如何用正则表达式忽略无IN_MEASUREMENT的FUNCTION数据集？

Absolutely! You can fix this with a single regex adjustment—no need for two separate searches. The issue with your original pattern is that it "leaks" across FUNCTION blocks: the .*? in your regex will match any characters (including across /end FUNCTION and into the next block) until it finds an IN_MEASUREMENT section, which is why it incorrectly associates the first FUNCTION with the second block's IN_MEASUREMENT content.

Fixed Regex Pattern #

Here's the adjusted pattern that ensures we only match FUNCTION blocks that actually contain an IN_MEASUREMENT section, and stay within the boundaries of that single FUNCTION:

sSearch = r'/begin FUNCTION\s*(\S*)\s*"[^"]*"(?:(?!/end FUNCTION).)*?/begin IN_MEASUREMENT\s*(.*?)\s*/end IN_MEASUREMENT(?:(?!/end FUNCTION).)*?/end FUNCTION'

Breakdown of Key Changes #

• (?:(?!/end FUNCTION).)*?: This non-capturing group uses a negative lookahead assertion. It tells the regex to match characters one at a time, but only as long as those characters don't start the /end FUNCTION sequence. This keeps the match confined to the current FUNCTION block, preventing it from spilling into the next one.
• We removed the extra space before /begin IN_MEASUREMENT (your original pattern had a space there which might cause issues with inconsistent formatting).
• The pattern now strictly looks for IN_MEASUREMENT inside the current FUNCTION block before matching the rest of the block up to /end FUNCTION.

Full Working Code #

import re
sFileContent = ''' /begin FUNCTION foo "a.1.2.3 footext" /begin SUB_FUNK fooo foobar foobar /end SUB_FUNK FUNK_VERSION "a.1.2.3" /end FUNCTION /begin FUNCTION foobarbarbar "SSV-TAG 3.1.0-3.1.0 foobarbarbar tex text_tex" /begin SUB_FUNK foobarbarbarfoo foobarbarbarfooo foobarbarbarfoooo /end SUB_FUNK /begin IN_MEASUREMENT barbar barfoo /end IN_MEASUREMENT /begin OUT_MEASUREMENT barfoobar barbarfoo /end OUT_MEASUREMENT /begin LOC_MEASUREMENT loc_barfoobar barbarfoo_loc /end LOC_MEASUREMENT FUNK_VERSION "3.1" /end FUNCTION'''
sSearch = r'/begin FUNCTION\s*(\S*)\s*"[^"]*"(?:(?!/end FUNCTION).)*?/begin IN_MEASUREMENT\s*(.*?)\s*/end IN_MEASUREMENT(?:(?!/end FUNCTION).)*?/end FUNCTION'
lMatches = re.findall(sSearch, sFileContent, re.S)
print(lMatches)

Output #

When you run this, you'll get the expected result:

[('foobarbarbar', 'barbar barfoo')]

This approach is efficient and clean—all logic is contained in one regex, so you don't need to loop through FUNCTION blocks and search each one separately.

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