Closing 10515 with patch from ferdy@ferdyx.org to convert HTML to

DocBook/XML and DocBook/SGML.

4 files changed, 944 insertions, 1 deletions

diff --git a/app-text/htmltidy/ChangeLog b/app-text/htmltidy/ChangeLog
index 2a569bb2d426..0be2538eca8c 100644
--- a/app-text/htmltidy/ChangeLog
+++ b/app-text/htmltidy/ChangeLog
@@ -1,6 +1,12 @@
 # ChangeLog for app-text/htmltidy
 # Copyright 2002 Gentoo Technologies, Inc.; Distributed under the GPL
-# $Header: /var/cvsroot/gentoo-x86/app-text/htmltidy/ChangeLog,v 1.4 2002/12/13 10:56:19 blizzy Exp $
+# $Header: /var/cvsroot/gentoo-x86/app-text/htmltidy/ChangeLog,v 1.5 2003/01/03 05:43:19 satai Exp $
+
+*htmltidy-2.7.18-r1 (03 Jan 2003)
+
+  03 Jan 2003; Matthew Turk <satai@gentoo.org> :
+  Added dbpatch to convert HTML to either Docbook/XML or Docbook/SGML.  Thanks
+  to Ferdy <ferdy@ferdyx.org> for the patch.
 
   06 Dec 2002; Rodney Rees <manson@gentoo.org> : changed sparc ~sparc keywords
  
diff --git a/app-text/htmltidy/files/digest-htmltidy-2.7.18-r1 b/app-text/htmltidy/files/digest-htmltidy-2.7.18-r1
new file mode 100644
index 000000000000..40d794f8de51
--- /dev/null
+++ b/app-text/htmltidy/files/digest-htmltidy-2.7.18-r1
@@ -0,0 +1 @@
+MD5 7694dcdb7d451b17477292d60186f477 tidy_src_020718.tgz 153771
diff --git a/app-text/htmltidy/files/htmltidy-dbpatch.diff b/app-text/htmltidy/files/htmltidy-dbpatch.diff
new file mode 100644
index 000000000000..f3322e273a72
--- /dev/null
+++ b/app-text/htmltidy/files/htmltidy-dbpatch.diff
@@ -0,0 +1,890 @@
+--- /cise/tmp/ppadala/tidy/include/html.h	Fri May 31 17:52:04 2002
++++ include/html.h	Tue Jul  2 15:04:28 2002
+@@ -653,6 +653,8 @@
+                     Lexer *lexer, Node *node);
+ void PPrintXMLTree(Out *fout, uint mode, uint indent,
+                     Lexer *lexer, Node *node);
++void PrintSgml(Out *fout, uint mode, uint indent, 
++                Lexer *lexer, Node *node);
+ void PFlushLine(Out *out, uint indent);
+ void PCondFlushLine(Out *out, uint indent);
+ void PrintBody(Out *fout, Lexer *lexer, Node *root); /* Feature request #434940 - fix by Dave Raggett/Ignacio Vazquez-Abrams 21 Jun 01 */
+@@ -908,6 +910,8 @@
+ extern Bool XmlOut;
+ extern Bool xHTML;
+ extern Bool HtmlOut;    /* Yes means set explicitly. */
++extern Bool DbSgml;
++extern Bool DbXml;
+ extern Bool XmlPi;      /* add <?xml?> */
+ extern Bool XmlPIs;     /* assume PIs end with ?> as per XML */
+ extern Bool XmlSpace;
+--- /cise/tmp/ppadala/tidy/src/config.c	Sun Jul  7 23:24:52 2002
++++ src/config.c	Sat Jul 13 18:14:47 2002
+@@ -81,6 +81,8 @@
+ Bool XmlOut = no;           /* create output as XML */
+ Bool xHTML = no;            /* output extensible HTML */
+ Bool HtmlOut = no;          /* output plain-old HTML, even for XHTML input. Yes means set explicitly. */
++Bool DbSgml = no;           /* output docbook SGML */
++Bool DbXml = no;            /* output docbook XML */
+ Bool XmlPi = no;            /* add <?xml?> for XML docs */
+ Bool RawOut = no;           /* avoid mapping values > 127 to entities: not used for anything yet */
+ Bool UpperCaseTags = no;    /* output tags in upper not lower case */
+--- /cise/tmp/ppadala/tidy/src/lexer.c	Sun Jul  7 23:25:47 2002
++++ src/lexer.c	Sat Jul 13 18:14:47 2002
+@@ -1674,6 +1674,35 @@
+     return doctype;
+ }
+ 
++Bool SetSgmlDocType(Lexer *lexer, Node *root)
++{   char *fpi, *sysid;
++    Node *doctype;
++
++    if (doctype_mode == doctype_user && doctype_str)
++    {
++        fpi = doctype_str;
++        sysid = "";
++    }
++    doctype = FindDocType(root);
++    if(!doctype) /* The html file doesn't contain doctype */
++        if ( !(doctype = NewXhtmlDocTypeNode( root )) )
++            return no;
++    
++    lexer->txtstart = lexer->txtend = lexer->lexsize;
++
++    /* add public identifier */
++    AddStringLiteral(lexer, fpi);
++    /* add system identifier */
++    AddStringLiteral(lexer, sysid);
++    
++    lexer->txtend = lexer->lexsize;
++
++    doctype->start = lexer->txtstart;
++    doctype->end = lexer->txtend;
++
++    return no; 
++}
++
+ Bool SetXHTMLDocType(Lexer *lexer, Node *root)
+ {
+     char *fpi = "", *sysid = "", *dtdsub, *name_space = XHTML_NAMESPACE; /* #578005 - fix by Anonymous 05 Jul 02 */
+--- /cise/tmp/ppadala/tidy/src/localize.c	Sun Jul  7 23:26:39 2002
++++ src/localize.c	Sat Jul 13 18:14:47 2002
+@@ -1054,6 +1054,8 @@
