replay.py

from helper import *
from ocr import custom_ocr

smallActionDelay = 0.05
actionDelay = 0.2
menuChangeDelay = 1

def getResolutionDependentData(resolution = pyautogui.size(), gamemode=''):
    requiredComparisonImages = [{'category': 'screens', 'name': 'startmenu'}, {'category': 'screens', 'name': 'map_selection'}, {'category': 'screens', 'name': 'difficulty_selection'}, {'category': 'screens', 'name': 'gamemode_selection'}, {'category': 'screens', 'name': 'hero_selection'}, {'category': 'screens', 'name': 'ingame'}, {'category': 'screens', 'name': 'ingame_paused'}, {'category': 'screens', 'name': 'victory_summary'}, {'category': 'screens', 'name': 'victory'}, {'category': 'screens', 'name': 'defeat'}, {'category': 'screens', 'name': 'overwrite_save'}, {'category': 'screens', 'name': 'levelup'}, {'category': 'screens', 'name': 'apopalypse_hint'}, {'category': 'screens', 'name': 'round_100_insta'}, {'category': 'game_state', 'name': 'game_paused'}, {'category': 'game_state', 'name': 'game_playing_slow'}, {'category': 'game_state', 'name': 'game_playing_fast'}]
    optionalComparisonImages = [{'category': 'screens', 'name': 'collection_claim_chest', 'for': [Mode.CHASE_REWARDS.name]}]
    requiredLocateImages = [{'name': 'remove_obstacle_confirm_button'}, {'name': 'button_home'}]
    optionalLocateImages = [{'name': 'unknown_insta', 'for': [Mode.CHASE_REWARDS.name]}, {'name': 'unknown_insta_mask', 'for': [Mode.CHASE_REWARDS.name]}]

    imagesDir = 'images/' + getResolutionString(resolution) + '/'

    comparisonImages = {}
    locateImages = {}

    if not exists(imagesDir):
        return None

    supportedModes = dict.fromkeys([e.name for e in Mode], True)

    for img in requiredComparisonImages:
        filename = (img['filename'] if 'filename' in img else img['name']) + '.png'
        if not exists(imagesDir + filename):
            print(filename + ' missing!')
            return None
        elif 'category' in img:
            if not img['category'] in comparisonImages:
                comparisonImages[img['category']] = {}
            comparisonImages[img['category']][img['name']] = cv2.imread(imagesDir + filename)
        else:
            comparisonImages[img['name']] = cv2.imread(imagesDir + filename)
    
    for img in optionalComparisonImages:
        filename = (img['filename'] if 'filename' in img else img['name']) + '.png'
        if not exists(imagesDir + filename):
            if 'for' in img:
                for mode in img['for']:
                    supportedModes.pop(mode, None)
        elif 'category' in img:
            if not img['category'] in comparisonImages:
                comparisonImages[img['category']] = {}
            comparisonImages[img['category']][img['name']] = cv2.imread(imagesDir + filename)
        else:
            comparisonImages[img['name']] = cv2.imread(imagesDir + filename)
    
    for img in requiredLocateImages:
        filename = (img['filename'] if 'filename' in img else img['name']) + '.png'
        if not exists(imagesDir + filename):
            print(filename + ' missing!')
            return None
        elif 'category' in img:
            if not img['category'] in locateImages:
                locateImages[img['category']] = {}
            locateImages[img['category']][img['name']] = cv2.imread(imagesDir + filename)
        else:
            locateImages[img['name']] = cv2.imread(imagesDir + filename)
    
    for img in optionalLocateImages:
        filename = (img['filename'] if 'filename' in img else img['name']) + '.png'
        if not exists(imagesDir + filename):
            if 'for' in img:
                for mode in img['for']:
                    supportedModes.pop(mode, None)
        elif 'category' in img:
            if not img['category'] in locateImages:
                locateImages[img['category']] = {}
            locateImages[img['category']][img['name']] = cv2.imread(imagesDir + filename)
        else:
            locateImages[img['name']] = cv2.imread(imagesDir + filename)

