Using csv to read SQL output

One of my devlings mistakenly broke the file paths for our image store (or she seems to think it was her), so we needed to go to the backups. I needed to convert the SQL backups to UPDATE statements to fix it. Python 2.3's new csv module made it really easy.

I've just realised I could've done this a lot more elegantly with iterators, but here's what I actually used:

#!/usr/bin/env python

import csv
import re
import sys

class SQLDialect(csv.Dialect):
    delimiter = ','
    quotechar = "'"
    escapechar = '\\'
    doublequote = False
    quoting = csv.QUOTE_NONNUMERIC
    skipinitialspace = True
    lineterminator = '\n'

dialect = SQLDialect()
path_match = re.compile("^/[0-9a-f]/[0-9a-f]$")

for line in sys.stdin:
    values = line.find(' VALUES ')
    line = line[values+9:]
    eor = 0
    while eor != -1:
        eor = line.find("'),(")
        if eor == -1:
            mine = line[:-3]
        else:
            mine = line[0:eor+1]
            line = line[eor+4:]
        cr = csv.reader([mine], dialect)
        data = list(cr)[0]
        print "UPDATE pictures SET file_path = '%(path)s' WHERE id = %(id)d" % ({"id": int(data[0]), "path": data[10]})