String.subproj/CFStringScanner.c

   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 /*      CFStringScanner.c
  26         Copyright 1999-2002, Apple, Inc. All rights reserved.
  27         Responsibility: Ali Ozer
  28 */
  29
  30 #include "CFInternal.h"
  31 #include <CoreFoundation/CFString.h>
  32 #if !defined(__MACOS8__)
  33 #include <sys/types.h>
  34 #endif
  35 #include <limits.h>
  36 #include <stdlib.h>
  37 #include <string.h>
  38
  39 CF_INLINE Boolean __CFCharacterIsADigit(UniChar ch) {
  40     return (ch >= '0' && ch <= '9') ? true : false;
  41 }
  42
  43 /* Returns -1 on illegal value */
  44 CF_INLINE SInt32 __CFCharacterNumericOrHexValue (UniChar ch) {
  45     if (ch >= '0' && ch <= '9') {
  46         return ch - '0';
  47     } else if (ch >= 'A' && ch <= 'F') {
  48         return ch + 10 - 'A';
  49     } else if (ch >= 'a' && ch <= 'f') {
  50         return ch + 10 - 'a';
  51     } else {
  52         return -1;
  53     }
  54 }
  55
  56 /* Returns -1 on illegal value */
  57 CF_INLINE SInt32 __CFCharacterNumericValue(UniChar ch) {
  58     return (ch >= '0' && ch <= '9') ? (ch - '0') : -1;
  59 }
  60
  61 CF_INLINE UniChar __CFStringGetFirstNonSpaceCharacterFromInlineBuffer(CFStringInlineBuffer *buf, SInt32 *indexPtr) {
  62     UniChar ch;
  63     while (__CFIsWhitespace(ch = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr))) (*indexPtr)++;
  64     return ch;
  65 }
  66
  67 /* result is int64_t or int, depending on doLonglong
  68 */
  69 __private_extern__ Boolean __CFStringScanInteger(CFStringInlineBuffer *buf, CFDictionaryRef locale, SInt32 *indexPtr, Boolean doLonglong, void *result) {
  70     Boolean doingLonglong = false;      /* Set to true if doLonglong, and we overflow an int... */
  71     Boolean neg = false;
  72     int intResult = 0;
  73     register int64_t longlongResult = 0;        /* ??? int64_t is slow when not in regs; I hope this does the right thing. */
  74     UniChar ch;
  75
  76     ch = __CFStringGetFirstNonSpaceCharacterFromInlineBuffer(buf, indexPtr);
  77
  78     if (ch == '-' || ch == '+') {
  79         neg = (ch == '-');
  80         (*indexPtr)++;
  81         ch = __CFStringGetFirstNonSpaceCharacterFromInlineBuffer(buf, indexPtr);
  82     }
  83
  84     if (! __CFCharacterIsADigit(ch)) return false;      /* No digits, bail out... */
  85     do {
  86         if (doingLonglong) {
  87             if ((longlongResult >= LLONG_MAX / 10) && ((longlongResult > LLONG_MAX / 10) || (__CFCharacterNumericValue(ch) - (neg ? 1 : 0) >= LLONG_MAX - longlongResult * 10))) {
  88                 /* ??? This might not handle LLONG_MIN correctly... */
  89                 longlongResult = neg ? LLONG_MIN : LLONG_MAX;
  90                 neg = false;
  91                 while (__CFCharacterIsADigit(ch = __CFStringGetCharacterFromInlineBufferAux(buf, ++(*indexPtr))));      /* Skip remaining digits */
  92             } else {
  93                 longlongResult = longlongResult * 10 + __CFCharacterNumericValue(ch);
  94                 ch = __CFStringGetCharacterFromInlineBufferAux(buf, ++(*indexPtr));
  95             }
  96         } else {
  97             if ((intResult >= INT_MAX / 10) && ((intResult > INT_MAX / 10) || (__CFCharacterNumericValue(ch) - (neg ? 1 : 0) >= INT_MAX - intResult * 10))) {
  98                 // Overflow, check for int64_t...
