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

  (c) 1998 (W3C) MIT, INRIA, Keio University
  See tidy.c 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 "platform.h"
#include "html.h"

Node *CleanNode(Lexer *lexer, Node *node);

void FreeStyleProps(StyleProp *props)
{
    StyleProp *next;

    while (props)
    {
        next = props->next;
        MemFree(props->name);
        MemFree(props->value);
        MemFree(props);
        props = next;
    }
}

StyleProp *InsertProperty(StyleProp *props, char *name, char *value)
{
    StyleProp *first, *prev, *prop;
    int cmp;

    prev = null;
    first = props;

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

        if (cmp == 0)
        {
            /* need to merge values */
            break;
        }

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

            prop = (StyleProp *)MemAlloc(sizeof(StyleProp));
            prop->name = wstrdup(name);
            prop->value = wstrdup(value);
            prop->next = props;

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

            return first;
        }

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

    prop = (StyleProp *)MemAlloc(sizeof(StyleProp));
    prop->name = wstrdup(name);
    prop->value = wstrdup(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.
*/
StyleProp *CreateProps(StyleProp *prop, char *style)
{
    char *name, *value, *name_end, *value_end;
    Bool more;

    style = wstrdup(style);
    name = style;

    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_end = value;
        more = no;

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

            ++value_end;
        }

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

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

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

        break;
    }

    MemFree(style);  /* free temporary copy */
    return prop;
}

char *CreatePropString(StyleProp *props)
{
    char *style, *p, *s;
    int len;
    StyleProp *prop;

    /* compute length */

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

    style = (char *)MemAlloc(len+1);

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

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

        *--p = ':';
        *++p = ' ';
        ++p;

        s = prop->value;
        while(*p++ = *s++);

        if (prop->next == null)
            break;

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

    return style;
}

/*
  create string with merged properties
*/
char *AddProperty(char *style, char *property)
{
    StyleProp *prop;

    prop = CreateProps(null, style);
    prop = CreateProps(prop, property);
    style = CreatePropString(prop);
    FreeStyleProps(prop);
    return style;
}

void FreeStyles(Lexer *lexer)
{
    Style *style, *next;

    for (style = lexer->styles; style; style = next)
    {
        next = style->next;
        
        MemFree(style->tag);
        MemFree(style->tag_class);
        MemFree(style->properties);
        MemFree(style);
    }
}

char *GensymClass(char *tag)
{
    static int n = 1;
    char buf[128];

    sprintf(buf, "c%d", n++);
    return wstrdup(buf);
}

char *FindStyle(Lexer *lexer, char *tag, char *properties)
{
    Style *style;

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

    style = (Style *)MemAlloc(sizeof(Style));
    style->tag = wstrdup(tag);
    style->tag_class = GensymClass(tag);
    style->properties = wstrdup(properties);
    style->next = lexer->styles;
    lexer->styles = style;
    return style->tag_class;
}

/*
 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
*/
void Style2Rule(Lexer *lexer, Node *node)
{
    AttVal *av;
    char *classname;

    for (av = node->attributes; av; av = av->next)
    {
        if (wstrcmp(av->attribute, "style") == 0)
            break;
    }

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

    if (av)
    {
        classname = FindStyle(lexer, node->element, av->value);
        MemFree(av->attribute);
        MemFree(av->value);
        av->attribute = wstrdup("class");
        av->value = wstrdup(classname);
    }
}

/* create style element using rules from dictionary */
void CreateStyleElement(Lexer *lexer, Node *doc)
{
    Node *node, *html, *head;
    Style *style;
    AttVal *av;

    if (lexer->styles == null)
        return;

    node = NewNode();
    node->type = StartTag;
    node->implicit = yes;
    node->element = wstrdup("STYLE");
    FindTag(node);

    /* insert type attribute */
    av = (AttVal *)MemAlloc(sizeof(AttVal));
    av->next = null; 
    av->attribute = wstrdup("type");
    av->value = wstrdup("text/css");
    av->delim = '"';
    av->dict = FindAttribute(av);
    node->attributes = av;

    lexer->txtstart = lexer->lexsize;

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

    lexer->txtend = lexer->lexsize;

    InsertNode(node, 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
    */

    for (html = doc->content; html; html = html->next)
    {
        if (html->tag == tag_html)
        {
            head = html->content;
            InsertNode(head, node);
            break;
        }
    }
}

