3 g~D@sdZddlZddlZddlZddlZddlZddlZddlZddlZ ddl Z ddl Z ddl mZddlmZmZmZmZmZmZmZmZddlmZmZddlmZddlmZmZddl m!Z!dd l"m#Z#dd l$m%Z%dd l&m'Z'dd l(m)Z)dd l*m+Z+ddl,m-Z-m.Z.ddl/m0Z0ddl1m2Z2m3Z3m4Z4ej5e6Z7e j8j9j:Z;eeGddde?Z@eddddZAGddde?ZBed7d7ZSdNe#eer<r;)urllibparseurlsplitr9headrr8)rr:r"netlocpathqueryfragmentrespr#r#r$_ensure_html_responsePs rGcCsRtt|jrt||dtjdt||j|dddd}t|t ||S)aAccess an HTML page with GET, and return the response. This consists of three parts: 1. If the URL looks suspiciously like an archive, send a HEAD first to check the Content-Type is HTML, to avoid downloading a large file. Raise `_NotHTTP` if the content type cannot be determined, or `_NotHTML` if it is not HTML. 2. Actually perform the request. Raise HTTP exceptions on network failures. 3. Check the Content-Type header to make sure we got HTML, and raise `_NotHTML` otherwise. )r:zGetting page %sz text/htmlz max-age=0)Acceptz Cache-Control)r4) rrfilenamerGloggerdebugrr5rr8)rr:rFr#r#r$_get_html_response`s   rL)r4rcCs2|r.d|kr.tj|d\}}d|kr.|dSdS)z=Determine if we have any encoding information in our headers.z Content-TypecharsetN)cgi parse_header)r4r'paramsr#r#r$_get_encoding_from_headerss  rQ)documentpage_urlrcCs.x(|jdD]}|jd}|dk r |Sq W|S)aDetermine the HTML document's base URL. This looks for a ```` tag in the HTML document. If present, its href attribute denotes the base URL of anchor tags in the document. If there is no such tag (or if it does not have a valid href attribute), the HTML file's URL is used as the base URL. :param document: An HTML document representation. The current implementation expects the result of ``html5lib.parse()``. :param page_url: The URL of the HTML document. z.//basehrefN)findallr5)rRrSbaserTr#r#r$_determine_base_urls  rW)partrcCstjjtjj|S)zP Clean a "part" of a URL path (i.e. after splitting on "@" characters). )r>r?quoteunquote)rXr#r#r$_clean_url_path_partsr[cCstjjtjj|S)z Clean the first part of a URL path that corresponds to a local filesystem path (i.e. the first part after splitting on "@" characters). )r>r6 pathname2url url2pathname)rXr#r#r$_clean_file_url_paths r^z(@|%2F))rC is_local_pathrcCsb|r t}nt}tj|}g}x:ttj|dgD]$\}}|j|||j|jq0Wdj |S)z* Clean the path portion of a URL. r3) r^r[_reserved_chars_resplitr itertoolschainappendupperjoin)rCr_Z clean_funcpartsZ cleaned_partsZto_cleanreservedr#r#r$_clean_url_paths ricCs6tjj|}|j }t|j|d}tjj|j|dS)z Make sure a link is fully quoted. For example, if ' ' occurs in the URL, it will be replaced with "%20", and without double-quoting other characters. )r_)rC)r>r?urlparserBrirC urlunparse_replace)rresultr_rCr#r#r$ _clean_links rn)anchorrSbase_urlrcCsL|jd}|sdSttjj||}|jd}|jd}t||||d}|S)zJ Convert an anchor element in a simple repository page to a Link. rTNzdata-requires-pythonz data-yanked) comes_fromrequires_python yanked_reason)r5rnr>r?urljoinr)rorSrprTrZ pyrequirerslinkr#r#r$_create_link_from_elements   rvc@s:eZdZdddddZeedddZed d d ZdS) CacheablePageContentHTMLPageN)pagercCs|js t||_dS)N)cache_link_parsingAssertionErrorry)r+ryr#r#r$r*s zCacheablePageContent.__init__)otherrcCst|t|o|jj|jjkS)N) isinstancetyperyr)r+r|r#r#r$__eq__ szCacheablePageContent.__eq__)rcCs t|jjS)N)hashryr)r+r#r#r$__hash__szCacheablePageContent.__hash__) r-r.r/r*objectboolrintrr#r#r#r$rwsrwrx)fnrcsLtjddtttdfdd tjdttdfdd }|S) z Given a function that parses an Iterable[Link] from an HTMLPage, cache the function's result (keyed by CacheablePageContent), unless the HTMLPage `page` has `page.cache_link_parsing == False`. N)maxsize)cacheable_pagercst|jS)N)listry)r)rr#r$wrappersz'with_cached_html_pages..wrapperrx)ryrcs|jrt|St|S)N)rzrwr)ry)rrr#r$wrapper_wrappers z/with_cached_html_pages..wrapper_wrapper) functools lru_cacherwrrwraps)rrr#)rrr$with_cached_html_pagess r)ryrccsZtj|j|jdd}|j}t||}x0|jdD]"}t|||d}|dkrLq0|Vq0WdS)zP Parse an HTML document, and yield its anchor elements as Link objects. F)Ztransport_encodingZnamespaceHTMLElementsz.