Make date parser handle all ES5 Date Time Strings correctly.

src/dateparser-inl.h

 // Copyright 2011 the V8 project authors. All rights reserved.
 // Redistribution and use in source and binary forms, with or without
 // modification, are permitted provided that the following conditions are
 // met:
 //
 //     * Redistributions of source code must retain the above copyright
 //       notice, this list of conditions and the following disclaimer.
 //     * Redistributions in binary form must reproduce the above
 //       copyright notice, this list of conditions and the following
 //       disclaimer in the documentation and/or other materials provided
@@ skipping 21 matching lines @@
 
 namespace v8 {
 namespace internal {
 
 template <typename Char>
 bool DateParser::Parse(Vector<Char> str,
                        FixedArray* out,
                        UnicodeCache* unicode_cache) {
   ASSERT(out->length() >= OUTPUT_SIZE);
   InputReader<Char> in(unicode_cache, str);
+  DateStringTokenizer<Char> scanner(&in);
   TimeZoneComposer tz;
   TimeComposer time;
   DayComposer day;
 
-  while (!in.IsEnd()) {
-    if (in.IsAsciiDigit()) {
-      // Parse a number (possibly with 1 or 2 trailing colons).
-      int n = in.ReadUnsignedNumber();
-      if (in.Skip(':')) {
-        if (in.Skip(':')) {
+  // Specification:
+  // Accept ES5 ISO 8601 date-time-strings or legacy dates compatible
+  // with Safari.
+  // ES5 ISO 8601 dates:
+  //   [('-'|'+')yy]yyyy[-MM[-DD]][THH:mm[:ss[.sss]][Z|(+|-)hh:mm]]
+  //   where yyyy is in the range 0000..9999 and
+  //         +/-yyyyyy is in the range -999999..+999999 -
+  //           but -000000 is invalid (year zero must be positive),
+  //         MM is in the range 01..12,
+  //         DD is in the range 01..31,
+  //         MM and DD defaults to 01 if missing,,
+  //         HH is generally in the range 00..23, but can be 24 if mm, ss
+  //           and sss are zero (or missing), representing midnight at the
+  //           end of a day,
+  //         mm and ss are in the range 00..59,
+  //         sss is in the range 000..999,
+  //         hh is in the range 00..23,
+  //         mm, ss, and sss default to 00 if missing, and
+  //         timezone defaults to Z if missing.
+  //  Extensions:
+  //   We also allow sss to have more or less than three digits (but at
+  //   least one).
+  //   We allow hh:mm to be specified as hhmm.
+  // Legacy dates:
+  //  Any unrecognized word before the first number is ignored.
+  //  Parenthesized text is ignored.
+  //  An unsigned number followed by ':' is a time value, and is
+  //  added to the TimeComposer. A number followed by '::' adds a second
+  //  zero as well. A number followed by '.' is also a time and must be
+  //  followed by milliseconds.
+  //  Any other number is a date component and is added to DayComposer.
+  //  A month name (or really: any word having the same first three letters
+  //  as a month name) is recorded as a named month in the Day composer.
+  //  A word recognizable as a time-zone is recorded as such, as is
+  //  '(+|-)(hhmm|hh:)'.
+  //  Legacy dates don't allow extra signs ('+' or '-') or umatched ')'
+  //  after a number has been read (before the first number, any garbage
+  //  is allowed).
+  // Intersection of the two:
+  //  A string that matches both formats (e.g. 1970-01-01) will be
+  //  parsed as an ES5 date-time string - which means it will default
+  //  to UTC time-zone. That's unavoidable if following the ES5
+  //  specification.
+  //  After a valid "T" has been read while scanning an ES5 datetime string,
+  //  the input can no longer be a valid legacy date, since the "T" is a
+  //  garbage string after a number has been read.
+
+  // First try getting as far as possible with as ES5 Date Time String.
+  DateToken next_unhandled_token = ParseES5DateTime(&scanner, &day, &time, &tz);
+  if (next_unhandled_token.IsInvalid()) return false;
+  bool has_read_number = !day.IsEmpty();
+  // If there's anything left, continue with the legacy parser.
+  for (DateToken token = next_unhandled_token;
+       !token.IsEndOfInput();
+       token = scanner.Next()) {
+    if (token.IsNumber()) {
+      has_read_number = true;
+      int n = token.number();
