Extract argument from general function code in python

Question

In this particular case, I want to extract the argument of a latex command, as:

... latex code ...
\bibitem{item1} 
... latex code ...
\bibitem{item2}
... latex code ...

Is there any function in python which will retrieve a list containing item1, item2, etc.?

I would imagine something like:

latex_text.extract_argument("\bibitem{","}")

a function which would scan the text and return a list of every argument contained within the two delimiters set.

• Lauritz V. Thaulow: Are you sure you're not better off with some bibtex command or output? Without knowing why you want to do this it's hard to say for sure, but I think you might want to look at other ways of achieving your aim. For one thing, if item1 contains commands which uses } then the regexp answers will not suffice, since you need to count opening and closing braces to find the one that closes the bibitem command. Have you looked at the .bbl file that bibtex produces?
• Lauritz V. Thaulow: Perhaps this is what you need?
• leandro: Thank you for your notes. Bibtex is too much work for most of our articles, which are short and do not contain many rereferences. We simply wanted some quick code to check if the references are all cited and cited in the correct order. You note on item1 having a "}" is correct, but this is not possible in this case, because the latex code will not accept that. Anyway, it would be nice to know if one can easily parse the argument between MATCHING brackets, but this is another question. My simple code is available here: leandro.iqm.unicamp.br/order_refs.py. I think it works fine.

Answer 1

import re
re.findall(r'\\bibitem\{(.*?)\}', latex_text) # ['item1', 'item2']

Answer 2

Use re.search:

import re
l = '\\bibitem{item1}'
m = re.search(r'\{([^}]*)\}', l)
print m.group(1) # => 'item1'

Explanation of the regular expression \{([^}]*)\}:

1. We are looking for a starting { which we must escape as \{.
2. We use a group (...) that contains
3. any number * of characters that are not }: ([^}]

The group from step 2 is m.group(1).

• leandro: Very nice, thanks for the explanation on the regular expressions, it will be very helpful.