//a)rSrpN)r r?contentencodingrrWrUrv)ryrRrrprorur#r#r$ parse_links(s rc@s:eZdZdZd eeeeeddddZeddd Z dS) rxz'Represents one page, along with its URLTN)rrrrzrcCs||_||_||_||_dS)am :param encoding: the encoding to decode the given content. :param url: the URL from which the HTML was downloaded. :param cache_link_parsing: whether links parsed from this page's url should be cached. PyPI index urls should have this set to False, for example. N)rrrrz)r+rrrrzr#r#r$r*CszHTMLPage.__init__)rcCs t|jS)N)rr)r+r#r#r$__str__VszHTMLPage.__str__)T) r-r.r/__doc__bytesrr0rr*rr#r#r#r$rx@s .)rureasonmethrcCs|dkrtj}|d||dS)Nz%Could not fetch URL %s: %s - skipping)rJrK)rurrr#r#r$_handle_get_page_failZsrT)r2rzrcCst|j}t|j||j|dS)N)rrrz)rQr4rxrr)r2rzrr#r#r$_make_html_pageds  r)rur:rc Cs |dkrtd|jjddd}t|}|r@tjd||dStjj|\}}}}}}|dkrt j j tj j |r|jds|d7}tjj|d}tjd |yt||d }WnFtk rtjd |Yn4tk r }ztjd ||j|jWYdd}~Xntk r6}zt||WYdd}~Xntk rb}zt||WYdd}~Xntk r}z&d } | t|7} t|| tjdWYdd}~Xndtjk r}zt|d|WYdd}~Xn0tjk rt|dYnXt||j dSdS)Nz?_get_html_page() missing 1 required keyword argument: 'session'#rrzICannot look at %s URL %s because it does not support lookup as web pages.file/z index.htmlz# file: URL is directory, getting %s)r:z`Skipping page %s because it looks like an archive, and cannot be checked by a HTTP HEAD request.ziSkipping page %s because the %s request got Content-Type: %s.The only supported Content-Type is text/htmlz4There was a problem confirming the ssl certificate: )rzconnection error: z timed out)rz)! TypeErrorrrar%rJwarningr>r?rjosrCisdirr6r]endswithrtrKrLr9r&r(r'rrrrr0infor ConnectionErrorTimeoutrrz) rur:rZ vcs_schemer"_rCrFexcrr#r#r$_get_html_pagensV      rc@s&eZdZUeeeeeedS)CollectedSourcesN)r-r.r/r rr find_links index_urlsr#r#r#r$rs  rc@sxeZdZdZeeddddZedeee dddd Z e e e d d d Zeeed ddZe eedddZdS) LinkCollectorz Responsible for collecting Link objects from all configured locations, making network requests as needed. The class's main method is its collect_sources() method. N)r: search_scopercCs||_||_dS)N)rr:)r+r:rr#r#r$r*szLinkCollector.__init__F)r:optionssuppress_no_indexrcCsb|jg|j}|jr:| r:tjddjdd|Dg}|jpBg}tj||d}t ||d}|S)z :param session: The Session to use to make requests. :param suppress_no_index: Whether to ignore the --no-index option when constructing the SearchScope object. zIgnoring indexes: %s,css|]}t|VqdS)N)r).0rr#r#r$ sz'LinkCollector.create..)rr)r:r) index_urlextra_index_urlsno_indexrJrKrfrrcreater)clsr:rrrrrlink_collectorr#r#r$rs   zLinkCollector.create)rcCs|jjS)N)rr)r+r#r#r$rszLinkCollector.find_links)locationrcCst||jdS)z> Fetch an HTML page containing package links. )r:)rr:)r+rr#r#r$ fetch_pageszLinkCollector.fetch_page) project_namecandidates_from_pagercstjfddjj|Dj}tjfddjDj}tjtj rddt j ||D}t |d|dg|}tj dj|tt|t|d S) Nc3s$|]}t|jjdddVqdS)F)rpage_validator expand_dirrzN)rr:is_secure_origin)rloc)rr+r#r$rsz0LinkCollector.collect_sources..c3s$|]}t|jjdddVqdS)T)rrrrzN)rr:r)rr)rr+r#r$rscSs*g|]"}|dk r|jdk rd|jqS)Nz* )ru)rsr#r#r$ sz1LinkCollector.collect_sources..z' location(s) to search for versions of : )rr) collections OrderedDictrZget_index_urls_locationsvaluesrrJ isEnabledForloggingDEBUGrbrcr!rKrfrr)r+rrZindex_url_sourcesZfind_links_sourceslinesr#)rr+r$collect_sourcess   zLinkCollector.collect_sources)F)r-r.r/rrrr* classmethodrrrpropertyrr0rrrrxrrrrr#r#r#r$rsr)N).N)T)N)YrrNrrrbrrre urllib.parser>urllib.requestZxml.etree.ElementTreexmloptparsertypingrrrrrrr r pip._vendorr r Zpip._vendor.requestsr Zpip._vendor.requests.exceptionsrrpip._internal.exceptionsrpip._internal.models.linkrZ!pip._internal.models.search_scoperpip._internal.network.sessionrZpip._internal.network.utilsrpip._internal.utils.filetypesrpip._internal.utils.miscrrpip._internal.vcsrsourcesrrr getLoggerr-rJetreeZ ElementTreeZElementZ HTMLElementr0ZResponseHeadersr% Exceptionr&r8r9rGrLrQrWr[r^compile IGNORECASEr`rrirnrvrwrrrxrrrrrr#r#r#r$sr (              2   =