    locateImages['collection'] = {}
    if exists(imagesDir + 'collection_events'):
        for filename in os.listdir(imagesDir + 'collection_events'):
            locateImages['collection'][filename.replace('.png', '')] = cv2.imread(imagesDir + 'collection_events/' + filename)
    
    return {'comparisonImages': comparisonImages, 'locateImages': locateImages, 'supportedModes': supportedModes, 'resolution': resolution}

class State(Enum):
    UNDEFINED = 0
    IDLE = 1
    INGAME = 2
    GOTO_HOME = 3
    GOTO_INGAME = 4
    SELECT_HERO = 5
    FIND_HARDEST_INCREASED_REWARDS_MAP = 6
    MANAGE_OBJECTIVES = 7
    EXIT = 8

class Screen(Enum):
    UNKNOWN = 0
    STARTMENU = 1
    MAP_SELECTION = 3
    DIFFICULTY_SELECTION = 4
    GAMEMODE_SELECTION = 5
    HERO_SELECTION = 6
    INGAME = 7
    INGAME_PAUSED = 8
    VICTORY_SUMMARY = 9
    VICTORY = 10
    DEFEAT = 11
    OVERWRITE_SAVE = 12
    LEVELUP = 13
    APOPALYPSE_HINT = 14
    ROUND_100_INSTA = 15
    COLLECTION_CLAIM_CHEST = 16
    BTD6_UNFOCUSED = 17

class Mode(Enum):
    ERROR = 0
    SINGLE_MAP = 1
    RANDOM_MAP = 2
    CHASE_REWARDS = 3
    DO_ACHIEVEMENTS = 4
    MISSING_MAPS = 5
    XP_FARMING = 6
    MM_FARMING = 7
    MISSING_STATS = 8
    VALIDATE_PLAYTHROUGHS = 9
    VALIDATE_COSTS = 10

def getGamemodePosition(gamemode):
    while not isinstance(imageAreas["click"]["gamemode_positions"][gamemode], list):
        gamemode = imageAreas["click"]["gamemode_positions"][gamemode]
    return imageAreas["click"]["gamemode_positions"][gamemode]

def getNextNonSellAction(steps):
    for step in steps:
        if step['action'] != 'sell' and step['action'] != 'await_round':
            return step
    return {'action': 'nop', 'cost': 0}

def getNextCostingAction(steps):
    for step in steps:
        if step.get('cost', 0) > 0:
            return step
    return {'action': 'nop', 'cost': 0}

def sumAdjacentSells(steps):
    gain = 0
    for step in steps:
        if step['action'] != 'sell':
            return gain
        gain += -step.get('cost', 0)
    return gain

exitAfterGame = False

def setExitAfterGame():
    global exitAfterGame
    activeWindow = ahk.get_active_window()
    if not activeWindow or not isBTD6Window(activeWindow.title):
        return
    customPrint("script will stop after finishing the current game!")
    exitAfterGame = True

def signalHandler(signum, frame):
    customPrint('received SIGINT! exiting!')
    sys.exit(0)


def main():
    signal.signal(signal.SIGINT, signalHandler)

    data = getResolutionDependentData()

    if not data:
        print('unsupported resolution! reference images missing!')
        return

    comparisonImages = data['comparisonImages']
    locateImages = data['locateImages']
    supportedModes = data['supportedModes']
    resolution = data['resolution']

    allAvailablePlaythroughs = getAllAvailablePlaythroughs(['own_playthroughs'], considerUserConfig=True)
    allAvailablePlaythroughsList = allPlaythroughsToList(allAvailablePlaythroughs)

    mode = Mode.ERROR
    logStats = True
    isContinue = False
    repeatObjectives = False
    doAllStepsBeforeStart = False
    listAvailablePlaythroughs = False
    handlePlaythroughValidation = ValidatedPlaythroughs.EXCLUDE_NON_VALIDATED
    usesAllAvailablePlaythroughsList = False

    collectionEvent = None
    valueUnit = ''

    originalObjectives = []
    objectives = []

    categoryRestriction = None
    gamemodeRestriction = None

    argv = np.array(sys.argv)

    parsedArguments = []


