
    o[we              
       0   d dl mZ d dlmZ d dlmZmZ d dlZd dlm	Z	 d dl
Z
d dlZd dlmZ d dlZd dlZd dlmZ d dlmc mZ d dlZd dlmZmZmZmZmZmZmZm Z m!Z! d dl"m#Z$ d d	l%m&Z&m'Z' d d
l(m)Z)  ej*        g d          d             Z+d Z,d Z-d Z.d Z/ej0        1                    d ej2        d ej3        d           ej3        d          g           ej2        d ej3        d                    g           G d d                      Z4dS )    )Iterator)partial)BytesIOStringION)Path)URLError)is_platform_windows)	NA	DataFrame
MultiIndexSeries	Timestamp
date_rangeread_csv	read_htmlto_datetime)ArrowStringArrayStringArray)file_path_to_url)zchinese_utf-16.htmlzchinese_utf-32.htmlzchinese_utf-8.htmlzletz_latin1.html)paramsc                 (     |ddd| j                   S )z6Parametrized fixture for HTML encoding test filenames.iodatahtml_encoding)param)requestdatapaths     9lib/python3.11/site-packages/pandas/tests/io/test_html.pyhtml_encoding_filer   (   s     8D&/7=AAA    c                 n   t          |           t          |          k    s*J dt          |            dt          |                       d}t          t          d | |                    }|s
J |            t          | |          D ],\  }}t	          j        ||g|R i | |j        r
J d            -d S )Nz*lists are not of equal size len(list1) == z, len(list2) == z$not all list elements are DataFramesc                 V    t          | t                    ot          |t                    S N)
isinstancer   )xys     r   <lambda>z(assert_framelist_equal.<locals>.<lambda>>   s    Ay11NjI6N6N r    zframes are both empty)lenallmapziptmassert_frame_equalempty)list1list2argskwargsmsgboth_framesframe_iframe_js           r   assert_framelist_equalr7   5   s    u::U###	&U	& 	&U	& 	& $##
 1CNN	
 	
 K u-- : :
gw@@@@@@@=99"99999: :r    c           	      &   t          j        d          }t          j        d           |                     |dd           t          j        t          d          5  t           |ddd	d
          d           d d d            d S # 1 swxY w Y   d S )Nbs4html5lib__version__z4.2zPandas requires versionmatchr   r   html	spam.htmlflavor)pytestimportorskipsetattrraisesImportErrorr   )monkeypatchr   r9   s      r   test_bs4_version_failsrH   I   s    

e
$
$C

###]E222	{*C	D	D	D M M((4==eLLLLM M M M M M M M M M M M M M M M M Ms   BB
B
c                      d} d}d|z   dz   }t          j        t          |          5  t          t	          |           d|           d d d            d S # 1 swxY w Y   d S )Nz
google.comzinvalid flavorz\{z \} is not a valid set of flavorsr<   googler=   rA   )rB   rE   
ValueErrorr   r   )urlrA   r3   s      r   test_invalid_flavorrN   R   s    
CF
&.>
>C	z	-	-	- @ @(3--x????@ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @s    AAAc                    t          j        d           t          j        d           t          j        d            | dddd          }t          |ddg	          }t          |ddg	          }t          ||           d S )
Nr9   lxmlr:   r   r   r>   valid_markup.htmlr   )	index_colrA   )rB   rC   r   r7   )r   filenamedfs_lxmldfs_bs4s       r   test_same_orderingrV   [   s    



###xff.ABBHQx@@@HAug>>>G8W-----r    rA   r9   r:   )marksrP   c            	       6   e Zd Zd Zej        d             Zej        d             Z ej        d          d             Zd Z	d Z
ej        j        ej        j        d	                         Zej        j        ej        j        d
                         Zej        j        d             Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Z d Z!d Z"d Z#d Z$d Z%ej        j        ej        j        d                          Z&ej        j        ej        j        ej        j        d!                                     Z'ej        j        d"             Z(ej        j        d#             Z)d$ Z*ej        j        d%             Z+ej        j        d&             Z,ej        j        d'             Z-ej        j        d(             Z.ej        j        d)             Z/ej        j        d*             Z0ej        j        d+             Z1d, Z2ej        d-             Z3ej        j        ej        j        d.                         Z4ej        j        ej        j        d/                         Z5d0 Z6d1 Z7d2 Z8d3 Z9d4 Z:d5 Z;ej        j        d6             Z<ej        j        d7             Z=d8 Z>d9 Z?d: Z@d; ZAd< ZBd= ZCd> ZDd? ZEd@ ZFdA ZGdB ZHdC ZIdD ZJej        K                    dEddFg          dG             ZLdH ZMdI ZNdJ ZOdK ZPdL ZQdM ZRdN ZSej        j        dO             ZTdP ZUdQ ZVej        K                    dRd eWdSg          dTfdF eWdUg           eWdSg          fg          dV             ZXej        K                    dWddFg          dX             ZYej        Z                    dY          dZ             Z[d[ Z\d\ Z]ej        j        ej        j        d]                         Z^d^ Z_d_ Z`ej        K                    dEg d`          da             Zadb Zbdc Zcdd Zdde ZedTS )fTestReadHtmlc                     d}t          j        t          |          5  |                     d           d d d            d S # 1 swxY w Y   d S )NzPassing literal html to 'read_html' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'StringIO' object.r<   a  <table>
                <thead>
                    <tr>
                        <th>A</th>
                        <th>B</th>
                    </tr>
                </thead>
                <tbody>
                    <tr>
                        <td>1</td>
                        <td>2</td>
                    </tr>
                </tbody>
                <tbody>
                    <tr>
                        <td>3</td>
                        <td>4</td>
                    </tr>
                </tbody>
            </table>)r,   assert_produces_warningFutureWarningr   selfr3   s     r   test_literal_html_deprecationz*TestReadHtml.test_literal_html_deprecationn   s    > 	 'SAAA 	 	NN  	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	s   AAAc                      |dddd          S )Nr   r   r>   r?    r^   r   s     r   	spam_datazTestReadHtml.spam_data   s    xffk:::r    c                      |dddd          S )Nr   r   r>   banklist.htmlra   rb   s     r   banklist_datazTestReadHtml.banklist_data   s    xffo>>>r    T)autousec              #   H   K   t          t          |          | _        d V  d S )Nr@   )r   r   )r^   rA   s     r   set_defaultszTestReadHtml.set_defaults   s'       6:::r    c                 F   t          j        ddd dd                              dj                                      t
                    }|                                }|                     t          |          ddid	
          d	         }t          j	        ||           d S )N      c                  d    t           j                            d                                          S )N   )nprandomdefault_rng)r1   s    r   r'   z2TestReadHtml.test_to_html_compat.<locals>.<lambda>   s"    )>)>q)A)A)H)H)J)J r    F)
data_gen_fc_idx_namesr_idx_namesz{:.3f}class	dataframer   )attrsrR   )
r,   makeCustomDataframer*   formatastypefloatto_htmlr   r   r-   )r^   dfoutress       r   test_to_html_compatz TestReadHtml.test_to_html_compat   s    "JJ!!   S!!&&-- 	 jjllnnXc]]7K2HTUnVV
 	c2&&&&&r    c                 *  	
 t          t          dt          j        dgd          t          g dd          t          dt          j        dgd          t          g d	d          g d
g dg dg dd          }|dk    ret	          t          j        g dt          j                            }t	          t          j        ddt          gt          j                            }n\t          j	        d          
t          
                    g d                    }t          
                    g d                    }|                    d          }t          j        d|          5  |                     t          |          |          d         }d d d            n# 1 swxY w Y   t          t          dt          j        dgd          t          g dd          t          dt          j        dgd          t          g d	d          t          ddt          gd          t          g dd          ||d          	|dk    r,dd l
ddlm t          	
fd	j        D                       	t)          j        |	           d S )N   rl   Int64)dtype)r   rn   rl         ?      @Float64)r   g       @r   )TFN)TFTabc)r   r   N)r   r   r   defghpythonr   r   pyarrowFindexzmode.string_storagedtype_backendr   Tboolean)ArrowExtensionArrayc           	      ^    i | ])}|                      |         d                     *S )T)from_pandas)array).0colr   expectedpas     r   
<dictcomp>z3TestReadHtml.test_dtype_backend.<locals>.<dictcomp>   sL        ,,RXXhsmQUX-V-VWW  r    )r   r   ro   nanr   r   object_r
   rB   rC   r   r|   pdoption_contextr   r   r   pandas.arraysr   columnsr,   r-   )r^   string_storager   r}   string_arraystring_array_nar~   resultr   r   r   s           @@@r   test_dtype_backendzTestReadHtml.test_dtype_backend   s   QN':::IIIW555S"&#.i@@@OOO9===(((((($__%%%	 	
 
