PNG  IHDRxsBIT|d pHYs+tEXtSoftwarewww.inkscape.org<,tEXtComment File Manager

File Manager

Path: /opt/cloudlinux/venv/lib64/python3.11/site-packages/lxml/includes/libxml/

Viewing File: HTMLparser.h

/*
 * Summary: interface for an HTML 4.0 non-verifying parser
 * Description: this module implements an HTML 4.0 non-verifying parser
 *              with API compatible with the XML parser ones. It should
 *              be able to parse "real world" HTML, even if severely
 *              broken from a specification point of view.
 *
 * Copy: See Copyright for the status of this software.
 *
 * Author: Daniel Veillard
 */

#ifndef __HTML_PARSER_H__
#define __HTML_PARSER_H__
#include <libxml/xmlversion.h>
#include <libxml/parser.h>

#ifdef LIBXML_HTML_ENABLED

#ifdef __cplusplus
extern "C" {
#endif

/*
 * Most of the back-end structures from XML and HTML are shared.
 */
typedef xmlParserCtxt htmlParserCtxt;
typedef xmlParserCtxtPtr htmlParserCtxtPtr;
typedef xmlParserNodeInfo htmlParserNodeInfo;
typedef xmlSAXHandler htmlSAXHandler;
typedef xmlSAXHandlerPtr htmlSAXHandlerPtr;
typedef xmlParserInput htmlParserInput;
typedef xmlParserInputPtr htmlParserInputPtr;
typedef xmlDocPtr htmlDocPtr;
typedef xmlNodePtr htmlNodePtr;

/*
 * Internal description of an HTML element, representing HTML 4.01
 * and XHTML 1.0 (which share the same structure).
 */
typedef struct _htmlElemDesc htmlElemDesc;
typedef htmlElemDesc *htmlElemDescPtr;
struct _htmlElemDesc {
    const char *name;	/* The tag name */
    char startTag;      /* Whether the start tag can be implied */
    char endTag;        /* Whether the end tag can be implied */
    char saveEndTag;    /* Whether the end tag should be saved */
    char empty;         /* Is this an empty element ? */
    char depr;          /* Is this a deprecated element ? */
    char dtd;           /* 1: only in Loose DTD, 2: only Frameset one */
    char isinline;      /* is this a block 0 or inline 1 element */
    const char *desc;   /* the description */

/* NRK Jan.2003
 * New fields encapsulating HTML structure
 *
 * Bugs:
 *	This is a very limited representation.  It fails to tell us when
 *	an element *requires* subelements (we only have whether they're
 *	allowed or not), and it doesn't tell us where CDATA and PCDATA
 *	are allowed.  Some element relationships are not fully represented:
 *	these are flagged with the word MODIFIER
 */
    const char** subelts;		/* allowed sub-elements of this element */
    const char* defaultsubelt;	/* subelement for suggested auto-repair
					   if necessary or NULL */
    const char** attrs_opt;		/* Optional Attributes */
    const char** attrs_depr;		/* Additional deprecated attributes */
    const char** attrs_req;		/* Required attributes */
};

/*
 * Internal description of an HTML entity.
 */
typedef struct _htmlEntityDesc htmlEntityDesc;
typedef htmlEntityDesc *htmlEntityDescPtr;
struct _htmlEntityDesc {
    unsigned int value;	/* the UNICODE value for the character */
    const char *name;	/* The entity name */
    const char *desc;   /* the description */
};

/*
 * There is only few public functions.
 */
XMLPUBFUN const htmlElemDesc * XMLCALL
			htmlTagLookup	(const xmlChar *tag);
XMLPUBFUN const htmlEntityDesc * XMLCALL
			htmlEntityLookup(const xmlChar *name);
XMLPUBFUN const htmlEntityDesc * XMLCALL
			htmlEntityValueLookup(unsigned int value);

XMLPUBFUN int XMLCALL
			htmlIsAutoClosed(htmlDocPtr doc,
					 htmlNodePtr elem);
XMLPUBFUN int XMLCALL
			htmlAutoCloseTag(htmlDocPtr doc,
					 const xmlChar *name,
					 htmlNodePtr elem);
XMLPUBFUN const htmlEntityDesc * XMLCALL
			htmlParseEntityRef(htmlParserCtxtPtr ctxt,
					 const xmlChar **str);
XMLPUBFUN int XMLCALL
			htmlParseCharRef(htmlParserCtxtPtr ctxt);
XMLPUBFUN void XMLCALL
			htmlParseElement(htmlParserCtxtPtr ctxt);

