')
    find(r'from `(\w+)`.`(\w+)`', 2)
    find(r'from\r\n    `(\w+)`.`(\w+)`', 2)
    find(r'from (\w+)')
    find(r'LEFT JOIN (\w+)')
    find(r'LEFT JOIN (\w+).(\w+)', 2)
    find(r'from (\w+)\.(\w+)')
    find(r'from\r\n\t(\w+)')
    find(r'inner join (\w+).(\w+)', 2)

    bi = list(set(bi))
    print(bi)
    r = list(set(s))
    print(r)
    # print(len(r))

    pass


file_name = 'all.sql'
text_file = codecs.open(file_name, "r")

# read whole file to a string
data = text_file.read()

splitter = '-- （'
parts = data.split(splitter)
print('parts', len(parts))
for part in parts:
    print(part.partition('\n')[0])
    find_table_names(part)


# close file
text_file.close()