V]c @ s d Z d d l m Z d d l Z d d l Z d d l Z d d l Z d d l m Z m Z d d l
m Z m Z m
Z
m Z d d l Z d d l m Z m Z m Z d d l m Z m Z m Z d d l Z y d d l m Z m Z e Z Wn e k
r Z e Z n Xe j d d
k o)e j d k Z! d e f d
YZ" d e f d YZ# d e f d YZ# d e f d YZ$ d e j% f d YZ& d e f d YZ' d e j% f d YZ( d e f d YZ) d e j% f d YZ* d S( s# Tests of Beautiful Soup as a whole.i( t set_traceN( t
BeautifulSoupt BeautifulStoneSoup( t CharsetMetaAttributeValuet ContentMetaAttributeValuet SoupStrainert NamespacedAttribute( t EntitySubstitutiont
UnicodeDammitt EncodingDetector( t default_buildert SoupTestt skipIf( t LXMLTreeBuildert LXMLTreeBuilderForXMLi i i t TestConstructorc B s5 e Z d Z d Z d Z d Z d Z RS( c C s/ d } | j | } | j d | j j d S( Nu
éé
u éé( t soupt assertEqualt h1t string( t selft dataR ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_short_unicode_input+ s c C s/ d } | j | } | j d | j j d S( Nu foo bar
u foo bar( R R R R ( R R R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_embedded_null0 s c C s> d j d } | j | d d g } | j d | j d S( Nu
Räksmörgåss utf-8t exclude_encodingss windows-1252( t encodeR R t original_encoding( R t utf8_dataR ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_exclude_encodings5 s c C s! d t f d Y} t d d d t } t j d t t d d | | } Wd QXt | j | sq t | j t d d | j j
| | } t j d t } t d d | d t } Wd QXt | d
j } | j
d s t | j | | j | j | | j
d S( Nt Mockc B s# e Z d Z d Z d Z RS( c [ s | | _ t | _ d S( N( t called_witht Truet is_xml( R t kwargs( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt __init__> s c S s d S( N( ( R R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt initialize_soupA s c _ s d S( Nt ( ( R t argsR! ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt prepare_markupC s ( t __name__t
__module__R" R# R& ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR = s t vart valuet convertEntitiest recordR$ t buildert
ignored_valuei sC Keyword arguments to the BeautifulSoup constructor will be ignored.( t objectt dictR t warningst catch_warningsR t
isinstanceR- t AssertionErrorR R t strt messaget
startswith( R R R! R R- t wt msg( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_custom_builder_class: s c C s d } | j | } | j } | j d | d | j d d g | d | j | d t d d } | j d | j d x i d d 6i d d 6f D]s } t j d
t " } | j | d d d | } Wd QX| j } | j d d g | d | j d | d q Wd S( Ns& s an id t idt at classR- t multi_valued_attributess a class t *R, t an( R R< R R
t NoneR1 R2 R ( R t markupR R< t
switcherooR8 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_cdata_list_attributes^ s !! ( R' R( R R R R: RD ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR ) s
$t TestWarningsc B sJ e Z e d Z d Z d Z d Z d Z d Z d Z RS( c C s' | j t j d } | j | d S( NiP ( R7 R t NO_PARSER_SPECIFIED_WARNINGt
assertTrue( R t st is_theret v( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt _no_parser_specified~ s c C sN t j d t } | j d } Wd QXt | d j } | j | d S( NR, s i ( R1 R2 R R R5 R6 t _assert_no_parser_specified( R R8 R R9 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt# test_warning_if_no_parser_specified s c C sQ t j d t } | j d d } Wd QXt | d j } | j | d S( NR, s t htmli ( R1 R2 R R R5 R6 RL ( R R8 R R9 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt* test_warning_if_parser_specified_too_vague s c C sA t j d t } | j d d } Wd QX| j g | d S( NR, s s html.parser( R1 R2 R R R ( R R8 R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt, test_no_warning_if_explicit_parser_specified s c C s t j d t " } | j d d t d } Wd QXt | d j } | j d | k | j d | k | j d | j d S( NR, s t parseOnlyTheset bi t
parse_onlys (
R1 R2 R R R R5 R6 RG R R ( R R8 R R9 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt) test_parseOnlyThese_renamed_to_parse_only s !c C s t j d t " } d } | j | d d } Wd QXt | d j } | j d | k | j d | k | j d | j d S( NR, s ét fromEncodingt utf8i t
