htmltidy/src/clean.c

/*
  clean.c -- clean up misuse of presentation markup

  (c) 1998-2008 (W3C) MIT, ERCIM, Keio University
  See tidy.h for the copyright notice.

  Filters from other formats such as Microsoft Word
  often make excessive use of presentation markup such
  as font tags, B, I, and the align attribute. By applying
  a set of production rules, it is straight forward to
  transform this to use CSS.

  Some rules replace some of the children of an element by
  style properties on the element, e.g.

  <p><b>...</b></p> -> <p style="font-weight: bold">...</p>

  Such rules are applied to the element's content and then
  to the element itself until none of the rules more apply.
  Having applied all the rules to an element, it will have
  a style attribute with one or more properties. 

  Other rules strip the element they apply to, replacing
  it by style properties on the contents, e.g.
  
  <dir><li><p>...</li></dir> -> <p style="margin-left 1em">...
      
  These rules are applied to an element before processing
  its content and replace the current element by the first
  element in the exposed content.

  After applying both sets of rules, you can replace the
  style attribute by a class value and style rule in the
  document head. To support this, an association of styles
  and class names is built.

  A naive approach is to rely on string matching to test
  when two property lists are the same. A better approach
  would be to first sort the properties before matching.

*/

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#include "tidy-int.h"
#include "clean.h"
#include "lexer.h"
#include "parser.h"
#include "attrs.h"
#include "message.h"
#include "tmbstr.h"
#include "utf8.h"

static Node* CleanNode( TidyDocImpl* doc, Node *node );

static void RenameElem( TidyDocImpl* doc, Node* node, TidyTagId tid )
{
    const Dict* dict = TY_(LookupTagDef)( tid );
    TidyDocFree( doc, node->element );
    node->element = TY_(tmbstrdup)( doc->allocator, dict->name );
    node->tag = dict;
}

static void FreeStyleProps(TidyDocImpl* doc, StyleProp *props)
{
    StyleProp *next;

    while (props)
    {
        next = props->next;
        TidyDocFree(doc, props->name);
        TidyDocFree(doc, props->value);
        TidyDocFree(doc, props);
        props = next;
    }
}

static StyleProp *InsertProperty( TidyDocImpl* doc, StyleProp* props, ctmbstr name, ctmbstr value )
{
    StyleProp *first, *prev, *prop;
    int cmp;

    prev = NULL;
    first = props;

    while (props)
    {
        cmp = TY_(tmbstrcmp)(props->name, name);

        if (cmp == 0)
        {
            /* this property is already defined, ignore new value */
            return first;
        }

        if (cmp > 0)
        {
            /* insert before this */

            prop = (StyleProp *)TidyDocAlloc(doc, sizeof(StyleProp));
            prop->name = TY_(tmbstrdup)(doc->allocator, name);
            prop->value = TY_(tmbstrdup)(doc->allocator, value);
            prop->next = props;

            if (prev)
                prev->next = prop;
            else
                first = prop;

            return first;
        }

        prev = props;
        props = props->next;
    }

    prop = (StyleProp *)TidyDocAlloc(doc, sizeof(StyleProp));
    prop->name = TY_(tmbstrdup)(doc->allocator, name);
    prop->value = TY_(tmbstrdup)(doc->allocator, value);
    prop->next = NULL;

    if (prev)
        prev->next = prop;
    else
        first = prop;

    return first;
}

/*
 Create sorted linked list of properties from style string
 It temporarily places nulls in place of ':' and ';' to
 delimit the strings for the property name and value.
 Some systems don't allow you to NULL literal strings,
 so to avoid this, a copy is made first.
*/
static StyleProp* CreateProps( TidyDocImpl* doc, StyleProp* prop, ctmbstr style )
{
    tmbstr name, value = NULL, name_end, value_end, line;
    Bool more;

    line = TY_(tmbstrdup)(doc->allocator, style);
    name = line;

    while (*name)
    {
        while (*name == ' ')
            ++name;

        name_end = name;

        while (*name_end)
        {
            if (*name_end == ':')
            {
                value = name_end + 1;
                break;
            }

