f M c @ sJ d Z d d l Z d d l Z d d l Z e j d Z e j d Z e j d Z e j d Z e j d Z e j d Z
e j d Z e j d
Z e j d Z
e j d Z e j d
Z e j d e j Z e j d e j Z e j d Z e j d Z Gd d d e Z Gd d d e j Z d S( u A parser for HTML and XHTML.i Nu [&<]u
&[a-zA-Z#]u% &([a-zA-Z][-.a-zA-Z0-9]*)[^a-zA-Z0-9]u) &#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]u <[a-zA-Z]u >u --\s*>u( ([a-zA-Z][-.a-zA-Z0-9:_]*)(?:\s|/(?!>))*u$ ([a-zA-Z][^
/> ]*)(?:\s|/(?!>))*uJ \s*([a-zA-Z_][-.:a-zA-Z_0-9]*)(\s*=\s*(\'[^\']*\'|"[^"]*"|[^\s"\'=<>`]*))?u] ((?<=[\'"\s/])[^\s/>][^\s/=>]*)(\s*=+\s*(\'[^\']*\'|"[^"]*"|(?![\'"])[^>\s]*))?(?:\s|/(?!>))*u
<[a-zA-Z][-.a-zA-Z0-9:_]* # tag name
(?:\s+ # whitespace before attribute name
(?:[a-zA-Z_][-.:a-zA-Z0-9_]* # attribute name
(?:\s*=\s* # value indicator
(?:'[^']*' # LITA-enclosed value
|\"[^\"]*\" # LIT-enclosed value
|[^'\">\s]+ # bare value
)
)?
)
)*
\s* # trailing whitespace
uF
<[a-zA-Z][^\t\n\r\f />\x00]* # tag name
(?:[\s/]* # optional whitespace before attribute name
(?:(?<=['"\s/])[^\s/>][^\s/=>]* # attribute name
(?:\s*=+\s* # value indicator
(?:'[^']*' # LITA-enclosed value
|"[^"]*" # LIT-enclosed value
|(?!['"])[^>\s]* # bare value
)
(?:\s*,)* # possibly followed by a comma
)?(?:\s|/(?!>))*
)*
)?
\s* # trailing whitespace
u# c B s5 | Ee Z d Z d Z d d d Z d d Z d S( u HTMLParseErroru& Exception raised for all parse errors.c C s3 | s t | | _ | d | _ | d | _ d S( Ni i ( u AssertionErroru msgu linenou offset( u selfu msgu position( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu __init__P s
u HTMLParseError.__init__c C sW | j } | j d k r, | d | j } n | j d k rS | d | j d } n | S( Nu , at line %du , column %di ( u msgu linenou Noneu offset( u selfu result( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu __str__V s u HTMLParseError.__str__N( NN( u __name__u
__module__u __qualname__u __doc__u Noneu __init__u __str__( u
__locals__( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu HTMLParseErrorM s u HTMLParseErrorc B sd | Ee Z d Z d Z d: Z d; d d Z d d Z d d Z d
d Z d d
Z
d9 Z d d Z
d d Z d d Z d d Z d d Z d d d Z d d Z d d Z d d Z d! d" Z d# d$ Z d% d& Z d' d( Z d) d* Z d+ d, Z d- d. Z d/ d0 Z d1 d2 Z d3 d4 Z d5 d6 Z d7 d8 Z! d9 S(< u
HTMLParseru Find tags and other markup and call handler functions.
Usage:
p = HTMLParser()
p.feed(data)
...
p.close()
Start tags are handled by calling self.handle_starttag() or
self.handle_startendtag(); end tags by self.handle_endtag(). The
data between tags is passed from the parser to the derived class
by calling self.handle_data() with the data as argument (the data
may be split up in arbitrary chunks). Entity references are
passed by calling self.handle_entityref() with the entity
reference as the argument. Numeric character references are
passed to self.handle_charref() with the string containing the
reference as the argument.
u scriptu stylec C s6 | r t j d t d d n | | _ | j d S( u Initialize and reset this instance.
If strict is set to False (the default) the parser will parse invalid
markup, otherwise it will raise an error. Note that the strict mode
is deprecated.
u The strict mode is deprecated.u
stackleveli N( u warningsu warnu DeprecationWarningu strictu reset( u selfu strict( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu __init__u s
u HTMLParser.__init__c C s8 d | _ d | _ t | _ d | _ t j j | d S( u1 Reset this instance. Loses all unprocessed data.u u ???N( u rawdatau lasttagu interesting_normalu interestingu Noneu
cdata_elemu _markupbaseu
ParserBaseu reset( u self( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu reset s
u HTMLParser.resetc C s! | j | | _ | j d d S( u Feed data to the parser.
