Ng%RdZddlmZddlmZmZmZmZmZm Z m Z ddl Z ddl m Z ddlmZddlmZddlmZdd lmZdd lmZmZdd lmZmZdd lmZmZdd lm Z m!Z!eej"e d-ddddidddddd d.d(Z#Gd)d*Z$Gd+d,Z%dS)/zProvides `partition_html().) annotations)IOAnyIteratorListLiteralOptionalcastN)etree)add_chunking_strategy)Element) read_txt_file)FileType)apply_metadataget_last_modified_date)Flow html_parser)!ontology_to_unstructured_elementsparse_html_to_ontology)is_temp_file_path lazypropertyTFv1to_text) filetextencodingurlheaders ssl_verifyskip_headers_and_footersdetection_originhtml_parser_versionimage_alt_modefilename Optional[str]rOptional[IO[bytes]]rrrrdict[str, str]rboolr r!r"Literal['v1', 'v2']r#Optional[Literal['to_text']]kwargsrreturn list[Element]c | |dkr|s|s|sgSt|||||||||| |  } tt| S)aPartitions an HTML document into its constituent elements. HTML source parameters ---------------------- The HTML to be partitioned can be specified four different ways: filename A string defining the target filename path. file A file-like object using "r" mode --> open(filename, "r"). text The string representation of the HTML document. url The URL of a webpage to parse. Only for URLs that return an HTML document. headers The HTTP headers to be used in the HTTP request when `url` is specified. ssl_verify If the URL parameter is set, determines whether or not SSL verification is performed on the HTTP request. encoding The encoding method used to decode the text input. If None, utf-8 will be used. skip_headers_and_footers If True, ignores any content that is within
or