 X%%&rxrz'R'R'RSSL)"(Cb>*T*T*TUUOO $Y//B+BHH___,E,EFFL.rxx8H8H8H/I/IJJOjjuj%%4nEE 	S 	S^^HSMM^OOPQRF	S 	S 	S 	S 	S 	S 	S 	S 	S 	S 	S 	S 	S 	S 	S QN':::IIIW555S"&#.i@@@OOO9===T5"-Y???///yAAA!$	 	
 
 I%%    999999      '/   H 	fh/////s   0+F''F+.F+c                 8   t          |d          5 }|                    |                                           |                     |j        d          }|                     |j        d          }d d d            n# 1 swxY w Y   t          ||           d S )Nutf-8encodingcontentFirst Federal Bank of Floridar<   Metcalf Bankopenserve_contentreadr   rM   r7   )r^   
httpserverrf   r   df1df2s         r   test_banklist_urlzTestReadHtml.test_banklist_url   s     -'222 	a$$QVVXX$666..5 !  C ..$ !  C	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	sC(((((   A!A??BBc                 8   t          |d          5 }|                    |                                           |                     |j        d          }|                     |j        d          }d d d            n# 1 swxY w Y   t          ||           d S )Nr   r   r   	.*Water.*r<   Unitr   )r^   r   rc   r   r   r   s         r   test_spam_urlzTestReadHtml.test_spam_url   s     )g... 	?!$$QVVXX$666..{.CCC..v.>>C	? 	? 	? 	? 	? 	? 	? 	? 	? 	? 	? 	? 	? 	? 	?
 	sC(((((r   c                     |                      |dddi          }|                      |dddi          }t          ||           d S )Nz.*Florida.*idtabler=   rw   r   r   r7   )r^   rf   r   r   s       r   test_banklistzTestReadHtml.test_banklist   sN    nn]-gnWWnn].wnXXsC(((((r    c                     |                      |d          }|                      |d          }t          ||           |d         j        d         dk    sJ |d         j        d         dk    sJ d S )Nr   r<   r   r   r   r   
ProximatesNutrient)r   r7   ilocr   r^   rc   r   r   s       r   	test_spamzTestReadHtml.test_spam  sz    nnYkn::nnYfn55sC(((1v{4 L00001v~a J......r    c                 h    |                      |          }|D ]}t          |t                    sJ d S r#   r   r$   r   )r^   rc   dfsr}   s       r   test_spam_no_matchzTestReadHtml.test_spam_no_match  sE    nnY'' 	- 	-Bb),,,,,,	- 	-r    c                 p    |                      |ddi          }|D ]}t          |t                    sJ d S )Nr   r   )rw   r   )r^   rf   r   r}   s       r   test_banklist_no_matchz#TestReadHtml.test_banklist_no_match  sM    nn]4/nBB 	- 	-Bb),,,,,,	- 	-r    c                 z    |                      |dd          d         }|j        d         dk    sJ |j        rJ d S )Nr   rn   r=   headerr   r   )r   r   r.   )r^   rc   r}   s      r   test_spam_headerzTestReadHtml.test_spam_header  sI    ^^I[^CCAFz!},,,,8r    c                     |                      |dd          }|                      |dd          }t          ||           d S Nr   r   r=   skiprowsr   r   r   s       r   test_skiprows_intzTestReadHtml.test_skiprows_int  F    nnYkAnFFnnYfqnAAsC(((((r    c                     |                      |dt          d                    }|                      |dt          d                    }t          ||           d S Nr   rn   r   r   )r   ranger7   r   s       r   test_skiprows_rangez TestReadHtml.test_skiprows_range$  R    nnYkE!HHnMMnnYfuQxxnHHsC(((((r    c                     |                      |dddg          }|                      |dddg          }t          ||           d S Nr   r   rn   r   r   r   r   s       r   test_skiprows_listzTestReadHtml.test_skiprows_list*  N    nnYkQFnKKnnYf1vnFFsC(((((r    c                     |                      |dddh          }|                      |dddh          }t          ||           d S r   r   r   s       r   test_skiprows_setzTestReadHtml.test_skiprows_set0  r   r    c                     |                      |dd          }|                      |dd          }t          ||           d S r   r   r   s       r   test_skiprows_slicez TestReadHtml.test_skiprows_slice6  r   r    c                     |                      |dt          d                    }|                      |dt          d                    }t          ||           d S r   r   slicer7   r   s       r   test_skiprows_slice_shortz&TestReadHtml.test_skiprows_slice_short<  r   r    c           	          |                      |dt          dd                    }|                      |dt          ddd                    }t          ||           d S )	Nr   rn      r   r   rk   r   r   r   s       r   test_skiprows_slice_longz%TestReadHtml.test_skiprows_slice_longB  sX    nnYkE!QKKnPPnnYfuQ2nOOsC(((((r    c                     |                      |dt          j        d                    }|                      |dt          j        d                    }t          ||           d S r   )r   ro   aranger7   r   s       r   test_skiprows_ndarrayz"TestReadHtml.test_skiprows_ndarrayH  sV    nnYkBIaLLnQQnnYfry||nLLsC(((((r    c                     t          j        t          d          5  |                     |dd           d d d            d S # 1 swxY w Y   d S )Nz%is not a valid type for skipping rowsr<   r   asdfr   )rB   rE   	TypeErrorr   )r^   rc   s     r   test_skiprows_invalidz"TestReadHtml.test_skiprows_invalidN  s    ]9-TVVV 	J 	JNN9K&NIII	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	J 	Js   AA	Ac                     |                      |dd          }|                      |dd          }t          ||           d S Nr   r   r=   rR   r   r   r   s       r   
test_indexzTestReadHtml.test_indexR  sF    nnYkQnGGnnYfnBBsC(((((r    c                     |                      |ddd          }|                      |ddd          }t          ||           d S Nr   r   r   )r=   r   rR   r   r   r   s       r   test_header_and_index_no_typesz+TestReadHtml.test_header_and_index_no_typesW  J    nnYk!qnQQnnYfQ!nLLsC(((((r    c                     |                      |ddd          }|                      |ddd          }t          ||           d S r   r   r   s       r    test_header_and_index_with_typesz-TestReadHtml.test_header_and_index_with_types\  r  r    c                     |                      |dd          }|                      |dd          }t          ||           d S r   r   r   s       r   test_infer_typeszTestReadHtml.test_infer_typesa  sF    nnYkQnGGnnYfnBBsC(((((r    c                    t          |d          5 }t          |                                          }d d d            n# 1 swxY w Y   t          |d          5 }t          |                                          }d d d            n# 1 swxY w Y   |                     |d          }|                     |d          }t	          ||           d S NzUTF-8r   r   r<   r   )r   r   r   r   r7   )r^   rc   r   data1data2r   r   s          r   test_string_iozTestReadHtml.test_string_iog  s:   )g... 	'!QVVXX&&E	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' )g... 	'!QVVXX&&E	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' 	' nnU+n66nnU&n11sC(((((s#   "A  AA"B