+     tidy_out(out, "  -asxml            to convert HTML to well formed XHTML\n");
+     tidy_out(out, "  -asxhtml          to convert HTML to well formed XHTML\n");
+     tidy_out(out, "  -ashtml           to force XHTML to well formed HTML\n");
++    tidy_out(out, "  -dbsgml           to convert HTML to Docbook SGML\n");
++    tidy_out(out, "  -dbxml            to convert HTML to Docbook XML\n");
+     tidy_out(out, "  -slides           to burst into slides on H2 elements\n");
+ 
+ /* TRT */
+--- /cise/tmp/ppadala/tidy/src/parser.c	Sun Jul  7 23:27:35 2002
++++ src/parser.c	Sat Jul 13 18:14:47 2002
+@@ -501,7 +501,7 @@
+         TrimTrailingSpace(lexer, element, text);
+ }
+ 
+-static Bool DescendantOf(Node *element, Dict *tag)
++Bool DescendantOf(Node *element, Dict *tag)
+ {
+     Node *parent;
+ 
+--- /cise/tmp/ppadala/tidy/src/pprint.c	Thu Jul 18 14:21:27 2002
++++ src/pprint.c	Tue Jul 30 13:50:53 2002
+@@ -6,9 +6,9 @@
+   
+   CVS Info :
+ 
+-    $Author: satai $ 
+-    $Date: 2003/01/03 05:43:19 $ 
+-    $Revision: 1.1 $ 
++    $Author: satai $ 
++    $Date: 2003/01/03 05:43:19 $ 
++    $Revision: 1.1 $ 
+ 
+ */
+ 
+@@ -36,6 +36,13 @@
+ static void PPrintPhp(Out *fout, uint indent,
+                    Lexer *lexer, Node *node);
+ 
++/* Tag types to distinguish printing */
++typedef enum {
++    SgmlTagStart,
++    SgmlTagEnd
++}SgmlTagType;
++
++extern Bool DescendantOf(Node *element, Dict *tag);
+ 
+ #define NORMAL        0
+ #define PREFORMATTED  1
+@@ -1769,6 +1776,634 @@
+     }
+ }
+ 
++void PrintSgmlDefault(Out *fout)
++{
++    char *str = "SGML cannot contain these elements";
++
++    fprintf(stderr, str);
++}
++
++void PrintSgmlBodyStart(Out *fout, uint indent)
++{
++    char *str = "<article>";
++    PPrintString(fout, indent, str);
++}
++
++#define DIGIT(c) (c - 48)
++#define TOTAL_H 6
++static Bool seen_h[TOTAL_H] = {no, no, no, no, no, no};
++
++/* Yuck ugly. FIXME */
++#define SECT(i) (i - startsect)
++static startsect = 0; /* We are at level 0(H1) initially */
++
++void PrintSgmlBodyEnd(Out *fout, uint indent)
++{   int i = TOTAL_H - 1;
++    char str[10];
++
++    while(i >= 0) {
++        if(seen_h[i] == yes) {
++            if(i == 5)
++                sprintf(str, "</simpleect>");
++            else
++                sprintf(str, "</sect%d>", SECT(i) + 1);
++            PPrintString(fout, indent, str);
++            seen_h[i] = no;
++        }
++        --i;
++    }
++    
++    sprintf(str, "</article>");
++    PPrintString(fout, indent, str);
++}
++
++char *GetContent(Lexer *lexer, Node *node)
++{   Node *content, *temp_node;
++    char *str, *temp, c;
++    Bool flag = no;
++    int i;
++
++    content = node->content;
++
++    /* Find the <a> tag */
++    for (temp_node = content;
++         temp_node && temp_node->tag != tag_a; 
++         temp_node = temp_node->next)
++         ;
++    
++    if(temp_node == NULL) { /* There is no <a> .. </a> tag */
++        /* Discard all elements which are not text nodes */
++        temp_node = content;
++        for (temp_node = content;
++             temp_node && temp_node->type != TextNode; 
++             temp_node = temp_node->next)
++           ;
++        if(temp_node == NULL) { /* There's no TextNode either */ 
++            str = MemAlloc(1);
++            str[0] = '\0';
++            return str;
++        }
++    }
++    content = temp_node;
++
++    if(content->type == TextNode)  {
++        int size = content->end - content->start;
++        
++        str = MemAlloc(size + 1);
++        str[size] = '\0';
++        wstrncpy(str, lexer->lexbuf + content->start, size);
++    }
++    else if(content->tag == tag_a){
++        AttVal *name;
++        int size;
++
++        name = GetAttrByName(content, "name");
++        if(name == NULL)
++            name = GetAttrByName(content, "href");
++
++        if(name == NULL) {  /* No href or name, let's take empty id */
++            size = 0;
++            str = MemAlloc(size + 1);
++            str[size] = '\0';
++        }
++        else {
++            size = wstrlen(name->value);
++            str = MemAlloc(size + 1);
++            str[size] = '\0';
++            wstrncpy(str, name->value, size);
++       }
++    }
++
++    temp = str;
++    if(str[0] == '#')
++        flag = yes;
++
++#define SGML_NAMELEN 44  /* Maximum id namelength */
++
++    i = 0;
