Merge pull request #1383 from sbalko/incoming

Implement strptime

3 files changed, 398 insertions, 1 deletions

diff --git a/AUTHORS b/AUTHORS
index 752e7593..021240e2 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -87,3 +87,4 @@ a license to everyone to use it as detailed in LICENSE.)
 * Manfred Manik Nerurkar <nerurkar*at*made-apps.biz> (copyright owned by MADE, GmbH)
 * Joseph Gentle <me@josephg.com>
 * Douglas T. Crosher <dtc-moz@scieneer.com> (copyright owned by Mozilla Founcation)
+* Soeren Balko <soeren.balko@gmail.com>
diff --git a/src/library.js b/src/library.js
index 20a163c6..24e35178 100644
--- a/src/library.js
+++ b/src/library.js
@@ -6215,10 +6215,282 @@ LibraryManager.library = {
   },
   strftime_l: 'strftime', // no locale support yet
 
+  strptime__deps: ['__tm_struct_layout'],
   strptime: function(buf, format, tm) {
     // char *strptime(const char *restrict buf, const char *restrict format, struct tm *restrict tm);
     // http://pubs.opengroup.org/onlinepubs/009695399/functions/strptime.html
-    // TODO: Implement.
+    var pattern = Pointer_stringify(format);
+
+    // escape special characters
+    // TODO: not sure we really need to escape all of these in JS regexps
+    var SPECIAL_CHARS = '\\!@#$^&*()+=-[]/{}|:<>?,.';
+    for (var i=0, ii=SPECIAL_CHARS.length; i<ii; ++i) {
+      pattern = pattern.replace(new RegExp('\\'+SPECIAL_CHARS[i], 'g'), '\\'+SPECIAL_CHARS[i]);
+    }
+
+    // reduce number of matchers
+    var EQUIVALENT_MATCHERS = {
+      '%A':  '%a',
+      '%B':  '%b',
+      '%c':  '%x\\s+%X',
+      '%D':  '%m\\/%d\\/%y',
+      '%e':  '%d',
+      '%h':  '%b',
+      '%R':  '%H\\:%M',
+      '%r':  '%I\\:%M\\:%S\\s%p',
+      '%T':  '%H\\:%M\\:%S',
+      '%x':  '%m\\/%d\\/(?:%y|%Y)',
+      '%X':  '%H\\:%M\\:%S'
+    };
+    for (var matcher in EQUIVALENT_MATCHERS) {
+      pattern = pattern.replace(matcher, EQUIVALENT_MATCHERS[matcher]);
+    }
+    
+    // TODO: take care of locale
+
+    var DATE_PATTERNS = {
+      /* weeday name */     '%a': '(?:Sun(?:day)?)|(?:Mon(?:day)?)|(?:Tue(?:sday)?)|(?:Wed(?:nesday)?)|(?:Thu(?:rsday)?)|(?:Fri(?:day)?)|(?:Sat(?:urday)?)',
+      /* month name */      '%b': '(?:Jan(?:uary)?)|(?:Feb(?:ruary)?)|(?:Mar(?:ch)?)|(?:Apr(?:il)?)|May|(?:Jun(?:e)?)|(?:Jul(?:y)?)|(?:Aug(?:ust)?)|(?:Sep(?:tember)?)|(?:Oct(?:ober)?)|(?:Nov(?:ember)?)|(?:Dec(?:ember)?)',
+      /* century */         '%C': '\\d\\d',
+      /* day of month */    '%d': '0[1-9]|[1-9](?!\\d)|1\\d|2\\d|30|31',
+      /* hour (24hr) */     '%H': '\\d(?!\\d)|[0,1]\\d|20|21|22|23',
+      /* hour (12hr) */     '%I': '\\d(?!\\d)|0\\d|10|11|12',
+      /* day of year */     '%j': '00[1-9]|0?[1-9](?!\\d)|0?[1-9]\\d(?!\\d)|[1,2]\\d\\d|3[0-6]\\d',
+      /* month */           '%m': '0[1-9]|[1-9](?!\\d)|10|11|12',
+      /* minutes */         '%M': '0\\d|\\d(?!\\d)|[1-5]\\d',
+      /* whitespace */      '%n': '\\s',
+      /* AM/PM */           '%p': 'AM|am|PM|pm|A\\.M\\.|a\\.m\\.|P\\.M\\.|p\\.m\\.',
+      /* seconds */         '%S': '0\\d|\\d(?!\\d)|[1-5]\\d|60',
+      /* week number */     '%U': '0\\d|\\d(?!\\d)|[1-4]\\d|50|51|52|53',
+      /* week number */     '%W': '0\\d|\\d(?!\\d)|[1-4]\\d|50|51|52|53',
+      /* weekday number */  '%w': '[0-6]',
+      /* 2-digit year */    '%y': '\\d\\d',
+      /* 4-digit year */    '%Y': '\\d\\d\\d\\d',
+      /* % */               '%%': '%',
+      /* whitespace */      '%t': '\\s',
+    };
+
+    var MONTH_NUMBERS = {JAN: 0, FEB: 1, MAR: 2, APR: 3, MAY: 4, JUN: 5, JUL: 6, AUG: 7, SEP: 8, OCT: 9, NOV: 10, DEC: 11};
+    var MONTH_DAYS_REGULAR = [31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31];
+    var MONTH_DAYS_LEAP = [31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31];
