parser/htmlparser/public/nsIParser.h

   1 /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
   2 /* vim: set ts=2 sw=2 et tw=78: */
   3 /* ***** BEGIN LICENSE BLOCK *****
   4  * Version: MPL 1.1/GPL 2.0/LGPL 2.1
   5  *
   6  * The contents of this file are subject to the Mozilla Public License Version
   7  * 1.1 (the "License"); you may not use this file except in compliance with
   8  * the License. You may obtain a copy of the License at
   9  * http://www.mozilla.org/MPL/
  10  *
  11  * Software distributed under the License is distributed on an "AS IS" basis,
  12  * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
  13  * for the specific language governing rights and limitations under the
  14  * License.
  15  *
  16  * The Original Code is mozilla.org code.
  17  *
  18  * The Initial Developer of the Original Code is
  19  * Netscape Communications Corporation.
  20  * Portions created by the Initial Developer are Copyright (C) 1998
  21  * the Initial Developer. All Rights Reserved.
  22  *
  23  * Contributor(s):
  24  *
  25  * Alternatively, the contents of this file may be used under the terms of
  26  * either of the GNU General Public License Version 2 or later (the "GPL"),
  27  * or the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
  28  * in which case the provisions of the GPL or the LGPL are applicable instead
  29  * of those above. If you wish to allow use of your version of this file only
  30  * under the terms of either the GPL or the LGPL, and not to allow others to
  31  * use your version of this file under the terms of the MPL, indicate your
  32  * decision by deleting the provisions above and replace them with the notice
  33  * and other provisions required by the GPL or the LGPL. If you do not delete
  34  * the provisions above, a recipient may use your version of this file under
  35  * the terms of any one of the MPL, the GPL or the LGPL.
  36  *
  37  * ***** END LICENSE BLOCK ***** */
  38 #ifndef NS_IPARSER___
  39 #define NS_IPARSER___
  40
  41
  42 /**
  43  * MODULE NOTES:
  44  *
  45  *  This class defines the iparser interface. This XPCOM
  46  *  inteface is all that parser clients ever need to see.
  47  *
  48  **/
  49
  50 #include "nsISupports.h"
  51 #include "nsIStreamListener.h"
  52 #include "nsIDTD.h"
  53 #include "nsStringGlue.h"
  54 #include "nsTArray.h"
  55
  56 // {506527cc-d832-420b-ba3a-80c05aa105f4}
  57 #define NS_IPARSER_IID \
  58 { 0x506527cc, 0xd832, 0x420b, \
  59   { 0xba, 0x3a, 0x80, 0xc0, 0x5a, 0xa1, 0x05, 0xf4 } }
  60
  61
  62 // {41421C60-310A-11d4-816F-000064657374}
  63 #define NS_IDEBUG_DUMP_CONTENT_IID \
  64 { 0x41421c60, 0x310a, 0x11d4, \
  65 { 0x81, 0x6f, 0x0, 0x0, 0x64, 0x65, 0x73, 0x74 } }
  66
  67 class nsIContentSink;
  68 class nsIRequestObserver;
  69 class nsIParserFilter;
  70 class nsString;
  71 class nsIURI;
  72 class nsIChannel;
  73
  74 enum eParserCommands {
  75   eViewNormal,
  76   eViewSource,
  77   eViewFragment,
  78   eViewErrors
  79 };
  80
  81 enum eParserDocType {
  82   ePlainText = 0,
  83   eXML,
  84   eHTML_Quirks,
  85   eHTML3_Quirks, // separate, for editor output, since HTML pre-4.0 lacks tbody
  86   eHTML_Strict
  87 };
  88
  89
  90 // define Charset source constants
  91 // note: the value order defines the priority; higher numbers take priority
  92 #define kCharsetUninitialized           0
  93 #define kCharsetFromWeakDocTypeDefault  1
  94 #define kCharsetFromUserDefault         2
  95 #define kCharsetFromDocTypeDefault      3
  96 #define kCharsetFromCache               4
  97 #define kCharsetFromParentFrame         5
  98 #define kCharsetFromBookmarks           6
  99 #define kCharsetFromAutoDetection       7
 100 #define kCharsetFromHintPrevDoc         8
 101 #define kCharsetFromMetaTag             9
 102 #define kCharsetFromByteOrderMark      10
 103 #define kCharsetFromChannel            11
 104 #define kCharsetFromOtherComponent     12
 105 // Levels below here will be forced onto childframes too
 106 #define kCharsetFromParentForced       13
 107 #define kCharsetFromUserForced         14
 108 #define kCharsetFromPreviousLoading    15
 109
 110 enum eStreamState {eNone,eOnStart,eOnDataAvail,eOnStop};
 111
 112 /**
 113  *  FOR DEBUG PURPOSE ONLY
 114  *
 115  *  Use this interface to query objects that contain content information.