    # Additional flags:
    # -ns: disable stats logging
    if len(np.where(argv == '-ns')[0]):
        customPrint('stats logging disabled!')
        parsedArguments.append('-ns')
        logStats = False
    else:
        customPrint('stats logging enabled!')
    # -r: after finishing all objectives the program restarts with the first objective
    if len(np.where(argv == '-r')[0]):
        customPrint('repeating objective indefinitely! cancel with ctrl + c!')
        parsedArguments.append('-r')
        repeatObjectives = True

    # -mk: after finishing all objectives the program restarts with the first objective
    if len(np.where(argv == '-mk')[0]):
        customPrint('including playthroughs with monkey knowledge enabled and adjusting prices according to userconfig.json!')
        parsedArguments.append('-mk')
        setMonkeyKnowledgeStatus(True)
    # -nomk: after finishing all objectives the program restarts with the first objective
    elif len(np.where(argv == '-nomk')[0]):
        customPrint('ignoring playthroughs with monkey knowledge enabled!')
        parsedArguments.append('-nomk')
        setMonkeyKnowledgeStatus(False)
    else:
        customPrint('"-mk" (for monkey knowledge enabled) or "-nomk" (for monkey knowledge disabled) must be specified! exiting!')
        return

    # -l: list all available playthroughs(only works with specific modes)
    if len(np.where(argv == '-l')[0]):
        parsedArguments.append('-l')
        listAvailablePlaythroughs = True

    # -nv: include non validated playthroughs. ignored when mode = validate
    if len(np.where(argv == '-nv')[0]):
        parsedArguments.append('-nv')
        handlePlaythroughValidation = ValidatedPlaythroughs.INCLUDE_ALL


    iArg = 1
    if len(argv) <= iArg:
        customPrint('arguments missing! Usage: py replay.py <mode> <mode arguments...> <flags>')
        return
    # py replay.py file <filename> [continue <(int start)|-> [until (int end)]]
    # replays the specified file
    # if continue is specified it is assumed you are already in game. the script starts with instruction start(0 for first instruction)
    #   if the value for continue equals "-" all instructions are executed before the game is started
    # if until is specified the script only executes instructions until instruction end(start=0, end=1 -> only first instruction is executed)
    # the continue option is mainly for creating/debugging playthroughs
    # -r for indefinite playing only works if continue is not set
    elif argv[iArg] == 'file':
        # run single map, next argument should be the filename
        iAdditionalStart = iArg + 2
        iAdditional = iAdditionalStart
        if len(argv) <= iArg + 1:
            customPrint('requested running a playthrough but no playthrough provided! exiting!')
            return

        parsedArguments.append(argv[iArg + 1])
        instructionOffset = -1
        instructionLast = -1
        gamemode = None

        if len(argv) > iAdditional and argv[iAdditional] in gamemodes:
            gamemode = argv[iAdditional]
            parsedArguments.append(argv[iAdditional])
            iAdditional += 1
        
        if len(argv) > iAdditional + 1 and argv[iAdditional] == 'continue':
            parsedArguments.append(argv[iAdditional])

            isContinue = True

            if str(argv[iAdditional + 1]) == '-':
                instructionOffset = 0
                doAllStepsBeforeStart = True
            elif str(argv[iAdditional + 1]).isdigit():
                instructionOffset = int(argv[iAdditional + 1])
            else:
                customPrint('continue of playthrough requested but no instruction offset provided!')
                return
            customPrint('stats logging disabled!')
            logStats = False
            parsedArguments.append(argv[iAdditional + 1])
            iAdditional += 2

