正则表达式匹配一个混合大写字母和\\ n的字符串

Question

I want to write a regex which will match a string only if it starts with an \\n, continues with at least one capital letter and ends with an \\n. The string could contain repetitions of this pattern, for example:

\\n[AZ]\\n[AZ]\\n.

I've tried this regular expression: \\n(([AZ]+\\n)+), on this input:

200LA 012F5421F2E8A172 164 XRAY 1.950 0.176 NA no Endolysin [Enterobacteria phage T4] ||1C63A 1C64A 1C65A MNIFEMLRIDEGLRLKIYKDTEGYYTIGIGHLLTKSPSLNAAKSELDKAIGRNTNGVITKDEAEKLFNQDVDAAVRGILR NAKLKPVYDSLDAVRRAALINMVFQMGETGVAGFTNSLRMAQQKRWDEAAVNLAKSRWYNQTPNRAKRVITTFRTGTWDA YKNL

I expected to get this result: ('MNIFEMLRIDEGLRLKIYKDTEGYYTIGIGHLLTKSPSLNAAKSELDKAIGRNTNGVITKDEAEKLFNQDVDAAVRGILRNAKLKPVYDSLDAVRRAALINMVFQMGETGVAGFTNSLRMAQQKRWDEAAVNLAKSRWYNQTPNRAKRVITTFRTGTWDAYKNL')

but instead, I got this one: ('MNIFEMLRIDEGLRLKIYKDTEGYYTIGIGHLLTKSPSLNAAKSELDKAIGRNTNGVITKDEAEKLFNQDVDAAVRGILR\\nNAKLKPVYDSLDAVRRAALINMVFQMGETGVAGFTNSLRMAQQKRWDEAAVNLAKSRWYNQTPNRAKRVITTFRTGTWDA\\nYKNL\\n', 'YKNL\\n')

Does anybody know what went wrong?

Thanks!

Answer 1

Your regex matches the (longest possible version of the) first section that matches your condition.

The first line in your text doesn't start with an \\n so it moves to look at the line right after the first \\n , and since it matches the condition, the groups specified by your regex are accepted as a result.

For your result, I would suggest matching with \\n(?:[AZ]+\\n)+ ( ?: stands for non-capturing group to prevent capturing of none but the last group), then replaceing newlines with empty strings:

>>> a = """>200LA 012F5421F2E8A172 164 XRAY  1.950  0.176 NA no Endolysin <ENLYS_BPT4(1-164)> [Enterobacteria phage T4] ||1C63A 1C64A 1C65A
... MNIFEMLRIDEGLRLKIYKDTEGYYTIGIGHLLTKSPSLNAAKSELDKAIGRNTNGVITKDEAEKLFNQDVDAAVRGILR
... NAKLKPVYDSLDAVRRAALINMVFQMGETGVAGFTNSLRMAQQKRWDEAAVNLAKSRWYNQTPNRAKRVITTFRTGTWDA
... YKNL
... """
>>> m = re.findall('\n(?:[A-Z]+\n)+', a)
>>> m[0].replace('\n', '')
'MNIFEMLRIDEGLRLKIYKDTEGYYTIGIGHLLTKSPSLNAAKSELDKAIGRNTNGVITKDEAEKLFNQDVDAAVRGILRNAKLKPVYDSLDAVRRAALINMVFQMGETGVAGFTNSLRMAQQKRWDEAAVNLAKSRWYNQTPNRAKRVITTFRTGTWDAYKNL'