XMLPUBFUN htmlParserCtxtPtr XMLCALL
			htmlNewParserCtxt(void);

XMLPUBFUN htmlParserCtxtPtr XMLCALL
			htmlCreateMemoryParserCtxt(const char *buffer,
						   int size);

XMLPUBFUN int XMLCALL
			htmlParseDocument(htmlParserCtxtPtr ctxt);
XMLPUBFUN htmlDocPtr XMLCALL
			htmlSAXParseDoc	(const xmlChar *cur,
					 const char *encoding,
					 htmlSAXHandlerPtr sax,
					 void *userData);
XMLPUBFUN htmlDocPtr XMLCALL
			htmlParseDoc	(const xmlChar *cur,
					 const char *encoding);
XMLPUBFUN htmlDocPtr XMLCALL
			htmlSAXParseFile(const char *filename,
					 const char *encoding,
					 htmlSAXHandlerPtr sax,
					 void *userData);
XMLPUBFUN htmlDocPtr XMLCALL
			htmlParseFile	(const char *filename,
					 const char *encoding);
XMLPUBFUN int XMLCALL
			UTF8ToHtml	(unsigned char *out,
					 int *outlen,
					 const unsigned char *in,
					 int *inlen);
XMLPUBFUN int XMLCALL
			htmlEncodeEntities(unsigned char *out,
					 int *outlen,
					 const unsigned char *in,
					 int *inlen, int quoteChar);
XMLPUBFUN int XMLCALL
			htmlIsScriptAttribute(const xmlChar *name);
XMLPUBFUN int XMLCALL
			htmlHandleOmittedElem(int val);

#ifdef LIBXML_PUSH_ENABLED
/**
 * Interfaces for the Push mode.
 */
XMLPUBFUN htmlParserCtxtPtr XMLCALL
			htmlCreatePushParserCtxt(htmlSAXHandlerPtr sax,
						 void *user_data,
						 const char *chunk,
						 int size,
						 const char *filename,
						 xmlCharEncoding enc);
XMLPUBFUN int XMLCALL
			htmlParseChunk		(htmlParserCtxtPtr ctxt,
						 const char *chunk,
						 int size,
						 int terminate);
#endif /* LIBXML_PUSH_ENABLED */

XMLPUBFUN void XMLCALL
			htmlFreeParserCtxt	(htmlParserCtxtPtr ctxt);

/*
 * New set of simpler/more flexible APIs
 */
/**
 * xmlParserOption:
 *
 * This is the set of XML parser options that can be passed down
 * to the xmlReadDoc() and similar calls.
 */
typedef enum {
    HTML_PARSE_RECOVER  = 1<<0, /* Relaxed parsing */
    HTML_PARSE_NODEFDTD = 1<<2, /* do not default a doctype if not found */
    HTML_PARSE_NOERROR	= 1<<5,	/* suppress error reports */
    HTML_PARSE_NOWARNING= 1<<6,	/* suppress warning reports */
    HTML_PARSE_PEDANTIC	= 1<<7,	/* pedantic error reporting */
    HTML_PARSE_NOBLANKS	= 1<<8,	/* remove blank nodes */
    HTML_PARSE_NONET	= 1<<11,/* Forbid network access */
    HTML_PARSE_NOIMPLIED= 1<<13,/* Do not add implied html/body... elements */
    HTML_PARSE_COMPACT  = 1<<16,/* compact small text nodes */
    HTML_PARSE_IGNORE_ENC=1<<21 /* ignore internal document encoding hint */
} htmlParserOption;

XMLPUBFUN void XMLCALL
		htmlCtxtReset		(htmlParserCtxtPtr ctxt);
XMLPUBFUN int XMLCALL
		htmlCtxtUseOptions	(htmlParserCtxtPtr ctxt,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlReadDoc		(const xmlChar *cur,
					 const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlReadFile		(const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlReadMemory		(const char *buffer,
					 int size,
					 const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlReadFd		(int fd,
					 const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlReadIO		(xmlInputReadCallback ioread,
					 xmlInputCloseCallback ioclose,
					 void *ioctx,
					 const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlCtxtReadDoc		(xmlParserCtxtPtr ctxt,
					 const xmlChar *cur,
					 const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlCtxtReadFile		(xmlParserCtxtPtr ctxt,
					 const char *filename,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlCtxtReadMemory		(xmlParserCtxtPtr ctxt,
					 const char *buffer,
					 int size,
					 const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlCtxtReadFd		(xmlParserCtxtPtr ctxt,
					 int fd,
					 const char *URL,
					 const char *encoding,
					 int options);
XMLPUBFUN htmlDocPtr XMLCALL
		htmlCtxtReadIO		(xmlParserCtxtPtr ctxt,
					 xmlInputReadCallback ioread,
					 xmlInputCloseCallback ioclose,
					 void *ioctx,
					 const char *URL,
					 const char *encoding,
					 int options);

