kvoctrain/kvoctrain/kvt-core/kvt-xml/XmlReader.cpp

   1 /* -*- C++ -*-
   2
   3   $Id$
   4
   5   This file is part of KIllustrator.
   6   Copyright (C) 1998 Kai-Uwe Sattler (kus@iti.cs.uni-magdeburg.de)
   7
   8   modified for kvoctrain by Ewald Arnold kvoctrain@ewald-arnold.dein April ´99
   9
  10   -----------------------------------------------------------------------
  11
  12   -----------------------------------------------------------------------
  13
  14   This program is free software; you can redistribute it and/or modify
  15   it under the terms of the GNU Library General Public License as
  16   published by
  17   the Free Software Foundation; either version 2 of the License, or
  18   (at your option) any later version.
  19
  20   This program is distributed in the hope that it will be useful,
  21   but WITHOUT ANY WARRANTY; without even the implied warranty of
  22   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  23   GNU General Public License for more details.
  24
  25   You should have received a copy of the GNU Library General Public License
  26   along with this program; if not, write to the Free Software
  27   Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  28
  29 */
  30
  31 #include <iostream>
  32 using namespace std;
  33 #include "XmlReader.h"
  34
  35 XmlReader::XmlReader (KOXML_ISTREAM& is)
  36   : tokenizer (is)
  37 {
  38 }
  39
  40 XmlReader::~XmlReader () {
  41 }
  42
  43 bool XmlReader::validHeader () {
  44   // Check for: <? xml version="1.0" ?>
  45   if (tokenizer.nextToken () != XmlTokenizer::Tok_Lt)
  46     return false;
  47   if (tokenizer.nextToken () != XmlTokenizer::Tok_QSign)
  48     return false;
  49
  50   if (tokenizer.nextToken () != XmlTokenizer::Tok_Symbol)
  51     return false;
  52   else if (tokenizer.element () != "xml")
  53     return false;
  54
  55   if (tokenizer.nextToken () != XmlTokenizer::Tok_Symbol)
  56     return false;
  57   else if (tokenizer.element () != "version")
  58     return false;
  59   if (tokenizer.nextToken () != XmlTokenizer::Tok_Eq)
  60     return false;
  61   if (tokenizer.nextToken () != XmlTokenizer::Tok_String)
  62     return false;
  63   else if (tokenizer.element () != "1.0")
  64     return false;
  65
  66   if (tokenizer.nextToken () != XmlTokenizer::Tok_QSign)
  67     return false;
  68   if (tokenizer.nextToken () != XmlTokenizer::Tok_Gt)
  69     return false;
  70
  71   // Check for: <(!)doctype symbol system string>
  72
  73   if (tokenizer.nextToken () != XmlTokenizer::Tok_Lt)
  74     return false;
  75
  76   XmlTokenizer::Token token = tokenizer.nextToken ();
  77   if (token == XmlTokenizer::Tok_Exclam)
  78     token = tokenizer.nextToken ();
  79
  80   if (token != XmlTokenizer::Tok_Symbol)
  81     return false;
  82   else if (tokenizer.element().lower() != "doctype")
  83     return false;
  84
  85   if (tokenizer.nextToken () != XmlTokenizer::Tok_Symbol)
  86     return false;
  87   s_dtype = tokenizer.element ();
  88
  89   if (tokenizer.nextToken () != XmlTokenizer::Tok_Symbol)
  90     return false;
  91   else if (tokenizer.element().lower() != "system")
  92     return false;
  93
  94   if (tokenizer.nextToken () != XmlTokenizer::Tok_String)
  95     return false;
  96   s_dtd = tokenizer.element ();
  97
  98   if (tokenizer.nextToken () != XmlTokenizer::Tok_Gt)
  99     return false;
 100
 101   return true;
 102 }
 103
 104 const KOXML_STRING& XmlReader::doctype () const {
 105   return s_dtype;
 106 }
 107
 108 const KOXML_STRING& XmlReader::dtd () const {
 109   return s_dtd;
 110 }
 111
 112 const KOXML_STRING& XmlReader::getText () {
 113   return text;
 114 }
 115
 116 bool XmlReader::readElement (XmlElement& elem) {
 117   XmlTokenizer::Token tok;
 118   bool result = false;
 119   bool ready = false;
 120
 121   elem.reset ();
 122
 123   while (! ready) {
 124     tok = tokenizer.nextToken ();
 125     if (tok == XmlTokenizer::Tok_Lt) {
 126       tok = tokenizer.nextToken ();
 127       if (tok == XmlTokenizer::Tok_Comment) {
 128         // skip comment
 129       }
 130       else {
 131         if (tok == XmlTokenizer::Tok_Slash)
 132           result = parseEndElement (elem);
 133         else if (tok == XmlTokenizer::Tok_Symbol)