/* ensure bidirectional links are consistent */
void FixNodeLinks(Node *node)
{
    if (node->prev)
        node->prev->next = node;
    else
        node->parent->content = node;

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

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

    child = node->content;
    node->content = child->content;
    node->last = child->last;
    child->content = null;
    FreeNode(child);

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

/*
 Add style property to element, creating style
 attribute as needed and adding ; delimiter
*/
void AddStyleProperty(Node *node, char *property)
{
    AttVal *av;

    for (av = node->attributes; av; av = av->next)
    {
        if (wstrcmp(av->attribute, "style") == 0)
            break;
    }

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

    if (av)
    {
        char *s;

        s = AddProperty(av->value, property);
        MemFree(av->value);
        av->value = s;
    }
    else /* else create new style attribute */
    {
        av = (AttVal *)MemAlloc(sizeof(AttVal));
        av->attribute = wstrdup("style");
        av->value = wstrdup(property);
        av->delim = '"';
        av->dict = FindAttribute(av);
        av->next = node->attributes;
        node->attributes = av;
    }
}

/*
  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.
*/
char *MergeProperties(char *s1, char *s2)
{
    char *s;
    StyleProp *prop;

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

void MergeStyles(Node *node, Node *child)
{
    AttVal *av;
    char *s1, *s2, *style;

    for (s2 = null, av = child->attributes; av; av = av->next)
    {
        if (wstrcmp(av->attribute, "style") == 0)
        {
            s2 = av->value;
            break;
        }
    }

    for (s1 = null, av = node->attributes; av; av = av->next)
    {
        if (wstrcmp(av->attribute, "style") == 0)
        {
            s1 = av->value;
            break;
        }
    }

    if (s1)
    {
        if (s2)  /* merge styles from both */
        {
            style = MergeProperties(s1, s2);
            MemFree(av->value);
            av->value = style;
        }
    }
    else if (s2)  /* copy style of child */
    {
        av = (AttVal *)MemAlloc(sizeof(AttVal));
        av->attribute = wstrdup("style");
        av->value = wstrdup(s2);
        av->delim = '"';
        av->dict = FindAttribute(av);
        av->next = node->attributes;
        node->attributes = av;
    }
}

char *FontSize2Name(char *size)
{
    char *sizes[7] =
      {
        "50%", "60%", "80%", null,
        "120%", "150%", "200%"
      };

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

    if (size[0] == '-')
        return "smaller";

    return "larger";
}

void AddFontFace(Node *node, char *face)
{
    char buf[1024];

    sprintf(buf, "font-family: %s", face);
    AddStyleProperty(node, buf);
}

void AddFontSize(Node *node, char *size)
{
    char *value, buf[1024];

    if (wstrcmp(size, "6") == 0 && node->tag == tag_p)
    {
        MemFree(node->element);
        node->element = wstrdup("H1");
        FindTag(node);
        return;
    }

    if (wstrcmp(size, "5") == 0 && node->tag == tag_p)
    {
        MemFree(node->element);
        node->element = wstrdup("H2");
        FindTag(node);
        return;
    }

    if (wstrcmp(size, "4") == 0 && node->tag == tag_p)
    {
        MemFree(node->element);
        node->element = wstrdup("H3");
        FindTag(node);
        return;
    }

    value = FontSize2Name(size);

    if (value)
    {
        sprintf(buf, "font-size: %s", value);
        AddStyleProperty(node, buf);
    }
}

void AddFontColor(Node *node, char *color)
{
    char buf[1024];

    sprintf(buf, "color: %s", color);
    AddStyleProperty(node, buf);
}

void AddAlign(Node *node, char *align)
{
    char buf[1024], *p, *q;

    /* force alignment value to lower case */
    for (p = buf, q = "text-align: "; *p++ = *q++;);
    for (p = p-1; *p++ = ToLower(*align++););
    AddStyleProperty(node, buf);
}

/*
 add style properties to node corresponding to
 the font face, size and color attributes
*/
void AddFontStyles(Node *node, AttVal *av)
{
    while (av)
    {
        if (wstrcmp(av->attribute, "face") == 0)
            AddFontFace(node, av->value);
        else if (wstrcmp(av->attribute, "size") == 0)
            AddFontSize(node, av->value);
        else if (wstrcmp(av->attribute, "color") == 0)
            AddFontColor(node, av->value);

        av = av->next;
    }
}

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

    prev = null;

    for (av = node->attributes; av; av = av->next)
    {
        if (wstrcmp(av->attribute, "align") == 0)
        {
            if (prev)
                prev->next = av->next;
            else
                node->attributes = null;