/* NRK/Jan2003: further knowledge of HTML structure
 */
typedef enum {
  HTML_NA = 0 ,		/* something we don't check at all */
  HTML_INVALID = 0x1 ,
  HTML_DEPRECATED = 0x2 ,
  HTML_VALID = 0x4 ,
  HTML_REQUIRED = 0xc /* VALID bit set so ( & HTML_VALID ) is TRUE */
} htmlStatus ;

/* Using htmlElemDesc rather than name here, to emphasise the fact
   that otherwise there's a lookup overhead
*/
XMLPUBFUN htmlStatus XMLCALL htmlAttrAllowed(const htmlElemDesc*, const xmlChar*, int) ;
XMLPUBFUN int XMLCALL htmlElementAllowedHere(const htmlElemDesc*, const xmlChar*) ;
XMLPUBFUN htmlStatus XMLCALL htmlElementStatusHere(const htmlElemDesc*, const htmlElemDesc*) ;
XMLPUBFUN htmlStatus XMLCALL htmlNodeStatus(const htmlNodePtr, int) ;
/**
 * htmlDefaultSubelement:
 * @elt: HTML element
 *
 * Returns the default subelement for this element
 */
#define htmlDefaultSubelement(elt) elt->defaultsubelt
/**
 * htmlElementAllowedHereDesc:
 * @parent: HTML parent element
 * @elt: HTML element
 *
 * Checks whether an HTML element description may be a
 * direct child of the specified element.
 *
 * Returns 1 if allowed; 0 otherwise.
 */
#define htmlElementAllowedHereDesc(parent,elt) \
	htmlElementAllowedHere((parent), (elt)->name)
/**
 * htmlRequiredAttrs:
 * @elt: HTML element
 *
 * Returns the attributes required for the specified element.
 */
#define htmlRequiredAttrs(elt) (elt)->attrs_req


