question on regular expressions

[Darren Dale]

Michael Fuhr wrote:

> Darren Dale <dd55 at cornell.edu> writes:
> 
>> I'm stuck. I'm trying to make this:
>>
>> file://C:%5Cfolder1%5Cfolder2%5Cmydoc1.pdf,file://C
>> %5Cfolderx%5Cfoldery%5Cmydoc2.pdf
>>
>> (no linebreaks) look like this:
>>
>> ./mydoc1.pdf,./mydoc2.pdf
>>
>> my regular expression abilities are dismal.
> 
> This works for the example string you gave:
> 
> newstring = re.sub(r'[^,]*%5[Cc]', './', examplestring)
> 
> This replaces all instances of zero or more non-commas that are
> followed by '%5C' or '%5c' with './'.  Greediness causes the pattern
> to replace everything up to the last '%5C' before a comma or the
> end of the string.
> 
> Regular expressions aren't the only way to do what you want.  Python
> has standard modules for parsing URLs and file paths -- take a look
> at urlparse, urllib/urllib2, and os.path.
> 

Thanks to both of you. I thought re's were appropriate because the string I
gave is buried in an xml file. A more representative example is:

[...snip...]<url>file://C:%5Cfolder1%5Cfolder2%5Cmydoc1.pdf</url>[...snip...
data]<url>file://C%5Cfolderx%5Cfoldery%5Cmydoc2.pdf</url>[...snip...]