 116  *  Ex. Parser can trigger dump content by querying the sink that has
 117  *      access to the content.
 118  *
 119  *  @update  harishd 05/25/00
 120  */
 121 class nsIDebugDumpContent : public nsISupports {
 122 public:
 123   NS_DECLARE_STATIC_IID_ACCESSOR(NS_IDEBUG_DUMP_CONTENT_IID)
 124   NS_IMETHOD DumpContentModel()=0;
 125 };
 126
 127 NS_DEFINE_STATIC_IID_ACCESSOR(nsIDebugDumpContent, NS_IDEBUG_DUMP_CONTENT_IID)
 128
 129 /**
 130  *  This class defines the iparser interface. This XPCOM
 131  *  inteface is all that parser clients ever need to see.
 132  */
 133 class nsIParser : public nsISupports {
 134   public:
 135
 136     NS_DECLARE_STATIC_IID_ACCESSOR(NS_IPARSER_IID)
 137
 138     /**
 139      * Select given content sink into parser for parser output
 140      * @update  gess5/11/98
 141      * @param   aSink is the new sink to be used by parser
 142      * @return
 143      */
 144     NS_IMETHOD_(void) SetContentSink(nsIContentSink* aSink)=0;
 145
 146
 147     /**
 148      * retrieve the sink set into the parser
 149      * @update  gess5/11/98
 150      * @return  current sink
 151      */
 152     NS_IMETHOD_(nsIContentSink*) GetContentSink(void)=0;
 153
 154     /**
 155      *  Call this method once you've created a parser, and want to instruct it
 156            *  about the command which caused the parser to be constructed. For example,
 157      *  this allows us to select a DTD which can do, say, view-source.
 158      *
 159      *  @update  gess 3/25/98
 160      *  @param   aCommand -- ptrs to string that contains command
 161      *  @return  nada
 162      */
 163     NS_IMETHOD_(void) GetCommand(nsCString& aCommand)=0;
 164     NS_IMETHOD_(void) SetCommand(const char* aCommand)=0;
 165     NS_IMETHOD_(void) SetCommand(eParserCommands aParserCommand)=0;
 166
 167     /**
 168      *  Call this method once you've created a parser, and want to instruct it
 169      *  about what charset to load
 170      *
 171      *  @update  ftang 4/23/99
 172      *  @param   aCharset- the charest of a document
 173      *  @param   aCharsetSource- the soure of the chares
 174      *  @return  nada
 175      */
 176     NS_IMETHOD_(void) SetDocumentCharset(const nsACString& aCharset, PRInt32 aSource)=0;
 177     NS_IMETHOD_(void) GetDocumentCharset(nsACString& oCharset, PRInt32& oSource)=0;
 178
 179     NS_IMETHOD_(void) SetParserFilter(nsIParserFilter* aFilter) = 0;
 180
 181     /**
 182      * Get the channel associated with this parser
 183      * @update harishd,gagan 07/17/01
 184      * @param aChannel out param that will contain the result
 185      * @return NS_OK if successful
 186      */
 187     NS_IMETHOD GetChannel(nsIChannel** aChannel) = 0;
 188
 189     /**
 190      * Get the DTD associated with this parser
 191      * @update vidur 9/29/99
 192      * @param aDTD out param that will contain the result
 193      * @return NS_OK if successful, NS_ERROR_FAILURE for runtime error
 194      */
 195     NS_IMETHOD GetDTD(nsIDTD** aDTD) = 0;
 196
 197     /**************************************************************************
 198      *  Parse methods always begin with an input source, and perform
 199      *  conversions until you wind up being emitted to the given contentsink
 200      *  (which may or may not be a proxy for the NGLayout content model).