Call this as often as you want, with as little or as much text
as you want (may include '\n').
i N( u rawdatau goahead( u selfu data( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu feed s u HTMLParser.feedc C s | j d d S( u Handle any buffered data.i N( u goahead( u self( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu close s u HTMLParser.closec C s t | | j d S( N( u HTMLParseErroru getpos( u selfu message( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu error s u HTMLParser.errorc C s | j S( u) Return full source of start tag: '<...>'.( u _HTMLParser__starttag_text( u self( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu get_starttag_text s u HTMLParser.get_starttag_textc C s2 | j | _ t j d | j t j | _ d S( Nu ( u loweru
cdata_elemu reu compileu Iu interesting( u selfu elem( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu set_cdata_mode s u HTMLParser.set_cdata_modec C s t | _ d | _ d S( N( u interesting_normalu interestingu Noneu
cdata_elem( u self( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu clear_cdata_mode s u HTMLParser.clear_cdata_modec
C s | j } d } t | } x| | k r| j j | | } | rT | j } n | j ra Pn | } | | k r | j | | | n | j | | } | | k r Pn | j } | d | rqt j
| | r | j | } n | d | r | j | } n | d | r-| j
| } n | d | rN| j | } ng | d | r| j rx| j | } q| j | } n+ | d | k r| j d | d } n P| d k r\| sPn | j r| j d n | j d | d } | d k r8| j d | d } | d k rB| d } qBn
| d 7} | j | | | n | j | | } q | d
| rMt j
| | } | r| j d d } | j | | j } | d | d s| d } n | j | | } q qd | | d k rI| j | | | d | j | | d } n Pq | d
| rt j
| | } | r| j d } | j | | j } | d | d s| d } n | j | | } q n t j
| | } | re| ra| j | | d k ra| j r'| j d qa| j } | | k rH| } n | j | | d } n Pq| d | k r| j d
| j | | d } qPq d s t d q W| r| | k r| j r| j | | | | j | | } n | | d | _ d S( Ni u u &#i u ;u &u# EOF in middle of entity or char refu interesting.search() liedi( u rawdatau lenu interestingu searchu startu
cdata_elemu handle_datau updateposu
startswithu starttagopenu matchu parse_starttagu parse_endtagu
parse_commentu parse_piu strictu parse_declarationu parse_html_declarationu erroru findu charrefu groupu handle_charrefu endu entityrefu handle_entityrefu
incompleteu AssertionError(
u selfu endu rawdatau iu nu matchu ju
startswithu ku name( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu goahead s
"
u HTMLParser.goaheadc C s | j } | | | d d k s/ t d | | | d d k rV | j | S| | | d d k r} | j | S| | | d j d k r | j d
| d } | d k r d
S| j | | d | | d S| j | Sd S( Ni u i ii( u rawdatau AssertionErroru
parse_commentu parse_marked_sectionu loweru findu handle_declu parse_bogus_comment( u selfu iu rawdatau gtpos( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu parse_html_declaration s &
u! HTMLParser.parse_html_declarationi c C s | j } | | | d d k s/ t d | j d | d } | d k rU d S| ry | j | | d | n | d S(
Ni u i ( u u />u
u junk characters in start tag: %ri iii( u >u />( u Noneu _HTMLParser__starttag_textu check_for_whole_start_tagu rawdatau strictu tagfindu matchu tagfind_tolerantu AssertionErroru endu groupu loweru lasttagu attrfindu attrfind_tolerantu unescapeu appendu stripu getposu countu lenu rfindu erroru handle_datau endswithu handle_startendtagu handle_starttagu CDATA_CONTENT_ELEMENTSu set_cdata_mode( u selfu iu endposu rawdatau attrsu matchu ku tagu mu attrnameu restu attrvalueu endu linenou offset( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu parse_starttagA s` 00 "u HTMLParser.parse_starttagc C sk | j } | j r' t j | | } n t j | | } | r[| j } | | | d } | d k rs | d S| d k r | j d | r | d S| j d | r d S| j r | j | | d | j d n | | k r | S| d Sn | d k rd S| d k rd
S| j r@| j | | | j d n | | k rP| S| d Sn t d
d S( Ni u >u /u />i u malformed empty start tagu u6 abcdefghijklmnopqrstuvwxyz=/ABCDEFGHIJKLMNOPQRSTUVWXYZu malformed start tagu we should not get here!iii(
u rawdatau strictu locatestarttagendu matchu locatestarttagend_tolerantu endu
startswithu updateposu erroru AssertionError( u selfu iu rawdatau mu ju next( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu check_for_whole_start_tag} s> u$ HTMLParser.check_for_whole_start_tagc C s | j } | | | d d k s/ t d t j | | d } | sO d S| j } t j | | } | sW| j d k r | j | | | | S| j
r | j d | | | f n t j | | d } | s| | | d d k r| d S| j
| Sn | j d j } | j d | j } | j | | d S| j d j } | j d k r| | j k r| j | | | | Sn | j | j | j | S(
Ni u u >i( u rawdatau AssertionErroru endendtagu searchu endu
endtagfindu matchu
cdata_elemu Noneu handle_datau strictu erroru tagfind_tolerantu parse_bogus_commentu groupu loweru findu
handle_endtagu clear_cdata_mode( u selfu iu rawdatau matchu gtposu namematchu tagnameu elem( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu parse_endtag s< & !