            ++name_end;
        }

        if (*name_end != ':')
            break;

        while ( value && *value == ' ')
            ++value;

        value_end = value;
        more = no;

        while (*value_end)
        {
            if (*value_end == ';')
            {
                more = yes;
                break;
            }

            ++value_end;
        }

        *name_end = '\0';
        *value_end = '\0';

        prop = InsertProperty(doc, prop, name, value);
        *name_end = ':';

        if (more)
        {
            *value_end = ';';
            name = value_end + 1;
            continue;
        }

        break;
    }

    TidyDocFree(doc, line);  /* free temporary copy */
    return prop;
}

static tmbstr CreatePropString(TidyDocImpl* doc, StyleProp *props)
{
    tmbstr style, p, s;
    uint len;
    StyleProp *prop;

    /* compute length */

    for (len = 0, prop = props; prop; prop = prop->next)
    {
        len += TY_(tmbstrlen)(prop->name) + 2;
        if (prop->value)
            len += TY_(tmbstrlen)(prop->value) + 2;
    }

    style = (tmbstr) TidyDocAlloc(doc, len+1);
    style[0] = '\0';

    for (p = style, prop = props; prop; prop = prop->next)
    {
        s = prop->name;

        while((*p++ = *s++))
            continue;

        if (prop->value)
        {
            *--p = ':';
            *++p = ' ';
            ++p;

            s = prop->value;
            while((*p++ = *s++))
                continue;
        }
        if (prop->next == NULL)
            break;

        *--p = ';';
        *++p = ' ';
        ++p;
    }

    return style;
}

/*
  create string with merged properties
static tmbstr AddProperty( ctmbstr style, ctmbstr property )
{
    tmbstr line;
    StyleProp *prop;

    prop = CreateProps(doc, NULL, style);
    prop = CreateProps(doc, prop, property);
    line = CreatePropString(doc, prop);
    FreeStyleProps(doc, prop);
    return line;
}
*/

void TY_(FreeStyles)( TidyDocImpl* doc )
{
    Lexer* lexer = doc->lexer;
    if ( lexer )
    {
        TagStyle *style, *next;
        for ( style = lexer->styles; style; style = next )
        {
            next = style->next;
            TidyDocFree( doc, style->tag );
            TidyDocFree( doc, style->tag_class );
            TidyDocFree( doc, style->properties );
            TidyDocFree( doc, style );
        }
    }
}

static tmbstr GensymClass( TidyDocImpl* doc )
{
    tmbchar buf[512];  /* CSSPrefix is limited to 256 characters */
    ctmbstr pfx = cfgStr(doc, TidyCSSPrefix);
    if ( pfx == NULL || *pfx == 0 )
      pfx = "c";

    TY_(tmbsnprintf)(buf, sizeof(buf), "%s%u", pfx, ++doc->nClassId );
    return TY_(tmbstrdup)(doc->allocator, buf);
}

static ctmbstr FindStyle( TidyDocImpl* doc, ctmbstr tag, ctmbstr properties )
{
    Lexer* lexer = doc->lexer;
    TagStyle* style;

    for (style = lexer->styles; style; style=style->next)
    {
        if (TY_(tmbstrcmp)(style->tag, tag) == 0 &&
            TY_(tmbstrcmp)(style->properties, properties) == 0)
            return style->tag_class;
    }

    style = (TagStyle *)TidyDocAlloc( doc, sizeof(TagStyle) );
    style->tag = TY_(tmbstrdup)(doc->allocator, tag);
    style->tag_class = GensymClass( doc );
    style->properties = TY_(tmbstrdup)( doc->allocator, properties );
    style->next = lexer->styles;
    lexer->styles = style;
    return style->tag_class;
}

/*
 Add class="foo" to node
*/
static void AddClass( TidyDocImpl* doc, Node* node, ctmbstr classname )
{
    AttVal *classattr = TY_(AttrGetById)(node, TidyAttr_CLASS);;