#ifdef __cplusplus
}
#endif

#endif /* LIBXML_HTML_ENABLED */
#endif /* __HTML_PARSER_H__ */
b IDATxytVսϓ22 A@IR :hCiZ[v*E:WũZA ^dQeQ @ !jZ'>gsV仿$|?g)&x-EIENT ;@xT.i%-X}SvS5.r/UHz^_$-W"w)Ɗ/@Z &IoX P$K}JzX:;` &, ŋui,e6mX ԵrKb1ԗ)DADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADADA݀!I*]R;I2$eZ#ORZSrr6mteffu*((Pu'v{DIߔ4^pIm'77WEEE;vƎ4-$]'RI{\I&G :IHJ DWBB=\WR޽m o$K(V9ABB.}jѢv`^?IOȅ} ڶmG}T#FJ`56$-ھ}FI&v;0(h;Б38CӧOWf!;A i:F_m9s&|q%=#wZprrrla A &P\\СC[A#! {olF} `E2}MK/vV)i{4BffV\|ۭX`b@kɶ@%i$K z5zhmX[IXZ` 'b%$r5M4º/l ԃߖxhʔ)[@=} K6IM}^5k㏷݆z ΗÿO:gdGBmyT/@+Vɶ纽z񕏵l.y޴it뭷zV0[Y^>Wsqs}\/@$(T7f.InݺiR$푔n.~?H))\ZRW'Mo~v Ov6oԃxz! S,&xm/yɞԟ?'uaSѽb,8GלKboi&3t7Y,)JJ c[nzӳdE&KsZLӄ I?@&%ӟ۶mSMMњ0iؐSZ,|J+N ~,0A0!5%Q-YQQa3}$_vVrf9f?S8`zDADADADADADADADADAdqP,تmMmg1V?rSI꒟]u|l RCyEf٢9 jURbztѰ!m5~tGj2DhG*{H9)꒟ר3:(+3\?/;TUݭʴ~S6lڧUJ*i$d(#=Yݺd{,p|3B))q:vN0Y.jkק6;SɶVzHJJЀ-utѹսk>QUU\޲~]fFnK?&ߡ5b=z9)^|u_k-[y%ZNU6 7Mi:]ۦtk[n X(e6Bb."8cۭ|~teuuw|ήI-5"~Uk;ZicEmN/:]M> cQ^uiƞ??Ңpc#TUU3UakNwA`:Y_V-8.KKfRitv޲* 9S6ֿj,ՃNOMߤ]z^fOh|<>@Å5 _/Iu?{SY4hK/2]4%it5q]GGe2%iR| W&f*^]??vq[LgE_3f}Fxu~}qd-ږFxu~I N>\;͗O֊:̗WJ@BhW=y|GgwܷH_NY?)Tdi'?խwhlmQi !SUUsw4kӺe4rfxu-[nHtMFj}H_u~w>)oV}(T'ebʒv3_[+vn@Ȭ\S}ot}w=kHFnxg S 0eޢm~l}uqZfFoZuuEg `zt~? b;t%>WTkķh[2eG8LIWx,^\thrl^Ϊ{=dž<}qV@ ⠨Wy^LF_>0UkDuʫuCs$)Iv:IK;6ֲ4{^6եm+l3>݆uM 9u?>Zc }g~qhKwڭeFMM~pМuqǿz6Tb@8@Y|jx](^]gf}M"tG -w.@vOqh~/HII`S[l.6nØXL9vUcOoB\xoǤ'T&IǍQw_wpv[kmO{w~>#=P1Pɞa-we:iǏlHo׈꒟f9SzH?+shk%Fs:qVhqY`jvO'ρ?PyX3lх]˾uV{ݞ]1,MzYNW~̈́ joYn}ȚF߾׮mS]F z+EDxm/d{F{-W-4wY듏:??_gPf ^3ecg ҵs8R2מz@TANGj)}CNi/R~}c:5{!ZHӋӾ6}T]G]7W6^n 9*,YqOZj:P?Q DFL|?-^.Ɵ7}fFh׶xe2Pscz1&5\cn[=Vn[ĶE鎀uˌd3GII k;lNmشOuuRVfBE]ۣeӶu :X-[(er4~LHi6:Ѻ@ԅrST0trk%$Č0ez" *z"T/X9|8.C5Feg}CQ%͞ˣJvL/?j^h&9xF`њZ(&yF&Iݻfg#W;3^{Wo^4'vV[[K';+mӍִ]AC@W?1^{එyh +^]fm~iԵ]AB@WTk̏t uR?l.OIHiYyԶ]Aˀ7c:q}ힽaf6Z~қm(+sK4{^6}T*UUu]n.:kx{:2 _m=sAߤU@?Z-Vކеz왍Nэ{|5 pڶn b p-@sPg]0G7fy-M{GCF'%{4`=$-Ge\ eU:m+Zt'WjO!OAF@ik&t݆ϥ_ e}=]"Wz_.͜E3leWFih|t-wZۍ-uw=6YN{6|} |*={Ѽn.S.z1zjۻTH]흾 DuDvmvK.`V]yY~sI@t?/ϓ. m&["+P?MzovVЫG3-GRR[(!!\_,^%?v@ҵő m`Y)tem8GMx.))A]Y i`ViW`?