Question

I’m trying to write a regex to parse an sfv file in python.
Basically, the lines are of the format

filename crc_bytes

but whitespaces can be added all over the place, including the file name. so the real format is

(whitespaces)filename(whitespaces)crc_bytes(whitespaces)

when filename can include whitespaces.

Now, I’m trying to extract filename and crc_bytes. So I’ve tried:

'\s*(.+)\s+([^\s]+)'

but it parsed

'   filename with spaces    crc  '

as

'filename with spaces   ', 'crc' 

//too much spaces————^

Any idea how to get rid of these spaces? probably, look-behind somehow?

bonus question:

Comments in sfv files are lines that start with ‘;’. If anyone would be able to treat comments in the regex I will forever be in his debt.

Thanks!!

Answer 1

Handling filenames with spaces

Using (.+\S) forces the filename to end with a non-whitespace ('\S) character.

>>> import re
>>> reg=re.compile('\s*(.+\S)\s+(\S+)')
>>> reg.findall(line)
[('filename with spaces', 'crc')]

Avoiding comments

You could use lookahead or add negation checks to the regex. I think, however, that adding another regex would be more readable:

>>> comment_line_regex=re.compile('\s*;.*')
>>> line1='   filename with spaces    crc  '
>>> line2=';  filename with spaces    crc  '
>>> line3='  ;  filename with spaces    crc  '
>>> lines = [line1, line2, line3]

Now we have three lines, two of which are comment line. The following parses only the lines which are not comments:

>>> [reg.findall(l) for l in lines if not comment_line_regex.match(l)]
[[('filename with spaces', 'crc')]]

Or, in a more verbose fashion:

>>> for line in lines:
...     if not comment_line_regex.match(line):
...             print reg.findall(line)
... 
[('filename with spaces', 'crc')]