parse.py

import csv
from parsley import makeGrammar
import sys

raw = r'''
squote = '\''
escaped_squote = '\\' squote
float = <digit+ '.' digit*>:f -> float(f)
int = <digit+>:i -> int(i)
string = squote <(escaped_squote | ~squote anything)*>:s squote -> s

value = int | float | string

values = value:head (ws ',' ws value)*:tail -> [head] + tail
tuple = '(' ws values:v ws ')' -> writer.writerow(v)
tuples = tuple:head (ws ',' ws tuple)*:tail -> [head] + tail

table_name = <(letter | '_' | '.')+>:t

table = ('`' table_name:t '`' -> t) | table_name

insert = token('INSERT') ws token('INTO') ws table:t ws token('VALUES') ws tuples:ts ws ';' ws

'''


def parse(fn):
    g = makeGrammar(raw, {'writer': csv.writer(sys.stdout)})
    with open(fn) as f:
        for line in f:
            if not line.startswith('INSERT'):
                continue
            try:
                g(line).insert()
            except Exception as e:
                a = repr(line[e.position-100:e.position])
                b = repr(line[e.position])
                c = repr(line[e.position+1:e.position+100])

                print '%s\033[91m%s\033[0m%s' % (a, b, c)
                print e.formatReason()

if __name__ == '__main__':
    import argparse

    ap = argparse.ArgumentParser(description='pick through a mysqldump')

    ap.add_argument('file',
                    help='mysqldump file that contains INSERTS')

    args = ap.parse_args()
    parse(args.file)