^~!S#^+ѽGZj?Vģ0.))A꨷lzL*]OXrY`DBBLOj{-MH'ii-ϰ ok7^ )쭡b]UXSְmռY|5*cֽk0B7镹%ڽP#8nȎq}mJr23_>lE5$iwui+ H~F`IjƵ@q \ @#qG0".0" l`„.0! ,AQHN6qzkKJ#o;`Xv2>,tێJJ7Z/*A .@fفjMzkg @TvZH3Zxu6Ra'%O?/dQ5xYkU]Rֽkق@DaS^RSּ5|BeHNN͘p HvcYcC5:y #`οb;z2.!kr}gUWkyZn=f Pvsn3p~;4p˚=ē~NmI] ¾ 0lH[_L hsh_ғߤc_њec)g7VIZ5yrgk̞W#IjӪv>՞y睝M8[|]\շ8M6%|@PZڨI-m>=k='aiRo-x?>Q.}`Ȏ:Wsmu u > .@,&;+!!˱tﭧDQwRW\vF\~Q7>spYw$%A~;~}6¾ g&if_=j,v+UL1(tWake:@Ș>j$Gq2t7S?vL|]u/ .(0E6Mk6hiۺzښOrifޱxm/Gx> Lal%%~{lBsR4*}{0Z/tNIɚpV^#Lf:u@k#RSu =S^ZyuR/.@n&΃z~B=0eg뺆#,Þ[B/?H uUf7y Wy}Bwegל`Wh(||`l`.;Ws?V@"c:iɍL֯PGv6zctM̠':wuW;d=;EveD}9J@B(0iհ bvP1{\P&G7D޴Iy_$-Qjm~Yrr&]CDv%bh|Yzni_ˆR;kg}nJOIIwyuL}{ЌNj}:+3Y?:WJ/N+Rzd=hb;dj͒suݔ@NKMԄ jqzC5@y°hL m;*5ezᕏ=ep XL n?מ:r`۵tŤZ|1v`V뽧_csج'ߤ%oTuumk%%%h)uy]Nk[n 'b2 l.=͜E%gf$[c;s:V-͞WߤWh-j7]4=F-X]>ZLSi[Y*We;Zan(ӇW|e(HNNP5[= r4tP &0<pc#`vTNV GFqvTi*Tyam$ߏWyE*VJKMTfFw>'$-ؽ.Ho.8c"@DADADADADADADADADA~j*֘,N;Pi3599h=goضLgiJ5փy~}&Zd9p֚ e:|hL``b/d9p? fgg+%%hMgXosج, ΩOl0Zh=xdjLmhݻoO[g_l,8a]٭+ӧ0$I]c]:粹:Teꢢ"5a^Kgh,&= =՟^߶“ߢE ܹS J}I%:8 IDAT~,9/ʃPW'Mo}zNƍ쨓zPbNZ~^z=4mswg;5 Y~SVMRXUյڱRf?s:w ;6H:ºi5-maM&O3;1IKeamZh͛7+##v+c ~u~ca]GnF'ټL~PPPbn voC4R,ӟgg %hq}@#M4IÇ Oy^xMZx ) yOw@HkN˖-Sǎmb]X@n+i͖!++K3gd\$mt$^YfJ\8PRF)77Wא!Cl$i:@@_oG I{$# 8磌ŋ91A (Im7֭>}ߴJq7ޗt^ -[ԩSj*}%]&' -ɓ'ꫯVzzvB#;a 7@GxI{j޼ƌ.LÇWBB7`O"I$/@R @eee@۷>}0,ɒ2$53Xs|cS~rpTYYY} kHc %&k.], @ADADADADADADADADA@lT<%''*Lo^={رc5h %$+CnܸQ3fҥK}vUVVs9G R,_{xˇ3o߾;TTTd}馛]uuuG~iԩ@4bnvmvfϞ /Peeeq}}za I~,誫{UWW뮻}_~YƍSMMMYχ֝waw\ďcxꩧtEƍկ_?۷5@u?1kNׯWzz/wy>}zj3 k(ٺuq_Zvf̘:~ ABQ&r|!%KҥKgԞ={<_X-z !CyFUUz~ ABQIIIjݺW$UXXDٳZ~ ABQƍecW$<(~<RSSvZujjjԧOZQu@4 8m&&&jԩg$ď1h ͟?_{768@g =@`)))5o6m3)ѣƌJ;wҿUTT /KZR{~a=@0o<*狔iFɶ[ˎ;T]]OX@?K.ۈxN pppppppppppppppppPfl߾] ,{ァk۶mڿo5BTӦMӴiӴ|r DB2e|An!Dy'tkΝ[A $***t5' "!駟oaDnΝ:t֭[gDШQ06qD;@ x M6v(PiizmZ4ew"@̴ixf [~-Fٱc&IZ2|n!?$@{[HTɏ#@hȎI# _m(F /6Z3z'\r,r!;w2Z3j=~GY7"I$iI.p_"?pN`y DD?: _  Gÿab7J !Bx@0 Bo cG@`1C[@0G @`0C_u V1 aCX>W ` | `!<S `"<. `#c`?cAC4 ?c p#~@0?:08&_MQ1J h#?/`7;I  q 7a wQ A 1 Hp !#<8/#@1Ul7=S=K.4Z?E_$i@!1!E4?`P_  @Bă10#: "aU,xbFY1 [n|n #'vEH:`xb #vD4Y hi.i&EΖv#O H4IŶ}:Ikh @tZRF#(tXҙzZ ?I3l7q@õ|ۍ1,GpuY Ꮿ@hJv#xxk$ v#9 5 }_$c S#=+"K{F*m7`#%H:NRSp6I?sIՖ{Ap$I$I:QRv2$Z @UJ*$]<FO4IENDB`