BBc                 0   t          |d          5 }|                                }d d d            n# 1 swxY w Y   |                     t          |          d          }|                     t          |          d          }t	          ||           d S r  )r   r   r   r   r7   )r^   rc   r   r   r   r   s         r   test_stringzTestReadHtml.test_stringr  s    )g... 	!6688D	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 nnXd^^;n??nnXd^^6n::sC(((((s   377c                 &   t          |d          5 }|                     |d          }d d d            n# 1 swxY w Y   t          |d          5 }|                     |d          }d d d            n# 1 swxY w Y   t          ||           d S r  )r   r   r7   )r^   rc   r   r   r   s        r   test_file_likezTestReadHtml.test_file_like{  s   )g... 	7!..+.66C	7 	7 	7 	7 	7 	7 	7 	7 	7 	7 	7 	7 	7 	7 	7 )g... 	2!..&.11C	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	sC(((((s   6::A66A:=A:c                     |                     dd           t          j        t          d          5  |                     dd           d d d            d S # 1 swxY w Y   d S )Nz#urlopen error unknown url type: git  coder<   zgit://github.comr   )r   rB   rE   r   r   r^   r   s     r   test_bad_url_protocolz"TestReadHtml.test_bad_url_protocol  s     	  !FS QQQ]8+PQQQ 	B 	BNN-[NAAA	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	Bs   AAAc                     |                     dd           t          j        t          t          fd          5  |                     |j        d           d d d            d S # 1 swxY w Y   d S )NzName or service not knownr  r  zHTTP Error 404: NOT FOUNDr<   r   )r   rB   rE   r   rL   r   rM   r  s     r   test_invalid_urlzTestReadHtml.test_invalid_url  s     	  !<3 GGG]Hj19TUUU 	> 	>NN:>N===	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	> 	>s   A$$A(+A(c                     |}|                      t          t          j                            |                    dddi          }t          |t                    sJ |D ]}t          |t                    sJ d S )NFirstr   r   r   )r   r   ospathabspathr$   listr   r^   rf   rM   r   r}   s        r   test_file_urlzTestReadHtml.test_file_url  s    nnRW__S1122'$PW  
 
 #t$$$$$ 	- 	-Bb),,,,,,	- 	-r    c                     |}t          j        t          d          5  |                     |dddi           d d d            d S # 1 swxY w Y   d S )NzNo tables foundr<   r   r   	tasdfabler   rB   rE   rL   r   )r^   rf   rM   s      r   test_invalid_table_attrsz%TestReadHtml.test_invalid_table_attrs  s    ]:->??? 	 	NN:4BU    	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	s   AA
A
c                 (     | j         |fdddid|S )NMetcalfr   r   r   r   )r^   r  r2   s      r   
_bank_datazTestReadHtml._bank_data  s'    t~dU)D'?UUfUUUr    c                 |    |                      |ddg          d         }t          |j        t                    sJ d S )Nr   r   r   r&  r$   r   r   r^   rf   r}   s      r   test_multiindex_headerz#TestReadHtml.test_multiindex_header  s>    __]Aq6_::1="*j1111111r    c                 |    |                      |ddg          d         }t          |j        t                    sJ d S )Nr   r   rR   )r&  r$   r   r   r*  s      r   test_multiindex_indexz"TestReadHtml.test_multiindex_index  s>    __]q!f_==a@"(J///////r    c                     |                      |ddgddg          d         }t          |j        t                    sJ t          |j        t                    sJ d S )Nr   r   )r   rR   )r&  r$   r   r   r   r*  s      r   test_multiindex_header_indexz)TestReadHtml.test_multiindex_header_index  s\    __]Aq6aV_LLQO"*j11111"(J///////r    c                 ~    |                      |ddgd          d         }t          |j        t                    sJ d S Nr   r   )r   r   r)  r*  s      r   &test_multiindex_header_skiprows_tuplesz3TestReadHtml.test_multiindex_header_skiprows_tuples  @    __]Aq6A_FFqI"*j1111111r    c                 ~    |                      |ddgd          d         }t          |j        t                    sJ d S r2  r)  r*  s      r   test_multiindex_header_skiprowsz,TestReadHtml.test_multiindex_header_skiprows  r4  r    c                     |                      |ddgddgd          d         }t          |j        t                    sJ t          |j        t                    sJ d S )Nr   r   )r   rR   r   )r&  r$   r   r   r   r*  s      r   %test_multiindex_header_index_skiprowsz2TestReadHtml.test_multiindex_header_index_skiprows  sk    __1a&QFQ  
 

 "(J/////"*j1111111r    c                 @   |}|                      t          t          j                            |                    t          j        t          j        d                    ddi          }t          |t                    sJ |D ]}t          |t                    sJ d S )NFloridar   r   r   )
r   r   r  r  r  recompiler$   r  r   r  s        r   test_regex_idempotencyz#TestReadHtml.test_regex_idempotency  s    nnRW__S1122*RZ	2233/  
 

 #t$$$$$ 	- 	-Bb),,,,,,	- 	-r    c                     d}t          j        t          |          5  |                     |dd           d d d            d S # 1 swxY w Y   d S )Nz\(you passed a negative value\)r<   Waterr   r   r!  r^   rc   r3   s      r   test_negative_skiprowsz#TestReadHtml.test_negative_skiprows  s    0]:S111 	B 	BNN9GbNAAA	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	B 	Bs   AAAc                     dS )Naf  
          <table class="contentstable" align="center"><tr>
            <td width="50%">
            <p class="biglink"><a class="biglink" href="whatsnew/2.7.html">What's new in Python 2.7?</a><br/>
                <span class="linkdescr">or <a href="whatsnew/index.html">all "What's new" documents</a> since 2.0</span></p>
            <p class="biglink"><a class="biglink" href="tutorial/index.html">Tutorial</a><br/>
                <span class="linkdescr">start here</span></p>
            <p class="biglink"><a class="biglink" href="library/index.html">Library Reference</a><br/>
                <span class="linkdescr">keep this under your pillow</span></p>
            <p class="biglink"><a class="biglink" href="reference/index.html">Language Reference</a><br/>
                <span class="linkdescr">describes syntax and language elements</span></p>
            <p class="biglink"><a class="biglink" href="using/index.html">Python Setup and Usage</a><br/>
                <span class="linkdescr">how to use Python on different platforms</span></p>
            <p class="biglink"><a class="biglink" href="howto/index.html">Python HOWTOs</a><br/>
                <span class="linkdescr">in-depth documents on specific topics</span></p>
            </td><td width="50%">
            <p class="biglink"><a class="biglink" href="installing/index.html">Installing Python Modules</a><br/>
                <span class="linkdescr">installing from the Python Package Index &amp; other sources</span></p>
            <p class="biglink"><a class="biglink" href="distributing/index.html">Distributing Python Modules</a><br/>
                <span class="linkdescr">publishing modules for installation by others</span></p>
            <p class="biglink"><a class="biglink" href="extending/index.html">Extending and Embedding</a><br/>
                <span class="linkdescr">tutorial for C/C++ programmers</span></p>
            <p class="biglink"><a class="biglink" href="c-api/index.html">Python/C API</a><br/>
                <span class="linkdescr">reference for C/C++ programmers</span></p>
            <p class="biglink"><a class="biglink" href="faq/index.html">FAQs</a><br/>
                <span class="linkdescr">frequently asked questions (with answers!)</span></p>
            </td></tr>
        </table>

