__wingettotalresttime2.py

#!/usr/bin/env python

import re
import sys
import subprocess
import time
import os
from enum import IntEnum
import copy
import datetime
import math
import texttable as tt

def print_table(list_of_rows):
    tab = tt.Texttable(max_width = 0)
    tab.set_chars(['', '|', '', ''])
    tab.set_deco(tt.Texttable.VLINES| tt.Texttable.BORDER)
    for row in list_of_rows:
        tab.add_row(row)

    print tab.draw()


date_re = re.compile(r'^[0-9-_]+')
rest_re = re.compile(r'^rest +(\d+) +min')
rest_d_re = re.compile(r'^rest *$')
rest_q_re = re.compile(r'^rest *[?]+ *$')

def DEFAULT_TIME_STEP_IN_SECONDS():
    return int(15)

def DEFAULT_TIME_STEP_AS_TIMEDELTA():
    return datetime.timedelta(seconds = DEFAULT_TIME_STEP_IN_SECONDS())

def get_index_for_time(cur_time, first_time):
    delta1 = cur_time - first_time
    assert(delta1.days == 0)
    return int(delta1.seconds) // DEFAULT_TIME_STEP_IN_SECONDS()

def get_time_for_index(cur_index, first_time):
    return first_time + ( DEFAULT_TIME_STEP_AS_TIMEDELTA() * cur_index )

def convert_index_to_another_time_array(cur_index, first_time_from, first_time_to):
    time1 = get_time_for_index(cur_index, first_time_from)
    return get_index_for_time(time1, first_time_to)

class LogFileHandling:
    @staticmethod
    def DEFAULT_FOLDER_WITH_LOGS():
        #return "/home/leonid/worklog/"
        return "/home/lbeynens//worklog/"

    @staticmethod
    def current_worklog_name():
        return "winlog_" + time.strftime("%Y-%m-%d")

    @staticmethod
    def current_worklog_path():
        return os.path.join(LogFileHandling.DEFAULT_FOLDER_WITH_LOGS(), LogFileHandling.current_worklog_name())

    @staticmethod
    def current_worklog_remote_path():
        return LogFileHandling.current_worklog_path()

    @staticmethod
    def current_worklog_path_from_remote():
        return LogFileHandling.current_worklog_path() + ".remote"

#    @staticmethod
#    def scp_from_remote(ip_final_part):
#        assert(int(ip_final_part))
#        subprocess.call(["scp", "lbeynens@192.168.0." + str(int(ip_final_part)) + ":" + LogFileHandling.current_worklog_remote_path(), LogFileHandling.current_worklog_path_from_remote()])




