On Jun 2, 11:42 am, "[EMAIL PROTECTED]" <Gilles@> wrote:
> Hello
>
> I need to go through each line of a CSV file, and extract some fields
> using a regex. Then, I need to check each retrieved field, and if it
> looks like "", turn this into NULL so that it's correct SQL.
>
> I tried a few things, but still can't it working:
> ========
> #Second field might be empty -> ""
> #"Col1",""
> #"Col1","Col2"
> p = re.compile('^"(.+?)","(.*?)"$')
>
> for line in textlines:
> m = p.search(line)
> if m:
> #Check each column : if '', then turn into NULL
>
> """
> for col in line:
> if col == "":
> col = "NULL"
> """
>
> """
> for col in m.group():
> if col == "":
> col="NULL"
> """
>
> """
> for col in m.group(0):
> if col == "":
> col="NULL"
> """
>
> """
> for i in range (0,len(line)):
> if line[i] == "":
> line[i]="NULL"
> """
>
> """
> for i in range(1,len(m.group(0))):
> if m.group(i) == "":
> m.group(i)="NULL"
> """
>
> sql = "INSERT INTO mytable (col1, col2) VALUES
> ('%s','%s')" % (m.group(1),m.group(2))
> print sql
> f.close()
> ========
>
> Does someone know the correct syntax?
>
> Thank you.
I think you want to use 'groups' instead of 'group'.
Here is a brief example. Note that this code is very insecure and
susceptible to a SQL injection attack. Hopefully these csv files are
from a trusted source.
sql = "INSERT INTO mytable (col1, col2) VALUES ('%s','%s')"%tuple(
(c, "NULL")[c == ''] for c in m.groups()
)
Also, check out the csv module for parsing your csv file.
Matt
--
http://mail.python.org/mailman/listinfo/python-list
