From: =?utf-8?b?T25kxZllaiBTdXLDvQ==?= Date: Sun, 27 Apr 2025 12:29:03 +0200 Subject: Backport patch for CVE-2021-33391 for tidy-html5 --- src/gdoc.c | 14 +- src/lexer.c | 122 +++++++++-- src/lexer.h | 695 ++++++++++++++++++++++++++++++++++++------------------------ 3 files changed, 530 insertions(+), 301 deletions(-) diff --git a/src/gdoc.c b/src/gdoc.c index 50cd9bc..3786746 100644 --- a/src/gdoc.c +++ b/src/gdoc.c @@ -96,14 +96,15 @@ static void DiscardContainer( TidyDocImpl* doc, Node *element, Node **pnode) static void CleanNode( TidyDocImpl* doc, Node *node ) { + Stack *stack = TY_(newStack)(doc, 16); Node *child, *next; - if (node->content) + if ( (child = node->content) ) { - for (child = node->content; child != NULL; child = next) + while (child) { next = child->next; - + if (TY_(nodeIsElement)(child)) { if (nodeIsSTYLE(child)) @@ -131,11 +132,16 @@ static void CleanNode( TidyDocImpl* doc, Node *node ) if (child->attributes) TY_(DropAttrByName)( doc, child, "class" ); - CleanNode(doc, child); + TY_(push)(stack,next); + child = child->content; + continue; } } + child = next ? next : TY_(pop)(stack); } } + TY_(freeStack)(stack); + TidyFree(doc->allocator, stack); } /* insert meta element to force browser to recognize doc as UTF8 */ diff --git a/src/lexer.c b/src/lexer.c index bc4e50a..d1cae84 100644 --- a/src/lexer.c +++ b/src/lexer.c @@ -877,15 +877,6 @@ static tmbchar LastChar( tmbstr str ) return 0; } -/* - node->type is one of these: - - #define TextNode 1 - #define StartTag 2 - #define EndTag 3 - #define StartEndTag 4 -*/ - Lexer* TY_(NewLexer)( TidyDocImpl* doc ) { Lexer* lexer = (Lexer*) TidyDocAlloc( doc, sizeof(Lexer) ); @@ -1545,13 +1536,7 @@ void TY_(FreeNode)( TidyDocImpl* doc, Node *node ) } } #endif - /* this is no good ;=(( - if (node && doc && doc->lexer) { - if (node == doc->lexer->token) { - doc->lexer->token = NULL; // TY_(NewNode)( doc->lexer->allocator, doc->lexer ); - } - } - ----------------- */ + while ( node ) { Node* next = node->next; @@ -4462,11 +4447,102 @@ static Node *ParseDocTypeDecl(TidyDocImpl* doc) return NULL; } -/* - * local variables: - * mode: c - * indent-tabs-mode: nil - * c-basic-offset: 4 - * eval: (c-set-offset 'substatement-open 0) - * end: + +/****************************************************************************//* + ** MARK: - Node Stack + ***************************************************************************/ + + +/** + * Create a new stack with a given starting capacity. If memory allocation + * fails, then the allocator will panic the program automatically. */ +Stack* TY_(newStack)(TidyDocImpl *doc, uint capacity) +{ + Stack *stack = (Stack *)TidyAlloc(doc->allocator, sizeof(Stack)); + stack->top = -1; + stack->capacity = capacity; + stack->firstNode = (Node **)TidyAlloc(doc->allocator, stack->capacity * sizeof(Node**)); + stack->allocator = doc->allocator; + return stack; +} + + +/** + * Increase the stack size. This will be called automatically when the + * current stack is full. If memory allocation fails, then the allocator + * will panic the program automatically. + */ +void TY_(growStack)(Stack *stack) +{ + uint new_capacity = stack->capacity * 2; + + Node **firstNode = (Node **)TidyAlloc(stack->allocator, new_capacity); + + memcpy( firstNode, stack->firstNode, sizeof(Node**) * (stack->top + 1) ); + TidyFree(stack->allocator, stack->firstNode); + + stack->firstNode = firstNode; + stack->capacity = new_capacity; +} + + +/** + * Stack is full when top is equal to the last index. + */ +Bool TY_(stackFull)(Stack *stack) +{ + return stack->top == stack->capacity - 1; +} + + +/** + * Stack is empty when top is equal to -1 + */ +Bool TY_(stackEmpty)(Stack *stack) +{ + return stack->top == -1; +} + + +/** + * Push an item to the stack. + */ +void TY_(push)(Stack *stack, Node *node) +{ + if (TY_(stackFull)(stack)) + TY_(growStack)(stack); + + if (node) + stack->firstNode[++stack->top] = node; +} + + +/** + * Pop an item from the stack. + */ +Node* TY_(pop)(Stack *stack) +{ + return TY_(stackEmpty)(stack) ? NULL : stack->firstNode[stack->top--]; +} + + +/** + * Peek at the stack. + */ +FUNC_UNUSED Node* TY_(peek)(Stack *stack) +{ + return TY_(stackEmpty)(stack) ? NULL : stack->firstNode[stack->top--]; +} + +/** + * Frees the stack when done. + */ +void TY_(freeStack)(Stack *stack) +{ + TidyFree( stack->allocator, stack->firstNode ); + stack->top = -1; + stack->capacity = 0; + stack->firstNode = NULL; + stack->allocator = NULL; +} diff --git a/src/lexer.h b/src/lexer.h index c181d4b..d9ae113 100644 --- a/src/lexer.h +++ b/src/lexer.h @@ -1,33 +1,46 @@ #ifndef __LEXER_H__ #define __LEXER_H__ -/* lexer.h -- Lexer for html parser - - (c) 1998-2008 (W3C) MIT, ERCIM, Keio University - See tidy.h for the copyright notice. - Given an input source, it returns a sequence of tokens. - - GetToken(source) gets the next token - UngetToken(source) provides one level undo - - The tags include an attribute list: - - - linked list of attribute/value nodes - - each node has 2 NULL-terminated strings. - - entities are replaced in attribute values - - white space is compacted if not in preformatted mode - If not in preformatted mode then leading white space - is discarded and subsequent white space sequences - compacted to single space characters. - - If XmlTags is no then Tag names are folded to upper - case and attribute names to lower case. - - Not yet done: - - Doctype subset and marked sections -*/ +/**************************************************************************//** + * @file + * Lexer for HTML and XML Parsers. + * + * Given an input source, it returns a sequence of tokens. + * + * GetToken(source) gets the next token + * UngetToken(source) provides one level undo + * + * The tags include an attribute list: + * + * - linked list of attribute/value nodes + * - each node has 2 NULL-terminated strings. + * - entities are replaced in attribute values + * + * white space is compacted if not in preformatted mode + * If not in preformatted mode then leading white space + * is discarded and subsequent white space sequences + * compacted to single space characters. + * + * If XmlTags is no then Tag names are folded to upper + * case and attribute names to lower case. + * + * Not yet done: + * - Doctype subset and marked sections + * + * @author HTACG, et al (consult git log) + * + * @copyright + * (c) 1998-2021 (W3C) MIT, ERCIM, Keio University, and HTACG. + * See tidy.h for the copyright notice. + * @par + * All Rights Reserved. + * @par + * See `tidy.h` for the complete license. + * + * @date Additional updates: consult git log + * + ******************************************************************************/ #ifdef __cplusplus extern "C" { @@ -35,8 +48,23 @@ extern "C" { #include "forward.h" -/* lexer character types -*/ +/** @addtogroup internal_api */ +/** @{ */ + + +/***************************************************************************//** + ** @defgroup lexer_h HTML and XML Lexing + ** + ** These functions and structures form the internal API for document + ** lexing. + ** + ** @{ + ******************************************************************************/ + + +/** + * Lexer character types. + */ #define digit 1u #define letter 2u #define namechar 4u @@ -47,8 +75,9 @@ extern "C" { #define digithex 128u -/* node->type is one of these values -*/ +/** + * node->type is one of these values + */ typedef enum { RootNode, @@ -68,9 +97,9 @@ typedef enum } NodeType; - -/* lexer GetToken states -*/ +/** + * Lexer GetToken() states. + */ typedef enum { LEX_CONTENT, @@ -88,7 +117,10 @@ typedef enum LEX_XMLDECL } LexerState; -/* ParseDocTypeDecl state constants */ + +/** + * ParseDocTypeDecl state constants. + */ typedef enum { DT_INTERMEDIATE, @@ -98,67 +130,43 @@ typedef enum DT_INTSUBSET } ParseDocTypeDeclState; -/* content model shortcut encoding - - Descriptions are tentative. -*/ +/** + * Content model shortcut encoding. + * Descriptions are tentative. + */ #define CM_UNKNOWN 0 -/* Elements with no content. Map to HTML specification. */ -#define CM_EMPTY (1 << 0) -/* Elements that appear outside of "BODY". */ -#define CM_HTML (1 << 1) -/* Elements that can appear within HEAD. */ -#define CM_HEAD (1 << 2) -/* HTML "block" elements. */ -#define CM_BLOCK (1 << 3) -/* HTML "inline" elements. */ -#define CM_INLINE (1 << 4) -/* Elements that mark list item ("LI"). */ -#define CM_LIST (1 << 5) -/* Elements that mark definition list item ("DL", "DT"). */ -#define CM_DEFLIST (1 << 6) -/* Elements that can appear inside TABLE. */ -#define CM_TABLE (1 << 7) -/* Used for "THEAD", "TFOOT" or "TBODY". */ -#define CM_ROWGRP (1 << 8) -/* Used for "TD", "TH" */ -#define CM_ROW (1 << 9) -/* Elements whose content must be protected against white space movement. - Includes some elements that can found in forms. */ -#define CM_FIELD (1 << 10) -/* Used to avoid propagating inline emphasis inside some elements - such as OBJECT or APPLET. */ -#define CM_OBJECT (1 << 11) -/* Elements that allows "PARAM". */ -#define CM_PARAM (1 << 12) -/* "FRAME", "FRAMESET", "NOFRAMES". Used in ParseFrameSet. */ -#define CM_FRAMES (1 << 13) -/* Heading elements (h1, h2, ...). */ -#define CM_HEADING (1 << 14) -/* Elements with an optional end tag. */ -#define CM_OPT (1 << 15) -/* Elements that use "align" attribute for vertical position. */ -#define CM_IMG (1 << 16) -/* Elements with inline and block model. Used to avoid calling InlineDup. */ -#define CM_MIXED (1 << 17) -/* Elements whose content needs to be indented only if containing one - CM_BLOCK element. */ -#define CM_NO_INDENT (1 << 18) -/* Elements that are obsolete (such as "dir", "menu"). */ -#define CM_OBSOLETE (1 << 19) -/* User defined elements. Used to determine how attributes wihout value - should be printed. */ -#define CM_NEW (1 << 20) -/* Elements that cannot be omitted. */ -#define CM_OMITST (1 << 21) +#define CM_EMPTY (1 << 0) /**< Elements with no content. Map to HTML specification. */ +#define CM_HTML (1 << 1) /**< Elements that appear outside of "BODY". */ +#define CM_HEAD (1 << 2) /**< Elements that can