Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 1 | // Copyright 2008 the V8 project authors. All rights reserved. |
| 2 | // Redistribution and use in source and binary forms, with or without |
| 3 | // modification, are permitted provided that the following conditions are |
| 4 | // met: |
| 5 | // |
| 6 | // * Redistributions of source code must retain the above copyright |
| 7 | // notice, this list of conditions and the following disclaimer. |
| 8 | // * Redistributions in binary form must reproduce the above |
| 9 | // copyright notice, this list of conditions and the following |
| 10 | // disclaimer in the documentation and/or other materials provided |
| 11 | // with the distribution. |
| 12 | // * Neither the name of Google Inc. nor the names of its |
| 13 | // contributors may be used to endorse or promote products derived |
| 14 | // from this software without specific prior written permission. |
| 15 | // |
| 16 | // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| 17 | // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| 18 | // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
| 19 | // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
| 20 | // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
| 21 | // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
| 22 | // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
| 23 | // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
| 24 | // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 25 | // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 26 | // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 27 | |
| 28 | #ifndef V8_DATEPARSER_H_ |
| 29 | #define V8_DATEPARSER_H_ |
| 30 | |
Shimeng (Simon) Wang | 8a31eba | 2010-12-06 19:01:33 -0800 | [diff] [blame] | 31 | #include "char-predicates-inl.h" |
| 32 | #include "scanner-base.h" |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 33 | |
| 34 | namespace v8 { |
| 35 | namespace internal { |
| 36 | |
| 37 | class DateParser : public AllStatic { |
| 38 | public: |
| 39 | |
| 40 | // Parse the string as a date. If parsing succeeds, return true after |
| 41 | // filling out the output array as follows (all integers are Smis): |
| 42 | // [0]: year |
| 43 | // [1]: month (0 = Jan, 1 = Feb, ...) |
| 44 | // [2]: day |
| 45 | // [3]: hour |
| 46 | // [4]: minute |
| 47 | // [5]: second |
Steve Block | 6ded16b | 2010-05-10 14:33:55 +0100 | [diff] [blame] | 48 | // [6]: millisecond |
| 49 | // [7]: UTC offset in seconds, or null value if no timezone specified |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 50 | // If parsing fails, return false (content of output array is not defined). |
| 51 | template <typename Char> |
| 52 | static bool Parse(Vector<Char> str, FixedArray* output); |
| 53 | |
| 54 | enum { |
Steve Block | 6ded16b | 2010-05-10 14:33:55 +0100 | [diff] [blame] | 55 | YEAR, MONTH, DAY, HOUR, MINUTE, SECOND, MILLISECOND, UTC_OFFSET, OUTPUT_SIZE |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 56 | }; |
| 57 | |
| 58 | private: |
| 59 | // Range testing |
| 60 | static inline bool Between(int x, int lo, int hi) { |
| 61 | return static_cast<unsigned>(x - lo) <= static_cast<unsigned>(hi - lo); |
| 62 | } |
| 63 | // Indicates a missing value. |
| 64 | static const int kNone = kMaxInt; |
| 65 | |
| 66 | // InputReader provides basic string parsing and character classification. |
| 67 | template <typename Char> |
| 68 | class InputReader BASE_EMBEDDED { |
| 69 | public: |
| 70 | explicit InputReader(Vector<Char> s) |
| 71 | : index_(0), |
| 72 | buffer_(s), |
Steve Block | 44f0eee | 2011-05-26 01:26:41 +0100 | [diff] [blame^] | 73 | has_read_number_(false), |
| 74 | scanner_constants_(Isolate::Current()->scanner_constants()) { |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 75 | Next(); |
| 76 | } |
| 77 | |
| 78 | // Advance to the next character of the string. |
| 79 | void Next() { ch_ = (index_ < buffer_.length()) ? buffer_[index_++] : 0; } |
| 80 | |