    /*
     if there already is a class attribute
     then append class name after a space.
    */
    if (classattr)
        TY_(AppendToClassAttr)( doc, classattr, classname );
    else /* create new class attribute */
        TY_(AddAttribute)( doc, node, "class", classname );
}

void TY_(AddStyleAsClass)( TidyDocImpl* doc, Node *node, ctmbstr stylevalue )
{
    ctmbstr classname;

    classname = FindStyle( doc, node->element, stylevalue );
    AddClass( doc, node, classname);
}

/*
 Find style attribute in node, and replace it
 by corresponding class attribute. Search for
 class in style dictionary otherwise gensym
 new class and add to dictionary.

 Assumes that node doesn't have a class attribute
*/
static void Style2Rule( TidyDocImpl* doc, Node *node)
{
    AttVal *styleattr, *classattr;
    ctmbstr classname;

    styleattr = TY_(AttrGetById)(node, TidyAttr_STYLE);

    if (styleattr)
    {
        /* fix for http://tidy.sf.net/bug/850215 */
        if (!styleattr->value)
        {
            TY_(RemoveAttribute)(doc, node, styleattr);
            return;
        }

        classname = FindStyle( doc, node->element, styleattr->value );
        classattr = TY_(AttrGetById)(node, TidyAttr_CLASS);

        /*
         if there already is a class attribute
         then append class name after an underscore
        */
        if (classattr)
        {
            TY_(AppendToClassAttr)( doc, classattr, classname );
            TY_(RemoveAttribute)( doc, node, styleattr );
        }
        else /* reuse style attribute for class attribute */
        {
            TidyDocFree(doc, styleattr->attribute);
            TidyDocFree(doc, styleattr->value);
            styleattr->attribute = TY_(tmbstrdup)(doc->allocator, "class");
            styleattr->value = TY_(tmbstrdup)(doc->allocator, classname);
        }
    }
}

static void AddColorRule( Lexer* lexer, ctmbstr selector, ctmbstr color )
{
    if ( selector && color )
    {
        TY_(AddStringLiteral)(lexer, selector);
        TY_(AddStringLiteral)(lexer, " { color: ");
        TY_(AddStringLiteral)(lexer, color);
        TY_(AddStringLiteral)(lexer, " }\n");
    }
}

/*
 move presentation attribs from body to style element

 background="foo" ->  body { background-image: url(foo) }
 bgcolor="foo"    ->  body { background-color: foo }
 text="foo"       ->  body { color: foo }
 link="foo"       ->  :link { color: foo }
 vlink="foo"      ->  :visited { color: foo }
 alink="foo"      ->  :active { color: foo }
*/
static void CleanBodyAttrs( TidyDocImpl* doc, Node* body )
{
    Lexer* lexer  = doc->lexer;
    tmbstr bgurl   = NULL;
    tmbstr bgcolor = NULL;
    tmbstr color   = NULL;
    AttVal* attr;
    
    if (NULL != (attr = TY_(AttrGetById)(body, TidyAttr_BACKGROUND)))
    {
        bgurl = attr->value;
        attr->value = NULL;
        TY_(RemoveAttribute)( doc, body, attr );
    }

    if (NULL != (attr = TY_(AttrGetById)(body, TidyAttr_BGCOLOR)))
    {
        bgcolor = attr->value;
        attr->value = NULL;
        TY_(RemoveAttribute)( doc, body, attr );
    }

    if (NULL != (attr = TY_(AttrGetById)(body, TidyAttr_TEXT)))
    {
        color = attr->value;
        attr->value = NULL;
        TY_(RemoveAttribute)( doc, body, attr );
    }

