1hE d dl Z d dlZd dlZd dlZd dlmZ d dlmZmZmZ G d de Z
G d de Z G d d e Z y)
N)
BeautifulSoup)EntitySubstitutionEncodingDetector
UnicodeDammitc e Zd ZdZd Zej j dg d d Zd Z d Z
d Z d Z d
Z
d Zy )
TestUnicodeDammitz"Standalone tests of UnicodeDammit.c @ d}t | }|j |k( sJ y )Nu I'm already Unicode! ☃)r unicode_markup)selfmarkupdammits S/var/www/skyplay_api_hub/venv/lib/python3.12/site-packages/bs4/tests/test_dammit.pytest_unicode_inputz$TestUnicodeDammit.test_unicode_input s% 3 v&$$... z smart_quotes_to,expect_converted))Nu ‘’“”)xmlz ‘’“”)htmlz ‘’“”)asciiz''""c f d}t |dg| j }|dj | k( sJ y)zbVerify the functionality of the smart_quotes_to argument
to the UnicodeDammit constructor.s windows-1252)known_definite_encodingssmart_quotes_toz
{}N)r r
format)r r expect_convertedr converteds r test_smart_quotes_toz&TestUnicodeDammit.test_smart_quotes_to sE 0! &4%5 +
.
O 2 23C DDDDr c ~ d}t | }|j j dk( sJ |j dk( sJ y )Ns Sacré bleu! ☃utf-8u Sacré bleu! ☃r original_encodinglowerr
)r utf8r
s r test_detect_utf8z"TestUnicodeDammit.test_detect_utf8* sC 1 t$''--/7:::$$(DDDDr c d}t |dg }|j j dk( sJ |j dk( sJ y )N
iso-8859-8u םולשr )r hebrewr
s r test_convert_hebrewz%TestUnicodeDammit.test_convert_hebrew0 sG $ v ~6''--/<???$$(BBBBr c d}t | }|j j dk( sJ |j j d |k( sJ y )Ns ケータイ Watchr )r r r r
encode)r utf_8r
s r /test_dont_see_smart_quotes_where_there_are_nonezATestUnicodeDammit.test_dont_see_smart_quotes_where_there_are_none6 sM I u%''--/7:::$$++G4===r c ~ dj d }t |dg }|j j dk( sJ y )N
Räksmörgåsr r% r) r r r r utf8_datar
s r test_ignore_inappropriate_codecsz2TestUnicodeDammit.test_ignore_inappropriate_codecs< s= #**73 y<.9''--/7:::r c dj d }dD ]. }t ||g }|j j dk( r.J y )Nr- r )z.utf8z...z
utF---16.!r. )r r0 bad_encodingr
s r test_ignore_invalid_codecsz,TestUnicodeDammit.test_ignore_invalid_codecsA sL #**73 : ?L"9|n=F++113w> >> ?r c dj d }t |dg }|j j dk( sJ t |ddg }|j J y )Nr- r )exclude_encodingsr r. r/ s r test_exclude_encodingsz(TestUnicodeDammit.test_exclude_encodingsG sd #**73 yWIF''--/>AAA yWn ;
? 0r r c B e Zd Zd Zd Zd Zd Zd Zd Zd Z d Z
d Z y
) TestEncodingDetectorc P t d }t |j }d|v sJ y )Ns' u utf-�)r list encodings)r detectedrD s r Ptest_encoding_detector_replaces_junk_in_encoding_name_with_replacement_characterzeTestEncodingDetector.test_encoding_detector_replaces_junk_in_encoding_name_with_replacement_characterV s- $$QR++, .);;;r c N dD ] }t |d }d|j k( r J y )N)s& s& s$ s# Tis_htmlzeuc-jp)r r r datar
s r test_detect_html5_style_meta_tagz5TestEncodingDetector.test_detect_html5_style_meta_tag] s5
8D #46Fv 7 77 77 8r c * d}t j j }t j t j
d }|t j _ t
| }d|j u sJ d|j v sJ t |d }|j sJ t j t j |t j _ y # t j t j |t j _ w xY w)NsT
بتر
ѐc y Nr? )strs r noopzETestEncodingDetector.test_last_ditch_entity_replacement..noop~ s r Tu �z html.parser) bs4r
_chardet_dammitloggingdisableWARNINGr contains_replacement_charactersr
r NOTSET)r docchardetrQ r
soups r "test_last_ditch_entity_replacementz7TestEncodingDetector.test_last_ditch_entity_replacementg s 2 **,,( 1
*.CJJ &"3'F6AAA AAv 4 44 44 m4D77 77 OOGNN +)0CJJ &
OOGNN +)0CJJ &s
A C :Dc b d}t | }d|j k( sJ d|j k( sJ y )N < a > < / a > u áéutf-16le)r r
r rJ s r test_byte_order_mark_removedz1TestEncodingDetector.test_byte_order_mark_removed s9 M t$ 5 5555V 5 5555r c d}t | }t |dg }d|j k( sJ t |dg }d|j k( sJ dg|j D cg c] }|d c}k( sJ d}t |dgd g
}d |j k( sJ dd g|j D cg c] }|d c}k( sJ y c c}w c c}w ) Nr^ zutf-16)r r )user_encodingsr_ r r$ r% )r rb )r r tried_encodings)r rK r
beforeafterxr& s r )test_known_definite_versus_user_encodingsz>TestEncodingDetector.test_known_definite_versus_user_encodings s
N t$ txjI633333
dG9=U 4 4444|f.D.D E! EEEE % gY ~
v 7 7777 &9O9O*PA1Q4*PPPP F +Qs B7# B<c d d}t j d 5 }t |dgdgdg }d d d \ }|j }t |t
sJ |j t k( sJ dj k( sJ g d|j D cg c] }|d c}k( sJ y # 1 sw Y vxY wc c}w )
Nr$ T)record shift-jisr r% )r override_encodingsrb )rj r r% r )
warningscatch_warningsr message
isinstanceDeprecationWarningfilename__file__r rc )r r& wr
warningrn rf s r "test_deprecated_override_encodingsz7TestEncodingDetector.test_deprecated_override_encodings s %
$
$D
1 Q"*5 $+9 ,~ F //'#5666 8+++v 7 7777 4!11 2aQqT 2
3s B!