++
++    while(*temp && i < SGML_NAMELEN) {
++        if(flag)
++            *temp = *(temp + 1);
++        if(*temp == ' ')
++            *temp = '_';
++        ++temp;
++        ++i;
++    }
++    *temp = '\0';
++    return str;
++}
++
++void PrintSectTag( Out *fout, uint indent, Lexer *lexer, Node *node, 
++                   uint startsect)
++{   char sectnum = node->element[1];
++    char str[100];
++
++    char *id = GetContent(lexer, node);
++
++    if(sectnum == '6')  /* there's no sect6. We can do variety of
++                           things here. may be <section> .. */
++        sprintf(str, "<simplesect id=\"%s\"><title>", id); 
++    else 
++        sprintf(str, "<sect%c id=\"%s\"><title>", SECT(sectnum), id); 
++    PPrintString(fout, indent, str);
++    MemFree(id);
++}
++
++Bool ImmediateDescendantOfHTags(Node *element)
++{   Node *parent = element->parent;
++
++    if (strlen(parent->element) == 2 && 
++            parent->element[0] == 'h' && 
++            IsDigit(parent->element[1]))
++        return yes;
++    return no;
++}
++
++void PrintSgmlLink(Out *fout, uint indent, Node *node)
++{   AttVal *addr;
++    char str[500];  /* FIXME allocate dynamically later */
++
++    addr = GetAttrByName(node, "name");
++    if(addr == NULL) {
++        addr = GetAttrByName(node, "href");
++        if(!ImmediateDescendantOfHTags(node)) {
++            if(addr->value[0] == '#') 
++                sprintf(str, "<link linkend=\"%s\">", addr->value + 1); 
++            else
++                sprintf(str, "<ulink url=\"%s\">", addr->value); 
++            if( !DescendantOf(node, tag_p) && 
++                 node->prev && node->prev->type == TextNode)
++                PPrintString(fout, indent, "<para>");
++            PPrintString(fout, indent, str);
++        }
++    }
++    else {
++        if(!ImmediateDescendantOfHTags(node)) {
++            if(!DescendantOf(node, tag_p)) 
++                sprintf(str, "<para id=\"%s\">", addr->value); 
++            else /* We cannnot have a <para> inside another <para> */
++                sprintf(str, "<anchor id=\"%s\"/>", addr->value); 
++            PPrintString(fout, indent, str);
++        }
++    }
++}
++
++void PrintSgmlLinkEnd(Out *fout, uint indent, Node *node)
++{   AttVal *addr;
++
++    addr = GetAttrByName(node, "name");
++    if(addr == NULL) {
++        addr = GetAttrByName(node, "href");
++        if(!ImmediateDescendantOfHTags(node)) {
++            if(addr->value[0] == '#') 
++                PPrintString(fout, indent, "</link>");
++            else
++                PPrintString(fout, indent, "</ulink>");
++            if( !DescendantOf(node, tag_p) && 
++                 node->prev && node->prev->type == TextNode)
++                PPrintString(fout, indent, "</para>");
++        }
++    }
++    else {
++        if(!ImmediateDescendantOfHTags(node)) {
++            if(!DescendantOf(node, tag_p))
++                PPrintString(fout, indent, "</para>");
++            /* else
++               <anchor ..  /> has already been placed. no need to
++               do any thing */
++        }
++    }
++}
++
++
++void PrintSgmlTagString(Out *fout, uint mode, uint indent, 
++                        SgmlTagType sgmltag_type, char *str)
++{   PPrintChar(str[0], mode | CDATA);
++    if(sgmltag_type == SgmlTagEnd)
++        PPrintChar('/', mode);
++    PPrintString(fout, indent, str + 1);
++}
++
++void PrintSgmlList(Lexer *lexer, Out *fout, 
++                      uint mode, uint indent,
++                      Node *node)
++{   if(node->tag == tag_ul)
++        PPrintString(fout, indent, "<itemizedlist>");
++    else if(node->tag == tag_ol)
++        PPrintString(fout, indent, "<orderedlist>");
++    else if(node->tag == tag_dl)
++        PPrintString(fout, indent, "<variablelist>");
++}
++
++void PrintSgmlListEnd(Lexer *lexer, Out *fout, 
++                      uint mode, uint indent,
++                      Node *node)
++{   if(node->tag == tag_ul)
++        PPrintString(fout, indent, "</itemizedlist>");
++    else if(node->tag == tag_ol)
++        PPrintString(fout, indent, "</orderedlist>");
++    else if(node->tag == tag_dl)
++        PPrintString(fout, indent, "</variablelist>");
++}
++
++void PrintSgmlListItem(Out *fout, uint indent, Node *node)
++{   if(node->tag ==  tag_li)
++        PPrintString(fout, indent, "<listitem>");