+    var DAY_NUMBERS_SUN_FIRST = {SUN: 0, MON: 1, TUE: 2, WED: 3, THU: 4, FRI: 5, SAT: 6};
+    var DAY_NUMBERS_MON_FIRST = {MON: 0, TUE: 1, WED: 2, THU: 3, FRI: 4, SAT: 5, SUN: 6};
+
+    var isLeapYear = function(year) {
+      return year%4===0 && (year%100!==0 || year%400===0);
+    };
+
+    var arraySum = function(array, index) {
+      var sum = 0;
+      for (var i=0; i<=index; sum += array[i++]);
+      return sum;
+    };
+
+    var addDays = function(date, days) {
+      while(days>0) {
+        var leap = isLeapYear(date.getFullYear());
+        var currentMonth = date.getMonth();
+        var daysInCurrentMonth = (leap ? MONTH_DAYS_LEAP : MONTH_DAYS_REGULAR)[currentMonth];
+
+        if (days>daysInCurrentMonth-date.getDate()) {
+          // we spill over to next month
+          days -= (daysInCurrentMonth-date.getDate()+1);
+          date.setDate(1);
+          if (currentMonth<11) {
+            date.setMonth(currentMonth+1)
+          } else {
+            date.setMonth(0);
+            date.setFullYear(date.getFullYear()+1);
+          }
+        } else {
+          // we stay in current month 
+          date.setDate(date.getDate()+days);
+          return date;
+        }
+      }
+      return date;
+    };
+
+    for (var datePattern in DATE_PATTERNS) {
+      pattern = pattern.replace(datePattern, '('+datePattern+DATE_PATTERNS[datePattern]+')');    
+    }
+
+    // take care of capturing groups
+    var capture = [];
+    for (var i=pattern.indexOf('%'); i>=0; i=pattern.indexOf('%')) {
+      capture.push(pattern[i+1]);
+      pattern = pattern.replace(new RegExp('\\%'+pattern[i+1], 'g'), '');
+    }
+
+    var matches = new RegExp('^'+pattern).exec(Pointer_stringify(buf))
+    // Module['print'](Pointer_stringify(buf)+ ' is matched by '+((new RegExp('^'+pattern)).source)+' into: '+JSON.stringify(matches));
+
+    var initDate = function() {
+      var fixup = function(value, min, max) {
+        return (typeof value !== 'number' || isNaN(value)) ? min : (value>=min ? (value<=max ? value: max): min);
+      };
+
+      return {
+        year: fixup({{{ makeGetValue('tm', '___tm_struct_layout.tm_year', 'i32') }}} + 1900 , 1970, 9999),
+        month: fixup({{{ makeGetValue('tm', '___tm_struct_layout.tm_mon', 'i32') }}}, 0, 11),
+        day: fixup({{{ makeGetValue('tm', '___tm_struct_layout.tm_mday', 'i32') }}}, 1, 31),
+        hour: fixup({{{ makeGetValue('tm', '___tm_struct_layout.tm_hour', 'i32') }}}, 0, 23),
+        min: fixup({{{ makeGetValue('tm', '___tm_struct_layout.tm_min', 'i32') }}}, 0, 59),
+        sec: fixup({{{ makeGetValue('tm', '___tm_struct_layout.tm_sec', 'i32') }}}, 0, 59)
+      };
+    };
+
+    if (matches) {
+      var date = initDate();
+      var value;
+
+      var getMatch = function(symbol) {
+        var pos = capture.indexOf(symbol);
+        // check if symbol appears in regexp
+        if (pos >= 0) {
+          // return matched value or null (falsy!) for non-matches
+          return matches[pos+1];
+        }
+        return;
+      }
+
+      // seconds
+      if ((value=getMatch('S'))) {
+        date.sec = parseInt(value);
+      }
+
+      // minutes
+      if ((value=getMatch('M'))) {
+        date.min = parseInt(value);
+      }
+
+      // hours
+      if ((value=getMatch('H'))) {
+        // 24h clock
+        date.hour = parseInt(value);
+      } else if ((value = getMatch('I'))) {
+        // AM/PM clock
+        var hour = parseInt(value);
+        if ((value=getMatch('p'))) {
+          hour += value.toUpperCase()[0] === 'P' ? 12 : 0;
+        }
+        date.hour = hour;
+      }
+
+      // year
+      if ((value=getMatch('Y'))) {
+        // parse from four-digit year
+        date.year = parseInt(value);
+      } else if ((value=getMatch('y'))) {
+        // parse from two-digit year...
