nightly commit

documents/Principle_zh-CN.md

@@ -8,7 +8,7 @@
 
 本日历系统基于UNIX时间戳进行处理。在设计上使用Int64进行存储以规避2038问题。同时，由于日历设计不需要精确到秒，因此本日历中与时间存储有关的时间戳使用标准UNIX时间戳/60来存储，构成粒度为分钟的时间戳。但是其余的，例如用户登录相关的时间戳，仍然为标准UNIX时间戳
 
-本日历目前无条件限定最小时间为1950年1月1日，最大时间为2200年12月31日。
+本日历目前无条件限定最小时间为1950年1月1日，最大时间为2199年12月31日。
 
 API在遇到当前接口产生错误时，使用外层结构来进行返回错误，例如token无效，应用程序错误，参数错误等。请注意捕获错误。如果没有错误，则表示data字段为有效数据。
 

src/database.py

@@ -4,6 +4,7 @@ import json
 import utils
 import threading
 import logging
+import dt
 
 def SafeDatabaseOperation(func):
     def wrapper(self, *args, **kwargs):

src/dt.py

@@ -0,0 +1,98 @@
+import datetime
+import time
+import re
+from functools import reduce
+
+MIN_TIMESTAMP = int(datetime.datetime(1950, 1, 1, 0, 0, 0, 0, tzinfo=datetime.timezone.utc).timestamp() / 60)
+MAX_TIMESTAMP = int(datetime.datetime(2200, 1, 1, 0, 0, 0, 0, tzinfo=datetime.timezone.utc).timestamp() / 60)
+DAY1_SPAN = 60 * 24
+DAY7_SPAN = 7 * DAY1_SPAN
+YEAR400_SPAN = DAY1_SPAN * 400 * 365
+
+def ResolveLoopStr(strl, starttime, tzoffset):
+    # check no loop
+    if strl == '':
+        return starttime + 1
+    
+    # try compute from loopStop
+    (loopRules, loopStopRules) = strl.split('-')
+    cache = precompiledLoopStopRules['infinity'].search(loopStopRules)
+    if cache is not None:
+        return MAX_TIMESTAMP
+    cache = precompiledLoopStopRules['datetime'].search(loopStopRules)
+    if cache is not None:
+        return int(cache.group(1)) # group 1 is datetime
+    cache = precompiledLoopStopRules['times'].search(loopStopRules)
+    if cache is not None:
+        loopTimes = cache   # for follwing calc
+    else:
+        raise Exception('Invalid loopStopRules')    # invalid rules
+
+    for rules in precompiledLoopRules:
+        cache = rules[0].search(loopRules)
+        if cache is not None:
+             return rules[1](cache, starttime, loopTimes, tzoffset)
+    else:
+        raise Exception('Invalid loopRules')
+
+
+def LoopHandle_Year(searchResult, starttime, times, tzoffset):
+    pass
+
+def LoopHandle_Month(searchResult, starttime, times, tzoffset):
+    pass
+
+def LoopHandle_Week(searchResult, starttime, times, tzoffset):
+    weekOccupied = tuple(map(lambda x: x == 'T', searchResult.group(1)))
+    weekEventCount = reduce(lambda x, y: x + (1 if y else 0), weekOccupied, 0)
+    if weekEventCount == 0:
+        raise Exception('Invalid week format')
+
+    weekSpan = int(searchResult.group(2))
+    nowDayOfWeek = datetime.datetime.fromtimestamp(starttime, UTCTimezone(tzoffset)).weekday()
+    if not weekOccupied[nowDayOfWeek]:
+        times+=1    # if first event is not suit for week loop rules, add one more event to suit it.
+    fullWeek = times / weekEventCount
+    remainEvent = times % weekEventCount
+    
+    val = DAY7_SPAN * fullWeek * weekSpan
+    if val > MAX_TIMESTAMP:
+        return MAX_TIMESTAMP    # return now, to reduce calc usage
+
+    while remainEvent != 0:
+        val += DAY1_SPAN
+        if weekOccupied[nowDayOfWeek % 7]:
+            remainEvent -= 1
+        nowDayOfWeek += 1
+
+    return val if val < MAX_TIMESTAMP else MAX_TIMESTAMP
+
+def LoopHandle_Day(searchResult, starttime, times, tzoffset):
+    val = starttime + DAY1_SPAN * times * int(searchResult.group(1))
+    return val if val < MAX_TIMESTAMP else MAX_TIMESTAMP
+
+precompiledLoopRules = (
+    (re.compile(r'^Y([RF]{1})([1-9]\d*)$'), LoopHandle_Year),
+    (re.compile(r'^M([RF]{1})([ABCD]{1})([1-9]\d*)$'), LoopHandle_Month),
+    (re.compile(r'^W([TF]{7})([1-9]\d*)$'), LoopHandle_Week),
+    (re.compile(r'^D([1-9]\d*)$'), LoopHandle_Day)
+)
+
+precompiledLoopStopRules = {
+    'infinity': re.compile(r'F')
+    'datetime': re.compile(r'^D([1-9]\d*|0)$')
+    'times': re.compile(r'^T([1-9]\d*)$')
+}
+
+class UTCTimezone(datetime.tzinfo):
+    def __init__(self, offset = 0):
+        self._offset = offset
+
+    def utcoffset(self, dt):
+        return datetime.timedelta(minutes=self._offset)
+
+    def tzname(self, dt):
+        return 'UTC {}'.format(self._offset)
+
+    def dst(self, dt):
+        return timedelta(0)