++    else if(node->tag == tag_dd)
++        PPrintString(fout, indent, "<listitem>");
++}
++
++void PrintSgmlListItemEnd(Out *fout, uint indent, Node *node)
++{   if(node->tag ==  tag_li)
++        PPrintString(fout, indent, "</listitem>");
++    else if(node->tag == tag_dd)
++        PPrintString(fout, indent, "</listitem></varlistentry>");
++}
++
++void PrintSgmlImage(Out *fout, uint indent, Node *node)
++{   AttVal *addr;
++    char str[100];
++
++    addr = GetAttrByName(node, "src");
++    /* We can get other attributes like width, height etc.. */
++    if(addr != NULL) {
++        PPrintString(fout, indent, "<inlinemediaobject><imageobject>");
++        PCondFlushLine(fout, indent);
++        sprintf(str, "<imagedata fileref=\"%s\">", addr->value); 
++        PPrintString(fout, indent, str);
++        PCondFlushLine(fout, indent);
++        PPrintString(fout, indent, "</imageobject></inlinemediaobject>");
++        PCondFlushLine(fout, indent);
++    }
++}
++
++int CountColumns(Node *node)
++{   Node *temp, *row_content;
++    int ncols = 0;
++
++    temp = node->content;
++
++    /* FIXME */
++    /* Perhaps this is not needed, check with HTML standard later */
++    while(temp->tag != tag_tr)
++        temp = temp->next;
++
++    /* This can contain th or td's */
++    row_content = temp->content;
++    while(row_content) {
++        if(row_content->tag == tag_th || row_content->tag == tag_td) {
++            AttVal *colspan;
++            
++            colspan = GetAttrByName(row_content, "colspan");
++            if(colspan)
++                ncols += atoi(colspan->value);
++            else
++                ++ncols;
++        }
++        else
++            fprintf(stderr, "PrintSgml: error in table processing\n");
++        row_content = row_content->next;
++   } 
++   return ncols;
++}
++
++void PrintSgmlTable(Out *fout, uint indent, Node *node)
++{   int ncols;
++    char str[100];
++
++    PPrintString(fout, indent, "<informaltable>");
++    ncols = CountColumns(node);
++    sprintf(str, "<tgroup cols=\"%d\"><tbody>", ncols);
++    PPrintString(fout, indent, str);
++}
++
++void PrintSgmlTableEnd(Out *fout, uint indent, Node *node)
++{   
++    PPrintString(fout, indent, "</tbody></tgroup></informaltable>");
++}
++
++Bool DescendantOfAddress(Node *element)
++{
++    Node *parent;
++    
++    for (parent = element->parent;
++            parent != null; parent = parent->parent)
++    {   if (parent->element && wstrcasecmp(parent->element, "address") == 0)
++            return yes;
++    }
++
++    return no;
++}
++
++void PrintSgmlTag( Out *fout, uint mode, uint indent, Lexer *lexer, Node *node,
++                    SgmlTagType sgmltag_type)
++{   static level = -1;
++
++    if(node->tag == tag_html) {
++        if(sgmltag_type == SgmlTagStart)
++            PrintSgmlBodyStart(fout, indent);
++        else if(sgmltag_type == SgmlTagEnd)
++            PrintSgmlBodyEnd(fout, indent);
++    }
++    else if(node->tag == tag_head)
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type,"<articleinfo>");
++    else if(node->tag == tag_title)
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type,"<title>");
++    /* May be we can replace with node->model & CM_LIST */
++    else if(node->tag == tag_ul || node->tag == tag_ol ||
++            node->tag == tag_dl) {
++        if(sgmltag_type == SgmlTagStart)
++            PrintSgmlList(lexer, fout, mode, indent, node);
++        else if(sgmltag_type == SgmlTagEnd)
++            PrintSgmlListEnd(lexer, fout, mode, indent, node);
++    }
++    else if(node->tag == tag_dt) {
++        if(sgmltag_type == SgmlTagStart)
++            PPrintString(fout, indent, "<varlistentry><term>");
++        else if(sgmltag_type == SgmlTagEnd)
++            PPrintString(fout, indent, "</term>");
++    }
++    else if(node->tag == tag_li || node->tag == tag_dd) {
++        if(sgmltag_type == SgmlTagStart)
++            PrintSgmlListItem(fout, indent, node);
++        else if(sgmltag_type == SgmlTagEnd)
++            PrintSgmlListItemEnd(fout, indent, node);
++    }
++    /* Later we should clean this before coming to PrintSgml */
++    else if(node->tag == tag_p && 
++            /* Table <entry> processing */
++            !DescendantOf(node, tag_th) && !DescendantOf(node, tag_td) && 
++            !DescendantOfAddress(node)) 