+        var year = parseInt(value);
+        if ((value=getMatch('C'))) {
+          // ...and century
+          year += parseInt(value)*100;
+        } else {
+          // ...and rule-of-thumb
+          year += year<69 ? 2000 : 1900;
+        }
+        date.year = year;
+      }
+
+      // month
+      if ((value=getMatch('m'))) {
+        // parse from month number
+        date.month = parseInt(value)-1;
+      } else if ((value=getMatch('b'))) {
+        // parse from month name
+        date.month = MONTH_NUMBERS[value.substring(0,3).toUpperCase()] || 0;
+        // TODO: derive month from day in year+year, week number+day of week+year 
+      }
+
+      // day
+      if ((value=getMatch('d'))) {
+        // get day of month directly
+        date.day = parseInt(value);
+      } else if ((value=getMatch('j'))) {
+        // get day of month from day of year ...
+        var day = parseInt(value);
+        var leapYear = isLeapYear(date.year);
+        for (var month=0; month<12; ++month) {
+          var daysUntilMonth = arraySum(leapYear ? MONTH_DAYS_LEAP : MONTH_DAYS_REGULAR, month-1);
+          if (day<=daysUntilMonth+(leapYear ? MONTH_DAYS_LEAP : MONTH_DAYS_REGULAR)[month]) {
+            date.day = day-daysUntilMonth;
+          }
+        }
+      } else if ((value=getMatch('a'))) {
+        // get day of month from weekday ...
+        var weekDay = value.substring(0,3).toUpperCase();
+        if ((value=getMatch('U'))) {
+          // ... and week number (Sunday being first day of week)
+          // Week number of the year (Sunday as the first day of the week) as a decimal number [00,53]. 
+          // All days in a new year preceding the first Sunday are considered to be in week 0.
+          var weekDayNumber = DAY_NUMBERS_SUN_FIRST[weekDay];
+          var weekNumber = parseInt(value);
+
+          // January 1st 
+          var janFirst = new Date(date.year, 0, 1);
+          var endDate;
+          if (janFirst.getDay() === 0) {
+            // Jan 1st is a Sunday, and, hence in the 1st CW
+            endDate = addDays(janFirst, weekDayNumber+7*(weekNumber-1));
+          } else {
+            // Jan 1st is not a Sunday, and, hence still in the 0th CW
+            endDate = addDays(janFirst, 7-janFirst.getDay()+weekDayNumber+7*(weekNumber-1));
+          }
+          date.day = endDate.getDate();
+          date.month = endDate.getMonth();
+        } else if ((value=getMatch('W'))) {
+          // ... and week number (Monday being first day of week)
+          // Week number of the year (Monday as the first day of the week) as a decimal number [00,53]. 
+          // All days in a new year preceding the first Monday are considered to be in week 0.