        <p><strong>Indices and tables:</strong></p>
        <table class="contentstable" align="center"><tr>
            <td width="50%">
            <p class="biglink"><a class="biglink" href="py-modindex.html">Python Global Module Index</a><br/>
                <span class="linkdescr">quick access to all modules</span></p>
            <p class="biglink"><a class="biglink" href="genindex.html">General Index</a><br/>
                <span class="linkdescr">all functions, classes, terms</span></p>
            <p class="biglink"><a class="biglink" href="glossary.html">Glossary</a><br/>
                <span class="linkdescr">the most important terms explained</span></p>
            </td><td width="50%">
            <p class="biglink"><a class="biglink" href="search.html">Search page</a><br/>
                <span class="linkdescr">search this documentation</span></p>
            <p class="biglink"><a class="biglink" href="contents.html">Complete Table of Contents</a><br/>
                <span class="linkdescr">lists all sections and subsections</span></p>
            </td></tr>
        </table>
        ra   r^   s    r   python_docszTestReadHtml.python_docs  s    - -r    c                     |                     |           |                     |j        d          }t          |          dk    sJ d S )Nr   Pythonr<   r   )r   r   rM   r(   )r^   rD  r   r   s       r   test_multiple_matchesz"TestReadHtml.test_multiple_matches  sI     	   555nnZ^8n<<3xx!||||||r    c                     |                     |           |                     |j        d          }d |D             }t          |          ddgk    sJ d S )Nr   rF  r<   c                 8    g | ]}|j         d          dd         S )r   r   rk   )r   )r   r}   s     r   
<listcomp>z7TestReadHtml.test_python_docs_table.<locals>.<listcomp>  s'    ///RbgdmAaC ///r    PythWhat)r   r   rM   sorted)r^   rD  r   r   zzs        r   test_python_docs_tablez#TestReadHtml.test_python_docs_table  sg     	   555nnZ^8n<<//3///bzzff-------r    c                 x    d}|                      t          |                    }t          |          dk    sJ dS )z@
        Make sure that read_html ignores empty tables.
        a  
            <table>
                <thead>
                    <tr>
                        <th>A</th>
                        <th>B</th>
                    </tr>
                </thead>
                <tbody>
                    <tr>
                        <td>1</td>
                        <td>2</td>
                    </tr>
                </tbody>
            </table>
            <table>
                <tbody>
                </tbody>
            </table>
        r   N)r   r   r(   )r^   r>   r   s      r   test_empty_tableszTestReadHtml.test_empty_tables  s?    ( //6{{ar    c                     |                      t          d                    d         }t          ddgddggddg	          }t          j        ||           d S )
Na  <table>
            <thead>
                <tr>
                    <th>A</th>
                    <th>B</th>
                </tr>
            </thead>
            <tbody>
                <tr>
                    <td>1</td>
                    <td>2</td>
                </tr>
            </tbody>
            <tbody>
                <tr>
                    <td>3</td>
                    <td>4</td>
                </tr>
            </tbody>
        </table>r   r   rn   rl   rk   ABr   r   r   r   r   r,   r-   r^   r   r   s      r   test_multiple_tbodyz TestReadHtml.test_multiple_tbody6  sq      
 
. /2 Aq6Aq6"2S#JGGG
fh/////r    c                     |                      t          d                    d         }t          ddidg          }t          j        ||           dS )zt
        Don't fail with bs4 when there is a header and only one column
        as described in issue #9178
        a3  <table>
                <thead>
                    <tr>
                        <th>Header</th>
                    </tr>
                </thead>
                <tbody>
                    <tr>
                        <td>first</td>
                    </tr>
                </tbody>
            </table>r   Headerfirstr   r   NrV  rW  s      r   test_header_and_one_columnz'TestReadHtml.test_header_and_one_columnV  sg    
  
 
 " 8W"5aSAAA
fh/////r    c                     |                      t          d                    d         }t          g dgg d          }t          j        ||           dS )zK
        Ensure parser adds <tr> within <thead> on malformed HTML.
        a  <table>
            <thead>
                <tr>
                    <th>Country</th>
                    <th>Municipality</th>
                    <th>Year</th>
                </tr>
            </thead>
            <tbody>
                <tr>
                    <td>Ukraine</td>
                    <th>Odessa</th>
                    <td>1944</td>
                </tr>
            </tbody>
        </table>r   )UkraineOdessa  )CountryMunicipalityYearrU  NrV  rW  s      r   test_thead_without_trz"TestReadHtml.test_thead_without_trp  sy      
 
& '* ---.777
 
 

 	fh/////r    c                    d}t          ddggddg          }t          ddgddggddg          }|                    d	
          }|                    d
          }|                     t          |                    d         }|                     t          |                    d         }t	          j        ||           t	          j        ||           dS )zh
        Make sure that read_html reads tfoot, containing td or th.
        Ignores empty tfoot
        a  <table>
            <thead>
                <tr>
                    <th>A</th>
                    <th>B</th>
                </tr>
            </thead>
            <tbody>
                <tr>
                    <td>bodyA</td>
                    <td>bodyB</td>
                </tr>
            </tbody>
            <tfoot>
                {footer}
            </tfoot>
        </table>bodyAbodyBrS  rT  rU  footAfootB )footerz%<tr><td>footA</td><th>footB</th></tr>r   N)r   ry   r   r   r,   r-   )r^   data_template	expected1	expected2r  r	  result1result2s           r   test_tfoot_readzTestReadHtml.test_tfoot_read  s    
$ Wg$6#7#sLLL	G$w&89C:
 
 
	 $$B$//$$,S$TT..%11!4..%11!4
gy111
gy11111r    c                     |                      t          d          d          d         }t          ddggd          }t          j        ||           d S )Na
  
            <table>
                <tr>
                    <td>S</td>
                    <td>I</td>
                </tr>
                <tr>
                    <td>text</td>
                    <td>1944</td>
                </tr>
            </table>
        r   r(  textra  )SIr   rV  rW  s      r   &test_parse_header_of_non_string_columnz3TestReadHtml.test_parse_header_of_non_string_column  sq          
 
  !$ vtn-zBBB
fh/////r    c                    ddl m fd}|                     |dddi          d         }t           |dd	d
d          t          t          d          }|j        |j        k    sJ g d}g d}|                    |                              ||          }|                    |          }	|}
ddg}|
|                             t                    |
|<   t          j        |
|	           d S )Nr   )_remove_whitespacec                 @    	  |           S # t           $ r | cY S w xY wr#   )AttributeError)r%   rz  s    r   try_remove_wsz8TestReadHtml.test_banklist_header.<locals>.try_remove_ws  s=    ))!,,,!   s   
 r$  r   r   r   r   r   csvzbanklist.csv)Updated DateClosing Date
converters)
z,First Vietnamese American Bank In Vietnamesez"Westernbank Puerto Rico En Espanolz*R-G Premier Bank of Puerto Rico En EspanolzEurobank En EspanolzSanderson State Bank En EspanolzLWashington Mutual Bank (Including its subsidiary Washington Mutual Bank FSB)zSilver State Bank En Espanolz%AmTrade International Bank En EspanolzHamilton Bank, NA En Espanolz6The Citizens Savings Bank Pioneer Community Bank, Inc.)
zFirst Vietnamese American BankzWesternbank Puerto RicozR-G Premier Bank of Puerto RicoEurobankzSanderson State BankzWashington Mutual BankzSilver State BankzAmTrade International BankzHamilton Bank, NAzThe Citizens Savings Bankr  r  )pandas.io.htmlrz  r   r   r   shaper*   replaceapplyr   r,   r-   )r^   rf   r   r}  r}   ground_trutholdnewdfnewgtnew	converted	date_colsrz  s               @r   test_banklist_headerz!TestReadHtml.test_banklist_header  s8   555555	 	 	 	 	 ^^M4/^RRSTUHT65.99(19MM
 
 
 x<-----
 
 