appear within HEAD. */ +#define CM_BLOCK (1 << 3) /**< HTML "block" elements. */ +#define CM_INLINE (1 << 4) /**< HTML "inline" elements. */ +#define CM_LIST (1 << 5) /**< Elements that mark list item ("LI"). */ +#define CM_DEFLIST (1 << 6) /**< Elements that mark definition list item ("DL", "DT"). */ +#define CM_TABLE (1 << 7) /**< Elements that can appear inside TABLE. */ +#define CM_ROWGRP (1 << 8) /**< Used for "THEAD", "TFOOT" or "TBODY". */ +#define CM_ROW (1 << 9) /**< Used for "TD", "TH" */ +#define CM_FIELD (1 << 10) /**< Elements whose content must be protected against white space movement. Includes some elements that can found in forms. */ +#define CM_OBJECT (1 << 11) /**< Used to avoid propagating inline emphasis inside some elements such as OBJECT or APPLET. */ +#define CM_PARAM (1 << 12) /**< Elements that allows "PARAM". */ +#define CM_FRAMES (1 << 13) /**< "FRAME", "FRAMESET", "NOFRAMES". Used in ParseFrameSet. */ +#define CM_HEADING (1 << 14) /**< Heading elements (h1, h2, ...). */ +#define CM_OPT (1 << 15) /**< Elements with an optional end tag. */ +#define CM_IMG (1 << 16) /**< Elements that use "align" attribute for vertical position. */ +#define CM_MIXED (1 << 17) /**< Elements with inline and block model. Used to avoid calling InlineDup. */ +#define CM_NO_INDENT (1 << 18) /**< Elements whose content needs to be indented only if containing one CM_BLOCK element. */ +#define CM_OBSOLETE (1 << 19) /**< Elements that are obsolete (such as "dir", "menu"). */ +#define CM_NEW (1 << 20) /**< User defined elements. Used to determine how attributes without value should be printed. */ +#define CM_OMITST (1 << 21) /**< Elements that cannot be omitted. */ -/* If the document uses just HTML 2.0 tags and attributes described -** it as HTML 2.0 Similarly for HTML 3.2 and the 3 flavors of HTML 4.0. -** If there are proprietary tags and attributes then describe it as -** HTML Proprietary. If it includes the xml-lang or xmlns attributes -** but is otherwise HTML 2.0, 3.2 or 4.0 then describe it as one of the -** flavors of Voyager (strict, loose or frameset). -*/ + +/** + * If the document uses just HTML 2.0 tags and attributes described + * it is HTML 2.0. Similarly for HTML 3.2 and the 3 flavors of HTML 4.0. + * If there are proprietary tags and attributes then describe it as + * HTML Proprietary. If it includes the xml-lang or xmlns attributes + * but is otherwise HTML 2.0, 3.2 or 4.0 then describe it as one of the + * flavors of Voyager (strict, loose or frameset). + */ /* unknown */ #define xxxx 0u @@ -220,8 +228,10 @@ typedef enum /* all proprietary types */ #define VERS_PROPRIETARY (VERS_NETSCAPE|VERS_MICROSOFT|VERS_SUN) -/* Linked list of class names and styles -*/ + +/** + * Linked list of class names and styles + */ struct _Style; typedef struct _Style TagStyle; @@ -234,8 +244,9 @@ struct _Style }; -/* Linked list of style properties -*/ +/** + * Linked list of style properties + */ struct _StyleProp; typedef struct _StyleProp StyleProp; @@ -247,11 +258,9 @@ struct _StyleProp }; - - -/* Attribute/Value linked list node -*/ - +/** + * Attribute/Value linked list node + */ struct _AttVal { AttVal* next; @@ -264,93 +273,89 @@ struct _AttVal }; - -/* - Mosaic handles inlines via a separate stack from other elements - We duplicate this to recover from inline markup errors such as: - - italic text -

