Python string.replace() not replacing characters

Question

Some background information: We have an ancient web-based document database system where I work, almost entirely consisting of MS Office documents with the "normal" extensions (.doc, .xls, .ppt). They are all named based on some sort of arbitrary ID number (i.e. 1245.doc). We're switching to SharePoint and I need to rename all of these files and sort them into folders. I have a CSV file with all sorts of information (like which ID number corresponds to which document's title), so I'm using it to rename these files. I've written a short Python script that renames the ID number title.

However, some of the titles of the documents have slashes and other possibly bad characters to have in a title of a file, so I want to replace them with underscores:

bad_characters = ["/", "\\", ":", "(", ")", "<", ">", "|", "?", "*"]
for letter in bad_characters:
    filename = line[2].replace(letter, "_")
    foldername = line[5].replace(letter, "_")

• Example of line[2]: "Blah blah boring - meeting 2/19/2008.doc"
• Example of line[5]: "Business meetings 2/2008"

When I add print letter inside of the for loop, it will print out the letter it's supposed to be replacing, but won't actually replace that character with an underscore like I want it to.

Is there anything I'm doing wrong here?

Answer 1

You are reassigning to the filename and foldername variables at every iteration of the loop. In effect, only * is being replaced.

Answer 2

That's because filename and foldername get thrown away with each iteration of the loop.

You should use:

filename = line[2]
foldername = line[5]

for letter in bad_characters:
    filename = filename.replace(letter, "_")
    foldername = foldername.replace(letter, "_")

But I would do it using regex. It's cleaner and (likely) faster:

p = re.compile('[/:()<>|?*]|(\\\)')
filename = p.sub('_', line[2])
folder = p.sub('_', line[5])

Answer 3

You should look at the python string method translate() http://docs.python.org/library/string.html#string.translate with http://docs.python.org/library/string.html#string.maketrans

Editing this to add an example as per comment suggestion below:

import string
toreplace=''.join(["/", "\\", ":", "(", ")", "<", ">", "|", "?", "*"]) 
underscore=''.join( ['_'] * len(toreplace))
transtable = string.maketrans(toreplace,underscore)
filename = filename.translate(transtable)
foldername = foldername.translate(transtable)

Can simplify by making the toreplace something like '/\\:,' etc, i just used what was given above

Answer 4

You are starting over with the base line instead of saving the replaced result, thus you are getting the equivalent to

filename = line[2].replace('*', '_')
foldername = line[5].replace('*', '_')

Try the following

bad_characters = ["/", "\\", ":", "(", ")", "<", ">", "|", "?", "*"]
filename = line[2]
foldername = line[5]
for letter in bad_characters:
    filename = filename.replace(letter, "_")
    foldername = foldername.replace(letter, "_")