 201      ************************************************************************/
 202
 203     // Call this method to resume the parser from the blocked state.
 204     NS_IMETHOD ContinueParsing() = 0;
 205
 206     // Call this method to resume the parser from an unblocked state.
 207     // This can happen, for example, if parsing was interrupted and then the
 208     // consumer needed to restart the parser without waiting for more data.
 209     // This also happens after loading scripts, which unblock the parser in
 210     // order to process the output of document.write() and then need to
 211     // continue on with the page load on an enabled parser.
 212     NS_IMETHOD ContinueInterruptedParsing() = 0;
 213
 214     // Stops parsing temporarily.
 215     NS_IMETHOD_(void) BlockParser() = 0;
 216
 217     // Open up the parser for tokenization, building up content
 218     // model..etc. However, this method does not resume parsing
 219     // automatically. It's the callers' responsibility to restart
 220     // the parsing engine.
 221     NS_IMETHOD_(void) UnblockParser() = 0;
 222
 223     NS_IMETHOD_(PRBool) IsParserEnabled() = 0;
 224     NS_IMETHOD_(PRBool) IsComplete() = 0;
 225
 226     NS_IMETHOD Parse(nsIURI* aURL,
 227                      nsIRequestObserver* aListener = nsnull,
 228                      void* aKey = 0,
 229                      nsDTDMode aMode = eDTDMode_autodetect) = 0;
 230     NS_IMETHOD Parse(const nsAString& aSourceBuffer,
 231                      void* aKey,
 232                      const nsACString& aMimeType,
 233                      PRBool aLastCall,
 234                      nsDTDMode aMode = eDTDMode_autodetect) = 0;
 235
 236     // Return a key, suitable for passing into one of the Parse methods above,
 237     // that will cause this parser to use the root context.
 238     NS_IMETHOD_(void *) GetRootContextKey() = 0;
 239
 240     NS_IMETHOD Terminate(void) = 0;
 241
 242     /**
 243      * This method gets called when you want to parse a fragment of HTML or XML
 244      * surrounded by the context |aTagStack|. It requires that the parser have
 245      * been given a fragment content sink.
 246      *
 247      * @param aSourceBuffer The XML or HTML that hasn't been parsed yet.
 248      * @param aKey The key used by the parser.
 249      * @param aTagStack The context of the source buffer.
 250      * @param aXMLMode Whether this is XML or HTML
 251      * @param aContentType The content-type of this document.
 252      * @param aMode The DTDMode that the parser should parse this fragment in.
 253      * @return Success or failure.
 254      */
 255     NS_IMETHOD ParseFragment(const nsAString& aSourceBuffer,
 256                              void* aKey,
 257                              nsTArray<nsString>& aTagStack,
 258                              PRBool aXMLMode,
 259                              const nsACString& aContentType,
 260                              nsDTDMode aMode = eDTDMode_autodetect) = 0;
 261
 262     /**
 263      * This method gets called when the tokens have been consumed, and it's time
 264      * to build the model via the content sink.
 265      * @update  gess5/11/98
 266      * @return  error code -- 0 if model building went well .
 267      */
 268     NS_IMETHOD BuildModel(void) = 0;
 269
 270
 271     /**
 272      *  Retrieve the parse mode from the parser...
 273      *
 274      *  @update  gess 6/9/98
 275      *  @return  ptr to scanner
 276      */
 277     NS_IMETHOD_(nsDTDMode) GetParseMode(void) = 0;
 278
 279     /**
 280      *  Call this method to cancel any pending parsing events.
 281      *  Parsing events may be pending if all of the document's content
 282      *  has been passed to the parser but the parser has been interrupted
 283      *  because processing the tokens took too long.
 284      *
 285      *  @update  kmcclusk 05/18/01
 286      *  @return  NS_OK if succeeded else ERROR.