| 81 | // Read a string of digits as an unsigned number (cap just below kMaxInt). |
| 82 | int ReadUnsignedNumber() { |
| 83 | has_read_number_ = true; |
| 84 | int n; |
| 85 | for (n = 0; IsAsciiDigit() && n < kMaxInt / 10 - 1; Next()) { |
| 86 | n = n * 10 + ch_ - '0'; |
| 87 | } |
| 88 | return n; |
| 89 | } |
| 90 | |
Shimeng (Simon) Wang | 8a31eba | 2010-12-06 19:01:33 -0800 | [diff] [blame] | 91 | // Read a string of digits, take the first three or fewer as an unsigned |
| 92 | // number of milliseconds, and ignore any digits after the first three. |
| 93 | int ReadMilliseconds() { |
| 94 | has_read_number_ = true; |
| 95 | int n = 0; |
| 96 | int power; |
| 97 | for (power = 100; IsAsciiDigit(); Next(), power = power / 10) { |
| 98 | n = n + power * (ch_ - '0'); |
| 99 | } |
| 100 | return n; |
| 101 | } |
| 102 | |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 103 | // Read a word (sequence of chars. >= 'A'), fill the given buffer with a |
| 104 | // lower-case prefix, and pad any remainder of the buffer with zeroes. |
| 105 | // Return word length. |
| 106 | int ReadWord(uint32_t* prefix, int prefix_size) { |
| 107 | int len; |
| 108 | for (len = 0; IsAsciiAlphaOrAbove(); Next(), len++) { |
Iain Merrick | 9ac36c9 | 2010-09-13 15:29:50 +0100 | [diff] [blame] | 109 | if (len < prefix_size) prefix[len] = AsciiAlphaToLower(ch_); |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 110 | } |
| 111 | for (int i = len; i < prefix_size; i++) prefix[i] = 0; |
| 112 | return len; |
| 113 | } |
| 114 | |
| 115 | // The skip methods return whether they actually skipped something. |
Shimeng (Simon) Wang | 8a31eba | 2010-12-06 19:01:33 -0800 | [diff] [blame] | 116 | bool Skip(uint32_t c) { |
| 117 | if (ch_ == c) { |
| 118 | Next(); |
| 119 | return true; |
| 120 | } |
| 121 | return false; |
| 122 | } |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 123 | |
| 124 | bool SkipWhiteSpace() { |
Steve Block | 44f0eee | 2011-05-26 01:26:41 +0100 | [diff] [blame^] | 125 | if (scanner_constants_->IsWhiteSpace(ch_)) { |
Shimeng (Simon) Wang | 8a31eba | 2010-12-06 19:01:33 -0800 | [diff] [blame] | 126 | Next(); |
| 127 | return true; |
| 128 | } |
| 129 | return false; |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 130 | } |
| 131 | |
| 132 | bool SkipParentheses() { |
| 133 | if (ch_ != '(') return false; |
| 134 | int balance = 0; |
| 135 | do { |
| 136 | if (ch_ == ')') --balance; |
| 137 | else if (ch_ == '(') ++balance; |
| 138 | Next(); |
| 139 | } while (balance > 0 && ch_); |
| 140 | return true; |
| 141 | } |
| 142 | |
| 143 | // Character testing/classification. Non-ASCII digits are not supported. |
| 144 | bool Is(uint32_t c) const { return ch_ == c; } |
| 145 | bool IsEnd() const { return ch_ == 0; } |
| 146 | bool IsAsciiDigit() const { return IsDecimalDigit(ch_); } |
| 147 | bool IsAsciiAlphaOrAbove() const { return ch_ >= 'A'; } |
| 148 | bool IsAsciiSign() const { return ch_ == '+' || ch_ == '-'; } |
| 149 | |
| 150 | // Return 1 for '+' and -1 for '-'. |
| 151 | int GetAsciiSignValue() const { return 44 - static_cast<int>(ch_); } |
| 152 | |
| 153 | // Indicates whether any (possibly empty!) numbers have been read. |
| 154 | bool HasReadNumber() const { return has_read_number_; } |
| 155 | |
| 156 | private: |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 157 | int index_; |
| 158 | Vector<Char> buffer_; |
| 159 | bool has_read_number_; |
| 160 | uint32_t ch_; |
Steve Block | 44f0eee | 2011-05-26 01:26:41 +0100 | [diff] [blame^] | 161 | ScannerConstants* scanner_constants_; |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 162 | }; |
| 163 | |
| 164 | enum KeywordType { INVALID, MONTH_NAME, TIME_ZONE_NAME, AM_PM }; |
| 165 | |
| 166 | // KeywordTable maps names of months, time zones, am/pm to numbers. |
| 167 | class KeywordTable : public AllStatic { |
| 168 | public: |
| 169 | // Look up a word in the keyword table and return an index. |
| 170 | // 'pre' contains a prefix of the word, zero-padded to size kPrefixLength |
