views:

111

answers:

5

I need to remove tags from a string in python.

<FNT name="Century Schoolbook" size="22">Title</FNT>

What is the most efficient way to remove the entire tag on both ends, leaving only "Title"? I've only seen ways to do this with HTML tags, and that hasn't worked for me in python. I'm using this particularly for ArcMap, a GIS program. It has it's own tags for its layout elements, and I just need to remove the tags for two specific title text elements. I believe regular expressions should work fine for this, but I'm open to any other suggestions.

+1  A: 

If it's only for parsing and retrieving value, you might take a look at BeautifulStoneSoup.

Eric Fortin
+2  A: 

This should work:

re.sub('<[^>]*>', '', mystring)

To everyone saying that regexes are not the correct tool for the job:

The context of the problem is such that all the objections regarding regular/context-free languages are invalid. His language essentially consists of three entities: a = <, b = >, and c = [^><]+. He wants to remove any occurrences of acb. This fairly directly characterizes his problem as one involving a context-free grammar, and it is not much harder to characterize it as a regular one.

I know everyone likes the "you can't parse HTML with regular expressions" answer, but the OP doesn't want to parse it, he just wants to perform a simple transformation.

Domenic
This didn't work. It returned the original string. Thanks though
Tanner
Sorry, I forgot the all-important `*` character. Try again?
Domenic
It worked! Thanks. That's all I needed.
Tanner
A: 

Searching this regex and replacing it with an empty string should work.

/<[A-Za-z\/][^>]*>/

Example (from python shell):

>>> import re
>>> my_string = '<FNT name="Century Schoolbook" size="22">Title</FNT>'
>>> print re.sub('<[A-Za-z\/][^>]*>', '', my_string)
Title
no
That didn't work either. Could you give me an example of how you would search and replace using this? I tried, and it returned the original string.
Tanner
Added an example. Did you forget `import re`?
no
A: 

If the source text is well-formed XML, you can use the stdlib module ElementTree:

import xml.etree.ElementTree as ET
mystring = """<FNT name="Century Schoolbook" size="22">Title</FNT>"""
element = ET.XML(mystring)
print element.text  # 'Title'

If the source isn't well-formed, BeautifulSoup is a good suggestion. Using regular expressions to parse tags is not a good idea, as several posters have pointed out.

ianmclaury
+1  A: 

Use an XML parser, such as ElementTree. Regular expressions are not the right tool for this job.

Nathan Davis
Unless the input is not guaranteed to be well-formed XML, in which case regex is arguably the _only_ reasonable tool for the job. I'm also willing to bet that regex will perform significantly faster than handling the string as an XML document.
no
If the input is not well-formed XML, then implementing a full parser would be the proper way to do this. The grammar is complex enough that regular expressions are not enough.
Nathan Davis