            MemFree(av->attribute);

            if (av->value)
            {
                AddAlign(node, av->value);
                MemFree(av->value);
            }

            MemFree(av);
            break;
        }

        prev = av;
    }
}

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

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

Bool Dir2Div(Lexer *lexer, Node *node, Node **pnode)
{
    Node *child;

    if (node->tag == tag_dir || node->tag == tag_ul || node->tag == tag_ol)
    {
        child = node->content;

        if (child == null)
            return no;

        /* check child has no peers */

        if (child->next)
            return no;

        if (child->tag != tag_li)
            return no;

        if (!child->implicit)
            return no;

        /* coerce dir to div */

        node->tag = tag_div;
        MemFree(node->element);
        node->element = wstrdup("DIV");
        AddStyleProperty(node, "margin-left: 2em");
        StripOnlyChild(node);
        return yes;

#if 0
        content = child->content;
        last = child->last;
        child->content = null;

        /* adjust parent and set margin on contents of <li> */

        for (child = content; child; child = child->next)
        {
            child->parent = node->parent;
            AddStyleProperty(child, "margin-left: 1em");
        }

        /* hook first/last into sequence */

        if (content)
        {
            content->prev = node->prev;
            last->next = node->next;
            FixNodeLinks(content);
            FixNodeLinks(last);
        }

        node->next = null;
        FreeNode(node);

        /* ensure that new node is cleaned */
        *pnode = CleanNode(lexer, content);
        return yes;
#endif
    }

    return no;
}

/*
    Symptom: implicit blockquote due to <ul>some text</ul>
    Action: replace by <div> with indent style
*/
Bool Blockquote2Div(Lexer *lexer, Node *node, Node **pnode)
{
    Node *child;

    if (node->tag == tag_blockquote && node->implicit)
    {
        child = node->content;

        if (child == null)
            return no;

        /* coerce blockquote to div */

        node->tag = tag_div;
        MemFree(node->element);
        node->element = wstrdup("DIV");
        AddStyleProperty(node, "margin-left: 2em");
        return yes;
    }

    return no;
}


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

Bool Center2Div(Lexer *lexer, Node *node, Node **pnode)
{
    if (node->tag == tag_center)
    {
        node->tag = tag_div;
        MemFree(node->element);
        node->element = wstrdup("DIV");
        AddStyleProperty(node, "text-align: center");
        return yes;
    }

    return no;
}

/*
    Symptom <div><div>...</div></div>
    Action: merge the two divs

  This is useful after nested <dir>s used by Word
  for indenting have been converted to <div>s
*/
Bool MergeDivs(Lexer *lexer, Node *node, Node **pnode)
{
    Node *child;

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

    child = node->content;

    if (!child)
        return no;

    if (child->tag != tag_div)
        return no;

    if (child->next != null)
        return no;

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

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

Bool NestedList(Lexer *lexer, Node *node, Node **pnode)
{
    Node *child, *list;

    if (node->tag == tag_ul || node->tag == tag_ol)
    {
        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;

        *pnode = node->next;

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

        /* get rid of outer ul and its li */
        child->content = null;
        node->content = null;
        node->next = null;
        FreeNode(node);

        /*
          If prev node was a list the chances are this node
          should be appended to that list. Word has no way of
          recognizing nested lists and just uses indents
        */

        if (list->prev)
        {
            node = list;
            list = node->prev;

            if (list->tag == tag_ul || list->tag == tag_ol)
            {
                list->next = node->next;

                if (list->next)
                    list->next->prev = list;

                child = list->last;  /* <li> */

                node->parent = child;
                node->next = null;
                node->prev = child->last;
                FixNodeLinks(node);
            }
        }

        CleanNode(lexer, node);
        return yes;
    }

    return no;
}

/*
    Symptom: the only child of a block-level element is a
    presentation element such as B, I or FONT

    Action: add style "font-weight: bold" to the block and
    strip the <b> element, leaving its children.

  example:

    <p>
      <b><font face="Arial" size="6">Draft Recommended Practice</font></b>
    </p>

  becomes:

      <p style="font-weight: bold; font-family: Arial; font-size: 6">
        Draft Recommended Practice
      </p>
*/
Bool BlockStyle(Lexer *lexer, Node *node, Node **pnode)
{
    Node *child;

    if (node->tag->model & (CM_BLOCK | CM_LIST | CM_DEFLIST | CM_TABLE))
    {
        /* check for align attribute */
        if (node->tag != tag_caption)
            TextAlign(lexer, node);

        child = node->content;

        if (child == null)
            return no;

        /* check child has no peers */

        if (child->next)
            return no;

        if (child->tag == tag_b)
        {
            AddStyleProperty(node, "font-weight: bold");
            StripOnlyChild(node);
            return yes;
        }

        if (child->tag == tag_i)
        {
            AddStyleProperty(node, "font-style: italic");
            StripOnlyChild(node);
            return yes;
        }

        if (child->tag == tag_font)
        {
            AddFontStyles(node, child->attributes);
            StripOnlyChild(node);
            return yes;
        }
    }

    return no;
}

/* the only child of an inline element such as em */
Bool InlineStyle(Lexer *lexer, Node *node, Node **pnode)
{
    Node *child;

    if (node->tag != tag_font && (node->tag->model & CM_INLINE))
    {
        child = node->content;

        if (child == null)
            return no;

        /* check child has no peers */

        if (child->next)
            return no;

        if (child->tag == tag_b)
        {
            AddStyleProperty(node, "font-weight: bold");
            StripOnlyChild(node);
            return yes;
        }

        if (child->tag == tag_i)
        {
            AddStyleProperty(node, "font-style: italic");
            StripOnlyChild(node);
            return yes;
        }

        if (child->tag == tag_font)
        {
            AddFontStyles(node, child->attributes);
            StripOnlyChild(node);
            return yes;
        }
    }

    return no;
}

/*
  Replace font elements by span elements, deleting
  the font element's attributes and replacing them
  by a single style attribute.
*/
Bool Font2Span(Lexer *lexer, Node *node, Node **pnode)
{
    AttVal *av, *style, *next;

    if (node->tag == tag_font)
    {
        /* if FONT is only child of parent element then leave alone */
        if (node->parent->content == node
            && node->next == null)
            return no;

        AddFontStyles(node, node->attributes);

        av = node->attributes;
        style = null;

        while (av)
        {
            next = av->next;

            if (wstrcmp(av->attribute, "style") == 0)
            {
                av->next = null;
                style = av;
            }
            else
            {
                if (av->attribute)
                    MemFree(av->attribute);
                if (av->value)
                    MemFree(av->value);

                MemFree(av);
            }

            av = next;
        }

        node->attributes = style;

        node->tag = tag_span;
        MemFree(node->element);
        node->element = wstrdup("SPAN");

        return yes;
    }

    return no;
}

Bool IsElement(Node *node)
{
    return (node->type == StartTag || node->type == StartEndTag ? yes : no);
}

/*
  Applies all matching rules to a node.
*/
Node *CleanNode(Lexer *lexer, Node *node)
{
    Node *next;

    for (next = node; IsElement(node); node = next)
    {
        if (Dir2Div(lexer, node, &next))
            continue;

        if (Blockquote2Div(lexer, node, &next))
            continue;

        if (NestedList(lexer, node, &next))
            continue;

        if (Center2Div(lexer, node, &next))
            continue;

        if (MergeDivs(lexer, node, &next))
            continue;

        if (BlockStyle(lexer, node, &next))
            continue;

        if (InlineStyle(lexer, node, &next))
            continue;

        if (Font2Span(lexer, node, &next))
            continue;

        break;
    }

    return next;
}

Node *CreateStyleProperties(Lexer *lexer, Node *node)
{
    Node *child;

    if (node->content)
    {
        for (child = node->content;
                child != null; child = child->next)
        {
            child = CreateStyleProperties(lexer, child);
        }
    }

    return CleanNode(lexer, node);
}

void DefineStyleRules(Lexer *lexer, Node *node)
{
    Node *child;

    if (node->content)
    {
        for (child = node->content;
                child != null; child = child->next)
        {
            DefineStyleRules(lexer, child);
        }
    }

    Style2Rule(lexer, node);
}

void CleanTree(Lexer *lexer, Node *doc)
{
    doc = CreateStyleProperties(lexer, doc);
    DefineStyleRules(lexer, doc);
    CreateStyleElement(lexer, doc);
}