from_encoding( R1 R2 R R R5 R6 RG R R ( R R8 RV R R9 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt* test_fromEncoding_renamed_to_from_encoding s c C s | j t | j d d t d S( Ns t no_such_argument( t assertRaisest TypeErrorR R ( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt" test_unrecognized_keyword_argument s (
R' R( R RK RM RO RP RT RX R\ ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyRE | s c B s5 e Z d Z d Z d Z d Z d Z RS( c C s t j } | j } zT t j d t } | j | } Wd QXt | d j } | j d | k Wd | j
Xt j d t } | j | } Wd QX| j d t | d S( NR, i s looks like a filename(
t tempfilet NamedTemporaryFilet nameR1 R2 R R R5 R6 RG t closeR t len( R t
filehandlet filenameR8 R R9 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_disk_file_warning s c C sK t j d t } | j d } Wd QX| j t d | D d S( NR, s http://www.crummybytes.com/c s s$ | ] } d t | j k Vq d S( s looks like a URLN( R5 R6 ( t .0R8 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pys s ( R1 R2 R R RG t any( R t warning_listR ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_url_warning_with_bytes_url s c C sK t j d t } | j d } Wd QX| j t d | D d S( NR, u http://www.crummyunicode.com/c s s$ | ] } d t | j k Vq d S( s looks like a URLN( R5 R6 ( Re R8 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pys s ( R1 R2 R R RG Rf ( R Rg R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt! test_url_warning_with_unicode_url s c C sK t j d t } | j d } Wd QX| j t d | D d S( NR, s$ http://www.crummybytes.com/ is greatc s s$ | ] } d t | j k Vq d S( s looks like a URLN( R5 R6 ( Re R8 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pys s ( R1 R2 R R t assertFalseRf ( R Rg R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt% test_url_warning_with_bytes_and_space s c C sK t j d t } | j d } Wd QX| j t d | D d S( NR, u% http://www.crummyuncode.com/ is greatc s s$ | ] } d t | j k Vq d S( s looks like a URLN( R5 R6 ( Re R8 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pys s ( R1 R2 R R Rj Rf ( R Rg R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt' test_url_warning_with_unicode_and_space s ( R' R( Rd Rh Ri Rk Rl ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyRE s
t TestSelectiveParsingc B s e Z d Z RS( c C sA d } t d } | j | d | } | j | j d d S( Ns& NoYesNoYes YesRR RS s YesYes Yes( R R R R ( R RB t strainerR ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_parse_with_soupstrainer s ( R' R( Ro ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyRm s t TestEntitySubstitutionc B s e Z d Z d Z d Z d Z d Z d Z d Z d Z d Z
d Z d
Z d Z
d Z d
Z RS( s1 Standalone tests of the EntitySubstitution class.c C s
t | _ d S( N( R t sub( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt setUp s c C s& d } | j | j j | d d S( Nu foo∀☃õbaru foo∀☃õbar( R Rq t substitute_html( R RH ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_simple_html_substitution s c C s5 d } t | } | j | j j | j d d S( Ns foos ‘’foo“”( R R Rq Rs RB ( R t quotest dammit( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_smart_quote_substitution s c C s) d } | j | j j | t | d S( Ns Welcome to "my bar"( R Rq t substitute_xmlt False( R RH ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pytI test_xml_converstion_includes_no_quotes_if_make_quoted_attribute_is_false s c C sB | j | j j d t d | j | j j d t d d S( Nt Welcomes "Welcome"s Bob's Bars "Bob's Bar"( R Rq Rx R ( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt6 test_xml_attribute_quoting_normally_uses_double_quotes s c C s) d } | j | j j | t d d S( Ns Welcome to "my bar"s 'Welcome to "my bar"'( R Rq Rx R ( R RH ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pytO test_xml_attribute_quoting_uses_single_quotes_when_value_contains_double_quotes s c C s) d } | j | j j | t d d S( Ns Welcome to "Bob's Bar"s" "Welcome to "Bob's Bar""( R Rq Rx R ( R RH ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pytb test_xml_attribute_quoting_escapes_single_quotes_when_value_contains_both_single_and_double_quotes s c C s& d } | j | j j | | d S( Ns Welcome to "Bob's Bar"( R Rq Rx ( R t quoted( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt< test_xml_quotes_arent_escaped_when_value_is_not_being_quoted s c C s | j | j j d d d S( Ns foos foo<bar>( R Rq Rx ( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt' test_xml_quoting_handles_angle_brackets