  99                 if (doLonglong) {
 100                     longlongResult = intResult;
 101                     doingLonglong = true;
 102                 } else {
 103                     /* ??? This might not handle INT_MIN correctly... */
 104                     intResult = neg ? INT_MIN : INT_MAX;
 105                     neg = false;
 106                     while (__CFCharacterIsADigit(ch = __CFStringGetCharacterFromInlineBufferAux(buf, ++(*indexPtr))));  /* Skip remaining digits */
 107                 }
 108             } else {
 109                 intResult = intResult * 10 + __CFCharacterNumericValue(ch);
 110                 ch = __CFStringGetCharacterFromInlineBufferAux(buf, ++(*indexPtr));
 111             }
 112         }
 113     } while (__CFCharacterIsADigit(ch));
 114
 115     if (result) {
 116         if (doLonglong) {
 117             if (!doingLonglong) longlongResult = intResult;
 118             *(int64_t *)result = neg ? -longlongResult : longlongResult;
 119         } else {
 120             *(int *)result = neg ? -intResult : intResult;
 121         }
 122     }
 123
 124     return true;
 125 }
 126
 127 __private_extern__ Boolean __CFStringScanHex(CFStringInlineBuffer *buf, SInt32 *indexPtr, unsigned *result) {
 128     UInt32 value = 0;
 129     SInt32 curDigit;
 130     UniChar ch;
 131
 132     ch = __CFStringGetFirstNonSpaceCharacterFromInlineBuffer(buf, indexPtr);
 133     /* Ignore the optional "0x" or "0X"; if it's followed by a non-hex, just parse the "0" and leave pointer at "x" */
 134     if (ch == '0') {
 135         ch = __CFStringGetCharacterFromInlineBufferAux(buf, ++(*indexPtr));
 136         if (ch == 'x' || ch == 'X') ch = __CFStringGetCharacterFromInlineBufferAux(buf, ++(*indexPtr));
 137         curDigit = __CFCharacterNumericOrHexValue(ch);
 138         if (curDigit == -1) {
 139             (*indexPtr)--;      /* Go back over the "x" or "X" */
 140             if (result) *result = 0;
 141             return true;        /* We just saw "0" */
 142         }
 143     } else {
 144         curDigit = __CFCharacterNumericOrHexValue(ch);
 145         if (curDigit == -1) return false;
 146     }
 147
 148     do {
 149         if (value > (UINT_MAX >> 4)) {
 150             value = UINT_MAX;   /* We do this over and over again, but it's an error case anyway */
 151         } else {
 152             value = (value << 4) + curDigit;
 153         }
 154         curDigit = __CFCharacterNumericOrHexValue(__CFStringGetCharacterFromInlineBufferAux(buf, ++(*indexPtr)));
 155     } while (curDigit != -1);
 156
 157     if (result) *result = value;
 158     return true;
 159 }
 160
 161 // Packed array of Boolean
 162 static const char __CFNumberSet[16] = {
 163     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  nul soh stx etx eot enq ack bel
 164     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  bs  ht  nl  vt  np  cr  so  si
 165     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  dle dc1 dc2 dc3 dc4 nak syn etb
 166     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  can em  sub esc fs  gs  rs  us
 167     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  sp   !   "   #   $   %   &   '
 168     0X28, // 0, 0, 0, 1, 0, 1, 0, 0, //  (   )   *   +   ,   -   .   /
 169     0XFF, // 1, 1, 1, 1, 1, 1, 1, 1, //  0   1   2   3   4   5   6   7
 170     0X03, // 1, 1, 0, 0, 0, 0, 0, 0, //  8   9   :   ;   <   =   >   ?
 171     0X20, // 0, 0, 0, 0, 0, 1, 0, 0, //  @   A   B   C   D   E   F   G
 172     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  H   I   J   K   L   M   N   O
 173     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  P   Q   R   S   T   U   V   W
 174     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  X   Y   Z   [   \   ]   ^   _
 175     0X20, // 0, 0, 0, 0, 0, 1, 0, 0, //  `   a   b   c   d   e   f   g
 176     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  h   i   j   k   l   m   n   o
 177     0X00, // 0, 0, 0, 0, 0, 0, 0, 0, //  p   q   r   s   t   u   v   w
 178     0X00, // 0, 0, 0, 0, 0, 0, 0, 0  //  x   y   z   {   |   }   ~  del
 179 };
 180
 181 __private_extern__ Boolean __CFStringScanDouble(CFStringInlineBuffer *buf, CFDictionaryRef locale, SInt32 *indexPtr, double *resultPtr) {
 182     #define STACK_BUFFER_SIZE 256
 183     #define ALLOC_CHUNK_SIZE 256 // first and subsequent malloc size.  Should be greater than STACK_BUFFER_SIZE