# Rules of parsing of the main file:
#
# * If there is a timestamp in a line, then this time will have state "must_be_work"
#
# * If there is a segment of lines without timestamps and "^rest" patterns, then this segment will have state "may_be_work"
#   The lines with pattern "^work" may be absent in this case.
#
# * If there is a segment of lines without timestamps, with only "^rest *(\d+) *min" patterns, and WITHOUT "^rest *???" and "^rest *$" patterns, then this time segment
#   will correspond to a record in the list of "special time segments".
#   The record will contain information on rest with the pointed number of minutes and on work with the remaining number of minutes. All the items in the time segment will
#   contain the reference to the index of the time segment (see the class SpecialTimeSegment below)
#
#   But if the total number of rest minutes in the special segment is greater or equal the length of the time segment, then the segment will be marked as "must_be_rest"
#   (and special warning will be printed).
#
# * If there is a segment of lines without timestamps, with only "^rest *$" patterns, and WITHOUT "^rest *???" and "rest \d+ min" patterns, then this time segment
#   will be marked as "must_be_rest"
#
# * If there is a segment of lines without timestamps, with "^rest *???" patterns, then this time segment
#   will be marked as "may_be_rest", and any information how much minutes are written in the lines with "^rest *(\d+) *min" will be IGNORED (TODO: think about that)
#
# * Particular case of the latter item: if there is a segment of lines without timestamps, with both
#       "^rest *(\d+) *min *[^?]*$" and
#       "^rest *???" patterns,
#   then a special warning will be printed.
#   Note that the lines with the pattern "^rest *\d+ *min *?+$" will be considered as pattern "^rest *\d+ *min", not as "^rest *???", so
#   if there are only these lines in a segment, then it will be considered either as special time segment or "must_be_rest"
#   (depending on the number of rest minutes)
#
# * The lines with "^work" pattern will be ignored
#
#  So, we have the labels "must_be_work", "may_be_work", "must_be_rest", "may_be_rest", and special records for mixed "special time segments".
#
#
# Rules of parsing of the REMOTE File:
#
# * If there is a timestamp in a line, then this time will have state "must_be_work"
#
# * If there is a segment of lines without timestamps and "^rest" patterns, then this segment will have state "may_be_work"
#   The lines with pattern "^work" may be absent in this case.
#
# * If there is a segment of lines without timestamps that contains "^rest.*" pattern, then this time segment will have state "may_be_rest"
#   (We should not edit the remote file with complicated rest messages)
#
# * The lines with "^work" pattern will be ignored
#
#  So, after parsing a remote file we will have the labels "must_be_work", "may_be_work", and "may_be_rest" only.
#
#
#
# Rules of merging:
#
# the first is local file, the second -- REMOTE
# * X + X => X for any X except special time segments
#
# * must_be_work + must_be_work => must_be_work
# * must_be_work + may_be_work => must_be_work
# * must_be_work + may_be_rest => must_be_work
#
# * may_be_work + must_be_work => must_be_work
# * may_be_work + may_be_work => may_be_work
# * may_be_work + may_be_rest => may_be_work
#
# * must_be_rest + must_be_work => must_be_work (and a warning should be printed)
# * must_be_rest + may_be_work => must_be_rest
# * must_be_rest + may_be_rest => must_be_rest
#
# * may_be_rest + must_be_work => must_be_work
# * may_be_rest + may_be_work => may_be_work
# * may_be_rest + may_be_rest => may_be_rest
#
# * special time segment + must_be_work => must_be_work, and the number of rest items in the special time segment is decreased by 1
# * special time segment + may_be_work => may_be_work, and the number of rest items in the special time segment is decreased by 1
# * special time segment + may_be_rest => this special time segment without changes!
#
#  So, it is sufficient to make order
#  may_be_rest < may_be_work < must_be_rest < must_be_work
#  and for adding "normal" states just make maximum.
#
#  Attention:
#  Note that these rules for merging special time segments shows that in the main log file we should not mark as "work" the time period
#  when we work on another computer with a log that will be used as a remote log later.
#
#
# Rules of calculation:
#
# * must_be_work and may_be_work add 1 to the counter of work items
# * must_be_rest and may_be_rest add 1 to the counter of rest items
# * all special time segments that have negative number of work minutes are counted as "may_be_rest" and add 1 to the counter of rest time items
# * all special time segments that have negative number of rest minutes are counted as "may_be_work" and add 1 to the counter of work time items
# * all special time segments that have non-negative numbers of work and rest minutes convert the minutes to time items and add them to the corresponding counters

class State(IntEnum):
    may_be_rest = 1
    may_be_work = 2
    must_be_rest = 3
    must_be_work = 4

class SpecialTimeSegment:
    def __init__(self, first_time, last_time, num_rest_minutes):
        self.first_time = first_time
        self.last_time = last_time
        delta_time = last_time - first_time
        self.num_items_total = int(delta_time.seconds) // DEFAULT_TIME_STEP_IN_SECONDS();
        self.num_items_rest = int(num_rest_minutes) * 60 // DEFAULT_TIME_STEP_IN_SECONDS();
        self.num_items_work = self.num_items_total - self.num_items_rest

#class ReferenceToSpecialTimeSegment:
#    def __init__(self, index):
#        self.index = index

class Data:
    def __init__(self, first_time, last_time):
        delta1 = last_time - first_time
        assert(delta1.days == 0)
        required_len = 1 + int(delta1.seconds) // DEFAULT_TIME_STEP_IN_SECONDS()

        self.first_time = first_time
        self.last_time = last_time
        self.items = [None] * required_len
        self.special_time_segments = []





class FirstLastTimeDetector:
    @staticmethod
    def round_sec_in_datetime(cur_time, should_floor): #cur_time should be datetime.datetime
        time1 = copy.deepcopy(cur_time)
        time_sec = time1.second
        index_float = float(time_sec) / DEFAULT_TIME_STEP_IN_SECONDS()
        if should_floor:
            index = math.floor(index_float)
        else:
            index = math.ceil(index_float)

        new_seconds = int(index) * DEFAULT_TIME_STEP_IN_SECONDS()
        if new_seconds < 60:
            time1.replace(second = new_seconds)
        else:
            time1.replace(second = new_seconds-60)
            time1 = time1 + datetime.timedelta(minutes = 1)
        return time1