more italic text normal text - - which for compatibility with Mosaic is mapped to: - - italic text -

more italic text normal text - - Note that any inline end tag pop's the effect of the current - inline start tag, so that pop's in the above example. +/** + * Mosaic handles inlines via a separate stack from other elements + * We duplicate this to recover from inline markup errors such as: + * ~~~ + * italic text + *

more italic text normal text + * ~~~ + * which for compatibility with Mosaic is mapped to: + * ~~~ + * italic text + *

more italic text normal text + * ~~~ + * Note that any inline end tag pop's the effect of the current + * inline start tag, so that `` pop's `` in the above example. */ struct _IStack { IStack* next; - const Dict* tag; /* tag's dictionary definition */ - tmbstr element; /* name (NULL for text nodes) */ + const Dict* tag; /**< tag's dictionary definition */ + tmbstr element; /**< name (NULL for text nodes) */ AttVal* attributes; }; -/* HTML/XHTML/XML Element, Comment, PI, DOCTYPE, XML Decl, -** etc. etc. -*/ - +/** + * HTML/XHTML/XML Element, Comment, PI, DOCTYPE, XML Decl, etc., etc. + */ struct _Node { - Node* parent; /* tree structure */ + Node* parent; /**< tree structure */ Node* prev; Node* next; Node* content; Node* last; AttVal* attributes; - const Dict* was; /* old tag when it was changed */ - const Dict* tag; /* tag's dictionary definition */ + const Dict* was; /**< old tag when it was changed */ + const Dict* tag; /**< tag's dictionary definition */ - tmbstr element; /* name (NULL for text nodes) */ + tmbstr element; /**< name (NULL for text nodes) */ - uint start; /* start of span onto text array */ - uint end; /* end of span onto text array */ - NodeType type; /* TextNode, StartTag, EndTag etc. */ + uint start; /**< start of span onto text array */ + uint end; /**< end of span onto text array */ + NodeType type; /**< TextNode, StartTag, EndTag etc. */ - uint line; /* current line of document */ - uint column; /* current column of document */ + uint line; /**< current line of document */ + uint column; /**< current column of document */ - Bool closed; /* true if closed by explicit end tag */ - Bool implicit; /* true if inferred */ - Bool linebreak; /* true if followed by a line break */ + Bool closed; /**< true if closed by explicit end tag */ + Bool implicit; /**< true if inferred */ + Bool linebreak; /**< true if followed by a line break */ }; -/* - The following are private to the lexer - Use NewLexer() to create a lexer, and - FreeLexer() to free it. -*/ - +/** + * The following are private to the lexer. + * Use `NewLexer()` to create a lexer, and `FreeLexer()` to free it. + */ struct _Lexer { - uint lines; /* lines seen */ - uint columns; /* at start of current token */ - Bool waswhite; /* used to collapse contiguous white space */ - Bool pushed; /* true after token has been pushed back */ - Bool insertspace; /* when space is moved after end tag */ - Bool excludeBlocks; /* Netscape compatibility */ - Bool exiled; /* true if moved out of table */ - Bool isvoyager; /* true if xmlns attribute on html element */ - uint versions; /* bit vector of HTML versions */ - uint doctype; /* version as given by doctype (if any) */ - uint versionEmitted; /* version of doctype emitted */ - Bool bad_doctype; /* e.g. if html or PUBLIC is missing */ - uint txtstart; /* start of current node */ - uint txtend; /* end of current node */ - LexerState state; /* state of lexer's finite state machine */ + uint lines; /**< lines seen */ + uint columns; /**< at start of current token */ + Bool waswhite; /**< used to collapse contiguous white space */ + Bool pushed; /**< true after token has been pushed back */ + Bool insertspace; /**< when space is moved after end tag */ + Bool excludeBlocks; /**< Netscape compatibility */ + Bool exiled; /**< true if moved out of table */ + Bool isvoyager; /**< true if xmlns attribute on html element (i.e., "Voyager" was the W3C codename for XHTML). */ + uint versions; /**< bit vector of HTML versions */ + uint doctype; /**< version as given by doctype (if any) */ + uint versionEmitted; /**< version of doctype emitted */ + Bool bad_doctype; /**< e.g. if html or PUBLIC is missing */ + uint txtstart; /**< start of current node */ + uint txtend; /**< end of current node */ + LexerState