+      if (scanner.SkipSymbol(':')) {
+        if (scanner.SkipSymbol(':')) {
           // n + "::"
           if (!time.IsEmpty()) return false;
           time.Add(n);
           time.Add(0);
         } else {
           // n + ":"
           if (!time.Add(n)) return false;
-          in.Skip('.');
+          if (scanner.Peek().IsSymbol('.')) scanner.Next();
         }
-      } else if (in.Skip('.') && time.IsExpecting(n)) {
+      } else if (scanner.SkipSymbol('.') && time.IsExpecting(n)) {
         time.Add(n);
-        if (!in.IsAsciiDigit()) return false;
-        int n = in.ReadMilliseconds();
+        if (!scanner.Peek().IsNumber()) return false;
+        int n = ReadMilliseconds(scanner.Next());
+        if (n < 0) return false;
         time.AddFinal(n);
       } else if (tz.IsExpecting(n)) {
         tz.SetAbsoluteMinute(n);
       } else if (time.IsExpecting(n)) {
         time.AddFinal(n);
         // Require end, white space, "Z", "+" or "-" immediately after
         // finalizing time.
-        if (!in.IsEnd() && !in.SkipWhiteSpace() && !in.Is('Z') &&
-            !in.IsAsciiSign()) return false;
+        DateToken peek = scanner.Peek();
+        if (!peek.IsEndOfInput() &&
+            !peek.IsWhiteSpace() &&
+            !peek.IsKeywordZ() &&
+            !peek.IsAsciiSign()) return false;
       } else {
         if (!day.Add(n)) return false;
-        in.Skip('-');  // Ignore suffix '-' for year, month, or day.
-        // Skip trailing 'T' for ECMAScript 5 date string format but make
-        // sure that it is followed by a digit (for the time).
-        if (in.Skip('T') && !in.IsAsciiDigit()) return false;
-      }
-    } else if (in.IsAsciiAlphaOrAbove()) {
+        scanner.SkipSymbol('-');
+      }
+    } else if (token.IsKeyword()) {
       // Parse a "word" (sequence of chars. >= 'A').
-      uint32_t pre[KeywordTable::kPrefixLength];
-      int len = in.ReadWord(pre, KeywordTable::kPrefixLength);
-      int index = KeywordTable::Lookup(pre, len);
-      KeywordType type = KeywordTable::GetType(index);
-
+      KeywordType type = token.keyword_type();
+      int value = token.keyword_value();
       if (type == AM_PM && !time.IsEmpty()) {
-        time.SetHourOffset(KeywordTable::GetValue(index));
+        time.SetHourOffset(value);
       } else if (type == MONTH_NAME) {
-        day.SetNamedMonth(KeywordTable::GetValue(index));
-        in.Skip('-');  // Ignore suffix '-' for month names
-      } else if (type == TIME_ZONE_NAME && in.HasReadNumber()) {
-        tz.Set(KeywordTable::GetValue(index));
+        day.SetNamedMonth(value);
+        scanner.SkipSymbol('-');
+      } else if (type == TIME_ZONE_NAME && has_read_number) {
+        tz.Set(value);
       } else {
         // Garbage words are illegal if a number has been read.
-        if (in.HasReadNumber()) return false;
-      }
-    } else if (in.IsAsciiSign() && (tz.IsUTC() || !time.IsEmpty())) {
+        if (has_read_number) return false;
+      }
+    } else if (token.IsAsciiSign() && (tz.IsUTC() || !time.IsEmpty())) {
       // Parse UTC offset (only after UTC or time).
-      tz.SetSign(in.GetAsciiSignValue());
-      in.Next();
-      int n = in.ReadUnsignedNumber();
-      if (in.Skip(':')) {
+      tz.SetSign(token.ascii_sign());
+      // The following number may be empty.
+      int n = 0;
+      if (scanner.Peek().IsNumber()) {
+        n = scanner.Next().number();
+      }
+      has_read_number = true;
+
+      if (scanner.Peek().IsSymbol(':')) {
         tz.SetAbsoluteHour(n);
         tz.SetAbsoluteMinute(kNone);
       } else {
         tz.SetAbsoluteHour(n / 100);
         tz.SetAbsoluteMinute(n % 100);
       }
-    } else if (in.Is('(')) {
-      // Ignore anything from '(' to a matching ')' or end of string.
-      in.SkipParentheses();
-    } else if ((in.IsAsciiSign() || in.Is(')')) && in.HasReadNumber()) {
+    } else if ((token.IsAsciiSign() || token.IsSymbol(')')) &&
+               has_read_number) {
       // Extra sign or ')' is illegal if a number has been read.
       return false;
     } else {
-      // Ignore other characters.
-      in.Next();
-    }
-  }
+      // Ignore other characters and whitespace.
+    }
+  }
+
   return day.Write(out) && time.Write(out) && tz.Write(out);
 }
 