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type, "<para>");
++    else if(node->tag == tag_blockquote)
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type, "<blockquote>");
++    else if(node->tag == tag_pre && 
++            /* Table <entry> processing */
++            !DescendantOf(node, tag_th) && !DescendantOf(node, tag_td))
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type, 
++                               "<programlisting>");
++    else if(node->tag == tag_a) {
++        if(sgmltag_type == SgmlTagStart)
++            PrintSgmlLink(fout, indent, node);
++        else if(sgmltag_type == SgmlTagEnd)
++            PrintSgmlLinkEnd(fout, indent, node);
++    }
++    /* Table would require more processing */
++    else if(node->tag == tag_table) {
++        if(sgmltag_type == SgmlTagStart)
++            PrintSgmlTable(fout, indent, node);
++        else if(sgmltag_type == SgmlTagEnd)
++            PrintSgmlTableEnd(fout, indent, node);
++    }
++    else if(node->tag == tag_tr)
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type, "<row>");
++    else if(node->tag == tag_td || node->tag == tag_th)
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type, "<entry>");
++    else if(node->tag == tag_img) { /* This is a StartEndTag */
++        if(sgmltag_type == SgmlTagStart)    
++            PrintSgmlImage(fout, indent, node);
++    }
++
++    else if(wstrcasecmp(node->element, "cite") == 0)
++        PrintSgmlTagString(fout, mode, indent, sgmltag_type, 
++                                "<citation>");
++    /* We should distinguish tag_strong and tag_em later 
++       haven't found proper docbook tag for <strong> */
++    else if(node->tag == tag_em || node->tag == tag_strong ||
++            wstrcasecmp(node->element, "address") == 0) {
++        if(sgmltag_type == SgmlTagStart) {
++            if(DescendantOf(node, tag_p) || DescendantOf(node, tag_pre))
++                PPrintString(fout, indent, "<emphasis>");
++            else
++                PPrintString(fout, indent, "<para><emphasis>");
++        }
++        else if(sgmltag_type == SgmlTagEnd) {
++            if(DescendantOf(node, tag_p) || DescendantOf(node, tag_pre))
++                PPrintString(fout, indent, "</emphasis>");
++            else 
++                PPrintString(fout, indent, "</emphasis></para>");
++        }
++    }
++    else {
++        if(wstrcasecmp(node->element, "code") == 0 &&
++            !(node->parent->tag == tag_dd || 
++            node->parent->tag == tag_li))
++                PrintSgmlTagString(fout, mode, indent, 
++                                sgmltag_type, "<literal>");
++        else if(strlen(node->element) == 2 && 
++                node->element[0] == 'h' && 
++                IsDigit(node->element[1])) {
++            if(sgmltag_type == SgmlTagStart) {        
++                int sectnum = DIGIT(node->element[1]) - 1;
++                char str[10];
++                if(seen_h[sectnum] == no)
++                    seen_h[sectnum] = yes;
++                else {
++                    int i = level;
++                    while(i > sectnum && seen_h[i] == yes) {
++                        if(i == 5)
++                            sprintf(str, "</simplesect>");
++                        else
++                            sprintf(str, "</sect%d>", SECT(i) + 1);
++                        PPrintString(fout, indent, str);
++                        seen_h[i] = no;
++                        --i;
++                    }
++                    if(sectnum == 5)
++                        sprintf(str, "</simplesect>");
++                    else
++                        sprintf(str, "</sect%d>", SECT(sectnum) + 1);
++                    PPrintString(fout, indent, str);
++                }
++                /* H1 is not the first level 
++                   like the curses man2html pages */
++                if(level == -1 && sectnum > 0) 
++                    startsect = sectnum;
++                
++                PrintSectTag(fout, indent, lexer, node, startsect);
++                level = sectnum;
++            }
++            else
++                PPrintString(fout, indent, "</title>");
++        }
++    }
++}
++
++void PrintSgml( Out *fout, uint mode, uint indent,
++                Lexer *lexer, Node *node)
++{   Node *content;
++