+          var weekDayNumber = DAY_NUMBERS_MON_FIRST[weekDay];
+          var weekNumber = parseInt(value);
+
+          // January 1st 
+          var janFirst = new Date(date.year, 0, 1);
+          var endDate;
+          if (janFirst.getDay()===1) {
+            // Jan 1st is a Monday, and, hence in the 1st CW
+             endDate = addDays(janFirst, weekDayNumber+7*(weekNumber-1));
+          } else {
+            // Jan 1st is not a Monday, and, hence still in the 0th CW
+            endDate = addDays(janFirst, 7-janFirst.getDay()+1+weekDayNumber+7*(weekNumber-1));
+          }
+
+          date.day = endDate.getDate();
+          date.month = endDate.getMonth();
+        }
+      }
+
+      /*
+      tm_sec  int seconds after the minute  0-61*
+      tm_min  int minutes after the hour  0-59
+      tm_hour int hours since midnight  0-23
+      tm_mday int day of the month  1-31
+      tm_mon  int months since January  0-11
+      tm_year int years since 1900  
+      tm_wday int days since Sunday 0-6
+      tm_yday int days since January 1  0-365
+      tm_isdst  int Daylight Saving Time flag 
+      */
+
+      var fullDate = new Date(date.year, date.month, date.day, date.hour, date.min, date.sec, 0);
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_sec', 'fullDate.getSeconds()', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_min', 'fullDate.getMinutes()', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_hour', 'fullDate.getHours()', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_mday', 'fullDate.getDate()', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_mon', 'fullDate.getMonth()', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_year', 'fullDate.getFullYear()-1900', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_wday', 'fullDate.getDay()', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_yday', 'arraySum(isLeapYear(fullDate.getFullYear()) ? MONTH_DAYS_LEAP : MONTH_DAYS_REGULAR, fullDate.getMonth()-1)+fullDate.getDate()-1', 'i32') }}}
+      {{{ makeSetValue('tm', '___tm_struct_layout.tm_isdst', '0', 'i32') }}}
+
+      // we need to convert the matched sequence into an integer array to take care of UTF-8 characters > 0x7F
+      // TODO: not sure that intArrayFromString handles all unicode characters correctly
+      return buf+intArrayFromString(matches[0]).length-1;
+    } 
+
     return 0;
   },
   strptime_l: 'strptime', // no locale support yet
diff --git a/tests/runner.py b/tests/runner.py
index ea702689..1f6b39f4 100755
--- a/tests/runner.py
+++ b/tests/runner.py
@@ -4958,6 +4958,130 @@ The current type of b is: 9
       '''
       self.do_run(src, 'time: ') # compilation check, mainly
 
+
+    def test_strptime_tm(self):
+      src=r'''
+        #include <time.h>
+        #include <stdio.h>
+        #include <string.h>
+
+        int main() {
+          struct tm tm;
+          char *ptr = strptime("17410105012000", "%H%M%S%d%m%Y", &tm);
+
+          printf("%s: %s, %d/%d/%d %d:%d:%d", 
+            (ptr != NULL && *ptr=='\0') ? "OK" : "ERR", 
+            tm.tm_wday == 0 ? "Sun" : (tm.tm_wday == 1 ? "Mon" : (tm.tm_wday == 2 ? "Tue" : (tm.tm_wday == 3 ? "Wed" : (tm.tm_wday == 4 ? "Thu" : (tm.tm_wday == 5 ? "Fri" : (tm.tm_wday == 6 ? "Sat" : "ERR")))))),
+            tm.tm_mon+1,
+            tm.tm_mday,
+            tm.tm_year+1900,
+            tm.tm_hour,
+            tm.tm_min,
+            tm.tm_sec            
+          );
+        }
+      ''' 
+      self.do_run(src, 'OK: Wed, 1/5/2000 17:41:1')
+
+    def test_strptime_days(self):
+      src = r'''
+        #include <time.h>
+        #include <stdio.h>
+        #include <string.h>
+
+        static const struct {
+          const char *input;
+          const char *format;
+        } day_tests[] = {
+          { "2000-01-01", "%Y-%m-%d"},