B-!B*c , dj d }dj d }||z |z }t j t 5 |j d d d d t j | }d|j d k( sJ y # 1 sw Y 5xY w)Nu ☃☃☃r! u “Hi, I like Windows!” windows_1252u+ ☃☃☃“Hi, I like Windows!”☃☃☃)r) r< raisesUnicodeDecodeErrordecoder detwingle)r r! rw rY fixeds r test_detwinglez#TestEncodingDetector.test_detwingle s !))&1
.
&
\!D( ]]-
. JJv '',< V@TTTT s B
Bc dD ]B }|j d }|j d sJ t j | }||k( rBJ y )N)u œu ₓu ðr! )r) endswithr r{ )r tricky_unicode_charinputoutputs r +test_detwingle_ignores_multibyte_charactersz@TestEncodingDetector.test_detwingle_ignores_multibyte_characters sT
$
#
(..v6E>>'* **",,U3FU? "? #r c d}|j d }d}|j d }t j } ||d J d ||d k( sJ d ||d k( sJ d || k( sJ d || k( sJ d d
z } |||z J |||z J |||z dd dk( sJ ||d dk( sJ |d |z d dk( sJ |d
|z d J y )Nz0r z,FrH r Tz
iso-8859-1 i )rI search_entire_document)r a)r) r find_declared_encoding)r html_unicode
html_bytes xml_unicode xml_bytesmspacers r test_find_declared_encodingz0TestEncodingDetector.test_find_declared_encoding s< J !((1
D &&w/ 3 3 u-555!L$7777!J5555q~---q|+++ *$%---)#$,,,
fz!4 MQX X
X 48LHHH !$?<OOO !$?GGGr N) r8 r9 r: rF rL r\ r` rg ru r} r r r? r r rA rA U s4 <8#1J6 Q@
.U0
# 'Hr rA c | e Zd ZdZd Zej j dddg d Zd Z d Z
d Z d
Z d Z
d Zd
Zd Zd Zd Zd Zd Zej j dg d d Zej j dg d d Zej j ddg d Zy ) TestEntitySubstitutionz1Standalone tests of the EntitySubstitution class.c t | _ y rO )r subr s r setup_methodz#TestEntitySubstitution.setup_method s %r zoriginal,substituted)u foo∀☃õbaru foo∀☃õbar)u ‘’foo“”z ‘’foo“”c D | j j | |k( sJ y rO r substitute_html)r original substituteds r test_substitute_htmlz+TestEntitySubstitution.test_substitute_html! s! xx''1[@@@r c p dD ]1 \ }}d}||z }||z }| j j | |k( r1J y )N) )z⊧u ⊧)z𝔑u 𝔑)z≧̸u ≧̸)z¬ ¬)z⫬u ⫬z||)fjr )z>>)z<zfoo<bar>r r s r 'test_xml_quoting_handles_angle_bracketsz>TestEntitySubstitution.test_xml_quoting_handles_angle_bracketsm s xx&&z26FFFFr c D | j j d dk( sJ y )NzAT&TzAT&Tr r s r #test_xml_quoting_handles_ampersandsz:TestEntitySubstitution.test_xml_quoting_handles_ampersandsp s xx&&v.*<<TestEntitySubstitution.test_unambiguous_ampersands_not_escaped s> xx''/3666xx,,V4>>>r zmarkup,html,html5,html5raw))÷ ÷r r )z
&nonesuch;&nonesuch;r r )z÷
÷r r )z¡
¡r r c | j j | |k( sJ | j j | |k( sJ | j j | |k( sJ y rO )r r substitute_html5r )r r r html5html5raws r 'test_when_entity_ampersands_are_escapedz>TestEntitySubstitution.test_when_entity_ampersands_are_escaped s[ , xx''/4777xx((0E999xx,,V4@@@r z
markup,expect)z&nosuchentity;z&nosuchentity;c | j j | |k( sJ | j j | |k( sJ y rO r )r r expects r !test_ambiguous_ampersands_escapedz8TestEntitySubstitution.test_ambiguous_ampersands_escaped s> xx''/6999xx,,V4>>>r N)r8 r9 r: r; r r< r= r> r r r r r r r r r r r r r r r r r? r r r r s ;& [[
S
C
A
AB4
86LM Y 9G=O
6
[[
? ? [[ $
AA" [[ BC??r r )
r< rT rl rR r
bs4.dammitr r r objectr rA r r? r r r sL
D0 D0NCH6 CHLV?V V?r