Sat, 03 Jan 2015 20:18:00 +0100
Conditionally enable double key logic according to:
private browsing mode or privacy.thirdparty.isolate preference and
implement in GetCookieStringCommon and FindCookie where it counts...
With some reservations of how to convince FindCookie users to test
condition and pass a nullptr when disabling double key logic.
michael@0 | 1 | /* |
michael@0 | 2 | ********************************************************************** |
michael@0 | 3 | * Copyright (c) 2001-2011, International Business Machines |
michael@0 | 4 | * Corporation and others. All Rights Reserved. |
michael@0 | 5 | ********************************************************************** |
michael@0 | 6 | * Date Name Description |
michael@0 | 7 | * 11/19/2001 aliu Creation. |
michael@0 | 8 | ********************************************************************** |
michael@0 | 9 | */ |
michael@0 | 10 | |
michael@0 | 11 | #include "unicode/uchar.h" |
michael@0 | 12 | #include "unicode/utf16.h" |
michael@0 | 13 | #include "patternprops.h" |
michael@0 | 14 | #include "util.h" |
michael@0 | 15 | |
michael@0 | 16 | U_NAMESPACE_BEGIN |
michael@0 | 17 | |
michael@0 | 18 | /** |
michael@0 | 19 | * Parse an integer at pos, either of the form \d+ or of the form |
michael@0 | 20 | * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex, |
michael@0 | 21 | * or octal format. |
michael@0 | 22 | * @param pos INPUT-OUTPUT parameter. On input, the first |
michael@0 | 23 | * character to parse. On output, the character after the last |
michael@0 | 24 | * parsed character. |
michael@0 | 25 | */ |
michael@0 | 26 | int32_t ICU_Utility::parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit) { |
michael@0 | 27 | int32_t count = 0; |
michael@0 | 28 | int32_t value = 0; |
michael@0 | 29 | int32_t p = pos; |
michael@0 | 30 | int8_t radix = 10; |
michael@0 | 31 | |
michael@0 | 32 | if (p < limit && rule.charAt(p) == 48 /*0*/) { |
michael@0 | 33 | if (p+1 < limit && (rule.charAt(p+1) == 0x78 /*x*/ || rule.charAt(p+1) == 0x58 /*X*/)) { |
michael@0 | 34 | p += 2; |
michael@0 | 35 | radix = 16; |
michael@0 | 36 | } |
michael@0 | 37 | else { |
michael@0 | 38 | p++; |
michael@0 | 39 | count = 1; |
michael@0 | 40 | radix = 8; |
michael@0 | 41 | } |
michael@0 | 42 | } |
michael@0 | 43 | |
michael@0 | 44 | while (p < limit) { |
michael@0 | 45 | int32_t d = u_digit(rule.charAt(p++), radix); |
michael@0 | 46 | if (d < 0) { |
michael@0 | 47 | --p; |
michael@0 | 48 | break; |
michael@0 | 49 | } |
michael@0 | 50 | ++count; |
michael@0 | 51 | int32_t v = (value * radix) + d; |
michael@0 | 52 | if (v <= value) { |
michael@0 | 53 | // If there are too many input digits, at some point |
michael@0 | 54 | // the value will go negative, e.g., if we have seen |
michael@0 | 55 | // "0x8000000" already and there is another '0', when |
michael@0 | 56 | // we parse the next 0 the value will go negative. |
michael@0 | 57 | return 0; |
michael@0 | 58 | } |
michael@0 | 59 | value = v; |
michael@0 | 60 | } |
michael@0 | 61 | if (count > 0) { |
michael@0 | 62 | pos = p; |
michael@0 | 63 | } |
michael@0 | 64 | return value; |
michael@0 | 65 | } |
michael@0 | 66 | |
michael@0 | 67 | /** |
michael@0 | 68 | * Parse a pattern string starting at offset pos. Keywords are |
michael@0 | 69 | * matched case-insensitively. Spaces may be skipped and may be |