    @staticmethod
    def get_first_last_time_from_file(file_path):
        first_time = None
        last_time = None

        if not os.path.isfile(file_path):
            return (None, None)

        with open(file_path, "r") as f:
            for line in f:
                date_match = date_re.match(line)
                if not date_match:
                    continue
                cur_time_str = date_match.group()
                cur_time = datetime.datetime.strptime(cur_time_str, "%Y-%m-%d_%H-%M-%S")
                if not first_time:
                    first_time = cur_time
                last_time = cur_time

        first_time = FirstLastTimeDetector.round_sec_in_datetime(first_time, should_floor = True)
        last_time = FirstLastTimeDetector.round_sec_in_datetime(last_time, should_floor = False)
        return (first_time, last_time)

    @staticmethod
    def get_first_last_time_from_files(file_path1, file_path2):
        (first_time1, last_time1) = FirstLastTimeDetector.get_first_last_time_from_file(file_path1)
        (first_time2, last_time2) = FirstLastTimeDetector.get_first_last_time_from_file(file_path2)
        if (first_time2 == None) or (last_time2 == None):
            return (first_time1, last_time1)
        return ( min(first_time1, first_time2), max(last_time1, last_time2) )

class Reader:
    def __init__(self, file_path, first_time, last_time, is_remote):
        self.file_path = file_path
        self.first_time = first_time
        self.last_time = last_time
        self.is_remote = is_remote;

        #state
        self.clean_state(prev_index = None, prev_time = None)

    def clean_state(self, prev_index, prev_time):
#        print "clean_state: prev_index =", prev_index, "prev_time =", prev_time
        self.prev_filled_index = prev_index
        self.prev_filled_time = prev_time
        self.is_rest = False
        self.is_rest_definitely = False
        self.is_rest_q = False
        self.total_sum_rest_minutes_for_block = 0

    def fill_indexes_in_data(self, data, cur_index, cur_time):
        assert( (cur_index >= 0) and isinstance(data, Data) and (cur_index < len(data.items)) )

        data.items[cur_index] = State.must_be_work

        if self.prev_filled_index == None:
            return

        if self.prev_filled_index == cur_index-1:
            return

        if self.prev_filled_index == cur_index:
            return #TODO: think about that

        prev = self.prev_filled_index #just alias
#        print "prev =", prev, "cur_index =", cur_index, "len(data.items) =", len(data.items)
        assert( (prev >= 0) and (prev < len(data.items)) and (prev < cur_index))

        cur_state = None
        if not self.is_rest and not self.is_rest_q and not self.is_rest_definitely:
            cur_state = State.may_be_work
        elif self.is_rest_q and self.is_rest_definitely:
            print ("WARNING: in the file '{}' in the time segment from {} to {} both 'definitely rest' and 'may be rest' marks are present " \
                    + "-- make the segment to be 'definitely rest'").format(self.file_path, self.prev_filled_time, self.cur_time)
            cur_state = State.must_be_rest
        elif self.is_rest and self.is_rest_definitely:
            print ("WARNING: in the file '{}' in the time segment from {} to {} both 'definitely rest' and 'rest' marks are present " \
                    + "-- make the segment to be 'definitely rest'").format(self.file_path, self.prev_filled_time, self.cur_time)
            cur_state = State.must_be_rest
        elif self.is_rest and self.is_rest_q:
            print ("WARNING: in the file '{}' in the time segment from {} to {} both 'may be rest' and 'rest' marks are present " \
                    + "-- make the segment to be 'may be rest'").format(self.file_path, self.prev_filled_time, cur_time)
            cur_state = State.may_be_rest
        elif  self.is_rest_definitely:
            cur_state = State.must_be_rest
        elif self.is_rest_q:
            cur_state = State.may_be_rest
        else:
            assert(self.is_rest and not self.is_rest_q and not self.is_rest_definitely)
            if self.is_remote:
                cur_state = State.may_be_rest
            else:
                spec_time_segment = SpecialTimeSegment(self.prev_filled_time, cur_time, self.total_sum_rest_minutes_for_block)