u HTMLParser.parse_endtagc C s! | j | | | j | d S( N( u handle_starttagu
handle_endtag( u selfu tagu attrs( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_startendtag s u HTMLParser.handle_startendtagc C s d S( N( ( u selfu tagu attrs( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_starttag s u HTMLParser.handle_starttagc C s d S( N( ( u selfu tag( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu
handle_endtag s u HTMLParser.handle_endtagc C s d S( N( ( u selfu name( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_charref s u HTMLParser.handle_charrefc C s d S( N( ( u selfu name( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_entityref s u HTMLParser.handle_entityrefc C s d S( N( ( u selfu data( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_data s u HTMLParser.handle_datac C s d S( N( ( u selfu data( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_comment s u HTMLParser.handle_commentc C s d S( N( ( u selfu decl( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_decl s u HTMLParser.handle_declc C s d S( N( ( u selfu data( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu handle_pi s u HTMLParser.handle_pic C s$ | j r | j d | f n d S( Nu unknown declaration: %r( u strictu error( u selfu data( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu unknown_decl s u HTMLParser.unknown_declc C s8 d | k r | Sd d } t j d | | d t j S( Nu &c
S sB | j d } yx | d d k r | d d } | d d k rh t | d d j d d } n t | j d } t | SWn t k
r d | SYn Xd d l m } | | k r | | S| j d r d
| Sx[ t d t | D]< } | d | | k r | | d | | | d Sq Wd
| Sd S(
Ni u #i u xu Xu ;i u &#( u html5u &i ( u xu X(
u groupsu intu rstripu chru
ValueErroru
html.entitiesu html5u endswithu rangeu len( u su cu html5u x( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu replaceEntities s& %
$u, HTMLParser.unescape..replaceEntitiesu& &(#?[xX]?(?:[0-9a-fA-F]+;|\w{1,32};?))u flags( u reu subu ASCII( u selfu su replaceEntities( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu unescape s
u HTMLParser.unescapeN( u scriptu styleF(" u __name__u
__module__u __qualname__u __doc__u CDATA_CONTENT_ELEMENTSu Falseu __init__u resetu feedu closeu erroru Noneu _HTMLParser__starttag_textu get_starttag_textu set_cdata_modeu clear_cdata_modeu goaheadu parse_html_declarationu parse_bogus_commentu parse_piu parse_starttagu check_for_whole_start_tagu parse_endtagu handle_startendtagu handle_starttagu
handle_endtagu handle_charrefu handle_entityrefu handle_datau handle_commentu handle_declu handle_piu unknown_declu unescape( u
__locals__( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu
HTMLParser_ s:
i < + * u
HTMLParser( u __doc__u _markupbaseu reu warningsu compileu interesting_normalu
incompleteu entityrefu charrefu starttagopenu picloseu commentcloseu tagfindu tagfind_tolerantu attrfindu attrfind_tolerantu VERBOSEu locatestarttagendu locatestarttagend_tolerantu endendtagu
endtagfindu Exceptionu HTMLParseErroru
ParserBaseu
HTMLParser( ( ( u0 /opt/alt/python33/lib64/python3.3/html/parser.pyu s0