++    if (node == null)
++        return;
++    
++    if (node->type == TextNode) {
++        if(DescendantOf(node, tag_dd) && !DescendantOf(node, tag_a) &&
++            !DescendantOf(node, tag_p) && 
++            /* We have to descide on this table stuff later
++             * <entry> processing is complex */
++            !DescendantOf(node, tag_td) && !DescendantOf(node, tag_th)) 
++        /*  && wstrcasecmp(node->parent->element, "code") != 0)
++         above line may be needed later to properly convert <code> stuff */
++        {
++            PPrintString(fout, indent, "<para>");
++            PPrintText(fout, mode, indent, lexer, node->start, node->end);
++            PPrintString(fout, indent, "</para>");
++        }
++        else {
++            if(DescendantOf(node, tag_style))
++                fprintf(stderr, "PrintSgml: skipping style elements\n\n");
++            else
++                PPrintText(fout, mode, indent, lexer, node->start, node->end);
++        }
++    }
++    else if(node->type == CDATATag && EscapeCdata)
++        PPrintText(fout, mode, indent, lexer, node->start, node->end);
++    else if (node->type == CommentTag)
++        PPrintComment(fout, indent, lexer, node);
++    else if (node->type == RootNode)
++    {
++        for (content = node->content;
++                content != null;
++                content = content->next)
++           PrintSgml(fout, mode, indent, lexer, content);
++    }
++    else if (node->type == DocTypeTag)
++        PPrintDocType(fout, indent, lexer, node);
++    else if (node->type == CDATATag)
++        PPrintCDATA(fout, indent, lexer, node);
++    else if (node->type == SectionTag)
++        PPrintSection(fout, indent, lexer, node);
++    else if (node->type == AspTag || 
++             node->type == JsteTag ||
++             node->type == PhpTag )
++        PrintSgmlDefault(fout);
++    else if (node->type == ProcInsTag)
++        PPrintPI(fout, indent, lexer, node);
++    else if (node->type == XmlDecl)// && DbXml May be this is needed
++        PPrintXmlDecl(fout, indent, lexer, node);
++    else if (node->tag->model & CM_EMPTY || 
++            (node->type == StartEndTag && !xHTML))
++    {
++        if (!(node->tag->model & CM_INLINE))
++            PCondFlushLine(fout, indent);
++
++        if (MakeClean && node->tag == tag_wbr)
++            PPrintString(fout, indent, " ");
++        else
++            PrintSgmlTag(fout, mode, indent, lexer, node, SgmlTagStart);
++    }
++    else {
++        if (node->type == StartEndTag)
++            node->type = StartTag;
++        
++        if (node->tag && node->tag->parser == ParsePre)
++        {
++            PCondFlushLine(fout, indent);
++
++            indent = 0;
++            PCondFlushLine(fout, indent);
++
++            PrintSgmlTag(fout, mode, indent, lexer, node, SgmlTagStart);
++            PFlushLine(fout, indent);
++
++            for (content = node->content;
++                    content != null;
++                    content = content->next)
++                PrintSgml(fout, (mode | PREFORMATTED | NOWRAP), 
++                            indent, lexer, content);
++            
++            PCondFlushLine(fout, indent);
++            PrintSgmlTag(fout, mode, indent, lexer, node, SgmlTagEnd);
++            PFlushLine(fout, indent);
++
++            if (IndentContent == no && node->next != null)
++                PFlushLine(fout, indent);
++        }
++        else if (node->tag->model & CM_INLINE)
++        {   PrintSgmlTag(fout, mode, indent, lexer, node, SgmlTagStart);
++
++            if (ShouldIndent(node))
++            {
++                PCondFlushLine(fout, indent);
++                indent += spaces;
++
++                for (content = node->content;
++                        content != null;
++                        content = content->next)
++                    PrintSgml(fout, mode, indent, lexer, content);
++
++                PCondFlushLine(fout, indent);
++                indent -= spaces;
++                PCondFlushLine(fout, indent);
++            }
++            else
++            {
++
++                for (content = node->content;
++                        content != null;
++                        content = content->next)
++                    PrintSgml(fout, mode, indent, lexer, content);
++            }
++