state; /**< state of lexer's finite state machine */ - Node* token; /* last token returned by GetToken() */ - Node* itoken; /* last duplicate inline returned by GetToken() */ - Node* root; /* remember root node of the document */ - Node* parent; /* remember parent node for CDATA elements */ - - Bool seenEndBody; /* true if a tag has been encountered */ - Bool seenEndHtml; /* true if a tag has been encountered */ + Node* token; /**< last token returned by GetToken() */ + Node* itoken; /**< last duplicate inline returned by GetToken() */ + Node* root; /**< remember root node of the document */ + Node* parent; /**< remember parent node for CDATA elements */ + + Bool seenEndBody; /**< true if a `` tag has been encountered */ + Bool seenEndHtml; /**< true if a `` tag has been encountered */ /* Lexer character buffer @@ -361,33 +366,57 @@ struct _Lexer lexsize must be reset for each file. */ - tmbstr lexbuf; /* MB character buffer */ - uint lexlength; /* allocated */ - uint lexsize; /* used */ + tmbstr lexbuf; /**< MB character buffer */ + uint lexlength; /**< allocated */ + uint lexsize; /**< used */ /* Inline stack for compatibility with Mosaic */ - Node* inode; /* for deferring text node */ - IStack* insert; /* for inferring inline tags */ + Node* inode; /**< for deferring text node */ + IStack* insert; /**< for inferring inline tags */ IStack* istack; - uint istacklength; /* allocated */ - uint istacksize; /* used */ - uint istackbase; /* start of frame */ + uint istacklength; /**< allocated */ + uint istacksize; /**< used */ + uint istackbase; /**< start of frame */ - TagStyle *styles; /* used for cleaning up presentation markup */ + TagStyle *styles; /**< used for cleaning up presentation markup */ - TidyAllocator* allocator; /* allocator */ + TidyAllocator* allocator; /**< allocator */ }; -/* Lexer Functions -*/ +/** + * modes for GetToken() + * + * MixedContent -- for elements which don't accept PCDATA + * Preformatted -- white space preserved as is + * IgnoreMarkup -- for CDATA elements such as script, style + */ +typedef enum +{ + IgnoreWhitespace, + MixedContent, + Preformatted, + IgnoreMarkup, + OtherNamespace, + CdataContent +} GetTokenMode; -/* choose what version to use for new doctype */ + +/** @name Lexer Functions + * @{ + */ + + +/** + * Choose what version to use for new doctype + */ TY_PRIVATE int TY_(HTMLVersion)( TidyDocImpl* doc ); -/* everything is allowed in proprietary version of HTML */ -/* this is handled here rather than in the tag/attr dicts */ +/** + * Everything is allowed in proprietary version of HTML. + * This is handled here rather than in the tag/attr dicts + */ TY_PRIVATE void TY_(ConstrainVersion)( TidyDocImpl* doc, uint vers ); TY_PRIVATE Bool TY_(IsWhite)(uint c); @@ -399,7 +428,6 @@ TY_PRIVATE Bool TY_(IsNamechar)(uint c); TY_PRIVATE Bool TY_(IsXMLLetter)(uint c); TY_PRIVATE Bool TY_(IsXMLNamechar)(uint c); -/* Bool IsLower(uint c); */ TY_PRIVATE Bool TY_(IsUpper)(uint c); TY_PRIVATE uint TY_(ToLower)(uint c); TY_PRIVATE uint TY_(ToUpper)(uint c); @@ -407,60 +435,82 @@ TY_PRIVATE uint TY_(ToUpper)(uint c); TY_PRIVATE Lexer* TY_(NewLexer)( TidyDocImpl* doc ); TY_PRIVATE void TY_(FreeLexer)( TidyDocImpl* doc ); -/* store character c as UTF-8 encoded byte stream */ + +/** + * Store character c as UTF-8 encoded byte stream + */ TY_PRIVATE void TY_(AddCharToLexer)( Lexer *lexer, uint c ); -/* - Used for elements and text nodes - element name is NULL for text nodes - start and end are offsets into lexbuf - which contains the textual content of - all elements in the parse tree. - parent and content allow traversal - of the parse tree in any direction. - attributes are represented as a linked - list of AttVal nodes which hold the - strings for attribute/value pairs. +/** + * Used for elements and text nodes. + * - Element name is NULL for text nodes. + * - start and end are offsets into lexbuf, + * which contains the textual content of + * all elements in the parse tree. + * - parent and content allow traversal + * of the parse tree in any direction. + * - attributes are represented as a linked + * list of AttVal nodes which hold the + * strings for attribute/value pairs. */ TY_PRIVATE Node* TY_(NewNode)( TidyAllocator* allocator, Lexer* lexer ); -/* used to clone heading nodes when split by an