    if ( bgurl || bgcolor || color )
    {
        TY_(AddStringLiteral)(lexer, " body {\n");
        if (bgurl)
        {
            TY_(AddStringLiteral)(lexer, "  background-image: url(");
            TY_(AddStringLiteral)(lexer, bgurl);
            TY_(AddStringLiteral)(lexer, ");\n");
            TidyDocFree(doc, bgurl);
        }
        if (bgcolor)
        {
            TY_(AddStringLiteral)(lexer, "  background-color: ");
            TY_(AddStringLiteral)(lexer, bgcolor);
            TY_(AddStringLiteral)(lexer, ";\n");
            TidyDocFree(doc, bgcolor);
        }
        if (color)
        {
            TY_(AddStringLiteral)(lexer, "  color: ");
            TY_(AddStringLiteral)(lexer, color);
            TY_(AddStringLiteral)(lexer, ";\n");
            TidyDocFree(doc, color);
        }

        TY_(AddStringLiteral)(lexer, " }\n");
    }

    if (NULL != (attr = TY_(AttrGetById)(body, TidyAttr_LINK)))
    {
        AddColorRule(lexer, " :link", attr->value);
        TY_(RemoveAttribute)( doc, body, attr );
    }

    if (NULL != (attr = TY_(AttrGetById)(body, TidyAttr_VLINK)))
    {
        AddColorRule(lexer, " :visited", attr->value);
        TY_(RemoveAttribute)( doc, body, attr );
    }

    if (NULL != (attr = TY_(AttrGetById)(body, TidyAttr_ALINK)))
    {
        AddColorRule(lexer, " :active", attr->value);
        TY_(RemoveAttribute)( doc, body, attr );
    }
}

static Bool NiceBody( TidyDocImpl* doc )
{
    Node* node = TY_(FindBody)(doc);
    if (node)
    {
        if (TY_(AttrGetById)(node, TidyAttr_BACKGROUND) ||
            TY_(AttrGetById)(node, TidyAttr_BGCOLOR)    ||
            TY_(AttrGetById)(node, TidyAttr_TEXT)       ||
            TY_(AttrGetById)(node, TidyAttr_LINK)       ||
            TY_(AttrGetById)(node, TidyAttr_VLINK)      ||
            TY_(AttrGetById)(node, TidyAttr_ALINK))
        {
            doc->badLayout |= USING_BODY;
            return no;
        }
    }

    return yes;
}

/* create style element using rules from dictionary */
static void CreateStyleElement( TidyDocImpl* doc )
{
    Lexer* lexer = doc->lexer;
    Node *node, *head, *body;
    TagStyle *style;
    AttVal *av;

    if ( lexer->styles == NULL && NiceBody(doc) )
        return;

    node = TY_(NewNode)( doc->allocator, lexer );
    node->type = StartTag;
    node->implicit = yes;
    node->element = TY_(tmbstrdup)(doc->allocator, "style");
    TY_(FindTag)( doc, node );

    /* insert type attribute */
    av = TY_(NewAttributeEx)( doc, "type", "text/css", '"' );
    TY_(InsertAttributeAtStart)( node, av );

    body = TY_(FindBody)( doc );
    lexer->txtstart = lexer->lexsize;
    if ( body )
        CleanBodyAttrs( doc, body );

    for (style = lexer->styles; style; style = style->next)
    {
        TY_(AddCharToLexer)(lexer, ' ');
        TY_(AddStringLiteral)(lexer, style->tag);
        TY_(AddCharToLexer)(lexer, '.');
        TY_(AddStringLiteral)(lexer, style->tag_class);
        TY_(AddCharToLexer)(lexer, ' ');
        TY_(AddCharToLexer)(lexer, '{');
        TY_(AddStringLiteral)(lexer, style->properties);
        TY_(AddCharToLexer)(lexer, '}');
        TY_(AddCharToLexer)(lexer, '\n');
    }

    lexer->txtend = lexer->lexsize;

    TY_(InsertNodeAtEnd)( node, TY_(TextToken)(lexer) );

    /*
     now insert style element into document head

     doc is root node. search its children for html node
     the head node should be first child of html node
    */
    if ( NULL != (head = TY_(FindHEAD)( doc )) )
        TY_(InsertNodeAtEnd)( head, node );
}