+          { "03/03/00", "%D"},
+          { "9/9/99", "%x"},
+          { "19990502123412", "%Y%m%d%H%M%S"},
+          { "2001 20 Mon", "%Y %U %a"},
+          { "2006 4 Fri", "%Y %U %a"},
+          { "2001 21 Mon", "%Y %W %a"},
+          { "2013 29 Wed", "%Y %W %a"},
+          { "2000-01-01 08:12:21 AM", "%Y-%m-%d %I:%M:%S %p"},
+          { "2000-01-01 08:12:21 PM", "%Y-%m-%d %I:%M:%S %p"},
+          { "2001 17 Tue", "%Y %U %a"},
+          { "2001 8 Thursday", "%Y %W %a"},
+        };
+
+        int main() {  
+          struct tm tm;
+
+          for (int i = 0; i < sizeof (day_tests) / sizeof (day_tests[0]); ++i) {
+            memset (&tm, '\0', sizeof (tm));
+            char *ptr = strptime(day_tests[i].input, day_tests[i].format, &tm);
+
+            printf("%s: %d/%d/%d (%dth DoW, %dth DoY)\n", (ptr != NULL && *ptr=='\0') ? "OK" : "ERR", tm.tm_mon+1, tm.tm_mday, 1900+tm.tm_year, tm.tm_wday, tm.tm_yday);
+          }
+        }
+      '''
+      self.do_run(src, 'OK: 1/1/2000 (6th DoW, 0th DoY)\n'\
+                       'OK: 3/3/2000 (5th DoW, 62th DoY)\n'\
+                       'OK: 9/9/1999 (4th DoW, 251th DoY)\n'\
+                       'OK: 5/2/1999 (0th DoW, 121th DoY)\n'\
+                       'OK: 5/21/2001 (1th DoW, 140th DoY)\n'\
+                       'OK: 1/27/2006 (5th DoW, 26th DoY)\n'\
+                       'OK: 5/21/2001 (1th DoW, 140th DoY)\n'\
+                       'OK: 7/24/2013 (3th DoW, 204th DoY)\n'\
+                       'OK: 1/1/2000 (6th DoW, 0th DoY)\n'\
+                       'OK: 1/1/2000 (6th DoW, 0th DoY)\n'\
+                       'OK: 5/1/2001 (2th DoW, 120th DoY)\n'\
+                       'OK: 2/22/2001 (4th DoW, 52th DoY)\n'\
+      )
+
+    def test_strptime_reentrant(self):
+      src=r'''
+        #include <time.h>
+        #include <stdio.h>
+        #include <string.h>
+        #include <stdlib.h>
+
+        int main () {
+          int result = 0;
+          struct tm tm;
+
+          memset (&tm, 0xaa, sizeof (tm));
+
+          /* Test we don't crash on uninitialized struct tm.
+             Some fields might contain bogus values until everything
+             needed is initialized, but we shouldn't crash.  */
+          if (strptime ("2007", "%Y", &tm) == NULL
+              || strptime ("12", "%d", &tm) == NULL
+              || strptime ("Feb", "%b", &tm) == NULL
+              || strptime ("13", "%M", &tm) == NULL
+              || strptime ("21", "%S", &tm) == NULL
+              || strptime ("16", "%H", &tm) == NULL) {
+            printf("ERR: returned NULL");
+            exit(EXIT_FAILURE);
+          }
+
+          if (tm.tm_sec != 21 || tm.tm_min != 13 || tm.tm_hour != 16
+              || tm.tm_mday != 12 || tm.tm_mon != 1 || tm.tm_year != 107
+              || tm.tm_wday != 1 || tm.tm_yday != 42) {
+            printf("ERR: unexpected tm content (1) - %d/%d/%d %d:%d:%d", tm.tm_mon+1, tm.tm_mday, tm.tm_year+1900, tm.tm_hour, tm.tm_min, tm.tm_sec);
+            exit(EXIT_FAILURE);
+          }
+
+          if (strptime ("8", "%d", &tm) == NULL) {
+            printf("ERR: strptime failed");
+            exit(EXIT_FAILURE);
+          }
+
+          if (tm.tm_sec != 21 || tm.tm_min != 13 || tm.tm_hour != 16
+              || tm.tm_mday != 8 || tm.tm_mon != 1 || tm.tm_year != 107
+              || tm.tm_wday != 4 || tm.tm_yday != 38) {
+            printf("ERR: unexpected tm content (2) - %d/%d/%d %d:%d:%d", tm.tm_mon+1, tm.tm_mday, tm.tm_year+1900, tm.tm_hour, tm.tm_min, tm.tm_sec);
+            exit(EXIT_FAILURE);
+          }
+
+          printf("OK");
+        }
+      '''
+      self.do_run(src, 'OK')
+
     def test_intentional_fault(self):
       # Some programs intentionally segfault themselves, we should compile that into a throw
       src = r'''