michael@0 | 70 | * optional or required. Integer values may be parsed, and if |
michael@0 | 71 | * they are, they will be returned in the given array. If |
michael@0 | 72 | * successful, the offset of the next non-space character is |
michael@0 | 73 | * returned. On failure, -1 is returned. |
michael@0 | 74 | * @param pattern must only contain lowercase characters, which |
michael@0 | 75 | * will match their uppercase equivalents as well. A space |
michael@0 | 76 | * character matches one or more required spaces. A '~' character |
michael@0 | 77 | * matches zero or more optional spaces. A '#' character matches |
michael@0 | 78 | * an integer and stores it in parsedInts, which the caller must |
michael@0 | 79 | * ensure has enough capacity. |
michael@0 | 80 | * @param parsedInts array to receive parsed integers. Caller |
michael@0 | 81 | * must ensure that parsedInts.length is >= the number of '#' |
michael@0 | 82 | * signs in 'pattern'. |
michael@0 | 83 | * @return the position after the last character parsed, or -1 if |
michael@0 | 84 | * the parse failed |
michael@0 | 85 | */ |
michael@0 | 86 | int32_t ICU_Utility::parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit, |
michael@0 | 87 | const UnicodeString& pattern, int32_t* parsedInts) { |
michael@0 | 88 | // TODO Update this to handle surrogates |
michael@0 | 89 | int32_t p; |
michael@0 | 90 | int32_t intCount = 0; // number of integers parsed |
michael@0 | 91 | for (int32_t i=0; i<pattern.length(); ++i) { |
michael@0 | 92 | UChar cpat = pattern.charAt(i); |
michael@0 | 93 | UChar c; |
michael@0 | 94 | switch (cpat) { |
michael@0 | 95 | case 32 /*' '*/: |
michael@0 | 96 | if (pos >= limit) { |
michael@0 | 97 | return -1; |
michael@0 | 98 | } |
michael@0 | 99 | c = rule.charAt(pos++); |
michael@0 | 100 | if (!PatternProps::isWhiteSpace(c)) { |
michael@0 | 101 | return -1; |
michael@0 | 102 | } |
michael@0 | 103 | // FALL THROUGH to skipWhitespace |
michael@0 | 104 | case 126 /*'~'*/: |
michael@0 | 105 | pos = skipWhitespace(rule, pos); |
michael@0 | 106 | break; |
michael@0 | 107 | case 35 /*'#'*/: |
michael@0 | 108 | p = pos; |
michael@0 | 109 | parsedInts[intCount++] = parseInteger(rule, p, limit); |
michael@0 | 110 | if (p == pos) { |
michael@0 | 111 | // Syntax error; failed to parse integer |
michael@0 | 112 | return -1; |
michael@0 | 113 | } |
michael@0 | 114 | pos = p; |
michael@0 | 115 | break; |
michael@0 | 116 | default: |
michael@0 | 117 | if (pos >= limit) { |
michael@0 | 118 | return -1; |
michael@0 | 119 | } |
michael@0 | 120 | c = (UChar) u_tolower(rule.charAt(pos++)); |
michael@0 | 121 | if (c != cpat) { |
michael@0 | 122 | return -1; |
michael@0 | 123 | } |
michael@0 | 124 | break; |
michael@0 | 125 | } |
michael@0 | 126 | } |
michael@0 | 127 | return pos; |
michael@0 | 128 | } |
michael@0 | 129 | |
michael@0 | 130 | /** |
michael@0 | 131 | * Parse a Unicode identifier from the given string at the given |
michael@0 | 132 | * position. Return the identifier, or an empty string if there |
michael@0 | 133 | * is no identifier. |
michael@0 | 134 | * @param str the string to parse |
michael@0 | 135 | * @param pos INPUT-OUPUT parameter. On INPUT, pos is the |
michael@0 | 136 | * first character to examine. It must be less than str.length(), |
michael@0 | 137 | * and it must not point to a whitespace character. That is, must |
michael@0 | 138 | * have pos < str.length(). On |
