Python中移除LaTeX字符：去除\textbf{}标签的实现问题

Ah, I see the problem with your current code—your regex is using greedy matching, which causes it to capture far more than just the content inside a single \textbf{} block.

Why Your Original Code Fails #

The .* quantifier matches as much text as possible. So when you run re.sub(r'\textbf{(.*)}', '\\1', s), it starts at the first \textbf{ and doesn't stop until it hits the very last } in the entire string. This means it captures everything from 1 + 1} \begin{center} \textbf{some text in here:} \end{center} and \textbf{2 as the group, leading to incorrect output with leftover braces and unprocessed \textbf tags.

The Solution #

You need to make the match non-greedy or explicitly match only characters that aren't closing braces. Here are two reliable approaches:

1. Non-Greedy Quantifier (.*?)

Add a ? after .* to make it match the shortest possible string between the opening { and closing }:

import re
s = '\textbf{1 + 1} \begin{center} \textbf{some text in here:} \end{center} and \textbf{2} etc'
result = re.sub(r'\\textbf{(.*?)}', r'\1', s)
print(result)

2. Match Non-Closing Braces ([^}]+)

This approach is more efficient if you know there are no nested braces inside your \textbf{} blocks. It matches any character except }:

result = re.sub(r'\\textbf{([^}]+)}', r'\1', s)

Output #

Both methods will produce your desired result:

1 + 1 \begin{center} some text in here: \end{center} and 2 etc

Note on Nested Braces #

If you ever need to handle nested braces (e.g., \textbf{text {with} nested braces}), neither of these simple regexes will work. For that, you'd need a recursive regex or a parser specifically designed for TeX, but since your example doesn't involve nested braces, the above solutions are perfect.

内容的提问来源于stack exchange，提问作者dimitris_ps