ǁij4ddlmZddlZddlZddlZddlZddlmZmZddlm Z ddl m Z ddl m Z ddlmZmZddlmZdd lmZmZdd lmZmZddlZdd lmZdd lmZdd lmZddlm Z m!Z!m"Z"ddl#m$Z$m%Z%m&Z&m'Z'm(Z(ejRdZ*GddZ+y)) annotationsN)datetimetimezone)cached_property)cycle)shuffle)sleeptime) TracebackType)AnyLiteral)parse_qsurlparse)_Element) HTMLParser)document_fromstring)DuckDuckGoSearchExceptionRatelimitExceptionTimeoutException)_expand_proxy_tb_alias _extract_vqd _normalize_normalize_url json_loadszduckduckgo_search.DDGSceZdZdZ d ddZddZ d ddZeddZdddZ d ddZ dd Z d dd Z d dd Z d dd Z d dd Z d d dZ d! d"dZ d# d$dZy)%DDGSzBDuckDuckgo_search class to get search results from duckduckgo.com.Nc ptjdtjdtdtj j d}|r|n t||_|j!t|jtsJd|sT|rRtjdd t|tr$|j d xs|j d n||_|r|ni|_ ||_ tj|j|jd d d d d||_d|_y)aInitialize the DDGS object. Args: headers (dict, optional): Dictionary of headers for the HTTP client. Defaults to None. proxy (str, optional): proxy for the HTTP client, supports http/https/socks5 protocols. example: "http://user:pass@example.com:3128". Defaults to None. timeout (int, optional): Timeout value for the HTTP client. Defaults to 10. verify (bool): SSL verification when making the request. Defaults to True. alwaysz^This package (`duckduckgo_search`) has been renamed to `ddgs`! Use `pip install ddgs` instead. stacklevel DDGS_PROXYNzproxy must be a strz-'proxies' is deprecated, use 'proxy' instead.rhttphttpsTrandomF)proxytimeout cookie_storereferer impersonateimpersonate_osfollow_redirectsverify)warnings simplefilterwarnRuntimeWarningosenvirongetrr' isinstancestrdictheadersr(primpClientclientsleep_timestamp)selfr:r'proxiesr(r. ddgs_proxys m/home/obispo/Crisostomo_bridge/mision_env/lib/python3.12/site-packages/duckduckgo_search/duckduckgo_search.py__init__z DDGS.__init__$s" h' l  "$ != /9?UV[?\ zz!Z C%@WBWW@ MMIVW XHRSZ\`HaV,D G0DgnDJ")wr  ll**LL #"    #c|SNr?s rB __enter__zDDGS.__enter__Ss rDcyrFrG)r?exc_typeexc_valexc_tbs rB__exit__z DDGS.__exit__Vs rDc tddddS)zGet HTML parser.TF)remove_blank_textremove_comments remove_pis collect_ids) LHTMLParserrHs rBparserz DDGS.parser^sT4TXfkllrDc|jsdnt|jz dk\rdn|}t|_t|y)zSleep between API requests.r/N)r>r r )r? sleeptimedelays rB_sleepz DDGS._sleepcs://DFTEYEY.>-?@A   s "K   !? ?