+
+template<typename CharType>
+DateParser::DateToken DateParser::DateStringTokenizer<CharType>::Scan() {
+  int pre_pos = in_->position();
+  if (in_->IsEnd()) return DateToken::EndOfInput();
+  if (in_->IsAsciiDigit()) {
+    int n = in_->ReadUnsignedNumeral();
+    int length = in_->position() - pre_pos;
+    return DateToken::Number(n, length);
+  }
+  if (in_->Skip(':')) return DateToken::Symbol(':');
+  if (in_->Skip('-')) return DateToken::Symbol('-');
+  if (in_->Skip('+')) return DateToken::Symbol('+');
+  if (in_->Skip('.')) return DateToken::Symbol('.');
+  if (in_->Skip(')')) return DateToken::Symbol(')');
+  if (in_->IsAsciiAlphaOrAbove()) {
+    ASSERT(KeywordTable::kPrefixLength == 3);
+    uint32_t buffer[3] = {0, 0, 0};
+    int length = in_->ReadWord(buffer, 3);
+    int index = KeywordTable::Lookup(buffer, length);
+    return DateToken::Keyword(KeywordTable::GetType(index),
+                              KeywordTable::GetValue(index),
+                              length);
+  }
+  if (in_->SkipWhiteSpace()) {
+    return DateToken::WhiteSpace(in_->position() - pre_pos);
+  }
+  if (in_->SkipParentheses()) {
+    return DateToken::Unknown();
+  }
+  in_->Next();
+  return DateToken::Unknown();
+}
+
+
+template <typename Char>
+DateParser::DateToken DateParser::ParseES5DateTime(
+    DateStringTokenizer<Char>* scanner,
+    DayComposer* day,
+    TimeComposer* time,
+    TimeZoneComposer* tz) {
+  ASSERT(day->IsEmpty());
+  ASSERT(time->IsEmpty());
+  ASSERT(tz->IsEmpty());
+
+  // Parse mandatory date string: [('-'|'+')yy]yyyy[':'MM[':'DD]]
+  if (scanner->Peek().IsAsciiSign()) {
+    // Keep the sign token, so we can pass it back to the legacy
+    // parser if we don't use it.
+    DateToken sign_token = scanner->Next();
+    if (!scanner->Peek().IsFixedLengthNumber(6)) return sign_token;
+    int sign = sign_token.ascii_sign();
+    int year = scanner->Next().number();
+    if (sign < 0 && year == 0) return sign_token;
+    day->Add(sign * year);
+  } else if (scanner->Peek().IsFixedLengthNumber(4)) {
+    day->Add(scanner->Next().number());
+  } else {
+    return scanner->Next();
+  }
+  if (scanner->SkipSymbol('-')) {
+    if (!scanner->Peek().IsFixedLengthNumber(2) ||
+        !DayComposer::IsMonth(scanner->Peek().number())) return scanner->Next();
+    day->Add(scanner->Next().number());
+    if (scanner->SkipSymbol('-')) {
+      if (!scanner->Peek().IsFixedLengthNumber(2) ||
+          !DayComposer::IsDay(scanner->Peek().number())) return scanner->Next();
+      day->Add(scanner->Next().number());
+    }
+  }
+  // Check for optional time string: 'T'HH':'mm[':'ss['.'sss]]Z
+  if (!scanner->Peek().IsKeywordType(TIME_SEPARATOR)) {
+    if (!scanner->Peek().IsEndOfInput()) return scanner->Next();
+  } else {
+    // ES5 Date Time String time part is present.
+    scanner->Next();