 
 
 }%%--c377  //	#^4	(399+FF	)
i/////r    c                     d}t          |d          5 }|                                }d d d            n# 1 swxY w Y   ||v sJ |                     |dddi          d         }||                                v sJ d S )NzGold Canyonr   r   r   r   r   r   )r   r   r   	to_string)r^   rf   gcr   raw_textr}   s         r   test_gold_canyonzTestReadHtml.test_gold_canyon  s    -'222 	 avvxxH	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  X~~~~^^MtWo^VV
 R\\^^######s   599c                     |                      t          d          d          d         }|                      t          d          d          d         }t          j        ||           d S )Na  <table>
                        <thead>
                            <tr style="text-align: right;">
                            <th></th>
                            <th>C_l0_g0</th>
                            <th>C_l0_g1</th>
                            <th>C_l0_g2</th>
                            <th>C_l0_g3</th>
                            <th>C_l0_g4</th>
                            </tr>
                        </thead>
                        <tbody>
                            <tr>
                            <th>R_l0_g0</th>
                            <td> 0.763</td>
                            <td> 0.233</td>
                            <td> nan</td>
                            <td> nan</td>
                            <td> nan</td>
                            </tr>
                            <tr>
                            <th>R_l0_g1</th>
                            <td> 0.244</td>
                            <td> 0.285</td>
                            <td> 0.392</td>
                            <td> 0.137</td>
                            <td> 0.222</td>
                            </tr>
                        </tbody>
                    </table>r   r-  a  <table>
                    <thead>
                        <tr style="text-align: right;">
                        <th></th>
                        <th>C_l0_g0</th>
                        <th>C_l0_g1</th>
                        <th>C_l0_g2</th>
                        <th>C_l0_g3</th>
                        <th>C_l0_g4</th>
                        </tr>
                    </thead>
                    <tbody>
                        <tr>
                        <th>R_l0_g0</th>
                        <td> 0.763</td>
                        <td> 0.233</td>
                        </tr>
                        <tr>
                        <th>R_l0_g1</th>
                        <td> 0.244</td>
                        <td> 0.285</td>
                        <td> 0.392</td>
                        <td> 0.137</td>
                        <td> 0.222</td>
                        </tr>
                    </tbody>
                 </table>)r   r   r,   r-   )r^   r   r   s      r   test_different_number_of_colsz*TestReadHtml.test_different_number_of_cols  s    >>  @ C " "
 "
D E"H  : =   
 
> ?B 	fh/////r    c                     |                      t          d                    d         }t          g dgg d          }t          j        ||           d S )NaZ  
            <table>
                <tr>
                    <th>A</th>
                    <th colspan="1">B</th>
                    <th rowspan="1">C</th>
                </tr>
                <tr>
                    <td>a</td>
                    <td>b</td>
                    <td>c</td>
                </tr>
            </table>
        r   r   )rS  rT  Crw  rV  rW  s      r   test_colspan_rowspan_1z#TestReadHtml.test_colspan_rowspan_1S  si     
 
" #& ooo.HHH
fh/////r    c                     |                      t          d          d          d         }t          g dgg d          }t          j        ||           d S )Na  
            <table>
                <tr>
                    <td colspan="2">X</td>
                    <td>Y</td>
                    <td rowspan="2">Z</td>
                    <td>W</td>
                </tr>
                <tr>
                    <td>A</td>
                    <td colspan="2">B</td>
                    <td>C</td>
                </tr>
            </table>
        r   r(  )rS  rT  rT  Zr  )XzX.1Yr  WrU  rV  rW  s      r    test_colspan_rowspan_copy_valuesz-TestReadHtml.test_colspan_rowspan_copy_valuesl  s      " %   
 
& '* +++,6Q6Q6Q
 
 
 	fh/////r    c                     |                      t          d          d          d         }t          g dgg d          }t          j        ||           d S )Na(  
            <table>
                <tr>
                    <td rowspan="2">A</td>
                    <td rowspan="2" colspan="3">B</td>
                    <td>C</td>
                </tr>
                <tr>
                    <td>D</td>
                </tr>
            </table>
        r   r(  )rS  rT  rT  rT  D)rS  rT  zB.1zB.2r  rU  rV  rW  s      r   test_colspan_rowspan_both_not_1z,TestReadHtml.test_colspan_rowspan_both_not_1  s          
 
  !$ +++,6S6S6S
 
 
 	fh/////r    c                     |                      t          d          d          d         }t          ddggddg          }t          j        ||           d S )Nz
            <table>
                <tr>
                    <td>A</td>
                    <td rowspan="2">B</td>
                </tr>
                <tr>
                    <td>C</td>
                </tr>
            </table>
        r   r(  r  rT  rS  rU  rV  rW  s      r   test_rowspan_at_end_of_rowz'TestReadHtml.test_rowspan_at_end_of_row  st     
     
 
 " C:,c
CCC
fh/////r    c                     |                      t          d          d          d         }t          ddgddggddg          }t          j        ||           d S )Nz
            <table>
                <tr>
                    <td rowspan="3">A</td>
                    <td rowspan="3">B</td>
                </tr>
            </table>
        r   r(  rS  rT  rU  rV  rW  s      r   test_rowspan_only_rowsz#TestReadHtml.test_rowspan_only_rows  s{     	 	    
 
  C:Sz":S#JOOO
fh/////r    c                     |                      t          d                    d         }t          ddgddggddgddgg          }t          dd	gg|
          }t	          j        ||           d S )Nam  
            <table>
                <tr>
                    <th>A</th>
                    <th>B</th>
                </tr>
                <tr>
                    <th>a</th>
                    <th>b</th>
                </tr>
                <tr>
                    <td>1</td>
                    <td>2</td>
                </tr>
            </table>
        r   rS  rT  r   r   r   levelscodesrn   rU  r   r   r   r   r,   r-   r^   r   r   r   s       r   +test_header_inferred_from_rows_with_only_thz8TestReadHtml.test_header_inferred_from_rows_with_only_th  s     
 
& '* c3Z#s$<aVaQRVDTUUUAq6(G<<<
fh/////r    c                 r   t          dt          dd          i          }|                                }|                     t	          |          dgd          }t          j        ||d                    |                     t	          |          dgd          }t          j        ||d                    d S )Ndate1/1/2001
   periodsr   r   parse_datesrR   )r   r   r|   r   r   r,   r-   )r^   r}   r   r   s       r   test_parse_dates_listz"TestReadHtml.test_parse_dates_list  s    
:r B B BCDD::<<nnXh//aSAnNN
b#a&)))nnXh//fXQRnSS
b#a&)))))r    c                    t          t          dd                    }t          |                    d           |                    d           d          }|                     t          |                                          ddd	gid
          }t          d|i          }t          j        ||d                    d S )Nr  r  r  c                 D    t          |                                           S r#   )strr  r%   s    r   r'   z7TestReadHtml.test_parse_dates_combine.<locals>.<lambda>	      AFFHH r    c                 D    t          |                                           S r#   )r  timer  s    r   r'   z7TestReadHtml.test_parse_dates_combine.<locals>.<lambda>
  r  r    )r  r  datetimer   rn   r  r   )	r   r   r   r*   r   r   r|   r,   r-   )r^   	raw_datesr}   r   newdfs        r   test_parse_dates_combinez%TestReadHtml.test_parse_dates_combine  s    :j"===>>	!&=&=>>!&=&=>> 
 
 nnRZZ\\""aV0DPQ  
 
 :y122
eSV,,,,,r    c                     |dddd          }t           j                            |          sJ t          |           d            t           j                            |          sJ t          |           d            |                     |dd	          d
         }|j        dk    sJ d|j        d         v sJ |d         j        t          j        d          k    sJ t          j
        |j        d         d          sJ d S )Nr   r   r>   wikipedia_states.htmlz is not a filez is an empty fileArizonar   r   r   )<      Unnamedr   sq mifloat64)r   r  HzPN$A)r  r  isfilereprgetsizer   r  r   r   ro   allcloselocr^   r   r   r   s       r   test_wikipedia_states_tablez(TestReadHtml.test_wikipedia_states_table  s   xff.EFFw~~d##BBT