*/ +/** + * Used to clone heading nodes when split by an `
` + */ TY_PRIVATE Node* TY_(CloneNode)( TidyDocImpl* doc, Node *element ); -/* free node's attributes */ + +/** + * Free node's attributes + */ TY_PRIVATE void TY_(FreeAttrs)( TidyDocImpl* doc, Node *node ); -/* doesn't repair attribute list linkage */ + +/** + * Doesn't repair attribute list linkage + */ TY_PRIVATE void TY_(FreeAttribute)( TidyDocImpl* doc, AttVal *av ); -/* detach attribute from node */ + +/** + * Detach attribute from node + */ TY_PRIVATE void TY_(DetachAttribute)( Node *node, AttVal *attr ); -/* detach attribute from node then free it -*/ + +/** + * Detach attribute from node then free it. + */ TY_PRIVATE void TY_(RemoveAttribute)( TidyDocImpl* doc, Node *node, AttVal *attr ); -/* - Free document nodes by iterating through peers and recursing - through children. Set next to NULL before calling FreeNode() - to avoid freeing peer nodes. Doesn't patch up prev/next links. + +/** + * Free document nodes by iterating through peers and recursing + * through children. Set `next` to `NULL` before calling `FreeNode()` + * to avoid freeing peer nodes. Doesn't patch up prev/next links. */ TY_PRIVATE void TY_(FreeNode)( TidyDocImpl* doc, Node *node ); + TY_PRIVATE Node* TY_(TextToken)( Lexer *lexer ); -/* used for creating preformatted text from Word2000 */ + +/** + * Used for creating preformatted text from Word2000. + */ TY_PRIVATE Node* TY_(NewLineNode)( Lexer *lexer ); -/* used for adding a   for Word2000 */ + +/** + * Used for adding a   for Word2000. + */ TY_PRIVATE Node* TY_(NewLiteralTextNode)(Lexer *lexer, ctmbstr txt ); -TY_PRIVATE void TY_(AddStringLiteral)( Lexer* lexer, ctmbstr str ); -/* TY_PRIVATE void AddStringLiteralLen( Lexer* lexer, ctmbstr str, int len ); */ -/* find element */ +TY_PRIVATE void TY_(AddStringLiteral)( Lexer* lexer, ctmbstr str ); TY_PRIVATE Node* TY_(FindDocType)( TidyDocImpl* doc ); TY_PRIVATE Node* TY_(FindHTML)( TidyDocImpl* doc ); TY_PRIVATE Node* TY_(FindHEAD)( TidyDocImpl* doc ); @@ -468,10 +518,16 @@ TY_PRIVATE Node* TY_(FindTITLE)(TidyDocImpl* doc); TY_PRIVATE Node* TY_(FindBody)( TidyDocImpl* doc ); TY_PRIVATE Node* TY_(FindXmlDecl)(TidyDocImpl* doc); -/* Returns containing block element, if any */ + +/** + * Returns containing block element, if any + */ TY_PRIVATE Node* TY_(FindContainer)( Node* node ); -/* add meta element for Tidy */ + +/** + * Add meta element for Tidy. + */ TY_PRIVATE Bool TY_(AddGenerator)( TidyDocImpl* doc ); TY_PRIVATE uint TY_(ApparentVersion)( TidyDocImpl* doc ); @@ -485,118 +541,209 @@ TY_PRIVATE Bool TY_(WarnMissingSIInEmittedDocType)( TidyDocImpl* doc ); TY_PRIVATE Bool TY_(SetXHTMLDocType)( TidyDocImpl* doc ); -/* fixup doctype if missing */ +/** + * Fixup doctype if missing. + */ TY_PRIVATE Bool TY_(FixDocType)( TidyDocImpl* doc ); -/* ensure XML document starts with */ -/* add encoding attribute if not using ASCII or UTF-8 output */ + +/** + * Ensure XML document starts with ,and + * add encoding attribute if not using ASCII or UTF-8 output. + */ TY_PRIVATE Bool TY_(FixXmlDecl)( TidyDocImpl* doc ); + TY_PRIVATE Node* TY_(InferredTag)(TidyDocImpl* doc, TidyTagId id); TY_PRIVATE void TY_(UngetToken)( TidyDocImpl* doc ); - -/* - modes for GetToken() - - MixedContent -- for elements which don't accept PCDATA - Preformatted -- white space preserved as is - IgnoreMarkup -- for CDATA elements such as script, style -*/ -typedef enum -{ - IgnoreWhitespace, - MixedContent, - Preformatted, - IgnoreMarkup, - OtherNamespace, - CdataContent -} GetTokenMode; - TY_PRIVATE Node* TY_(GetToken)( TidyDocImpl* doc, GetTokenMode mode ); TY_PRIVATE void TY_(InitMap)(void); -/* create a new attribute */ +/** + * Create a new attribute. + */ TY_PRIVATE AttVal* TY_(NewAttribute)( TidyDocImpl* doc ); -/* create a new attribute with given name and value */ + +/** + * Create a new attribute with given name and value. + */ TY_PRIVATE AttVal* TY_(NewAttributeEx)( TidyDocImpl* doc, ctmbstr name, ctmbstr value, int delim ); -/* insert attribute at the end of attribute list of a node */ + +/** + * Insert attribute at the end of attribute list of a node. + */ TY_PRIVATE void TY_(InsertAttributeAtEnd)( Node *node, AttVal *av ); -/* insert attribute at the start of attribute list of a node */ +/** + * Insert attribute at the start of attribute list of a node. + */ TY_PRIVATE void TY_(InsertAttributeAtStart)( Node *node, AttVal *av ); -/************************************* - In-line Stack functions -*************************************/ + +/** @} + * @name Inline Stack Functions + * @{ + */ -/* duplicate attributes */ +/** + * Duplicate attributes. + */ TY_PRIVATE AttVal* TY_(DupAttrs)( TidyDocImpl* doc, AttVal* attrs ); -/* - push a copy of an inline node onto stack - but don't push if implicit or OBJECT or APPLET - (implicit tags are ones generated from the istack) - One issue arises with pushing inlines when - the tag is already pushed. For instance: - -