++            PrintSgmlTag(fout, mode, indent, lexer, node, SgmlTagEnd);
++        }
++        else 
++        {   PCondFlushLine(fout, indent);
++            if (SmartIndent && node->prev != null)
++                PFlushLine(fout, indent);
++
++            PrintSgmlTag(fout, mode ,indent, lexer, node, SgmlTagStart);
++            if (ShouldIndent(node))
++                PCondFlushLine(fout, indent);
++            else if (node->tag->model & CM_HTML || 
++                     node->tag == tag_noframes ||
++                    (node->tag->model & CM_HEAD && !(node->tag == tag_title)))
++                PFlushLine(fout, indent);
++
++            if (ShouldIndent(node))
++            {   PCondFlushLine(fout, indent);
++                indent += spaces;
++
++                for (content = node->content;
++                        content != null;
++                        content = content->next)
++                    PrintSgml(fout, mode, indent, lexer, content);
++                PCondFlushLine(fout, indent);
++                indent -= spaces;
++                PCondFlushLine(fout, indent);
++            }
++            else
++            {   Node *last;
++                last = null;
++                for (content = node->content;
++                        content != null;
++                        content = content->next) {
++                    /* kludge for naked text before block level tag */
++                    if (last && !IndentContent && last->type == TextNode &&
++                        content->tag && !(content->tag->model & CM_INLINE) )
++                    {
++                        /* PFlushLine(fout, indent); */
++                        PFlushLine(fout, indent);
++                    }
++
++                    PrintSgml(fout, mode, 
++                        (ShouldIndent(node) ? indent+spaces : indent), 
++                        lexer, content);
++                    last = content;
++                }
++            }
++            PrintSgmlTag(fout, mode, indent, lexer, node, SgmlTagEnd);
++            PFlushLine(fout, indent);
++            if (IndentContent == no &&
++                node->next != null &&
++                HideEndTags == no &&
++                (node->tag->model & (CM_BLOCK|CM_LIST|CM_DEFLIST|CM_TABLE)))
++                PFlushLine(fout, indent);
++        }
++    }
++}
++
+ void PPrintTree(Out *fout, uint mode, uint indent,
+                     Lexer *lexer, Node *node)
+ {
+@@ -2034,17 +2669,14 @@
+         PPrintJste(fout, indent, lexer, node);
+     else if (node->type == PhpTag)
+         PPrintPhp(fout, indent, lexer, node);
+-    else if ( node->tag->model & CM_EMPTY 
+-              || (node->type == StartEndTag && !xHTML) )
++    else if (node->tag->model & CM_EMPTY || (node->type == StartEndTag && !xHTML))
+     {
+         PCondFlushLine(fout, indent);
+         PPrintTag(lexer, fout, mode, indent, node);
+         PFlushLine(fout, indent);
+ 
+-        /* CPR: folks don't want so much vertical spacing in XML
+         if (node->next)
+             PFlushLine(fout, indent);
+-        */
+     }
+     else /* some kind of container element */
+     {
+@@ -2076,7 +2708,7 @@
+ 
+         PPrintTag(lexer, fout, mode, indent, node);
+ 
+-        if ( !mixed && node->content )
++        if (!mixed)
+             PFlushLine(fout, indent);
+  
+         for (content = node->content;
+@@ -2084,16 +2716,14 @@
+                 content = content->next)
+             PPrintXMLTree(fout, mode, cindent, lexer, content);
+ 
+-        if ( !mixed && node->content )
++        if (!mixed)
+             PCondFlushLine(fout, cindent);
+ 
+         PPrintEndTag(fout, mode, indent, node);
+         PCondFlushLine(fout, indent);
+ 
+-        /* CPR: folks don't want so much vertical spacing in XML
+         if (node->next)
+             PFlushLine(fout, indent);
+-        */
+     }
+ }
+ 
+--- /cise/tmp/ppadala/tidy/src/tab2space.c	Wed Feb  6 04:09:37 2002
++++ src/tab2space.c	Sat Jul  6 23:50:55 2002
+@@ -2,7 +2,7 @@
+ #include <stdlib.h> 
+ #include <string.h>
+ 
+-#ifndef __BEOS__
++#if !(defined(__BEOS__) || defined(linux))
+ typedef unsigned int uint;
+ #endif
+ typedef unsigned char byte;
+--- /cise/tmp/ppadala/tidy/src/tidy.c	Sun Jul  7 23:29:25 2002
++++ src/tidy.c	Fri Jul 19 01:22:54 2002
+@@ -1853,6 +1853,10 @@
+                 IndentContent = yes;