%B%B%BBBBwt$$FFd&F&F&FFFFIa@@C|x''''FN2.....g$(;(;;;;;{6:j19=======r    c                     |dddd          }|                      |dd          d         }|j        dk    sJ d	|j        d
         d         v sJ |j        j        dk    sJ t	          j        |j        d         d          sJ d S )Nr   r   r>   r  r  r   r   )r     r  r   r   rn   )Alaska)zTotal area[2]r  r  )r   r  r   nlevelsro   r  r  r  s       r    test_wikipedia_states_multiindexz-TestReadHtml.test_wikipedia_states_multiindex  s    xff.EFFICCAF|x''''FN2.q11111~%****{6:&JKYWWWWWWWr    c                     |                      t          d          ddg          }t          ddggt          j        ddg          	          }t          j        |d         |           d S )
NaK  
                <table>
                    <thead>
                        <tr><th></th><th></tr>
                        <tr><th>A</th><th>B</th></tr>
                    </thead>
                    <tbody>
                        <tr><td>a</td><td>b</td></tr>
                    </tbody>
                </table>
            r   r   r(  r   r   )Unnamed: 0_level_0rS  )zUnnamed: 1_level_0rT  rw  )r   r   r   r   from_tuplesr,   r-   rW  s      r   %test_parser_error_on_empty_header_rowz2TestReadHtml.test_parser_error_on_empty_header_row%  s    
  q6   
 
  3ZL*,.IJ 
 
 
 	fQi22222r    c                     |                      t          d          d          d         }t          ddidg          }|d         j        t	          j        d          k    sJ t          j        ||           d S )	Na  <html>
            <body>
             <table>
                <thead>
                    <tr>
                        <th>Header</th>
                    </tr>
                </thead>
                <tbody>
                    <tr>
                        <td>1100#101</td>
                    </tr>
                </tbody>
            </table>
            </body>
        </html>#)decimalr   rZ  gClg0@r\  r  )r   r   r   r   ro   r,   r-   rW  s      r   test_decimal_rowszTestReadHtml.test_decimal_rows>  s     $ '   
 
( ), 8X"6qcBBBh%))<)<<<<<
fh/////r    argFc                     t          j        d          }t          j        t          |          5  |                     ||           d d d            d S # 1 swxY w Y   d S )NzPassing a bool to header is invalid. Use header=None for no header or header=int or list-like of ints to specify the row(s) making up the column namesr<   r(  )r;  escaperB   rE   r   r   )r^   rc   r  r3   s       r   test_bool_header_argz!TestReadHtml.test_bool_header_arg[  s     i
 

 ]9C000 	2 	2NN9SN111	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2 	2s   AAAc                     |                      t          d          dt          i          d         }t          dddgi          }t	          j        ||           d S )Na  <table>
                 <thead>
                   <tr>
                     <th>a</th>
                    </tr>
                 </thead>
                 <tbody>
                   <tr>
                     <td> 0.763</td>
                   </tr>
                   <tr>
                     <td> 0.244</td>
                   </tr>
                 </tbody>
               </table>r   r  r   z0.763z0.244)r   r   r  r   r,   r-   rW  s      r   test_converterszTestReadHtml.test_convertersf  sq     " Sz%   
 
& '* cGW#5677
fh/////r    c                     |                      t          d          dg          d         }t          ddt          j        gi          }t          j        ||           d S )Na  <table>
                 <thead>
                   <tr>
                     <th>a</th>
                   </tr>
                 </thead>
                 <tbody>
                   <tr>
                     <td> 0.763</td>
                   </tr>
                   <tr>
                     <td> 0.244</td>
                   </tr>
                 </tbody>
               </table>gZd;?)	na_valuesr   r   g"~j?r   r   r   ro   r   r,   r-   rW  s      r   test_na_valueszTestReadHtml.test_na_values  sp     " g%   
 
& '* cE26?344
fh/////r    c                 z   d}t          dddgi          }|                     t          |          d          d         }t          j        ||           t          dt
          j        t
          j        gi          }|                     t          |          d          d         }t          j        ||           d S )	Na  <table>
                        <thead>
                            <tr>
                            <th>a</th>
                            </tr>
                        </thead>
                        <tbody>
                            <tr>
                            <td> N/A</td>
                            </tr>
                            <tr>
                            <td> NA</td>
                            </tr>
                        </tbody>
                    </table>r   zN/Ar
   F)keep_default_nar   T)r   r   r   r,   r-   ro   r   )r^   	html_dataexpected_dfhtml_dfs       r   test_keep_default_naz!TestReadHtml.test_keep_default_na  s     	   udm 455..)!4!4e.LLQO
k7333rvrv&6 788..)!4!4d.KKAN
k733333r    c                     |                      t          d                    d         }t          ddgt          j        t          j        ggddg          }t          j        ||           d S )Nak  
            <table>
                <tr>
                    <th>A</th>
                    <th>B</th>
                </tr>
                <tr>
                    <td>a</td>
                    <td>b</td>
                </tr>
                <tr>
                    <td></td>
                    <td></td>
                </tr>
            </table>
        r   r   r   rS  rT  rU  r  rW  s      r   test_preserve_empty_rowsz%TestReadHtml.test_preserve_empty_rows  su     
 
& '* C:/?"@3PS*UUU
fh/////r    c                     |                      t          d                    d         }t          ddgddggddgddgg          }t          dd	gg|
          }t	          j        ||           d S )NaU  
            <table>
                <thead>
                    <tr><th></th><th></tr>
                    <tr><th>A</th><th>B</th></tr>
                    <tr><th>a</th><th>b</th></tr>
                </thead>
                <tbody>
                    <tr><td>1</td><td>2</td></tr>
                </tbody>
            </table>
        r   rS  rT  r   r   r   r  rn   rU  r  r  s       r   ,test_ignore_empty_rows_when_inferring_headerz9TestReadHtml.test_ignore_empty_rows_when_inferring_header  s     
 
 " c3Z#s$<aVaQRVDTUUUAq6(G<<<
fh/////r    c                     t          g d          }g dg dg|_        |                    d          }|                     t	          |                    d         }t          j        ||           d S )N))HillaryD   r  )BernieJ   r  )DonaldE   R)r   )r  AgeParty)NamezUnnamed: 1_level_1zUnnamed: 2_level_1Fr   r   )r   r   r|   r   r   r,   r-   )r^   r  r>   r  s       r   test_multiple_header_rowsz&TestReadHtml.test_multiple_header_rows  s    QQQ
 