| 171 | // and 'len' is the word length. |
| 172 | static int Lookup(const uint32_t* pre, int len); |
| 173 | // Get the type of the keyword at index i. |
| 174 | static KeywordType GetType(int i) { |
| 175 | return static_cast<KeywordType>(array[i][kTypeOffset]); |
| 176 | } |
| 177 | // Get the value of the keyword at index i. |
| 178 | static int GetValue(int i) { return array[i][kValueOffset]; } |
| 179 | |
| 180 | static const int kPrefixLength = 3; |
| 181 | static const int kTypeOffset = kPrefixLength; |
| 182 | static const int kValueOffset = kTypeOffset + 1; |
| 183 | static const int kEntrySize = kValueOffset + 1; |
| 184 | static const int8_t array[][kEntrySize]; |
| 185 | }; |
| 186 | |
| 187 | class TimeZoneComposer BASE_EMBEDDED { |
| 188 | public: |
| 189 | TimeZoneComposer() : sign_(kNone), hour_(kNone), minute_(kNone) {} |
| 190 | void Set(int offset_in_hours) { |
| 191 | sign_ = offset_in_hours < 0 ? -1 : 1; |
| 192 | hour_ = offset_in_hours * sign_; |
| 193 | minute_ = 0; |
| 194 | } |
| 195 | void SetSign(int sign) { sign_ = sign < 0 ? -1 : 1; } |
| 196 | void SetAbsoluteHour(int hour) { hour_ = hour; } |
| 197 | void SetAbsoluteMinute(int minute) { minute_ = minute; } |
| 198 | bool IsExpecting(int n) const { |
| 199 | return hour_ != kNone && minute_ == kNone && TimeComposer::IsMinute(n); |
| 200 | } |
| 201 | bool IsUTC() const { return hour_ == 0 && minute_ == 0; } |
| 202 | bool Write(FixedArray* output); |
| 203 | private: |
| 204 | int sign_; |
| 205 | int hour_; |
| 206 | int minute_; |
| 207 | }; |
| 208 | |
| 209 | class TimeComposer BASE_EMBEDDED { |
| 210 | public: |
| 211 | TimeComposer() : index_(0), hour_offset_(kNone) {} |
| 212 | bool IsEmpty() const { return index_ == 0; } |
| 213 | bool IsExpecting(int n) const { |
Steve Block | 6ded16b | 2010-05-10 14:33:55 +0100 | [diff] [blame] | 214 | return (index_ == 1 && IsMinute(n)) || |
| 215 | (index_ == 2 && IsSecond(n)) || |
| 216 | (index_ == 3 && IsMillisecond(n)); |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 217 | } |
| 218 | bool Add(int n) { |
| 219 | return index_ < kSize ? (comp_[index_++] = n, true) : false; |
| 220 | } |
| 221 | bool AddFinal(int n) { |
| 222 | if (!Add(n)) return false; |
| 223 | while (index_ < kSize) comp_[index_++] = 0; |
| 224 | return true; |
| 225 | } |
| 226 | void SetHourOffset(int n) { hour_offset_ = n; } |
| 227 | bool Write(FixedArray* output); |
| 228 | |
| 229 | static bool IsMinute(int x) { return Between(x, 0, 59); } |
| 230 | private: |
| 231 | static bool IsHour(int x) { return Between(x, 0, 23); } |
| 232 | static bool IsHour12(int x) { return Between(x, 0, 12); } |
| 233 | static bool IsSecond(int x) { return Between(x, 0, 59); } |
Steve Block | 6ded16b | 2010-05-10 14:33:55 +0100 | [diff] [blame] | 234 | static bool IsMillisecond(int x) { return Between(x, 0, 999); } |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 235 | |
Steve Block | 6ded16b | 2010-05-10 14:33:55 +0100 | [diff] [blame] | 236 | static const int kSize = 4; |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 237 | int comp_[kSize]; |
| 238 | int index_; |
| 239 | int hour_offset_; |
| 240 | }; |
| 241 | |
| 242 | class DayComposer BASE_EMBEDDED { |
| 243 | public: |
| 244 | DayComposer() : index_(0), named_month_(kNone) {} |
| 245 | bool IsEmpty() const { return index_ == 0; } |
| 246 | bool Add(int n) { |
| 247 | return index_ < kSize ? (comp_[index_++] = n, true) : false; |
| 248 | } |
| 249 | void SetNamedMonth(int n) { named_month_ = n; } |
| 250 | bool Write(FixedArray* output); |
| 251 | private: |
| 252 | static bool IsMonth(int x) { return Between(x, 1, 12); } |
| 253 | static bool IsDay(int x) { return Between(x, 1, 31); } |
| 254 | |
| 255 | static const int kSize = 3; |
| 256 | int comp_[kSize]; |
| 257 | int index_; |
| 258 | int named_month_; |
| 259 | }; |
| 260 | }; |
| 261 | |
| 262 | |
| 263 | } } // namespace v8::internal |
| 264 | |
| 265 | #endif // V8_DATEPARSER_H_ |