+                 SmartIndent = yes;
+             }
++            else if (wstrcasecmp(arg, "dbsgml") == 0)
++                DbSgml = yes;
++            else if(wstrcasecmp(arg, "dbxml") == 0)
++                DbXml = yes;
+             else if (wstrcasecmp(arg, "omit") == 0)
+                 HideEndTags = yes;
+             else if (wstrcasecmp(arg, "upper") == 0)
+@@ -2180,6 +2184,28 @@
+             else
+             {
+                 lexer->warnings = 0;
++                
++                if (DbSgml || DbXml) {
++                    char *str;
++
++                    if(DbSgml) 
++                        str = "article PUBLIC \"-//OASIS//DTD DocBook V4.1//EN\"";
++                    else 
++                        str = "article PUBLIC \"-//OASIS//DTD DocBk XML V4.1.2 //EN\"";
++
++                    EncloseBodyText = yes; /* We want those <p>s */
++                    EncloseBlockText = yes;
++                    LogicalEmphasis = yes;
++                    DropFontTags = yes; /* <font> .. </font> are not needed */
++                    
++                    /* May be this should be decided by user */
++                    QuoteMarks = yes; 
++
++                    doctype_mode = doctype_user;
++                    /* TidyDeInit does MemFree(doctype_str) if it's != NULL */
++                    doctype_str = MemAlloc(wstrlen(str));
++                    wstrcpy(doctype_str, str);
++                }
+ 
+                 document = ParseDocument(lexer);
+ 
+@@ -2226,6 +2252,10 @@
+                 {
+                     if (xHTML)
+                         SetXHTMLDocType(lexer, document);
++                    else if(DbSgml)
++                        SetSgmlDocType(lexer, document);
++                    else if(DbXml)
++                        SetSgmlDocType(lexer, document);
+                     else
+                         FixDocType(lexer, document);
+ 
+@@ -2247,7 +2277,7 @@
+                 }
+ 
+                 /* ensure presence of initial <?XML version="1.0"?> */
+-                if (XmlOut && XmlPi)
++                if ((XmlOut && XmlPi) || DbXml)
+                     FixXmlDecl(lexer, document);
+ 
+                 /*
+@@ -2381,9 +2411,12 @@
+                     /* Feature request #434940 - fix by Dave Raggett/Ignacio Vazquez-Abrams 21 Jun 01 */
+                     else if (BodyOnly)
+                         PrintBody(&out, lexer, document);
+-                    else
+-                        PPrintTree(&out, null, 0, lexer, document);
+-
++                    else {
++                        if(DbSgml || DbXml) 
++                            PrintSgml(&out, null, 0, lexer, document);
++                        else 
++                            PPrintTree(&out, null, 0, lexer, document);
++                    }
+                     PFlushLine(&out, 0);
+                 }
+ 
diff --git a/app-text/htmltidy/htmltidy-2.7.18-r1.ebuild b/app-text/htmltidy/htmltidy-2.7.18-r1.ebuild
new file mode 100644
index 000000000000..e1b76f8ed50a
--- /dev/null
+++ b/app-text/htmltidy/htmltidy-2.7.18-r1.ebuild
@@ -0,0 +1,46 @@
+# Copyright 1999-2002 Gentoo Technologies, Inc.
+# Distributed under the terms of the GNU General Public License v2
+# $Header: /var/cvsroot/gentoo-x86/app-text/htmltidy/htmltidy-2.7.18-r1.ebuild,v 1.1 2003/01/03 05:43:19 satai Exp $ /home/cvsroot/gentoo-x86/app-text/htmltidy/htmltidy-2.7.18.ebuild, v 1.5 2002/07/25 11:07:00 cybersystem Exp $
+
+# convert from normalized gentoo version number to htmltidy's wacky date thing
+
+SLOT="0"
+LICENSE="GPL-2"
+KEYWORDS="x86 ppc sparc"
+
+parts=(${PV//./ })
+vers=$(printf "%02d%02d%02d" ${parts[0]} ${parts[1]} ${parts[2]})
+MY_P=tidy_src_${vers}
+S=${WORKDIR}/tidy
+
+DESCRIPTION="fix mistakes and tidy up sloppy editing in HTML and XML"
+SRC_URI="http://tidy.sourceforge.net/src/${MY_P}.tgz"
+HOMEPAGE="http://tidy.sourceforge.net/"
+DEPEND=""
+
+src_unpack() {
+	unpack ${MY_P}.tgz
+	cd ${S}
+	( use xml || use sgml ) &&
+		patch -p0 < ${FILESDIR}/htmltidy-dbpatch.diff
+
+	# skip "chown"
+	t=Makefile
+	cp $t $t.orig
+	sed 's:chgrp\|chown:#&:' $t.orig > $t
+}
+
+src_compile() {
+	emake \
+		OTHERCFLAGS="${CFLAGS}" \
+		|| die "emake failed"
+}
+
+src_install() {
+	dodir /usr/bin
+	dodir /usr/share/man/man1
+	make \
+		INSTALLDIR="${D}/usr/" \
+		MANPAGESDIR="${D}/usr/share/man" \
+		install || die
+}