Python remove items from list

Question

I'm having some troubles with removing items from a list. I'm looking for a more elegant solution. Preferably a solution in one for-loop or filter.

The objective of the piece of code: remove all empty entries and all entries starting with a '#' from the config handle.

At the moment i'm using:

# Read the config file and put every line in a seperate entry in a list
configHandle = [item.rstrip('\n') for item in open('config.conf')]

# Strip comment items from the configHandle
for item in configHandle:
    if item.startswith('#'):
        configHandle.remove(item)

# remove all empty items in handle
configHandle = filter(lambda a: a != '', configHandle)
print configHandle

This works but I think it is a bit of a nasty solution.

When I try:

# Read the config file and put every line in a seperate entry in a list
configHandle = [item.rstrip('\n') for item in open('config.conf')]

# Strip comment items and empty items from the configHandle
for item in configHandle:
    if item.startswith('#'):
        configHandle.remove(item)
    elif len(item) == 0:
        configHandle.remove(item)

This, however, fails. I cannot figure out why.

Can someone push me in the right direction?

Mazdak · Answer 1 · 2016-08-04T21:49:56.247

1

Because You're changing the list while iterating over it. You can use a list comprehension to get ride of this problem:

configHandle = [i for i in configHandle if i and not i.startswith('#')]

Also for opening a file you better to use a with statement that close the file at the end of the block automatically¹:

with open('config.conf') as infile :
   configHandle = infile.splitlines()
   configHandle = [line for line in configHandle if line and not line.startswith('#')]

_{1. Because there is no guarantee for external links to be collected by garbage-collector. And you need to close them explicitly, which can be done by calling the close() method of a file object, or as mentioned as a more pythonic way use a with statement.}

edited Aug 04 '16 at 21:49

answered Aug 04 '16 at 21:43

Mazdak

105,000
18
159
188

Thanks for your reply! I do not understand the first solution (using a comprehension) completely. How does that line else remove the empty entries in the list? Because this also does the trick for me configHandle = [i for i in configHandle if len(i)>0 and not i.startswith('#')] – DCB Aug 05 '16 at 07:14
@DCB When you write `if line` python checks the validity of `line` if it evaluate as True python run your condition, and in this case an empty string evaluates as False. Read more here https://docs.python.org/3.5/library/stdtypes.html#truth-value-testing – Mazdak Aug 05 '16 at 10:48

score 1 · Answer 2 · edited Sep 27 '17 at 04:45

1

Don't remove items while you iterating, it's a common pitfall

edited Sep 27 '17 at 04:45

Graham

7,431
18
59
84

answered Aug 04 '16 at 21:43

Ohad Eytan

8,114
1
22
31

score 0 · Answer 3 · answered Aug 04 '16 at 21:42

0

You aren't allowed to modify an item that you're iterating over.

Instead you should use things like filter or list comprehensions.

configHandle = filter(lambda a: (a != '') and not a.startswith('#'), configHandle)

answered Aug 04 '16 at 21:42

Jonathon Reinhart

132,704
33
254
328

score 0 · Answer 4 · edited May 23 '17 at 11:44

Your filter expression is good; just include the additional condition you're looking for:

configHandle = filter(lambda a: a != '' and not a.startswith('#'), configHandle)

There are other options if you don't want to use filter, but, as has been stated in other answers, it is a very bad idea to attempt to modify a list while you are iterating through it. The answers to this stackoverflow question provides alternatives to using filter to remove from a list based on a condition.

Python remove items from list

4 Answers4