                if spec_time_segment.num_items_work <= 1 + 60.0 / DEFAULT_TIME_STEP_IN_SECONDS():
                    cur_state = State.may_be_rest
                elif spec_time_segment.num_items_rest <= 1 + 60.0 / DEFAULT_TIME_STEP_IN_SECONDS():
                    cur_state = State.may_be_work
                else:
                    data.special_time_segments.append(spec_time_segment)
                    last_index = len(data.special_time_segments) - 1
                    cur_state = data.special_time_segments[last_index]

        if self.is_remote and (cur_state == State.must_be_rest):
            cur_state = State.may_be_rest

        for index in xrange(prev+1, cur_index):
            data.items[index] = cur_state

        pass


    def parse(self):
        data = Data(self.first_time, self.last_time)
        self.clean_state(prev_index = None, prev_time = None)

        with open(self.file_path, "r") as f:
            for line in f:
                is_empty_line = (len(line.strip()) == 0)
                if is_empty_line:
                    continue

                date_match = date_re.match(line)
                rest_match = rest_re.match(line)
                rest_d_match = rest_d_re.match(line)
                rest_q_match = rest_q_re.match(line)

                if date_match:
                    cur_time_str = date_match.group()
                    cur_time = datetime.datetime.strptime(cur_time_str, "%Y-%m-%d_%H-%M-%S")
                    cur_index = get_index_for_time(cur_time, self.first_time)
                    assert( (cur_index >= 0) and (cur_index < len(data.items)) )
                    self.fill_indexes_in_data(data, cur_index, cur_time)

                    self.clean_state(prev_index = cur_index, prev_time = cur_time)

                    continue

                if rest_match:
                    self.is_rest = True
                    self.total_sum_rest_minutes_for_block += int(rest_match.group(1))

                if rest_q_match:
                    self.is_rest_q = True

                if rest_d_match:
                    self.is_rest_definitely = True

        return data

def merge_data(data1, data2):
    if (data2 == None):
        return data1

    data_loc = copy.deepcopy(data1)
    data_rem = copy.deepcopy(data2)
    #data_loc is local
    #data_rem is remote
    assert(isinstance(data_loc, Data))
    assert(isinstance(data_rem, Data))
    assert(len(data_loc.items) == len(data_rem.items))
    assert(data_loc.first_time == data_rem.first_time)
    assert(data_loc.last_time == data_rem.last_time)

    merged_data = Data(data_loc.first_time, data_loc.last_time)
    merged_data.special_time_segments = data_loc.special_time_segments

    for index in xrange(0, len(merged_data.items)):
        item1 = data_loc.items[index]
        item2 = data_rem.items[index]

        if item1 == None:
            merged_data.items[index] = item2
            continue
        if item2 == None:
            merged_data.items[index] = item1
            continue

        assert(isinstance(item2, State))
        assert(item2 != State.must_be_rest)

        if (isinstance(item1, State) and isinstance(item2, State)): #simple case

            if (item1 == item2):
                merged_data.items[index] = item1
                continue

            if (item1 == State.must_be_rest) and (item2 == State.must_be_work):
                print ("Warning: conflict during merging, setting must_be_work state: in the time item with index {} " \
                        + "(corresponds {}) item1 is {} whereas item2 is {}").format(index, get_time_for_index(index, merged_data.first_time), item1, item2)
                merged_data.items[index] = State.must_be_work
                continue


            merged_data.items[index] = max(item1, item2)
            continue


        if isinstance(item1, SpecialTimeSegment):
            if item2 == State.may_be_rest:
                merged_data.items[index] = item1
                continue
            if (item2 == State.must_be_work) or (item2 == State.may_be_work):
                merged_data.items[index] = item2
                item1.num_items_rest -= 1
                continue
            raise RuntimeError()
        pass
    return merged_data