/* ensure bidirectional links are consistent */
void TY_(FixNodeLinks)(Node *node)
{
    Node *child;

    if (node->prev)
        node->prev->next = node;
    else
        node->parent->content = node;

    if (node->next)
        node->next->prev = node;
    else
        node->parent->last = node;

    for (child = node->content; child; child = child->next)
        child->parent = node;
}

/*
 used to strip child of node when
 the node has one and only one child
*/
static void StripOnlyChild(TidyDocImpl* doc, Node *node)
{
    Node *child;

    child = node->content;
    node->content = child->content;
    node->last = child->last;
    child->content = NULL;
    TY_(FreeNode)(doc, child);

    for (child = node->content; child; child = child->next)
        child->parent = node;
}

/*
  used to strip font start and end tags.
  Extricate "element", replace it by its content and delete it.
*/
static void DiscardContainer( TidyDocImpl* doc, Node *element, Node **pnode)
{
    if (element->content)
    {
        Node *node, *parent = element->parent;

        element->last->next = element->next;

        if (element->next)
        {
            element->next->prev = element->last;
        }
        else
            parent->last = element->last;

        if (element->prev)
        {
            element->content->prev = element->prev;
            element->prev->next = element->content;
        }
        else
            parent->content = element->content;

        for (node = element->content; node; node = node->next)
            node->parent = parent;

        *pnode = element->content;

        element->next = element->content = NULL;
        TY_(FreeNode)(doc, element);
    }
    else
    {
        *pnode = TY_(DiscardElement)(doc, element);
    }
}

/*
  Create new string that consists of the
  combined style properties in s1 and s2

  To merge property lists, we build a linked
  list of property/values and insert properties
  into the list in order, merging values for
  the same property name.
*/
static tmbstr MergeProperties( TidyDocImpl* doc, ctmbstr s1, ctmbstr s2 )
{
    tmbstr s;
    StyleProp *prop;

    prop = CreateProps(doc, NULL, s1);
    prop = CreateProps(doc, prop, s2);
    s = CreatePropString(doc, prop);
    FreeStyleProps(doc, prop);
    return s;
}

/*
 Add style property to element, creating style
 attribute as needed and adding ; delimiter
*/
void TY_(AddStyleProperty)(TidyDocImpl* doc, Node *node, ctmbstr property )
{
    AttVal *av = TY_(AttrGetById)(node, TidyAttr_STYLE);

    /* if style attribute already exists then insert property */

    if ( av )
    {
        if (av->value != NULL)
        {
            tmbstr s = MergeProperties( doc, av->value, property );
            TidyDocFree( doc, av->value );
            av->value = s;
        }
        else
        {
            av->value = TY_(tmbstrdup)( doc->allocator, property );
        }
    }
    else /* else create new style attribute */
    {
        av = TY_(NewAttributeEx)( doc, "style", property, '"' );
        TY_(InsertAttributeAtStart)( node, av );
    }
}

static void MergeClasses(TidyDocImpl* doc, Node *node, Node *child)
{
    AttVal *av;
    tmbstr s1, s2, names;

    for (s2 = NULL, av = child->attributes; av; av = av->next)
    {
        if (attrIsCLASS(av))
        {
            s2 = av->value;
            break;
        }
    }

    for (s1 = NULL, av = node->attributes; av; av = av->next)
    {
        if (attrIsCLASS(av))
        {
            s1 = av->value;
            break;
        }
    }

    if (s1)
    {
        if (s2)  /* merge class names from both */
        {
            uint l1, l2;
            l1 = TY_(tmbstrlen)(s1);
            l2 = TY_(tmbstrlen)(s2);
            names = (tmbstr) TidyDocAlloc(doc, l1 + l2 + 2);
            TY_(tmbstrcpy)(names, s1);
            names[l1] = ' ';
            TY_(tmbstrcpy)(names+l1+1, s2);
            TidyDocFree(doc, av->value);
            av->value = names;
        }
    }
    else if (s2)  /* copy class names from child */
    {
        av = TY_(NewAttributeEx)( doc, "class", s2, '"' );
        TY_(InsertAttributeAtStart)( node, av );
    }
}

static void MergeStyles(TidyDocImpl* doc, Node *node, Node *child)
{
    AttVal *av;
    tmbstr s1, s2, style;