s c C s | j | j j d d d S( Ns AT&Ts AT&T( R Rq Rx ( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt# test_xml_quoting_handles_ampersands s c C s | j | j j d d d S( Ns ÁT&Ts ÁT&T( R Rq Rx ( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pytE test_xml_quoting_including_ampersands_when_they_are_part_of_an_entity s c C s | j | j j d d d S( Ns ÁT&Ts ÁT&T( R Rq t" substitute_xml_containing_entities( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pytD test_xml_quoting_ignoring_ampersands_when_they_are_part_of_an_entity s c C s& d } | j | j j | | d S( s: There's no need to do this except inside attribute values.s Bob's "bar"N( R Rq Rs ( R t text( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_quotes_not_html_substituted s ( R' R( t __doc__Rr Rt Rw Rz R| R} R~ R R R R R R ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyRp s t TestEncodingConversionc B sM e Z d Z d Z d Z d Z d Z e e d d Z RS( c C sH t t | j d | _ | j j d | _ | j | j d d S( NuU Sacré bleu!s utf-8sU Sacré bleu!( t superR Rr t unicode_dataR R R ( R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyRr ) s c C s t j j } t j t j z d } | t j _ d } | j | } | j } | j t | t
| j | | j | j | j | j
j d Wd t j t j | t j _ Xd S( Nc S s d S( N( RA ( R5 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt noop8 s s as utf-8( t bs4Rv t chardet_dammitt loggingt disablet WARNINGR t decodeRG R3 t unicodeR t document_forR t lowert NOTSET( R t chardetR t asciit soup_from_asciit unicode_output( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_ascii_in_unicode_out2 s c C sX | j | j } | j | j | j | j | j j d | j | j d d S( Nu Sacré bleu!( R R R R t fooR R RA ( R t soup_from_unicode( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_unicode_in_unicode_outF s c C sE | j | j } | j | j | j | j | j j d d S( Nu Sacré bleu!( R R R R R R R ( R t soup_from_utf8( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_utf8_in_unicode_outN s c C s2 | j | j } | j | j d | j d S( Ns utf-8( R R R R R ( R R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt
test_utf8_outU s sQ Bad HTMLParser detected; skipping test of non-ASCII characters in attribute name.c C s8 d } | j | j | j j d | j d d S( Nu RV ( R R t divR ( R RB ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt1 test_attribute_name_containing_unicode_charactersZ s (
R' R( Rr R R R R R t PYTHON_3_PRE_3_2R ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR % s t TestUnicodeDammitc B s e Z d Z d Z d Z d Z d Z d Z d Z d Z d Z
d Z d
Z d Z
d Z d
Z d Z d Z d Z d Z RS( s" Standalone tests of UnicodeDammit.c C s) d } t | } | j | j | d S( Nu I'm already Unicode! ☃( R R t unicode_markup( R RB Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_unicode_inputd s c C s) d } t | } | j | j d d S( Ns u ‘’“”( R R R ( R RB Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_smart_quotes_to_unicodei s c C s/ d } t | d d } | j | j d d S( Ns t smart_quotes_tot xmls+ ‘’“”( R R R ( R RB Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt! test_smart_quotes_to_xml_entitieso s c C s/ d } t | d d } | j | j d d S( Ns R RN s' ‘’“”( R R R ( R RB Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt" test_smart_quotes_to_html_entitiesu s c C s/ d } t | d d } | j | j d d S( Ns R R s ''""( R R R ( R RB Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_smart_quotes_to_ascii{ s c C sB d } t | } | j | j j d | j | j d d S( Ns Sacré bleu! ☃s utf-8u Sacré bleu! ☃( R R R R R ( R RV Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_detect_utf8 s c C sH d } t | d g } | j | j j d | j | j d d S( Ns s