 134           result = parseElement (tokenizer.element (), elem);
 135         else if (tok == XmlTokenizer::Tok_EOF)
 136           result = false;
 137         ready = true;
 138       }
 139     }
 140     else if (tok == XmlTokenizer::Tok_Text) {
 141       elem.tagId = "#PCDATA";
 142       text = tokenizer.element ();
 143       result = true;
 144       ready = true;
 145     }
 146     else
 147       return false;
 148   }
 149   return result;
 150 }
 151
 152 bool XmlReader::parseEndElement (XmlElement& elem) {
 153   bool result = false;
 154
 155   XmlTokenizer::Token tok = tokenizer.nextToken ();
 156   if (tok == XmlTokenizer::Tok_Symbol) {
 157     KOXML_STRING tag = tokenizer.element ();
 158     tok = tokenizer.nextToken ();
 159     if (tok == XmlTokenizer::Tok_Gt) {
 160       elem.tagId = tag;
 161       elem.endTag = true;
 162       elem.closed = true;
 163       result = true;
 164     }
 165   }
 166   return result;
 167 }
 168
 169 bool XmlReader::parseElement (const KOXML_STRING& id, XmlElement& elem) {
 170   KOXML_STRING tag = id;
 171   bool closed = false;
 172   std::list<XmlAttribute> attrib_list;
 173
 174   XmlTokenizer::Token tok = tokenizer.nextToken ();
 175   while (tok != XmlTokenizer::Tok_Gt) {
 176     if (tok == XmlTokenizer::Tok_Slash) {
 177       if (closed)
 178         return false;
 179       else
 180         closed = true;
 181     }
 182     else if (tok == XmlTokenizer::Tok_Symbol) {
 183       if (! attrib_list.empty () || closed)
 184         return false;
 185       tokenizer.unget ();
 186       if (! readAttributes (attrib_list))
 187         return false;
 188     }
 189     else
 190       return false;
 191
 192     tok = tokenizer.nextToken ();
 193   }
 194   elem.tagId = tag;
 195   elem.endTag = false;
 196   elem.closed = closed;
 197   elem.attribs = attrib_list;
 198   return true;
 199 }
 200
 201 bool XmlReader::readAttributes (std::list<XmlAttribute>& attrib_list) {
 202   XmlTokenizer::Token tok = tokenizer.nextToken ();
 203
 204   while (tok != XmlTokenizer::Tok_Gt) {
 205     if (tok == XmlTokenizer::Tok_Symbol) {
 206       KOXML_STRING id = tokenizer.element ();
 207       if (tokenizer.nextToken () == XmlTokenizer::Tok_Eq) {
 208         if (tokenizer.nextToken () == XmlTokenizer::Tok_String) {
 209
 210           // un-escape dangerous characters in reverted order
 211           KOXML_STRING val = tokenizer.element();
 212           int pos = 0;
 213           while ((pos = val.find("&quot;", pos)) >= 0) {
 214             KOXML_STRING_REMOVE( val, pos, 6);
 215             KOXML_STRING_INSERT( val, pos, "\"");
 216             pos += 1; // skip "
 217           }
 218           pos = 0;
 219           while ((pos = val.find("&lf;", pos)) >= 0) {
 220             KOXML_STRING_REMOVE( val, pos, 4);
 221             KOXML_STRING_INSERT( val, pos, "\r");
 222             pos += 1; // skip \r
 223           }
 224           pos = 0;
 225           while ((pos = val.find("&nl;", pos)) >= 0) {
 226             KOXML_STRING_REMOVE( val, pos, 4);
 227             KOXML_STRING_INSERT( val, pos, "\n");
 228             pos += 1; // skip \n
 229           }
 230           pos = 0;
 231           while ((pos = val.find("&lt;", pos)) >= 0) {
 232             KOXML_STRING_REMOVE( val, pos, 4);
 233             KOXML_STRING_INSERT( val, pos, "<");
 234             pos += 1; // skip <
 235           }
 236           pos = 0;
 237           while ((pos = val.find ("&amp;", pos)) >= 0) {
 238             KOXML_STRING_REMOVE( val, pos+1, 4);
 239             pos += 1; // skip &
 240           }
 241
 242           XmlAttribute attrib (id, val);
 243           attrib_list.push_back (attrib);
 244         }
 245         else {
 246           cerr << "invalid attribute value" << endl;
 247           return false;
 248         }
 249       }
 250       else {
 251         cerr << "missing '='" << endl;
 252         return false;
 253       }
 254     }
 255     else if (tok == XmlTokenizer::Tok_Slash) {
 256       break;
 257     }
 258     else {
 259       cerr << "invalid attribute name: " << (int) tok << endl;
 260       return false;
 261     }
 262     tok = tokenizer.nextToken ();
 263   }
 264   tokenizer.unget ();
 265   return true;
 266 }