 287      */
 288
 289     NS_IMETHOD CancelParsingEvents() = 0;
 290
 291     virtual void Reset() = 0;
 292 };
 293
 294 NS_DEFINE_STATIC_IID_ACCESSOR(nsIParser, NS_IPARSER_IID)
 295
 296 /* ===========================================================*
 297   Some useful constants...
 298  * ===========================================================*/
 299
 300 #include "prtypes.h"
 301 #include "nsError.h"
 302
 303 #define NS_ERROR_HTMLPARSER_EOF                            NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1000)
 304 #define NS_ERROR_HTMLPARSER_UNKNOWN                        NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1001)
 305 #define NS_ERROR_HTMLPARSER_CANTPROPAGATE                  NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1002)
 306 #define NS_ERROR_HTMLPARSER_CONTEXTMISMATCH                NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1003)
 307 #define NS_ERROR_HTMLPARSER_BADFILENAME                    NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1004)
 308 #define NS_ERROR_HTMLPARSER_BADURL                         NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1005)
 309 #define NS_ERROR_HTMLPARSER_INVALIDPARSERCONTEXT           NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1006)
 310 #define NS_ERROR_HTMLPARSER_INTERRUPTED                    NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1007)
 311 #define NS_ERROR_HTMLPARSER_BLOCK                          NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1008)
 312 #define NS_ERROR_HTMLPARSER_BADTOKENIZER                   NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1009)
 313 #define NS_ERROR_HTMLPARSER_BADATTRIBUTE                   NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1010)
 314 #define NS_ERROR_HTMLPARSER_UNRESOLVEDDTD                  NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1011)
 315 #define NS_ERROR_HTMLPARSER_MISPLACEDTABLECONTENT          NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1012)
 316 #define NS_ERROR_HTMLPARSER_BADDTD                         NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1013)
 317 #define NS_ERROR_HTMLPARSER_BADCONTEXT                     NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1014)
 318 #define NS_ERROR_HTMLPARSER_STOPPARSING                    NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1015)
 319 #define NS_ERROR_HTMLPARSER_UNTERMINATEDSTRINGLITERAL      NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1016)
 320 #define NS_ERROR_HTMLPARSER_HIERARCHYTOODEEP               NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1017)
 321 #define NS_ERROR_HTMLPARSER_FAKE_ENDTAG                    NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1018)
 322 #define NS_ERROR_HTMLPARSER_INVALID_COMMENT                NS_ERROR_GENERATE_FAILURE(NS_ERROR_MODULE_HTMLPARSER,1019)
 323
 324 #define NS_ERROR_HTMLPARSER_CONTINUE              NS_OK
 325
 326
 327 const PRUint32  kEOF              = NS_ERROR_HTMLPARSER_EOF;
 328 const PRUint32  kUnknownError     = NS_ERROR_HTMLPARSER_UNKNOWN;
 329 const PRUint32  kCantPropagate    = NS_ERROR_HTMLPARSER_CANTPROPAGATE;
 330 const PRUint32  kContextMismatch  = NS_ERROR_HTMLPARSER_CONTEXTMISMATCH;
 331 const PRUint32  kBadFilename      = NS_ERROR_HTMLPARSER_BADFILENAME;
 332 const PRUint32  kBadURL           = NS_ERROR_HTMLPARSER_BADURL;
 333 const PRUint32  kInvalidParserContext = NS_ERROR_HTMLPARSER_INVALIDPARSERCONTEXT;
 334 const PRUint32  kBlocked          = NS_ERROR_HTMLPARSER_BLOCK;