michael@0 | 139 | * OUTPUT, the position after the last parsed character. |
michael@0 | 140 | * @return the Unicode identifier, or an empty string if there is |
michael@0 | 141 | * no valid identifier at pos. |
michael@0 | 142 | */ |
michael@0 | 143 | UnicodeString ICU_Utility::parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos) { |
michael@0 | 144 | // assert(pos < str.length()); |
michael@0 | 145 | UnicodeString buf; |
michael@0 | 146 | int p = pos; |
michael@0 | 147 | while (p < str.length()) { |
michael@0 | 148 | UChar32 ch = str.char32At(p); |
michael@0 | 149 | if (buf.length() == 0) { |
michael@0 | 150 | if (u_isIDStart(ch)) { |
michael@0 | 151 | buf.append(ch); |
michael@0 | 152 | } else { |
michael@0 | 153 | buf.truncate(0); |
michael@0 | 154 | return buf; |
michael@0 | 155 | } |
michael@0 | 156 | } else { |
michael@0 | 157 | if (u_isIDPart(ch)) { |
michael@0 | 158 | buf.append(ch); |
michael@0 | 159 | } else { |
michael@0 | 160 | break; |
michael@0 | 161 | } |
michael@0 | 162 | } |
michael@0 | 163 | p += U16_LENGTH(ch); |
michael@0 | 164 | } |
michael@0 | 165 | pos = p; |
michael@0 | 166 | return buf; |
michael@0 | 167 | } |
michael@0 | 168 | |
michael@0 | 169 | /** |
michael@0 | 170 | * Parse an unsigned 31-bit integer at the given offset. Use |
michael@0 | 171 | * UCharacter.digit() to parse individual characters into digits. |
michael@0 | 172 | * @param text the text to be parsed |
michael@0 | 173 | * @param pos INPUT-OUTPUT parameter. On entry, pos[0] is the |
michael@0 | 174 | * offset within text at which to start parsing; it should point |
michael@0 | 175 | * to a valid digit. On exit, pos[0] is the offset after the last |
michael@0 | 176 | * parsed character. If the parse failed, it will be unchanged on |
michael@0 | 177 | * exit. Must be >= 0 on entry. |
michael@0 | 178 | * @param radix the radix in which to parse; must be >= 2 and <= |
michael@0 | 179 | * 36. |
michael@0 | 180 | * @return a non-negative parsed number, or -1 upon parse failure. |
michael@0 | 181 | * Parse fails if there are no digits, that is, if pos[0] does not |
michael@0 | 182 | * point to a valid digit on entry, or if the number to be parsed |
michael@0 | 183 | * does not fit into a 31-bit unsigned integer. |
michael@0 | 184 | */ |
michael@0 | 185 | int32_t ICU_Utility::parseNumber(const UnicodeString& text, |
michael@0 | 186 | int32_t& pos, int8_t radix) { |
michael@0 | 187 | // assert(pos[0] >= 0); |
michael@0 | 188 | // assert(radix >= 2); |
michael@0 | 189 | // assert(radix <= 36); |
michael@0 | 190 | int32_t n = 0; |
michael@0 | 191 | int32_t p = pos; |
michael@0 | 192 | while (p < text.length()) { |
michael@0 | 193 | UChar32 ch = text.char32At(p); |
michael@0 | 194 | int32_t d = u_digit(ch, radix); |
michael@0 | 195 | if (d < 0) { |
michael@0 | 196 | break; |
michael@0 | 197 | } |
michael@0 | 198 | n = radix*n + d; |
michael@0 | 199 | // ASSUME that when a 32-bit integer overflows it becomes |
michael@0 | 200 | // negative. E.g., 214748364 * 10 + 8 => negative value. |
michael@0 | 201 | if (n < 0) { |
michael@0 | 202 | return -1; |
michael@0 | 203 | } |
michael@0 | 204 | ++p; |
michael@0 | 205 | } |
michael@0 | 206 | if (p == pos) { |
michael@0 | 207 | return -1; |
michael@0 | 208 | } |
michael@0 | 209 | pos = p; |
michael@0 | 210 | return n; |
michael@0 | 211 | } |
michael@0 | 212 | |
michael@0 | 213 | U_NAMESPACE_END |
michael@0 | 214 |