text -

more text - - Shouldn't be mapped to - -

text

-

more text -*/ +/** + * Push a copy of an inline node onto stack, but don't push if + * implicit or OBJECT or APPLET (implicit tags are ones generated + * from the istack). + * + * One issue arises with pushing inlines when the tag is already pushed. + * For instance: + * ~~~ + *

text + *

more text + * ~~~ + * Shouldn't be mapped to + * ~~~ + *

text

+ *

more text + * ~~~ + */ TY_PRIVATE void TY_(PushInline)( TidyDocImpl* doc, Node* node ); -/* pop inline stack */ + +/** + * Pop inline stack. + */ TY_PRIVATE void TY_(PopInline)( TidyDocImpl* doc, Node* node ); + TY_PRIVATE Bool TY_(IsPushed)( TidyDocImpl* doc, Node* node ); TY_PRIVATE Bool TY_(IsPushedLast)( TidyDocImpl* doc, Node *element, Node *node ); -/* - This has the effect of inserting "missing" inline - elements around the contents of blocklevel elements - such as P, TD, TH, DIV, PRE etc. This procedure is - called at the start of ParseBlock. when the inline - stack is not empty, as will be the case in: -

italic heading

- - which is then treated as equivalent to - -

italic heading

- - This is implemented by setting the lexer into a mode - where it gets tokens from the inline stack rather than - from the input stream. -*/ +/** + * This has the effect of inserting "missing" inline elements around the + * contents of blocklevel elements such as P, TD, TH, DIV, PRE etc. This + * procedure is called at the start of `ParseBlock`, when the inline + * stack is not empty, as will be the case in: + * ~~~ + *

italic heading

 + * ~~~ + * which is then treated as equivalent to + * ~~~ + *

italic heading

+ * ~~~ + * This is implemented by setting the lexer into a mode where it gets + * tokens from the inline stack rather than from the input stream. + */ TY_PRIVATE int TY_(InlineDup)( TidyDocImpl* doc, Node *node ); -/* - defer duplicates when entering a table or other - element where the inlines shouldn't be duplicated -*/ + +/** + * Fefer duplicates when entering a table or other + * element where the inlines shouldn't be duplicated. + */ TY_PRIVATE void TY_(DeferDup)( TidyDocImpl* doc ); + + TY_PRIVATE Node* TY_(InsertedToken)( TidyDocImpl* doc ); -/* stack manipulation for inline elements */ +/** + * Stack manipulation for inline elements + */ TY_PRIVATE Bool TY_(SwitchInline)( TidyDocImpl* doc, Node* element, Node* node ); + + TY_PRIVATE Bool TY_(InlineDup1)( TidyDocImpl* doc, Node* node, Node* element ); + +/** @} + * @name Generic stack of nodes. + * @{ + */ + + +/** + * This typedef represents a stack of addresses to nodes. Tidy uses these to + * try to limit recursion by pushing nodes to a stack when possible instead + * of recursing. + */ +typedef struct _Stack { + int top; /**< Current top position. */ + unsigned capacity; /**< Current capacity. Can be expanded. */ + Node **firstNode; /** A pointer to the first pointer to a Node in an array of node addresses. */ + TidyAllocator* allocator; /**< Tidy's allocator, used at instantiation and expanding. */ +} Stack; + + +/** + * Create a new stack with a given starting capacity. If memory allocation + * fails, then the allocator will panic the program automatically. + */ +TY_PRIVATE Stack* TY_(newStack)(TidyDocImpl *doc, uint capacity); + + +/** + * Increase the stack size. This will be called automatically when the + * current stack is full. If memory allocation fails, then the allocator + * will panic the program automatically. + */ +TY_PRIVATE void TY_(growStack)(Stack *stack); + + +/** + * Stack is full when top is equal to the last index. + */ +TY_PRIVATE Bool TY_(stackFull)(Stack *stack); + + +/** + * Stack is empty when top is equal to -1 + */ +TY_PRIVATE Bool TY_(stackEmpty)(Stack *stack); + + +/** + * Push an item to the stack. + */ +TY_PRIVATE void TY_(push)(Stack *stack, Node *node); + + +/** + * Pop an item from the stack. + */ +TY_PRIVATE Node* TY_(pop)(Stack *stack); + + +/** + * Peek at the stack. + */ +TY_PRIVATE Node* TY_(peek)(Stack *stack); + +/** + * Frees the stack when done. + */ +TY_PRIVATE void TY_(freeStack)(Stack *stack); + + +/** @} + */ + + #ifdef __cplusplus } #endif +/** @} end parser_h group */ +/** @} end internal_api group */ + #endif /* __LEXER_H__ */