$z43C3C2DJ%OP P'488*4J6):WJV]X\W^(_`` XR(&#aR0A0A/B"RD'IJPRR+se1T"X5F5F4Gr",NOUW W Xs2C E A)E  EcZ|jddd|ij}t||S)z!Get vqd value for a search query.GETzhttps://duckduckgo.comqr\)rpr]r)r?keywords resp_contents rB_get_vqdz DDGS._get_vqds/}}U,DcS[_}]ee L(33rDc|dvrtjd|ddd}|dk(rddgn|g}t|d g}gd } }|D]S} | dk(r|j||||}n3| dk(r|j ||||}n| d k(r|j ||||}|cSt| #t $r'} tjd | d | | } Yd } ~ d } ~ wwxYw) aDuckDuckGo text search. Query params: https://duckduckgo.com/params. Args: keywords: keywords for query. region: us-en, uk-en, ru-ru, etc. Defaults to None. safesearch: on, moderate, off. Defaults to "moderate". timelimit: d, w, m, y. Defaults to None. backend: auto, html, lite. Defaults to auto. auto - try all backends in random order, html - collect data from https://html.duckduckgo.com, lite - collect data from https://lite.duckduckgo.com, bing - collect data from https://www.bing.com. max_results: max number of results. If None, returns results only from the first response. Defaults to None. Returns: List of dictionaries with search results, or None if there was an error. Raises: DuckDuckGoSearchException: Base exception for duckduckgo_search errors. RatelimitException: Inherits from DuckDuckGoSearchException, raised for exceeding API request rate limits. TimeoutException: Inherits from DuckDuckGoSearchException, raised for API request timeouts. )apiecosiazbackend=z$ is deprecated, using backend='auto'r r!autohtmllitebingNzError to search using z backend: ) r0r2r _text_html _text_lite _text_bingreriinfor) r?ruregion safesearch timelimitbackend max_resultsbackendsresultserrbros rBtextz DDGS.texts> ' ' MMXWJ&JKXY ZG'.&'8FF#wi84 A ;"ooh ;WG&["ooh ;WG&["ooh ;WG (,,   4QCz"FG sAB## C,CCc Z|sJdddd}|jj||dd}|r||d<|r||d<t}g}td D]M} |j d d | j } d | vr|cSt | |j} | jd} t| ts|cS| D]*} t| ts| jd}|rt|trt|dnd}|sK||vsP|jdrb|j|| jd}|rt|trt|dnd}| jd}|r(t|trdjd|Dnd}|j!t#|t%|t#|d|st'||k\s'|ccS| jd}|r|s|cSt|tr|dnd}t|ts|jd}|jd}t|ts t|tst)||Dcic]\}}t|t|}}}P|Scc}}w)Nkeywords is mandatoryzhttps://html.duckduckgo.com/?1RefererzSec-Fetch-UserrsrkldfPOSTz https://html.duckduckgo.com/htmlr^s No results.z //div[h2]z ./a/@hrefrzhttp://www.google.com/search?q=z%https://duckduckgo.com/y.js?ad_domainz ./h2/a/text()z ./a//text()c32K|]}t|ywrFr8.0xs rB z"DDGS._text_html..&A!s1v&Atitlehrefbodyz.//div[@class="nav-link"].//input[@type="hidden"]/@name.//input[@type="hidden"]/@value)r=headers_updatesetrangerpr]rrUxpathr7listrr8 startswithaddjoinappendrrlenzip)r?rurrrr:payloadcacher_rvtreeelementse hrefxpathr titlexpathr bodyxpathrnpx next_pagenamesvaluesnvs rBrzDDGS._text_htmls000x6"  ""7+  "GDM %GDM(*q, NA==1SZa=bjjL,.&|T[[ADzz+.Hh- +a* ! 4I09jTX>Y3y|,_cD - $h! $%&WW_%= 6@ZPZ\`EaJqM 2gi$%GGM$: ENS]^gimSnrww&Ay&AAtv)3E):(6t(<(24(8'3w<;+F#*N1 +4**89Ck#-c4#8BdI)X.!(HI")JKeT*z&$/G:=eV:LM$!Qs1vs1v~MGMY, N\NsJ'c &|sJdddd}|jj||dd}|r||d<|r||d<t}g}td D]} |j d d | j } d | vr|cSt | |j} | jd} t| ts|cStttdd | } | D]l\}}t|ts|dk(r|jd}|rt|trt|dnd}|||vs|jdr$tdD cgc]} t!