    /*
       the child may have a class attribute used
       for attaching styles, if so the class name
       needs to be copied to node's class
    */
    MergeClasses(doc, node, child);

    for (s2 = NULL, av = child->attributes; av; av = av->next)
    {
        if (attrIsSTYLE(av))
        {
            s2 = av->value;
            break;
        }
    }

    for (s1 = NULL, av = node->attributes; av; av = av->next)
    {
        if (attrIsSTYLE(av))
        {
            s1 = av->value;
            break;
        }
    }

    if (s1)
    {
        if (s2)  /* merge styles from both */
        {
            style = MergeProperties(doc, s1, s2);
            TidyDocFree(doc, av->value);
            av->value = style;
        }
    }
    else if (s2)  /* copy style of child */
    {
        av = TY_(NewAttributeEx)( doc, "style", s2, '"' );
        TY_(InsertAttributeAtStart)( node, av );
    }
}

static ctmbstr FontSize2Name(ctmbstr size)
{
    static const ctmbstr sizes[7] =
    {
        "60%", "70%", "80%", NULL,
        "120%", "150%", "200%"
    };

    /* increment of 0.8 */
    static const ctmbstr minussizes[] =
    {
        "100%", "80%", "64%", "51%",
        "40%", "32%", "26%"
    };

    /* increment of 1.2 */
    static const ctmbstr plussizes[] =
    {
        "100%", "120%", "144%", "172%",
        "207%", "248%", "298%"
    };

    if (size[0] == '\0')
        return NULL;

    if ('0' <= size[0] && size[0] <= '6')
    {
        int n = size[0] - '0';
        return sizes[n];
    }

    if (size[0] == '-')
    {
        if ('0' <= size[1] && size[1] <= '6')
        {
            int n = size[1] - '0';
            return minussizes[n];
        }
        return "smaller"; /*"70%"; */
    }

    if ('0' <= size[1] && size[1] <= '6')
    {
        int n = size[1] - '0';
        return plussizes[n];
    }

    return "larger"; /* "140%" */
}

static void AddFontFace( TidyDocImpl* doc, Node *node, ctmbstr face )
{
    tmbchar buf[256];
    TY_(tmbsnprintf)(buf, sizeof(buf), "font-family: %s", face );
    TY_(AddStyleProperty)( doc, node, buf );
}

static void AddFontSize( TidyDocImpl* doc, Node* node, ctmbstr size )
{
    ctmbstr value = NULL;

    if (nodeIsP(node))
    {
        if (TY_(tmbstrcmp)(size, "6") == 0)
            value = "h1";
        else if (TY_(tmbstrcmp)(size, "5") == 0)
            value = "h2";
        else if (TY_(tmbstrcmp)(size, "4") == 0)
            value = "h3";

        if (value)
        {
            TidyDocFree(doc, node->element);
            node->element = TY_(tmbstrdup)(doc->allocator, value);
            TY_(FindTag)(doc, node);
            return;
        }
    }

    value = FontSize2Name(size);

    if (value)
    {
        tmbchar buf[64];
        TY_(tmbsnprintf)(buf, sizeof(buf), "font-size: %s", value);
        TY_(AddStyleProperty)( doc, node, buf );
    }
}

static void AddFontColor( TidyDocImpl* doc, Node *node, ctmbstr color)
{
    tmbchar buf[128];
    TY_(tmbsnprintf)(buf, sizeof(buf), "color: %s", color);
    TY_(AddStyleProperty)( doc, node, buf );
}

/* force alignment value to lower case */
static void AddAlign( TidyDocImpl* doc, Node *node, ctmbstr align )
{
    uint i;
    tmbchar buf[128];

