cQj(dZddlZddlZddlmZmZddlmZddlmZm Z ddl m Z ddl m Z mZdd lmZdd lmZdd lmZmZdd lmZdd lmZddlmZddlmZej<eZ GddZ!y)zDDGS class implementation.N)ThreadPoolExecutorwait)ceil)randomshuffle) TracebackType)AnyClassVar)BaseSearchEngine)ENGINES) DDGSExceptionTimeoutException) HttpClient)ResultsAggregator)SimpleFilterRanker)_expand_proxy_tb_aliasceZdZUdZdZeedzed< d*dddedzdedzde ezd dfd Z d+d Z d,d e e dzd e dzdedzd dfdZdeded eeefdZ d-ddddddddedededzdedededzdedzdeded ed eeeeffd!Zded ed eeeeffd"Zded ed eeeeffd#Zded ed eeeeffd$Zded ed eeeeffd%Zded ed eeeeffd&Zd.d'ed(ed eeeezffd)Zy)/DDGSaDDGS | Dux Distributed Global Search. A metasearch library that aggregates results from diverse web search services. Args: proxy: The proxy to use for the search. Defaults to None. timeout: The timeout for the search. Defaults to 5. verify: bool (True to verify, False to skip) or str path to a PEM file. Defaults to True. Attributes: threads: The maximum number of threads per search. Defaults to None (automatic, based on max_results). Raises: DDGSException: If an error occurs during the search. Example: >>> from ddgs import DDGS >>> results = DDGS().search("python") NthreadsT)verifyproxytimeoutrreturnct|xstjjd|_||_||_i|_y)N DDGS_PROXY)rosenvironget_proxy_timeout_verify_engines_cache)selfrrrs N/home/oem/CyberBrain26-unified/.venv/lib/python3.12/site-packages/ddgs/ddgs.py__init__z DDGS.__init__.s=-U3Srzz~~l7S     c|S)z7Enter the context manager and return the DDGS instance.)r$s r% __enter__zDDGS.__enter__<s r'exc_typeexc_valexc_tbcy)zExit the context manager.Nr))r$r+r,r-s r%__exit__z DDGS.__exit__@sr'categorybackendc t|trdj|}|jdDcgc]}|j }}tt |j }t|d|vsd|vr"|}|dk(rddg|Dcgc] }|dvs| c}z}n|}g}g} |D]?} t |j| x} r|j| /| j| A| rGtjdd jt| d jt|g} |D]x} | |jvr| j|j| 0| |j|j|j } | |j| <| j| z| s'tjd |j#|dS| j%d d | Scc}wcc}w)aRetrieve a list of search engine instances for a given category and backend. Args: category: The category of search engines (e.g., 'text', 'images', etc.). backend: A single or comma-delimited backends. Defaults to "auto". Returns: A list of initialized search engine instances corresponding to the specified category and backend. Instances are cached for reuse. ,autoalltext wikipedia grokipedia)r7r8z9%s - backends do not exist or are disabled. Available: %sz, rrrz backend is not set. Using 'auto'c&|jtfSN)priorityr)es r%z#DDGS._get_engines..sajj&%9r'T)keyreverse) isinstancelistjoinsplitstripr keysrrappendloggerwarningsortedr#r r!r" _get_enginessort)r$r0r1x backend_list engine_keysrFkengine_classes invalid_keysr? engine_class instancesengine_instances r%rKzDDGS._get_enginesHs gt $hhw'G+2==+=>a > >78,1134   \ !Ul%:D6!#\25nARmIma5nnD  )C&x044S99|9%%l3##C(  )  NNK &./ &-.  * 2Lt222  !4!4\!BC#/T[[$--`d`l`l"m4C##L1  1 2 NN= >$$Xv6 6 94HW? 6osG4 G9G9zus-enmoderate r r4)region safesearch timelimit max_resultspager1querykeywordsrXrYrZr[r\kwargsc |xs|}|s d} t| |j|| } t| D chc]} | jc} }t }t hd}|rt |t|dz dzn|}tjrt |tj}id}}t|d5}tg| ddD]\}} | r7| j|vr|j| j|f||||d | }| ||<t||k\s||k\rt||jd \}}|j!D]I\}}||vs |j#x}r,|j%||j'|jK|Dcic]}||| }}|st||k\snddd|j1}t3}|j5||}|r |r|d|S|Sd |vr t7|t|xsdcc} w#t($r.}|}t*j-d |j.|Yd}~ d}~wwxYwcc}w#1swYxYw)aPerform a search across engines in the given category. Args: category: The category of search engines (e.g., 'text', 'images', etc.). query: The search query. keywords: Deprecated alias for `query`. region: The region to use for the search (e.g., us-en, uk-en, ru-ru, etc.). safesearch: The safesearch setting (e.g., on, moderate, off). timelimit: The timelimit for the search (e.g., d, w, m, y) or custom date range. max_results: The maximum number of results to return. Defaults to 10. page: The page of results to return. Defaults to 1. backend: A single or comma-delimited backends. Defaults to "auto". **kwargs: Additional keyword arguments to pass to the search engines. Returns: A list of dictionaries containing the search results. zquery is mandatory.>urlhrefimage embed_urlrWr Nr) max_workersthread_name_prefix)start)rXrYrZr\FIRST_EXCEPTION)r return_whenzError in engine %s: %rz timed outzNo results found.)rrKlenprovidersetrminrrrr enumeratesubmitsearchrr!itemsresultextendadd ExceptionrHinfoname extract_dictsrrankr)r$r0r]r^rXrYrZr[r\r1r_msgenginesenginelen_unique_providersseen_providersresults_aggregatorrefutureserrexecutorifuturedonenot_doneff_enginerexresultsrankers r% _search_synczDDGS._search_syncs@!E'C$ $##Hg6"'#JFOO#JK#&5;LLq:rOZc.[25E0F0JK`t <<k4<<8K4 KF S W_&'7'7$'7qA  6.8 ,X__  &#-"+! !F'-GFOw<;.!{2B%)'4==Vg%hND('.}}Y 89Y() ?1?$6$=$=a$@$2$6$6x7H7H$I Y7??q'!*}?G?3'9#:k#I;  @%224#%++gu- ,77H"I! I(I!-I!<I!" I +#I I!I I!!I*c *|jd|fi|S)zPerform a text search.r6rr$r]r_s r%r6z DDGS.text t  9&99r'c *|jd|fi|S)zPerform an image search.imagesrrs r%rz DDGS.images t  5;F;;r'c *|jd|fi|S)zPerform a news search.newsrrs r%rz DDGS.newsrr'c *|jd|fi|S)zPerform a video search.videosrrs r%rz DDGS.videosrr'c *|jd|fi|S)zPerform a book search.booksrrs r%rz DDGS.bookss t  %:6::r'rafmtct|j|j|j}|j |}|j dk7rd|d|j }t ||dk(r |j}nB|dk(r |j}n0|dk(r |j}n|dk(r |j}n |j}||d S) a9Fetch a URL and extract its content. Args: url: The URL to fetch and extract content from. fmt: Output format: "text_markdown", "text_plain", "text_rich", "text" (raw HTML), "content" (raw bytes). Returns: A dictionary with 'url' and 'content' keys. r9zFailed to fetch z: HTTP text_plain text_richr6content)rar) rr r!r"r status_coderrrr6r text_markdown)r$rarclientresprzrs r%extractz DDGS.extracts$++t}}T\\Zzz#   s "$SE1A1A0BCC$ $ , ooG K nnG F]iiG I llG((Gw//r')N)rr)NNNr;)r)__name__ __module__ __qualname____doc__rr int__annotations__strboolr&r*type BaseExceptionrr/rBr r rKdictrr6rrrrbytesrr)r'r%rrs*%)GXcDj !(!  " Tz t  s    04(,'+ (}%,(%($ (  (=== s# $ =F $ Y8 $ $"$Y8Y8Y8* Y8  Y8Y8:Y84ZY8Y8Y8Y8 d38n Y8v:#::d38n1E:rsP  7" "7#&*)   8 ${0{0r'