 184     char localCharBuffer[STACK_BUFFER_SIZE];
 185     char *charPtr = localCharBuffer;
 186     char *endCharPtr;
 187     UniChar decimalChar = '.';
 188     SInt32 numChars = 0;
 189     SInt32 capacity = STACK_BUFFER_SIZE;        // in chars
 190     double result;
 191     UniChar ch;
 192     CFAllocatorRef tmpAlloc = NULL;
 193
 194 #if 0
 195     if (locale != NULL) {
 196         CFStringRef decimalSeparator = [locale objectForKey: NSDecimalSeparator];
 197         if (decimalSeparator != nil) decimalChar = [decimalSeparator characterAtIndex:0];
 198     }
 199 #endif
 200     ch = __CFStringGetFirstNonSpaceCharacterFromInlineBuffer(buf, indexPtr);
 201     // At this point indexPtr points at the first non-space char
 202 #if 0
 203 #warning need to allow, case insensitively, all of: "nan", "inf", "-inf", "+inf", "-infinity", "+infinity", "infinity";
 204 #warning -- strtod() will actually do most or all of that for us
 205 #define BITSFORDOUBLENAN        ((uint64_t)0x7ff8000000000000)
 206 #define BITSFORDOUBLEPOSINF     ((uint64_t)0x7ff0000000000000)
 207 #define BITSFORDOUBLENEGINF     ((uint64_t)0xfff0000000000000)
 208     if ('N' == ch || 'n' == ch) {       // check for "NaN", case insensitively
 209         UniChar next1 = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + 1);
 210         UniChar next2 = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + 2);
 211         if (('a' == next1 || 'A' == next1) &&
 212             ('N' == next2 || 'n' == next2)) {
 213             *indexPtr += 3;
 214             if (resultPtr) *(uint64_t *)resultPtr = BITSFORDOUBLENAN;
 215             return true;
 216         }
 217     }
 218     if ('I' == ch || 'i' == ch) {       // check for "Inf", case insensitively
 219         UniChar next1 = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + 1);
 220         UniChar next2 = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + 2);
 221         if (('n' == next1 || 'N' == next1) &&
 222             ('f' == next2 || 'F' == next2)) {
 223             *indexPtr += 3;
 224             if (resultPtr) *(uint64_t *)resultPtr = BITSFORDOUBLEPOSINF;
 225             return true;
 226         }
 227     }
 228     if ('+' == ch || '-' == ch) {       // check for "+/-Inf", case insensitively
 229         UniChar next1 = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + 1);
 230         UniChar next2 = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + 2);
 231         UniChar next3 = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + 3);
 232         if (('I' == next1 || 'i' == next1) &&
 233             ('n' == next2 || 'N' == next2) &&
 234             ('f' == next3 || 'F' == next3)) {
 235             *indexPtr += 4;
 236             if (resultPtr) *(uint64_t *)resultPtr = ('-' == ch) ? BITSFORDOUBLENEGINF : BITSFORDOUBLEPOSINF;
 237             return true;
 238         }
 239     }
 240 #endif 0
 241     do {
 242         if (ch >= 128 || (__CFNumberSet[ch >> 3] & (1 << (ch & 7))) == 0) {
 243             // Not in __CFNumberSet
 244             if (ch != decimalChar) break;
 245             ch = '.';   // Replace the decimal character with something strtod will understand
 246         }
 247         if (numChars >= capacity - 1) {
 248             capacity += ALLOC_CHUNK_SIZE;
 249             if (tmpAlloc == NULL) tmpAlloc = __CFGetDefaultAllocator();
 250             if (charPtr == localCharBuffer) {
 251                 charPtr = CFAllocatorAllocate(tmpAlloc, capacity * sizeof(char), 0);
 252                 memmove(charPtr, localCharBuffer, numChars * sizeof(char));
 253             } else {
 254                 charPtr = CFAllocatorReallocate(tmpAlloc, charPtr, capacity * sizeof(char), 0);
 255             }
 256         }
 257         charPtr[numChars++] = (char)ch;
 258         ch = __CFStringGetCharacterFromInlineBufferAux(buf, *indexPtr + numChars);
 259     } while (true);
 260     charPtr[numChars] = 0;      // Null byte for strtod
 261     result = strtod(charPtr, &endCharPtr);
 262     if (tmpAlloc) CFAllocatorDeallocate(tmpAlloc, charPtr);
 263     if (charPtr == endCharPtr) return false;
 264     *indexPtr += (endCharPtr - charPtr);
 265     if (resultPtr) *resultPtr = result; // only store result if we succeed
 266
 267     return true;
 268 }
 269
 270
 271 #undef STACK_BUFFER_SIZE
 272 #undef ALLOC_CHUNK_SIZE
 273
 274