    TY_(tmbstrcpy)( buf, "text-align: " );
    for ( i = 12; i < sizeof(buf)/sizeof(buf[0])-1; ++i )
    {
        if ( (buf[i] = (tmbchar)TY_(ToLower)(*align++)) == '\0' )
            break;
    }
    buf[i] = '\0';
    TY_(AddStyleProperty)( doc, node, buf );
}

/*
 add style properties to node corresponding to
 the font face, size and color attributes
*/
static void AddFontStyles( TidyDocImpl* doc, Node *node, AttVal *av)
{
    while (av)
    {
        if (AttrHasValue(av))
        {
            if (attrIsFACE(av))
                AddFontFace( doc, node, av->value );
            else if (attrIsSIZE(av))
                AddFontSize( doc, node, av->value );
            else if (attrIsCOLOR(av))
                AddFontColor( doc, node, av->value );
        }
        av = av->next;
    }
}

/*
    Symptom: <p align=center>
    Action: <p style="text-align: center">
*/
static void TextAlign( TidyDocImpl* doc, Node* node )
{
    AttVal *av, *prev;

    prev = NULL;

    for (av = node->attributes; av; av = av->next)
    {
        if (attrIsALIGN(av))
        {
            if (prev)
                prev->next = av->next;
            else
                node->attributes = av->next;

            if (av->value)
                AddAlign( doc, node, av->value );

            TY_(FreeAttribute)(doc, av);
            break;
        }

        prev = av;
    }
}

/*
    Symptom: <table bgcolor="red">
    Action: <table style="background-color: red">
*/
static void TableBgColor( TidyDocImpl* doc, Node* node )
{
    AttVal* attr;
    tmbchar buf[256];

    if (NULL != (attr = TY_(AttrGetById)(node, TidyAttr_BGCOLOR)))
    {
        TY_(tmbsnprintf)(buf, sizeof(buf), "background-color: %s", attr->value );
        TY_(RemoveAttribute)( doc, node, attr );
        TY_(AddStyleProperty)( doc, node, buf );
    }
}

/*
   The clean up rules use the pnode argument to return the
   next node when the original node has been deleted
*/

/*
    Symptom: <dir> <li> where <li> is only child
    Action: coerce <dir> <li> to <div> with indent.
*/

static Bool Dir2Div( TidyDocImpl* doc, Node *node, Node **ARG_UNUSED(pnode))
{
    Node *child;

    if ( nodeIsDIR(node) || nodeIsUL(node) || nodeIsOL(node) )
    {
        child = node->content;

        if (child == NULL)
            return no;

        /* check child has no peers */

        if (child->next)
            return no;

        if ( !nodeIsLI(child) )
            return no;

        if ( !child->implicit )
            return no;

        /* coerce dir to div */
        node->tag = TY_(LookupTagDef)( TidyTag_DIV );
        TidyDocFree( doc, node->element );
        node->element = TY_(tmbstrdup)(doc->allocator, "div");
        TY_(AddStyleProperty)( doc, node, "margin-left: 2em" );
        StripOnlyChild( doc, node );
        return yes;
    }

    return no;
}

/*
    Symptom: <center>
    Action: replace <center> by <div style="text-align: center">
*/

static Bool Center2Div( TidyDocImpl* doc, Node *node, Node **pnode)
{
    if ( nodeIsCENTER(node) )
    {
#if 0 // 00000000 what is this doing inside an nodeIsCENTER(node)??? 0000000
        if ( cfgBool(doc, TidyDropFontTags) )
        {
            if (node->content)
            {
                Node *last = node->last;
                DiscardContainer( doc, node, pnode );

                node = TY_(InferredTag)(doc, TidyTag_BR);
                TY_(InsertNodeAfterElement)(last, node);
            }
            else
            {
                Node *prev = node->prev, *next = node->next,
                     *parent = node->parent;
                DiscardContainer( doc, node, pnode );

