m�
�=5Dc@s�dZd
k
Z
dklZ
ddgZde
if
d��YZde
if
d��YZe	d�
Z
edjo
e
�
n

d
S(	svHTML 2.0 parser.

See the HTML 2.0 specification:
http://www.w3.org/hypertext/WWW/MarkUp/html-spec/html-spec_toc.html
N(
sAS_ISt
HTMLParsertHTMLParseErrorc
BstZ
dZRS(
s3Error raised when an HTML document can't be parsed.(t__name__t
__module__t__doc__(((t/usr/lib/python2.4/htmllib.pyR
s
cBs0tZ
dZd
klZ
dd�
Zd�Zd�Zd�Zd�Z	d�Z
d	�Zd
�Zd�Z
d�Zd
�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�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.�Z0d/�Z1d0�Z2d1�Z3d2�Z4d3�Z5d4�Z6d5�Z7d6�Z8d7�Z9d8�Z:d9�Z;d:�Z<d;�Z=d<�Z>dd=�
Z?d>�Z@d?�ZAd@�ZBdA�ZCdB�ZDdC�ZEdD�ZFdE�ZGdF�ZHdG�ZIdH�ZJdI�ZKdJ�ZLdK�ZMdL�ZNdM�ZOdN�ZPdO�ZQdP�ZRdQ�ZSdR�ZTdS�ZUdT�ZVdU�ZWdV�ZXdW�ZYdX�ZZdY�Z[RS(Zs�This is the basic HTML parser class.

    It supports all entity names required by the XHTML 1.0 Recommendation.
    It also defines handlers for all HTML 2.0 and many HTML 3.0 and 3.2
    elements.

    (
s
entitydefsicCs ti
i||�
|
|_d
S(s�Creates an instance of the HTMLParser class.

        The formatter parameter is the formatter instance associated with
        the parser.

        N(tsgmllibt
SGMLParsert__init__tselftverboset	formatter(R	RR
((RRs

cCst|
�
�
dS(
N(R
tmessage(R	R((Rterror's
c

Cs\ti
i|�

d|_d
|_d|_d|_d|_	g|_
d
|_g|_dS(Ni(
RRtresetR	tNonetsavedatatisindexttitletbasetanchort
anchorlisttnofillt
list_stack(
R	((RR*s

	
	
	
	
	
	
	
cCsV|i
dj	o
|i
|
|_
n/
|io
|ii|
�

n
|ii|
�

dS(
N(R	RRtdataRRtadd_literal_datatadd_flowing_data(R	R((Rthandle_data:s




c

Cs
d
|_
dS(s�Begins saving character data in a buffer instead of sending it
        to the formatter object.

        Retrieve the stored data via the save_end() method.  Use of the
        save_bgn() / save_end() pair may not be nested.

        tN(R	R(
R	((Rtsave_bgnEs
c
Cs9|i
}
d|_
|ip
d
i|
i��
}
n

|
S(sH
Ends buffering character data and returns all data saved since
        the preceding call to the save_bgn() method.

        If the nofill flag is false, whitespace is collapsed to single
        spaces.  A call to this method without a preceding call to the
        save_bgn() method will raise a TypeError exception.

        t
 N(R	RRRRtjointsplit(R	R((Rtsave_endOs
	
	



cCs+|
|_|io
|ii|
�

n

d
S(s}
This method is called at the start of an anchor region.

        The arguments correspond to the attributes of the <A> tag with
        the same names.  The default implementation maintains a list of
        hyperlinks (defined by the HREF attribute for <A> tags) within
        the document.  The list of hyperlinks is available as the data
        attribute anchorlist.

        N(threfR	RRtappend(R	R"tnamettype((Rt
anchor_bgn`s	
	


c

Cs5|i
o'
|id
t|i�
�

d|_
n

dS(s�This method is called at the end of an anchor region.

        The default implementation adds a textual footnote marker using an
        index into the list of hyperlinks created by the anchor_bgn()method.

        s[%d]N(R	RRtlenRR(
R	((Rt
anchor_endns



cGs|i
|�

d
S(s�This method is called to handle images.

        The default implementation simply passes the alt value to the
        handle_data() method.

        N(R	Rtalt(R	tsrcR)targs((Rthandle_image{s
c
CsdS(
N((R	tattrs((Rt
start_html�sc


CsdS(
N((
R	((Rtend_html�sc
CsdS(
N((R	R-((Rt
start_head�sc


CsdS(
N((
R	((Rtend_head�sc
CsdS(
N((R	R-((Rt
start_body�sc


CsdS(
N((
R	((Rtend_body�sc
Cs|i
�
dS(
N(R	R(R	R-((Rtstart_title�s
c

Cs|i
�|_dS(
N(R	R!R(
R	((Rt	end_title�s
cCs5x.|
D]&\}}|d
jo

||_q
qWdS(NR"(R-t
at
vR	R(R	R-R6R7((Rtdo_base�s



cCs
d
|_
dS(Ni
(R	R(R	R-((Rt
do_isindex�s
c
CsdS(
N((R	R-((Rtdo_link�s
c
CsdS(
N((R	R-((Rtdo_meta�s
c
CsdS(
N((R	R-((Rt	do_nextid�s
cCs$|i
id
�

|i
id�

dS(Ni
th1i(R=ii
i(R	Rt
end_paragrapht	push_font(R	R-((Rtstart_h1�s

c

Cs!|i
id
�

|i
i�
dS(Ni
(R	RR>tpop_font(
R	((Rtend_h1�s

cCs$|i
id
�

|i
id�

dS(Ni
th2i(RCii
i(R	RR>R?(R	R-((Rtstart_h2�s

c

Cs!|i
id
�

|i
i�
dS(Ni
(R	RR>RA(
R	((Rtend_h2�s

cCs$|i
id
�

|i
id�

dS(Ni
th3i(RFii
i(R	RR>R?(R	R-((Rtstart_h3�s

c

Cs!|i
id
�

|i
i�
dS(Ni
(R	RR>RA(
R	((Rtend_h3�s

cCs$|i
id
�

|i
id�

dS(Ni
th4i(RIii
i(R	RR>R?(R	R-((Rtstart_h4�s

c

Cs!|i
id
�

|i
i�
dS(Ni
(R	RR>RA(
R	((Rtend_h4�s

cCs$|i
id
�

|i
id�

dS(Ni
th5i(RLii
i(R	RR>R?(R	R-((Rtstart_h5�s

c

Cs!|i
id
�

|i
i�
dS(Ni
(R	RR>RA(
R	((Rtend_h5�s

cCs$|i
id
�

|i
id�

dS(Ni
th6i(ROii
i(R	RR>R?(R	R-((Rtstart_h6�s

c

Cs!|i
id
�

|i
i�
dS(Ni
(R	RR>RA(
R	((Rtend_h6�s

cCs|i
id
�

dS(Ni
(R	RR>(R	R-((Rtdo_p�s
cCs@|i
id
�

|i
itttd
f�

|id
|_dS(Ni
(R	RR>R?tAS_ISR(R	R-((Rt	start_pre�s


c

Cs:|i
id
�

|i
i�
td|id
�|_dS(Ni
i(R	RR>RAtmaxR(
R	((Rtend_pre�s



cCs|i
|
�

|id
�

dS(Ntxmp(R	RTR-t
setliteral(R	R-((Rt	start_xmp�s


c


Cs|i
�
dS(
N(R	RV(
R	((Rtend_xmp�s
cCs|i
|
�

|id
�

dS(Ntlisting(R	RTR-RX(R	R-((Rt
start_listing�s


c


Cs|i
�
dS(
N(R	RV(
R	((Rtend_listing�s
cCs0|i
id
�

|i
itdttf�

dS(Nii
(R	RR>R?RS(R	R-((Rt
start_address�s

c

Cs!|i
id
�

|i
i�
dS(Ni(R	RR>RA(
R	((Rtend_address�s

cCs$|i
id
�

|i
id�

dS(Ni
t
blockquote(R	RR>tpush_margin(R	R-((Rtstart_blockquote

s

c

Cs!|i
id
�

|i
i�
dS(Ni
(R	RR>t
pop_margin(
R	((Rtend_blockquote
s

cCsA|i
i|i�

|i
id
�

|iid
ddg�

dS(Ntult
*i(R	RR>RRaR#(R	R-((Rtstart_ul
s


c

Cs=|i
o
|i
d
=n

|ii|i
�

|ii�
dS(Ni����(R	RRR>Rc(
R	((Rtend_ul
s



cCsm|i
id
�

|io0
|id\}}}}|d|d<}n

d\}}|i
i||�
dS(Nii����i
iRf(Rfi(	R	RR>Rtdummytlabeltcounterttoptadd_label_data(R	R-RiRlRkRj((Rtdo_li
s