| dc} |j#||jd}|rt|trt|dnd}|dk(s|jd}|r6t|tr&dj%d|Dj'nd}s%|j)t+t-|t+|d|sYt/||k\si|ccS| jd}|r|s|cSt|tr|dnd}t|ts8|jd}|jd}t|tslt|ts~t||Dcic]\}}t|t|}}}|Scc} wcc}}w)Nrzhttps://lite.duckduckgo.com/rrrrrrrrz!https://lite.duckduckgo.com/lite/rsNo more results.z//table[last()]//trrz .//a//@hrefrrz .//a//text()r z&.//td[@class='result-snippet']//text()c32K|]}t|ywrFrrs rBrz"DDGS._text_lite..Ms#>qCF#>rrz(//form[./input[contains(@value, 'ext')]]rrr)r=rrrrpr]rrUrr7rrrrr8rnextrrstriprrrr)r?rurrrr:rrrrrvrrr^irrrrrrrrrrrrrs rBrzDDGS._text_lites000x6"  ""7+  "GDM %GDM(*q6 NA==1T[b=ckkL"l2&|T[[ADzz"78Hh-uU1a[)84D! /1a*Av$%GGM$: 4=*YX\B]s9Q<0cg L#u}# l 8=Qx@!T$-@!IIdO)*)@J:DT^`dIeC 1 $6kmEa$%GG,T$U  )Z 4-HGG#>I#>>DDF!#  #NN-7->,:4,@,6t,<!" +s7|{/J'.C! /F**GHCk#-c4#8BdI)X.!(HI")JKeT*z&$/G:=eV:LM$!Qs1vs1v~MGMm6 NpCA>Ns L!L c F|sJd|r-d|d|d|d|d}|jjd|d|i}|r;ttd z}|d k(r d |d z d |n ddddd|z}d|d|d<t } g} t dD] } |j dd|j} d| vr| cSt| |j} | jd}t|ts| cS|D]~}t|ts|jd}|rt|trt|dnd}|rJ|jdr9dt!t#|j$j'd d!gd}|s|| vs| j)||jd"}|r(t|trd!j+d#|Dnd!}|jd$}|r(t|trd!j+d%|Dnd!}| j-t/|t1|t/|j3d&d'd(|skt5| |k\s{| ccS|s| cS| d)zd*zd)z|d+<d,| dkDr| nd!|d-<| S).Nrzu=z&m=zui=z&mkt=)_EDGE_CD_EDGE_Szhttps://www.bing.comrsiQyez5_imrez123)dwmzex1:""filtersrrrzhttps://www.bing.com/searchrtzThere are no results forz //li[contains(@class, 'b_algo')]z8./h2/a/@href | ./div[contains(@class, 'header')]/a/@hrefrzhttps://www.bing.com/ck/a?c|rIt|dkDr;tj|ddx}dt| dzzzjSdS)Nr =)rbase64urlsafe_b64decodedecode)urs rBz!DDGS._text_bing..sV SVaZ'-&>&>QqrU cWZ[\W]V]abUbNc?c&d&k&k&m!%rDrrz?./h2/a//text() | ./div[contains(@class, 'header')]/a/h2//text()c32K|]}t|ywrFrrs rBrz"DDGS._text_bing..s#?qCF#?rz .//p//text()c32K|]}t|ywrFrrs rBrz"DDGS._text_bing..rr rarr firstPEREFORM)r= set_cookiesintr rrrprrrUrr7rrr8rrrqueryr6rrrrrreplacer)r?rurrrr_rrcoderrpagervrrrrrrrrrs rBrzDDGS._text_binghs000x  F84 fX6G KK # #$:G D   DFeO$A+4+;T!c'!A3'cX[beHfgpHqAqD#(aGI (*!H* @D==0MV]=^ccL)\9&|T[[ADzz"DEHh- +a* !(b cI09jTX>Y3y|,_cD0L M&#8D>#7#78<?GFOU* @XrDc |sJd|j|} ddi} |jj| dddd} |rd|nd}|rd |nd}|rd |nd}|rd |nd}|rd |nd}| rd | nd} d||| | |j|d|d|d|d|d| d}t }g}t dD]}|j dd|ddij}t|}|jdg}|D]}|jd}|s||vs|j||dt|t|dt|d|d|d|dd}|j|| sxt|| k\s|ccS|jd }|| s|cS|jd!d"jd#d$|d%<|S)&a'DuckDuckGo images search. Query params: https://duckduckgo.com/params. Args: keywords: keywords for query. region: us-en, uk-en, ru-ru, etc. Defaults to "us-en". safesearch: on, moderate, off. Defaults to "moderate". timelimit: Day, Week, Month, Year. Defaults to None. size: Small, Medium, Large, Wallpaper. Defaults to None. color: color, Monochrome, Red, Orange, Yellow, Green, Blue, Purple, Pink, Brown, Black, Gray, Teal, White. Defaults to None. type_image: photo, clipart, gif, transparent, line. Defaults to None. layout: Square, Tall, Wide. Defaults to None. license_image: any (All Creative Commons), Public (PublicDomain), Share (Free to Share and Use), ShareCommercially (Free to Share and Use Commercially), Modify (Free to Modify, Share, and Use), ModifyCommercially (Free to Modify, Share, and Use Commercially). Defaults to None. max_results: max number of results. If None, returns results only from the first response. Defaults to None. Returns: List of dictionaries with images search results. Raises: DuckDuckGoSearchException: Base exception for duckduckgo_search errors. RatelimitException: Inherits from DuckDuckGoSearchException, raised for exceeding API request rate limits. TimeoutException: Inherits from DuckDuckGoSearchException, raised for API request timeouts. rzSec-Fetch-Modecorsr-1onmoderateoffztime:rzsize:zcolor:ztype:zlayout:zlicense:r`,)orslvqdpfrrrzhttps://duckduckgo.com/i.jsrzhttps://duckduckgo.com/)r\r:rimager thumbnailrkheightwidthsource)rrrrkrrrrs=r&rs)rwr=rrfrrrpr]rr6rrrrsplit)r?rurrrsizecolor type_imagelayout license_imagerrr:safesearch_baserrrrrv resp_json page_datarow image_urlresultrs rBimagesz DDGS.imagessBP000xmmH% f  ""7+!$#dC+4eI;'" !%tf~2$)& r-7uZL)R '-76(#26C(=/2  !1!1!34;avQugQzl!F81]OT  (*q >A==4WyZsNt)g #<0I! i4I  'GGG, %!7IIi(!$W!/ !:%3C 4D%E-c%j9"%h-!$W"%h-FNN6*"s7|{'B& ' ==(D|;::d+B/55c:1=GCL5 >8rDc |sJd|j|} dddd} |rd|nd}|rd|nd}|rd |nd}|rd |nd}|d || |d |d |d || |jd } t} g} tdD]}|j dd| j }t |}|jdg}|D]F}|d| vs | j|d| j||s3t| |k\sB| ccS|jd}||s| cS|jddjdd| d<| S)a DuckDuckGo videos search. Query params: https://duckduckgo.com/params. Args: keywords: keywords for query. region: us-en, uk-en, ru-ru, etc. Defaults to "us-en". safesearch: on, moderate, off. Defaults to "moderate". timelimit: d, w, m. Defaults to None. resolution: high, standart. Defaults to None. duration: short, medium, long. Defaults to None. license_videos: creativeCommon, youtube. Defaults to None. max_results: max number of results. If None, returns results only from the first response. Defaults to None. Returns: List of dictionaries with videos search results. Raises: DuckDuckGoSearchException: Base exception for duckduckgo_search errors. RatelimitException: Inherits from DuckDuckGoSearchException, raised for exceeding API request rate limits. TimeoutException: Inherits from DuckDuckGoSearchException, raised for API request timeouts. rrr-2rzpublishedAfter:rzvideoDefinition:zvideoDuration:z videoLicense:r`r)rrrsrrrrrzhttps://duckduckgo.com/v.jsrtrr]rrrrrr) rwrfrrrpr]rr6rrrr)r?rurrr resolutiondurationlicense_videosrrrrrrrrvrr r rs rBvideosz DDGS.videoss>000xmmH%!