                node = TY_(InferredTag)(doc, TidyTag_BR);
                if (next)
                    TY_(InsertNodeBeforeElement)(next, node);
                else if (prev)
                    TY_(InsertNodeAfterElement)(prev, node);
                else
                    TY_(InsertNodeAtStart)(parent, node);
            }

            return yes;
        }
#endif // 00000000 what is this doing inside an nodeIsCENTER(node)??? 0000000
        RenameElem( doc, node, TidyTag_DIV );
        TY_(AddStyleProperty)( doc, node, "text-align: center" );
        return yes;
    }

    return no;
}

/* Copy child attributes to node. Duplicate attributes are overwritten.
   Unique attributes (such as ID) disable the action.
   Attributes style and class are not dealt with. A call to MergeStyles
   will do that.
*/
static Bool CopyAttrs( TidyDocImpl* doc, Node *node, Node *child)
{
    AttVal *av1, *av2;
    TidyAttrId id;

    /* Detect attributes that cannot be merged or overwritten. */
    if (TY_(AttrGetById)(child, TidyAttr_ID) != NULL
        && TY_(AttrGetById)(node, TidyAttr_ID) != NULL)
        return no;

    /* Move child attributes to node. Attributes in node
     can be overwritten or merged. */
    for (av2 = child->attributes; av2; )
    {
        /* Dealt by MergeStyles. */
        if (attrIsSTYLE(av2) || attrIsCLASS(av2))
        {
            av2 = av2->next;
            continue;
        }
        /* Avoid duplicates in node */
        if ((id=AttrId(av2)) != TidyAttr_UNKNOWN
            && (av1=TY_(AttrGetById)(node, id))!= NULL)
            TY_(RemoveAttribute)( doc, node, av1 );

        /* Move attribute from child to node */
        TY_(DetachAttribute)( child, av2 );
        av1 = av2;
        av2 = av2->next;
        av1->next = NULL;
        TY_(InsertAttributeAtEnd)( node, av1 );
    }

    return yes;
}

/*
    Symptom <XX><XX>...</XX></XX>
    Action: merge the two XXs

  For instance, this is useful after nested <dir>s used by Word
  for indenting have been converted to <div>s

  If state is "no", no merging.
  If state is "yes", inner element is discarded. Only Style and Class
  attributes are merged using MergeStyles().
  If state is "auto", atttibutes are merged as described in CopyAttrs().
  Style and Class attributes are merged using MergeStyles().
*/
static Bool MergeNestedElements( TidyDocImpl* doc,
                                 TidyTagId Id, TidyTriState state, Node *node,
                                 Node **ARG_UNUSED(pnode))
{
    Node *child;

    if ( state == TidyNoState
         || !TagIsId(node, Id) )
        return no;

    child = node->content;

    if ( child == NULL
         || child->next != NULL
         || !TagIsId(child, Id) )
        return no;

    if ( state == TidyAutoState
         && CopyAttrs(doc, node, child) == no )
        return no;

    MergeStyles( doc, node, child );
    StripOnlyChild( doc, node );
    return yes;
}

/*
    Symptom: <ul><li><ul>...</ul></li></ul>
    Action: discard outer list
*/

static Bool NestedList( TidyDocImpl* doc, Node *node, Node **pnode )
{
    Node *child, *list;

    if ( nodeIsUL(node) || nodeIsOL(node) )
    {
        child = node->content;

        if (child == NULL)
            return no;

        /* check child has no peers */

        if (child->next)
            return no;

        list = child->content;

        if (!list)
            return no;

        if (list->tag != node->tag)
            return no;

        /* check list has no peers */
        if (list->next)
            return no;

        *pnode = list;  /* Set node to resume iteration */

        /* move inner list node into position of outer node */
        list->prev = node->prev;
        list->next = node->next;
        list->parent = node->parent;
        TY_(FixNodeLinks)(list);

        /* get rid of outer ul and its li */
        child->content = NULL;
        TY_(FreeNode)( doc, child ); /* See test #427841. */