 
 322@@@
 """//..$003
k733333r    c                      |dddd          }|                      |d          }t          |t                    sJ t          |d         t                    sJ d S )Nr   r   r>   rQ   r   r-  )r   r$   r  r   )r^   r   rS   r   s       r   test_works_on_valid_markupz'TestReadHtml.test_works_on_valid_markup  sc    8D&&2EFFnnXn33#t$$$$$#a&),,,,,,,r    c                 V     |dddd          }|                      |dddg           d S )	Nr   r   r>   re   r   rP   r:   rK   r%  )r^   r   rf   s      r   test_fallback_successz"TestReadHtml.test_fallback_success  s:     vvGG}K@TUUUUUr    c                     t          dd          }t          t          j                            d                              d          |          }|                                }d|v sJ d S )Nz
2000-01-01r  r  rn   )r  rk   r   )r   r   ro   rp   rq   standard_normalr|   )r^   rngr}   r   s       r   test_to_html_timestampz#TestReadHtml.test_to_html_timestamp  si    r222ry,,Q//??HHPSTTTv%%%%%%r    c                    t          dddg          }|                                }|                    d          }|                    d          }|                    d          }|                    d          }|                    d          }d|v sJ ||k    sJ ||k    sJ ||k    sJ d	|v sJ d
|vsJ d|vsJ ||k    sJ d S )Nr   rn   rS  rT  T)borderr   Fz border="1"z border="2"z border="0"z border)r   r|   )r^   r}   out_border_defaultout_border_trueout_border_explicit_defaultout_border_nondefaultout_border_zeroout_border_falses           r   test_to_html_borderlessz$TestReadHtml.test_to_html_borderless  s   aa(()**ZZ\\**D*11&(jjj&:&:# "

!
 4 4**A*..::U:33 22222"44444!%@@@@@!%::::: 55555O3333 00000"2222222r    zdisplayed_only,exp0,exp1fooNzfoo  bar  baz  quxc                     d}|                      t          |          |          }t          j        |d         |           |t          j        |d         |           d S t	          |          dk    sJ d S )Na  <html>
          <body>
            <table>
              <tr>
                <td>
                  foo
                  <span style="display:none;text-align:center">bar</span>
                  <span style="display:none">baz</span>
                  <span style="display: none">qux</span>
                </td>
              </tr>
            </table>
            <table style="display: none">
              <tr>
                <td>foo</td>
              </tr>
            </table>
          </body>
        </html>displayed_onlyr   r   )r   r   r,   r-   r(   )r^   r  exp0exp1r   r   s         r   test_displayed_onlyz TestReadHtml.test_displayed_only  sw    ( nnXd^^NnKK
c!fd+++!#a&$/////s88q======r    r  c                     d}t          t          |          |          d         }t          ddgddgd          }t          j        ||           d S )	NaW  
        <table>
            <tr>
                <th>A</th>
                <th>B</th>
            </tr>
            <tr>
                <td>1</td>
                <td>2</td>
            </tr>
            <tr>
                <td><span style="display:none"></span>4</td>
                <td>5</td>
            </tr>
        </table>
        r  r   r   rk   rn   r   r  rV  )r^   r  
html_tabler   r   s        r   &test_displayed_only_with_many_elementsz3TestReadHtml.test_displayed_only_with_many_elements>  sa    
  8J//OOOPQRAq6A7788
fh/////r    z\ignore:You provided Unicode markup but also provided a value for from_encoding.*:UserWarningc                    t           j                            |          }t           j                            |          d         }|                    d          \  }}	 t          |d          5 }|                     |                                |d                                          }d d d            n# 1 swxY w Y   t          |d          5 }|                     t          |                                          |d                                          }d d d            n# 1 swxY w Y   |                     ||d                                          }	t          j        ||           t          j        ||	           d S # t          $ r+ t                      rd|v sd|v rt          j                      w xY w)Nr   _rb)r   rR   1632)r  r  basenamesplitextsplitr   r   r   popr   r,   r-   	Exceptionr	   rB   skip)
r^   r   	base_pathrootr  r   fobjfrom_stringfrom_file_likefrom_filenames
             r   test_encodezTestReadHtml.test_encodeT  s)   
 G$$%788	w	**1-jjoo8	($// 4"nnIIKK(a -  #%%               
 ($// 4!%DIIKK((8q "0 " "#%%               
 !NN"X +  cee  !+~>>>!+}===== 	 	 	"$$ "8##tx'7'7KMMM	sV   F	 .=B7+F	 7B;;F	 >B;?F	 A
D(F	 (D,,F	 /D,0AF	 	5F>c                 p   | j         j                            d          dk    rt          j        d            G d dt
                    } |d          }|                      |          sJ t          j        t          d          5  |                      |           d d d            d S # 1 swxY w Y   d S )	NrA   rP   zNot applicable for lxmlc                       e Zd Zd ZdS )FTestReadHtml.test_parse_failure_unseekable.<locals>.UnseekableStringIOc                     dS NFra   rC  s    r   seekablezOTestReadHtml.test_parse_failure_unseekable.<locals>.UnseekableStringIO.seekable{  s    ur    N)__name__
__module____qualname__r2  ra   r    r   UnseekableStringIOr/  z  s#            r    r6  z?
            <table><tr><td>spam<foobr />eggs</td></tr></table>z#passed a non-rewindable file objectr<   )r   keywordsgetrB   r%  r   rE   rL   )r^   r6  bads      r   test_parse_failure_unseekablez*TestReadHtml.test_parse_failure_unseekablet  s    >"&&x00F::K1222	 	 	 	 	 	 	 	 ! B
 

 ~~c"""""]:-RSSS 	  	 NN3	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	 s   B++B/2B/c                      G d d          } |d          } |d          }|                      |          sJ |                      |          sJ d S )Nc                   <    e Zd Zd	dZd
dZd Zd Zd ZdefdZ	dS )9TestReadHtml.test_parse_failure_rewinds.<locals>.MockFilereturnNc                 "    || _         d| _        d S r1  )r   at_end)r^   r   s     r   __init__zBTestReadHtml.test_parse_failure_rewinds.<locals>.MockFile.__init__  s     	#r    c                 4    | j         rdn| j        }d| _         |S )Nrk  T)r@  r   )r^   sizer   s      r   r   z>TestReadHtml.test_parse_failure_rewinds.<locals>.MockFile.read  s     ![7rrdi"r    c                     d| _         d S r1  )r@  )r^   offsets     r   seekz>TestReadHtml.test_parse_failure_rewinds.<locals>.MockFile.seek  s    #r    c                     dS )NTra   rC  s    r   r2  zBTestReadHtml.test_parse_failure_rewinds.<locals>.MockFile.seekable  s    tr    c                     d S r#   ra   rC  s    r   __next__zBTestReadHtml.test_parse_failure_rewinds.<locals>.MockFile.__next__  s    r    c                     | S r#   ra   rC  s    r   __iter__zBTestReadHtml.test_parse_failure_rewinds.<locals>.MockFile.__iter__  s	     r    )r>  Nr#   )
r3  r4  r5  rA  r   rF  r2  rI  r   rK  ra   r    r   MockFiler=    s        $ $ $ $   
$ $ $    (      r    rL  z/<table><tr><td>spam<br />eggs</td></tr></table>z2<table><tr><td>spam<foobr />eggs</td></tr></table>r%  )r^   rL  goodr9  s       r   test_parse_failure_rewindsz'TestReadHtml.test_parse_failure_rewinds  s    	 	 	 	 	 	 	 	2 xIJJhKLL~~d#####~~c"""""""r    c                     G d dt           j                  } |dddd          } || j        |f          } || j        |f          }|                                 |                                 |                                s|                                r)	 |                                |                                )d |j        cxu r	|j        u sn J d S )Nc                        e Zd Z fdZ xZS )@TestReadHtml.test_importcheck_thread_safety.<locals>.ErrorThreadc                     	 t                                                       d | _        d S # t          $ r}|| _        Y d }~d S d }~ww xY wr#   )superrunerrr$  )r^   rU  	__class__s     r   rT  zDTestReadHtml.test_importcheck_thread_safety.<locals>.ErrorThread.run  sZ    $GGKKMMM  $DHHH ! # # #"DHHHHHHH#s    , 
AAA)r3  r4  r5  rT  __classcell__)rV  s   @r   ErrorThreadrQ    s8        $ $ $ $ $ $ $ $ $r    rX  r   r   r>   rQ   )targetr1   )	threadingThreadr   startis_aliverU  )r^   r   rX  rS   helper_thread1helper_thread2s         r   test_importcheck_thread_safetyz+TestReadHtml.test_importcheck_thread_safety  s   
	$ 	$ 	$ 	$ 	$)* 	$ 	$ 	$ 8D&&2EFF$DN(MMM$DN(MMM%%'' 	>+B+B+D+D 	 %%'' 	>+B+B+D+D 	~)????^-?????????r    c                      |dddd          }t          |          }|                     |          d         }|                     |          d         }t          j        ||           d S )Nr   r   r>   r?   r   )r   r   r,   r-   )r^   r   file_path_string	file_pathr   r   s         r   test_parse_path_objectz#TestReadHtml.test_parse_path_object  sl    #8D&&+FF)**	nn-..q1nnY''*
c3'''''r    c                     |                      t          d                    d         }t          dggdg          }t          j        ||           d S )Nz
            <table>
                <tr>
                    <th>A</th>
                </tr>
                <tr>
                    <td>word1<br>word2</td>
                </tr>
            </table>
        r   zword1 word2rS  rU  rV  rW  s      r   test_parse_br_as_spacez#TestReadHtml.test_parse_br_as_space  se    	 
 