$$tD5>oi[1B 8B' |4 2:^H:.=K=(89QS;a |1XJa7GH !1!1!34  (*q >A==0MV]=^ffL"<0I! i4I  'y>.IIc)n-NN3'"s7|{'B&  ' ==(D|;::d+B/55c:1=GCL > rDc >|sJd|j|}dddd}|dd||||jd}|r||d<t} g} td D]@} |j d d | j } t | } | jd g}|D]}|d| vs | j|d|jdd}tj|dtjj|dt|dt|dt||dd}| j!||st#| |k\s| ccS| jd}||s| cS|j%ddj%dd|d<C| S)aHDuckDuckGo news search. Query params: https://duckduckgo.com/params. Args: keywords: keywords for query. region: us-en, uk-en, ru-ru, etc. Defaults to "us-en". safesearch: on, moderate, off. Defaults to "moderate". timelimit: d, w, m. Defaults to None. max_results: max number of results. If None, returns results only from the first response. Defaults to None. Returns: List of dictionaries with news search results. Raises: DuckDuckGoSearchException: Base exception for duckduckgo_search errors. RatelimitException: Inherits from DuckDuckGoSearchException, raised for exceeding API request rate limits. TimeoutException: Inherits from DuckDuckGoSearchException, raised for API request timeouts. rrrrrr`)rrnoamprsrrrrrrzhttps://duckduckgo.com/news.jsrtrrkrNdaterexcerptr)rrrrkrrrrrrrr)rwrfrrrpr]rr6rr fromtimestamprutc isoformatrrrrr)r?rurrrrrrrrrrrvrr r r r rs rBnewsz DDGS.newsYs2000xmmH%!$$tD !1!1!34   %GDM(*q >A==0PY`=aiiL"<0I! i4I  'u:U*IIc%j) # 6I ( 6 6s6{HLL Q [ [ ]!$W *3y> :-c%j9!/ !:"%h- FNN6*"s7|{'B& ' ==(D|;::d+B/55c:1=GCL1 >4rD)NNNrT) r:dict[str, str] | Noner' str | Noner@zdict[str, str] | str | Noner( int | Noner.boolreturnNone)r!r)NNN)rKztype[BaseException] | NonerLzBaseException | NonerMzTracebackType | Noner!r")r!rT)g?)rXfloatr!r")NNNNNNN)rmzCLiteral['GET', 'HEAD', 'OPTIONS', 'DELETE', 'POST', 'PUT', 'PATCH']rkr8r\rr]z bytes | Noner^rr:rr_rr`r r(z float | Noner!r )rur8r!r8)NrNr{N)rur8rrrr8rrrr8rrr!list[dict[str, str]]) rur8rrrrrrr!r$) us-enrNNNNNNN)rur8rr8rr8rrrrrrrrrrrrrrr!r$)r%rNNNNN)rur8rr8rr8rrrrrrrrrrr!r$)r%rNN) rur8rr8rr8rrrrr!r$)rh __module__ __qualname____doc__rCrIrNrrUrZrprwrrrrr rrrGrDrBrr!sL*. /3  -#&-#-#- -#  -#  -# -#^ 04(,'+  , & %    mm)- $&*)-)- $"aS"a"a& "a  "a $ "a'"a'"a"a"a "aH4"$ $"&4-4-4- 4-  4-  4- 4- 4-r" $"& III I  I  I\" $"& SSS S  S  Sp" $"& GGG G  G  GX$ $ !%!$("&``` `  `  ````"` ` `J$ $!%#%)"&DDD D  D  DD#D D DR$ $"& FFF F  F  F FrDr), __future__rrloggingr4r0rr functoolsr itertoolsrr&rr r typesr typingr r urllib.parserrr; lxml.etreer lxml.htmlrrTr exceptionsrrrutilsrrrrr getLoggerrirrGrDrBr5si"  '%+ /)WW   3 4~ ~ rD