def calculate_time_info(data):
    num_rest_items_from_list = 0;
    num_work_items_from_list = 0;

    for item in data.items:
        if item == None:
            continue

        if item in (State.may_be_work, State.must_be_work):
            num_work_items_from_list += 1
            continue

        if item in (State.may_be_rest, State.must_be_rest):
            num_rest_items_from_list += 1
            continue

        if isinstance(item, SpecialTimeSegment):
            if (item.num_items_rest >=0) and (item.num_items_work >= 0):
                continue #will be handled below
            if (item.num_items_rest <=0) and (item.num_items_work <= 0):
                continue #should not be handled
            if (item.num_items_rest < 0):
                num_work_items_from_list += 1
                continue
            if (item.num_items_work < 0):
                num_rest_items_from_list += 1
                continue

        print "ERROR: item =", item
        raise RuntimeError()

    num_rest_items = num_rest_items_from_list
    num_work_items = num_work_items_from_list

    for item in data.special_time_segments:
        if (item.num_items_rest < 0) or (item.num_items_work < 0):
            continue #are handled above
        num_rest_items += item.num_items_rest
        num_work_items += item.num_items_work

    return { "num_rest_items_from_list": num_rest_items_from_list,
            "num_work_items_from_list": num_work_items_from_list,
            "num_rest_items": num_rest_items,
            "num_work_items": num_work_items}

def str_timedelta(td):
    return ':'.join(str(td).split(':')[:2])

def print_time_info(time_info):
    print time_info
    print "rest from list =", str_timedelta(int(time_info["num_rest_items_from_list"]) * DEFAULT_TIME_STEP_AS_TIMEDELTA())
    print "work from list =", str_timedelta(int(time_info["num_work_items_from_list"]) * DEFAULT_TIME_STEP_AS_TIMEDELTA())
    print "total rest =", str_timedelta(int(time_info["num_rest_items"]) * DEFAULT_TIME_STEP_AS_TIMEDELTA())
    print "total work =", str_timedelta(int(time_info["num_work_items"]) * DEFAULT_TIME_STEP_AS_TIMEDELTA())
    print "speed of rest ={:.2}".format( float(time_info["num_rest_items"]) / float(time_info["num_rest_items"] + time_info["num_work_items"]) )

    time_to_should_work_for_target = datetime.timedelta(hours = 6) - datetime.timedelta(seconds = int(time_info["num_work_items"]) * DEFAULT_TIME_STEP_IN_SECONDS())
    ideal_time_of_target = datetime.datetime.now() + time_to_should_work_for_target
    print "TIME_TO_SHOULD_WORK_FOR_TARGET =", time_to_should_work_for_target
    print "IDEAL_TIME_OF_TARGET =", ideal_time_of_target.strftime("%H:%M:%S")

def print_data_as_list(data, should_shorten = False):
    if data == None:
        print "No data"
        return

    assert(isinstance(data, Data))
    prev_item = data.items[0]
    prev_index = 0

    list_rests = []

    for (cur_index, cur_item) in enumerate(data.items):
        if (cur_item == prev_item) and (cur_index+1 != len(data.items)):
            continue
        prev_time = get_time_for_index(prev_index, data.first_time)
        cur_time = get_time_for_index(cur_index-1, data.first_time)
        diff_time = cur_time - prev_time
        diff_time_min = diff_time.seconds // 60

        str_prev_time = prev_time.strftime("%H:%M:%S")
        str_cur_time = cur_time.strftime("%H:%M:%S")

        str_prev_item = str(prev_item)
        if isinstance(prev_item, SpecialTimeSegment):
            w = prev_item.num_items_work * DEFAULT_TIME_STEP_IN_SECONDS() // 60
            r = prev_item.num_items_rest * DEFAULT_TIME_STEP_IN_SECONDS() // 60
            str_prev_item = "work: {} rest: {}".format(w, r)
        else:
            if prev_item == State.must_be_work:
                str_prev_item = "W"
            elif prev_item == State.must_be_rest:
                str_prev_item = "R"
            elif prev_item == State.may_be_work:
                str_prev_item = "W?"
            elif prev_item == State.may_be_rest:
                str_prev_item = "R?"