 335 const PRUint32  kBadStringLiteral = NS_ERROR_HTMLPARSER_UNTERMINATEDSTRINGLITERAL;
 336 const PRUint32  kHierarchyTooDeep = NS_ERROR_HTMLPARSER_HIERARCHYTOODEEP;
 337 const PRUint32  kFakeEndTag       = NS_ERROR_HTMLPARSER_FAKE_ENDTAG;
 338 const PRUint32  kNotAComment      = NS_ERROR_HTMLPARSER_INVALID_COMMENT;
 339
 340 const PRUnichar  kNewLine          = '\n';
 341 const PRUnichar  kCR               = '\r';
 342 const PRUnichar  kLF               = '\n';
 343 const PRUnichar  kTab              = '\t';
 344 const PRUnichar  kSpace            = ' ';
 345 const PRUnichar  kQuote            = '"';
 346 const PRUnichar  kApostrophe       = '\'';
 347 const PRUnichar  kLessThan         = '<';
 348 const PRUnichar  kGreaterThan      = '>';
 349 const PRUnichar  kAmpersand        = '&';
 350 const PRUnichar  kForwardSlash     = '/';
 351 const PRUnichar  kBackSlash        = '\\';
 352 const PRUnichar  kEqual            = '=';
 353 const PRUnichar  kMinus            = '-';
 354 const PRUnichar  kPlus             = '+';
 355 const PRUnichar  kExclamation      = '!';
 356 const PRUnichar  kSemicolon        = ';';
 357 const PRUnichar  kHashsign         = '#';
 358 const PRUnichar  kAsterisk         = '*';
 359 const PRUnichar  kUnderbar         = '_';
 360 const PRUnichar  kComma            = ',';
 361 const PRUnichar  kLeftParen        = '(';
 362 const PRUnichar  kRightParen       = ')';
 363 const PRUnichar  kLeftBrace        = '{';
 364 const PRUnichar  kRightBrace       = '}';
 365 const PRUnichar  kQuestionMark     = '?';
 366 const PRUnichar  kLeftSquareBracket  = '[';
 367 const PRUnichar  kRightSquareBracket = ']';
 368 const PRUnichar kNullCh           = '\0';
 369
 370 // XXXbz these type defines should really just go away....  Until they
 371 // do, changes here should be reflected in nsContentDLF.cpp
 372 #define kHTMLTextContentType  "text/html"
 373 #define kXMLTextContentType   "text/xml"
 374 #define kXMLApplicationContentType "application/xml"
 375 #define kXHTMLApplicationContentType "application/xhtml+xml"
 376 #define kXULTextContentType   "application/vnd.mozilla.xul+xml"
 377 #define kRDFTextContentType   "text/rdf"
 378 #define kRDFApplicationContentType "application/rdf+xml"
 379 #define kXIFTextContentType   "text/xif"
 380 #define kPlainTextContentType "text/plain"
 381 #define kViewSourceCommand    "view-source"
 382 #define kViewFragmentCommand  "view-fragment"
 383 #define kTextCSSContentType   "text/css"
 384 #define kApplicationJSContentType "application/javascript"
 385 #define kApplicationXJSContentType "application/x-javascript"
 386 #define kTextECMAScriptContentType "text/ecmascript"
 387 #define kApplicationECMAScriptContentType "application/ecmascript"
 388 #define kTextJSContentType    "text/javascript"
 389 #define kSGMLTextContentType   "text/sgml"
 390 #define kSVGTextContentType   "image/svg+xml"
 391
 392 #define NS_IPARSER_FLAG_UNKNOWN_MODE         0x00000000
 393 #define NS_IPARSER_FLAG_QUIRKS_MODE          0x00000002
 394 #define NS_IPARSER_FLAG_STRICT_MODE          0x00000004
 395 #define NS_IPARSER_FLAG_AUTO_DETECT_MODE     0x00000010
 396 #define NS_IPARSER_FLAG_VIEW_NORMAL          0x00000020
 397 #define NS_IPARSER_FLAG_VIEW_SOURCE          0x00000040
 398 #define NS_IPARSER_FLAG_VIEW_ERRORS          0x00000080
 399 #define NS_IPARSER_FLAG_PLAIN_TEXT           0x00000100
 400 #define NS_IPARSER_FLAG_XML                  0x00000200
 401 #define NS_IPARSER_FLAG_HTML                 0x00000400
 402 #define NS_IPARSER_FLAG_SCRIPT_ENABLED       0x00000800
 403 #define NS_IPARSER_FLAG_FRAMES_ENABLED       0x00001000
 404
 405 #endif