iso-8859-8u םולש( R R R R R ( R t hebrewRv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_convert_hebrew s c C sK d } t | } | j | j j d | j | j j d | d S( Ns ケータイ Watchs utf-8( R R R R R R ( R t utf_8Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt/ test_dont_see_smart_quotes_where_there_are_none s c C s> d j d } t | d g } | j | j j d d S( Nu
Räksmörgåss utf-8s
iso-8859-8( R R R R R ( R R Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_ignore_inappropriate_codecs s c C sX d j d } xB d d d g D]1 } t | | g } | j | j j d q Wd S( Nu
Räksmörgåss utf-8s .utf8s ...s
utF---16.!( R R R R R ( R R t bad_encodingRv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_ignore_invalid_codecs s c C sl d j d } t | d d g } | j | j j d t | d d d g } | j | j d d S( Nu
Räksmörgåss utf-8R s windows-1252( R R R R R RA ( R R Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR s c C s1 t d } t | j } d | k s- t d S( Ns' u utf-�( R t listt encodingsR4 ( R t detectedR ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pytP test_encoding_detector_replaces_junk_in_encoding_name_with_replacement_character s c C s: x3 d D]+ } t | d t } | j d | j q Wd S( Ns& s& s$ s# t is_htmls euc-jp( s& s& s$ s# ( R R R R ( R R Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_detect_html5_style_meta_tag s
c C s d } t j j } t j t j zm d } | t j _ t | } | j t | j | j
d | j k t | d } | j
| j Wd t j t j
| t j _ Xd S( NsT
بتر
ѐc S s d S( N( RA ( R5 ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR s u �s html.parser( R Rv R R R R R R R t contains_replacement_charactersRG R R R ( R t docR R Rv R ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt" test_last_ditch_entity_replacement s c C s< d } t | } | j d | j | j d | j d S( Ns < a > < / a > u áés utf-16le( R R R R ( R R Rv ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_byte_order_mark_removed s c C sn d j d } d j d } | | | } | j t | j d t j | } | j d | j d d S( Nu ☃i RV u “Hi, I like Windows!”t windows_1252u+ ☃☃☃“Hi, I like Windows!”☃☃☃u ☃☃☃( R RZ t UnicodeDecodeErrorR R t detwingleR ( R RV R R t fixed( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_detwingle s c C sY xR d D]J } | j d } | j | j d t j | } | j | | q Wd S( Nu œu ₓu ðRV s ( u œu ₓu ð( R RG t endswithR R R ( R t tricky_unicode_chart inputt output( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt+ test_detwingle_ignores_multibyte_characters s
( R' R( R R R R R R R R R R R R R R R R R R ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR a s$ # t TestNamedspacedAttributec B s# e Z d Z d Z d Z RS( c C s# t d d } | j | d d S( Nt xmlns( R RA R ( R R< ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt test_name_may_be_none s c C s# t d d } | j d | d S( NR< RR s a:b( R R ( R R< ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt6 test_attribute_is_equivalent_to_colon_separated_string s c C s t d d d } t d d d } | j | | t d d d } | j | | t d d d } | j | | t d d d } | j | | d S( NR< RR t ct z( R R RA t assertNotEqual( R R< RR R t dt e( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt; test_attributes_are_equivalent_if_prefix_and_name_identical s ( R' R( R R R ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR s t) TestAttributeValueWithCharsetSubstitutionc B s e Z d Z d Z RS( c C sL t d } | j d | | j d | j | j d | j d d S( Ns euc-jpRV ( R R t original_valueR ( R R* ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt! test_content_meta_attribute_value, s c C sL t d } | j d | | j d | j | j d | j d d S( Ns text/html; charset=euc-jps text/html; charset=utf8RV ( R R R R ( R R* ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR 3 s ( R' R( R ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyR * s ( i i (+ R t pdbR R t unittestt sysR] R R R t bs4.elementR R R R t
bs4.dammitR R R t bs4.testingR
R R R1 t bs4.builderR
R R t LXML_PRESENTt ImportErrorR Ry t version_infoR R RE Rm t TestCaseRp R R R R ( ( ( s2 lib/python2.7/site-packages/bs4/tests/test_soup.pyt s4 "
"S,/ E<