        if (not should_shorten) or (diff_time_min > 2):
            print "{} => {}  ({:3} min): {}".format(str_prev_time, str_cur_time, diff_time_min, str_prev_item)

        if ( (prev_item == State.must_be_rest) or (prev_item == State.may_be_rest) ) and (diff_time_min > 2):
            list_rests.append(diff_time_min)

        prev_index = cur_index
        prev_item = cur_item

    print "rests from lists:", list_rests
    rests_from_spec_time_segments = [ int(x.num_items_rest) * DEFAULT_TIME_STEP_IN_SECONDS() // 60 for x in data.special_time_segments ]
    print "rests from special time segments:", rests_from_spec_time_segments


def print_data_as_table(data1, data2 = None, data3 = None):
    def str_for_item(item):
        str_item = str(item)
        if isinstance(item, SpecialTimeSegment):
            w = item.num_items_work * DEFAULT_TIME_STEP_IN_SECONDS() // 60
            r = item.num_items_rest * DEFAULT_TIME_STEP_IN_SECONDS() // 60
            str_item = "w: {} r: {}".format(w, r)
        else:
            if item == State.must_be_work:
                str_item = "W"
            elif item == State.must_be_rest:
                str_item = "R"
            elif item == State.may_be_work:
                str_item = "W?"
            elif item == State.may_be_rest:
                str_item = "R?"
        return str_item

    assert(isinstance(data1, Data))
    if data2:
        assert(isinstance(data2, Data))
    if data3:
        assert(isinstance(data3, Data))

    first_time = data1.first_time
    last_time = data1.last_time

    if data2:
        first_time = min(first_time, data2.first_time)
        last_time = max(last_time, data2.last_time)
    if data3:
        first_time = min(first_time, data3.first_time)
        last_time = max(last_time, data3.last_time)

    num_items = get_index_for_time(last_time, first_time) + 1

    table_rows = []
    for cur_index in xrange(num_items):
        index1 = convert_index_to_another_time_array(cur_index, first_time, data1.first_time)
        item1 = data1.items[index1]
        item2 = None
        item3 = None
        if data2:
            index2 = convert_index_to_another_time_array(cur_index, first_time, data2.first_time)
            item2 = data2.items[index2]
        if data3:
            index3 = convert_index_to_another_time_array(cur_index, first_time, data3.first_time)
            item3 = data3.items[index3]

        cur_time = get_time_for_index(cur_index, first_time)
        str_cur_time = cur_time.strftime("%H:%M:%S")

        table_rows.append([str_cur_time, str_for_item(item1), str_for_item(item2), str_for_item(item3)])
    print_table(table_rows)


if __name__ == "__main__":
#    ip_final_part = None
#    if len(sys.argv) > 1:
#        ip_final_part = int(sys.argv[1])
#
#    if ip_final_part:
#        LogFileHandling.scp_from_remote(ip_final_part)

    should_print_whole_table = False
    if (len(sys.argv) > 1):
        if (sys.argv[1] == "all"):
            should_print_whole_table = True

    file1 = LogFileHandling.current_worklog_path()
    file2 = LogFileHandling.current_worklog_path_from_remote()

    (total_first_time, total_last_time) = FirstLastTimeDetector.get_first_last_time_from_files(file1, file2)

    reader1 = Reader(file1, total_first_time, total_last_time, is_remote=False)
    if os.path.isfile(file2):
        reader2 = Reader(file2, total_first_time, total_last_time, is_remote=True)
    else:
        reader2 = None

    data1 = reader1.parse()

    if reader2:
        data2 = reader2.parse()
    else:
        data2 = None

    data_merged = merge_data(data1, data2)

    time_info = calculate_time_info(data_merged)

    print ("=" * 80)
    print "Local"
    print_data_as_list(data1)
    print ""
    print "Remote"
    print_data_as_list(data2)
    print ""
    print "Merged"
    print_data_as_list(data_merged)

    print ("=" * 80)
    print "Merged shortened"
    print_data_as_list(data_merged, True)

    if should_print_whole_table:
        print ("=" * 80)
        print_data_as_table(data1, data2, data_merged)

    print ("=" * 80)
    print_time_info(time_info)