  M?"3cUCCC
fh/////r    )r)   bodyr   rl  c                    d}g dg dg dg dg dg dd}|d	         }|d
         }|d         }|dk    r|d         }|d         }|d         }n,|dk    r	|d         }n|dk    r	|d         }n|dk    r|d         }|                      t          |          |          d         }t          ||g|          }|                    t          j                  }t          j        ||           d S )Na  
          <table>
            <tr>
              <th>HTTP</th>
              <th>FTP</th>
              <th><a href="https://en.wiktionary.org/wiki/linkless">Linkless</a></th>
            </tr>
            <tr>
              <td><a href="https://en.wikipedia.org/">Wikipedia</a></td>
              <td>SURROUNDING <a href="ftp://ftp.us.debian.org/">Debian</a> TEXT</td>
              <td>Linkless</td>
            </tr>
            <tfoot>
              <tr>
                <td><a href="https://en.wikipedia.org/wiki/Page_footer">Footer</a></td>
                <td>
                  Multiple <a href="1">links:</a> <a href="2">Only first captured.</a>
                </td>
              </tr>
            </tfoot>
          </table>
          )HTTPFTPLinkless))ri  N)rj  N)rk  z'https://en.wiktionary.org/wiki/linkless)	WikipediaSURROUNDING Debian TEXTrk  ))rl  zhttps://en.wikipedia.org/)rm  zftp://ftp.us.debian.org/)rk  N)Footer$Multiple links: Only first captured.N))rn  z)https://en.wikipedia.org/wiki/Page_footer)ro  1N)head_ignorehead_extractbody_ignorebody_extractfooter_ignorefooter_extractrs  ru  rq  r)   rt  rv  rr  rg  rl  r   extract_linksr   rw  )r   r   r   fillnaro   r   r,   r-   )	r^   r  gh_13141_datagh_13141_expecteddata_expfoot_exphead_expr   r   s	            r   test_extract_linkszTestReadHtml.test_extract_links  sV   0 766  
 POO  
  
  %
 
2 %]3$_5$]3%<<(8H()9:H(8HHF]](8HHH__()9:HHH__(8H 7 7sKKANh18DDD??26**
fh/////r    c                     d}t          j        t          |          5  t          |d           d d d            d S # 1 swxY w Y   d S )NzY`extract_links` must be one of {None, "header", "footer", "body", "all"}, got "incorrect"r<   	incorrectrw  r!  r@  s      r   test_extract_links_badz#TestReadHtml.test_extract_links_bad%  s    I 	 ]:S111 	< 	<i{;;;;	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	< 	<   =AAc                     d}|                      t          |          d          d         }t          dgg          }t          j        ||           d S )Nz
        <table>
          <tr>
            <td>
              <a href='https://google.com'>Google.com</a>
            </td>
          </tr>
        </table>
        r)   rw  r   )z
Google.comzhttps://google.comrV  r^   r   r   r   s       r    test_extract_links_all_no_headerz-TestReadHtml.test_extract_links_all_no_header-  sX     eDDQGCDEFF
fh/////r    c                     d}t          j        t          |          5  t          dd           d d d            d S # 1 swxY w Y   d S )NzPdtype_backend numpy is invalid, only 'numpy_nullable' and 'pyarrow' are allowed.r<   testnumpyr   r!  r]   s     r   test_invalid_dtype_backendz'TestReadHtml.test_invalid_dtype_backend<  s    % 	 ]:S111 	5 	5fG4444	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5 	5r  c                     d}|                      t          |                    d         }t          ddgddggddg	          }t          j        ||           d S )
Na  
        <table>
            <tr>
                <th>
                    <style>.style</style>
                    A
                    </th>
                <th>B</th>
            </tr>
            <tr>
                <td>A1</td>
                <td>B1</td>
            </tr>
            <tr>
                <td>A2</td>
                <td>B2</td>
            </tr>
        </table>
        r   A1B1A2B2rS  rT  rU  rV  r  s       r   test_style_tagzTestReadHtml.test_style_tagD  sc    & //2D$<$">c
SSS
fh/////r    )fr3  r4  r5  r_   rB   fixturerc   rf   ri   r   r   marknetwork
single_cpur   r   slowr   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.  r0  r3  r6  r8  r=  rA  rD  rG  rO  rQ  rX  r]  re  rr  rx  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  parametrizer  r  r  r  r  r  r  r   r  r  r  r   r  r  filterwarningsr,  r:  rN  r`  rd  rf  r  r  r  r  r  ra   r    r   rY   rY   f   s         @ ^; ; ^; ^? ? ^? V^D!!!  "!' ' '$50 50 50n [[) )  )  [[) )  ) [) ) )/ / /- - -
- - -
  
) ) )) ) )) ) )) ) )) ) )) ) )) ) )) ) )J J J) ) )
) ) )
) ) )
) ) )	) 	) 	)) ) )) ) ) [[B B  B
 [[[> >   >
 [- - - [  V V V [2 2 2 [0 0 0 [0 0 0
 [2 2 2 [2 2 2 [2 2 2 [	- 	- 	-B B B
 ^. . ^.` [[   
 [[. .  .     60 0 0@0 0 040 0 0@$2 $2 $2L0 0 02 [-0 -0 -0^ [	$ 	$ 	$F0 F0 F0P0 0 02!0 !0 !0F0 0 0@0 0 0:0 0 0*0 0 08* * *- - -> > >X X X3 3 320 0 0: [UT5M222 2 3220 0 060 0 064 4 420 0 040 0 0.4 4 4- - - [V V V
& & &3 3 3& ["99eW%%t,II3455yy%7I7IJ	
 ! ! !< [-e}==0 0 >=0* [	&  	 8     ( #  #  #D [[@ @  @.( ( (0 0 0* [U$G$G$GHHB0 B0 IHB0H< < <0 0 05 5 50 0 0 0 0r    rY   )5collections.abcr   	functoolsr   r   r   r   r  pathlibr   r;  rZ  urllib.errorr   r  ro   rB   pandas.compatr	   pandas.util._test_decoratorsutil_test_decoratorstdpandasr   r
   r   r   r   r   r   r   r   r   pandas._testing_testingr,   pandas.core.arraysr   r   pandas.io.commonr   r  r   r7   rH   rN   rV   r  r  r   
skip_if_norY   ra   r    r   <module>r     s   $ $ $ $ $ $              
			       				     ! ! ! ! ! !      - - - - - - ) ) ) ) ) ) ) ) )    
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
             
 . - - - - -     B B B
: : :(M M M@ @ @. . . U=2=#7#7z9R9R"STTTV=2=#8#8999 n0 n0 n0 n0 n0 n0 n0 n0 n0 n0r    