cCs�|i
i|i�

|i
id
�

d}xL|
D]D\}}|djo+
t	|�
djo
|d}n

|}q1
q1W|ii
d
|dg�

dS(Ntols1.R%i
t
.i(R	RR>RRaRjR-R6R7R'R#(R	R-R6RjR7((Rtstart_ol
s








c

Cs=|i
o
|i
d
=n

|ii|i
�

|ii�
dS(Ni����(R	RRR>Rc(
R	((Rtend_ol(
s



cCs|i
|
�

dS(
N(R	RgR-(R	R-((Rt
start_menu-
s
c


Cs|i
�
dS(
N(R	Rh(
R	((Rtend_menu0
s
cCs|i
|
�

dS(
N(R	RgR-(R	R-((Rt	start_dir3
s
c


Cs|i
�
dS(
N(R	Rh(
R	((Rtend_dir6
s
cCs-|i
id
�

|iidddg�

dS(Ni
tdlRi(R	RR>RR#(R	R-((Rtstart_dl9
s

c

Cs)|i
d
�

|io
|id=n

dS(Ni
i����(R	tddpopR(
R	((Rtend_dl=
s



c
Cs|i
�
dS(
N(R	Ry(R	R-((Rtdo_dtA
s
cCs7|i
�
|iid
�

|iid
ddg�

dS(NtddRi(R	RyRRaRR#(R	R-((Rtdo_ddD
s



cCsU|i
i|
�

|io7
|id
ddjo
|id
=|i
i�
qQ
n

dS(Ni����iR|(R	RR>tblRRc(R	R~((RRyI
s







cCs|i
|
�

dS(
N(R	tstart_iR-(R	R-((Rt
start_citeT
sc


Cs|i
�
dS(
N(R	tend_i(
R	((Rtend_citeU
scCs|i
|
�

dS(
N(R	tstart_ttR-(R	R-((Rt
start_codeW
sc


Cs|i
�
dS(
N(R	tend_tt(
R	((Rtend_codeX
scCs|i
|
�

dS(
N(R	RR-(R	R-((Rtstart_emZ
sc


Cs|i
�
dS(
N(R	R�(
R	((Rtend_em[
scCs|i
|
�

dS(
N(R	R�R-(R	R-((Rt	start_kbd]
sc


Cs|i
�
dS(
N(R	R�(
R	((Rtend_kbd^
scCs|i
|
�

dS(
N(R	R�R-(R	R-((Rt
start_samp`
sc


Cs|i
�
dS(
N(R	R�(
R	((Rtend_sampa
scCs|i
|
�

dS(
N(R	tstart_bR-(R	R-((Rtstart_strongc
sc


Cs|i
�
dS(
N(R	tend_b(
R	((Rt
end_strongd
scCs|i
|
�

dS(
N(R	RR-(R	R-((Rt	start_varf
sc


Cs|i
�
dS(
N(R	R�(
R	((Rtend_varg
scCs |i
itd
ttf�

dS(Ni
(R	RR?RS(R	R-((RRk
s
c


Cs|i
i�
dS(
N(R	RRA(
R	((RR�m
s
cCs |i
ittd
tf�

dS(Ni
(R	RR?RS(R	R-((RR�p
s
c


Cs|i
i�
dS(
N(R	RRA(
R	((RR�r
s
cCs |i
itttd
f�

dS(Ni
(R	RR?RS(R	R-((RR�u
s
c


Cs|i
i�
dS(
N(R	RRA(
R	((RR�w
s
cCs�d
}d
}d
}xk|
D]c\}}|i�}|djo

|}n

|djo

|}n

|djo
|i�}q
qW|i	|||�
dS(NRR"R$R%(
R"R$R%R-tattrnametvaluetstriptlowerR	R&(R	R-R$R�R�R"R%((Rtstart_az
s
















c


Cs|i
�
dS(
N(R	R((
R	((Rtend_a�
s
c
Cs|i
i�
dS(
N(R	Rtadd_line_break(R	R-((Rtdo_br�
s
c
Cs|i
i�
dS(
N(R	Rtadd_hor_rule(R	R-((Rtdo_hr�
s
c
Cs%
d
}d}d
}d
}d}d}	x�|
D]�\}}|djo

|}n

|djo

|}n

|djo

|}n

|djo

|}n

|djo*
yt	|�
}Wq�t
j
o



q�
Xn

|d	jo*
yt	|�
}	Wq

t
j
o



q


Xq+
q+W|i||||||	�
dS(
NRs(image)italignR)tismapR*twidththeight(
R�R)R�R*R�R�R-R�R�tintt
ValueErrorR	R,(
R	R-R*R�R�R�R�R�R)R�((Rtdo_img�
s6


























	





cCs|i
|
�

|i�
dS(
N(R	RTR-t
setnomoretags(R	R-((Rtdo_plaintext�
s


c
CsdS(
N((R	ttagR-((Rtunknown_starttag�
s
c
CsdS(
N((R	R�((Rtunknown_endtag�
s
(\RRRthtmlentitydefst
entitydefsRR
RRRR!R&R(R,R.R/R0R1R2R3R4R5R8R9R:R;R<R@RBRDRERGRHRJRKRMRNRPRQRRRTRVRYRZR\R]R^R_RbRdRgRhRnRqRrRsRtRuRvRxRzR{R}RyR�R�R�R�R�R�R�R�R�R�R�R�R�R�RR�R�R�R�R�R�R�R�R�R�R�R�R�(((RRs�

				
			
		
		
		
																																				
										
		
		
		
		
		
		
														c
	CsJ
dk}dk
}|p
|id
}n

|o
|ddj}
|
o
|d=n

|o
|d}n
d}|djo

|i}nF
yt|d�}Wn/t	j
o#

}
|GdG|GH|id
�

n
X|i�}||ij	o
|i�
n

|
o
|i�}n
|i|i��
}t|�
}|i|�

|i�
dS(Ni
is-ss	test.htmlt
-t
rt
:(tsysRR+targvtsilenttfiletstdint
ftopentIOErrortmsgtexittreadRtcloset
NullFormattertAbstractFormattert
DumbWriterRt
ptfeed(	R+R�R�RR�R�R�R�R((Rttest�
s2

















t__main__(RRRRSt__all__tSGMLParseErrorR
RRRR�R(R
RR�RR�RS((Rt
?s	
��'