+    if (!scanner->Peek().IsFixedLengthNumber(2) ||
+        !Between(scanner->Peek().number(), 0, 24)) {
+      return DateToken::Invalid();
+    }
+    // Allow 24:00[:00[.000]], but no other time starting with 24.
+    bool hour_is_24 = (scanner->Peek().number() == 24);
+    time->Add(scanner->Next().number());
+    if (!scanner->SkipSymbol(':')) return DateToken::Invalid();
+    if (!scanner->Peek().IsFixedLengthNumber(2) ||
+        !TimeComposer::IsMinute(scanner->Peek().number()) ||
+        (hour_is_24 && scanner->Peek().number() > 0)) {
+      return DateToken::Invalid();
+    }
+    time->Add(scanner->Next().number());
+    if (scanner->SkipSymbol(':')) {
+      if (!scanner->Peek().IsFixedLengthNumber(2) ||
+          !TimeComposer::IsSecond(scanner->Peek().number()) ||
+          (hour_is_24 && scanner->Peek().number() > 0)) {
+        return DateToken::Invalid();
+      }
+      time->Add(scanner->Next().number());
+      if (scanner->SkipSymbol('.')) {
+        if (!scanner->Peek().IsNumber() ||
+            (hour_is_24 && scanner->Peek().number() > 0)) {
+          return DateToken::Invalid();
+        }
+        // Allow more or less than the mandated three digits.
+        time->Add(ReadMilliseconds(scanner->Next()));
+      }
+    }
+    // Check for optional timezone designation: 'Z' | ('+'|'-')hh':'mm
+    if (scanner->Peek().IsKeywordZ()) {
+      scanner->Next();
+      tz->Set(0);
+    } else if (scanner->Peek().IsSymbol('+') ||
+               scanner->Peek().IsSymbol('-')) {
+      tz->SetSign(scanner->Next().symbol() == '+' ? 1 : -1);
+      if (scanner->Peek().IsFixedLengthNumber(4)) {
+        // hhmm extension syntax.
+        int hourmin = scanner->Next().number();
+        int hour = hourmin / 100;
+        int min = hourmin % 100;
+        if (!TimeComposer::IsHour(hour) || !TimeComposer::IsMinute(min)) {
+          return DateToken::Invalid();
+        }
+        tz->SetAbsoluteHour(hour);
+        tz->SetAbsoluteMinute(min);
+      } else {
+        // hh:mm standard syntax.
+        if (!scanner->Peek().IsFixedLengthNumber(2) ||
+            !TimeComposer::IsHour(scanner->Peek().number())) {
+          return DateToken::Invalid();
+        }
+        tz->SetAbsoluteHour(scanner->Next().number());
+        if (!scanner->SkipSymbol(':')) return DateToken::Invalid();
+        if (!scanner->Peek().IsFixedLengthNumber(2) ||
+            !TimeComposer::IsMinute(scanner->Peek().number())) {
+          return DateToken::Invalid();
+        }
+        tz->SetAbsoluteMinute(scanner->Next().number());
+      }
+    }
+    if (!scanner->Peek().IsEndOfInput()) return DateToken::Invalid();
+  }
+  // Successfully parsed ES5 Date Time String. Default to UTC if no TZ given.
+  if (tz->IsEmpty()) tz->Set(0);
+  day->set_iso_date();
+  return DateToken::EndOfInput();
+}
+
+
 } }  // namespace v8::internal
 
 #endif  // V8_DATEPARSER_INL_H_