Parsing.subproj/CFXMLInputStream.h

   1 /*
   2  * Copyright (c) 2003 Apple Computer, Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_HEADER_START@
   5  *
   6  * Copyright (c) 1999-2003 Apple Computer, Inc.  All Rights Reserved.
   7  *
   8  * This file contains Original Code and/or Modifications of Original Code
   9  * as defined in and that are subject to the Apple Public Source License
  10  * Version 2.0 (the 'License'). You may not use this file except in
  11  * compliance with the License. Please obtain a copy of the License at
  12  * http://www.opensource.apple.com/apsl/ and read it before using this
  13  * file.
  14  *
  15  * The Original Code and all software distributed under the License are
  16  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  17  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  18  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  19  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  20  * Please see the License for the specific language governing rights and
  21  * limitations under the License.
  22  *
  23  * @APPLE_LICENSE_HEADER_END@
  24  */
  25 /*      CFXMLInputStream.h
  26         Copyright (c) 2000-2003, Apple, Inc. All rights reserved.
  27 */
  28
  29 #if !defined(__COREFOUNDATION_CFXMLINPUTSTREAM__)
  30 #define __COREFOUNDATION_CFXMLINPUTSTREAM__ 1
  31
  32 #include <CoreFoundation/CFBase.h>
  33 #include "CFInternal.h"
  34 #include <CoreFoundation/CFString.h>
  35 #include <CoreFoundation/CFSet.h>
  36 #include <CoreFoundation/CFXMLNode.h>
  37
  38 struct __CFXMLNode {
  39     // additionalData currently always points off the bottom of this struct; we could just eliminate it.  Also, we may want to add a flags/version argument, and could use it to mark whether the node was the special one that CFXMLParser mucks with, as well as whether the allocator was "special" (could save us the alloc instance variable, below) -- REW, 3/8/2000
  40     CFRuntimeBase _cfBase;
  41     CFIndex version;
  42     CFXMLNodeTypeCode dataTypeID;
  43     CFStringRef dataString;
  44     void *additionalData;
  45 };
  46
  47 struct __CFXMLInputStream {
  48     CFDataRef data;               // The XML data
  49     CFURLRef url;                 // the source URL for the data
  50     CFStringEncoding encoding;    // the data's encoding
  51     const UInt8 *currentByte;     // pointer into data at the first byte not yet translated to a character
  52
  53     UniChar *charBuffer;          // the buffer of characters translated from data
  54     UniChar *currentChar;         // pointer into charBuffer at the current stream location.  MUST be NULL if there are no more characters in charBuffer to consume.
  55     UniChar *mark;                // The point at which the mark was dropped.  NULL if the mark is currently unset.
  56     UniChar *parserMark;          // mark available only for the parser's use
  57     CFIndex bufferLength;         // The number of meaningful characters in charBuffer
  58     CFIndex bufferCapacity;       // The current maximum capacity of charBuffer in UniChars
  59
  60     CFIndex charIndex, lineNum;   // location in the file
  61     UInt32 flags;                 // See #defines below for bit flag meanings
  62     CFMutableSetRef nameSet;             // set of all names we've encountered; used for uniquing
  63     CFMutableStringRef tempString;
  64
  65     CFAllocatorRef allocator; // This is unfortunate; this is always the same as the parser's allocator.  We'd like to get rid of it at some point, but that would mean adding an allocator to all the function calls, which means risking that the allocator passed in gets out-of-sync.  Maybe once we have CFStreams, we can encapsulate it all in that.  REW, 5/22/2000
  66 };
  67
  68 // whether the stream has been opened for reading
  69 #define STREAM_OPEN            0x1
  70 // whether the encoding matches ASCII over 0x0-0x7F
  71 #define ENCODING_MATCHES_ASCII 0x2
  72 // whether the encoding is Unicode with the "natural" byte ordering
  73 #define ENCODING_IS_UNICODE_NATURAL    0x4
  74 // whether the encoding is Unicode with the bytes swapped
  75 #define ENCODING_IS_UNICODE_SWAPPED    0x8
  76 // whether the stream has encountered an error in encodings.
  77 #define ENCODING_COMPOSITION_ERROR      0x10
  78
  79 typedef struct __CFXMLInputStream _CFXMLInputStream;
  80
  81 void _initializeInputStream(_CFXMLInputStream *stream, CFAllocatorRef alloc, CFURLRef dataSource, CFDataRef xmlData);
  82 Boolean _openInputStream(_CFXMLInputStream *stream); // None of the subsequent calls will work until the input stream has been opened
  83 void _freeInputStream(_CFXMLInputStream *stream);
  84
  85 CFStringEncoding _inputStreamGetEncoding(_CFXMLInputStream *stream);
  86 CFIndex _inputStreamCurrentLocation(_CFXMLInputStream *stream);
  87 CFIndex _inputStreamCurrentLine(_CFXMLInputStream *stream);
  88 Boolean _inputStreamAtEOF(_CFXMLInputStream *stream);
  89 Boolean _inputStreamComposingErrorOccurred(_CFXMLInputStream *stream);
  90
  91 Boolean _inputStreamPeekCharacter(_CFXMLInputStream *stream, UniChar *ch);
  92 Boolean _inputStreamGetCharacter(_CFXMLInputStream *stream, UniChar *ch);
  93 Boolean _inputStreamReturnCharacter(_CFXMLInputStream *stream, UniChar ch);
  94 void _inputStreamSetMark(_CFXMLInputStream *stream);
  95 void _inputStreamClearMark(_CFXMLInputStream *stream);
  96 void _inputStreamGetCharactersFromMark(_CFXMLInputStream *stream, CFMutableStringRef string);
  97 void _inputStreamBackUpToMark(_CFXMLInputStream *stream);
  98 void _inputStringInitialize(_CFXMLInputStream *stream, UniChar *characters, CFIndex length);
  99 CFIndex _inputStreamSkipWhitespace(_CFXMLInputStream *stream, CFMutableStringRef str);
 100 Boolean _inputStreamScanToCharacters(_CFXMLInputStream *stream, const UniChar *scanChars, CFIndex numChars, CFMutableStringRef str);
 101 Boolean _inputStreamMatchString(_CFXMLInputStream *stream, const UniChar *stringToMatch, CFIndex length);
 102 Boolean _inputStreamScanQuotedString(_CFXMLInputStream *stream, CFMutableStringRef str);
 103 Boolean _inputStreamScanXMLName(_CFXMLInputStream *stream, Boolean isNMToken, CFStringRef *str);
 104
 105 /* Returns the character index within the current line of the current parse location */
 106 /* To add someday -- CF_EXPORT
 107 CFIndex CFXMLParserGetOffsetInCurrentLine(CFXMLParserRef parser); */
 108
 109 #endif /* ! __COREFOUNDATION_CFXMLINPUTSTREAM__ */
 110