Ƒi SrSSKJr SSKrSSKrSSKJrJr Sr\R"S\R5r \R"S\R5r \R"S\R5r S S jrS S jrSS jrg)z2Stage 1b: HTML/XML charset declaration extraction.) annotationsN)DETERMINISTIC_CONFIDENCEDetectionResultis*<\?xml[^>]+encoding\s*=\s*['"]([^'"]+)['"]s,]+charset\s*=\s*['"]?\s*([^\s'">;]+)s6]+content\s*=\s*['"][^'"]*charset=([^\s'">;]+)cURS5R5R5n[R"U5 U$![ [ [4a gf=f)a=Validate encoding name via codecs and return the lowercased original name. We use ``codecs.lookup()`` to verify the encoding is recognized by Python, but return the original (lowercased) name rather than the codec's canonical name so that common aliases like ``iso-8859-1`` and ``windows-1252`` are preserved as-is. asciiN)decodestriplowercodecslookup LookupErrorUnicodeDecodeError ValueError)nametexts W/var/www/html/banglarbhumi/venv/lib/python3.13/site-packages/chardet/pipeline/markup.py_normalize_encodingrsV{{7#))+113 d  +Z 8sAAAAc U(dgUS[n[[[4H]nUR U5nU(dM[ UR S55nUcM<[X5(dMN[U[SS9s $ g)aDScan the first bytes of *data* for an HTML/XML charset declaration. Checks for: 1. ```` 2. ```` 3. ```` :param data: The raw byte data to scan. :returns: A :class:`DetectionResult` with confidence 0.95, or ``None``. N)encoding confidencelanguage) _SCAN_LIMIT_XML_ENCODING_RE_HTML5_CHARSET_RE_HTML4_CONTENT_TYPE_REsearchrgroup_validate_bytesrr)dataheadpatternmatchrs rdetect_markup_charsetr$(sz    D$&79OPt$ 5*5;;q>:H#(G(G&%7! Q cbUS[RU5 g![[4a gf=f)zCheck that *data* can be decoded under *encoding* without errors. Only validates the first ``_SCAN_LIMIT`` bytes to avoid decoding a full 200 kB input just to verify a charset declaration found in the header. NFT)rrrr )r rs rrrGs: \k!!(+   ,s ..)rbytesreturnz str | None)r r'r(zDetectionResult | None)r r'rstrr(bool)__doc__ __future__rr rechardet.pipelinerrrcompile IGNORECASErrrrr$rr%rr2st8" F ::6 JJ8"--BBMM "> r%