            if len(argv) >= iAdditional + 1 and argv[iAdditional] == 'until':
                if str(argv[iAdditional + 1]).isdigit():
                    instructionLast = int(argv[iAdditional + 1])
                else:
                    customPrint('cutting of instructions for playthrough requested but no index provided!')
                    return
                parsedArguments.append(argv[iAdditional])
                parsedArguments.append(argv[iAdditional + 1])
                iAdditional += 2
        if not parseBTD6InstructionFileName(argv[iArg + 1]):
            customPrint('"' + str(argv[iArg + 1]) + '" can\'t be recognized as a playthrough filename! exiting!')
            return
        elif str(argv[iArg + 1]).count('/') or str(argv[iArg + 1]).count('\\') and exists(argv[iArg + 1]):
            filename = argv[iArg + 1]
        elif exists('own_playthroughs/' + argv[iArg + 1]):
            filename = 'own_playthroughs/' + argv[iArg + 1]
        elif exists('playthroughs/' + argv[iArg + 1]):
            filename = 'playthroughs/' + argv[iArg + 1]
        elif exists('unvalidated_playthroughs/' + argv[iArg + 1]):
            filename = 'unvalidated_playthroughs/' + argv[iArg + 1]
        else:
            customPrint('requested playthrough ' + str(argv[iArg + 1]) + ' not found! exiting!')
            return
        mapConfig = parseBTD6InstructionsFile(filename, gamemode=gamemode)

        mode = Mode.SINGLE_MAP
        if instructionOffset == -1:
            originalObjectives.append({'type': State.GOTO_HOME})
            if 'hero' in mapConfig:
                originalObjectives.append({'type': State.SELECT_HERO, 'mapConfig': mapConfig})
                originalObjectives.append({'type': State.GOTO_HOME})
            originalObjectives.append({'type': State.GOTO_INGAME, 'mapConfig': mapConfig})
        else:
            if instructionOffset >= len(mapConfig['steps']) or (instructionLast != -1 and instructionOffset >= instructionLast):
                customPrint('instruction offset > last instruction (' + (str(instructionLast) if instructionLast != -1 else str(len(mapConfig['steps']))) + ')')
                return

            if instructionLast != -1:
                mapConfig['steps'] = mapConfig['steps'][(instructionOffset + mapConfig['extrainstructions']):instructionLast]
            else:
                mapConfig['steps'] = mapConfig['steps'][(instructionOffset + mapConfig['extrainstructions']):]
            customPrint('continuing playthrough. first instruction:')
            customPrint(mapConfig['steps'][0])
        originalObjectives.append({'type': State.INGAME, 'mapConfig': mapConfig})
        originalObjectives.append({'type': State.MANAGE_OBJECTIVES})
    # py replay.py random [category] [gamemode]
    # plays a random game from all available playthroughs (which fullfill the category and gamemode requirement if specified)
    elif argv[iArg] == 'random':
        iAdditional = iArg + 1
        if len(argv) > iAdditional and argv[iAdditional] in mapsByCategory:
            categoryRestriction = argv[iAdditional]
            parsedArguments.append(argv[iAdditional])
            iAdditional += 1
        
        if len(argv) > iAdditional and argv[iAdditional] in gamemodes:
            gamemodeRestriction = argv[iAdditional]
            parsedArguments.append(argv[iAdditional])
            iAdditional += 1
        
        customPrint('Mode: playing random games' + (f' on {gamemodeRestriction}' if gamemodeRestriction else '') + (f' in {categoryRestriction} category' if categoryRestriction else '') + '!')

        allAvailablePlaythroughs = filterAllAvailablePlaythroughs(allAvailablePlaythroughs, getMonkeyKnowledgeStatus(), handlePlaythroughValidation, categoryRestriction, gamemodeRestriction)
        allAvailablePlaythroughsList = allPlaythroughsToList(allAvailablePlaythroughs)

        originalObjectives.append({'type': State.MANAGE_OBJECTIVES})
        mode = Mode.RANDOM_MAP
        usesAllAvailablePlaythroughsList = True
    # py replay.py chase <event> [category] [gamemode]
    # chases increased rewards for the specified event
    # if category is not provided it finds the map with increased rewards in expert category and plays the most valuable available playthrough and downgrades category if no playthrough is available
    # use -r to farm indefinitely
    elif argv[iArg] == 'chase':
        if len(argv) <= iArg + 1 or not argv[iArg + 1] in locateImages['collection']:
            customPrint('requested chasing event rewards but no event specified or unknown event! exiting!')
            return
        
        collectionEvent = argv[iArg + 1]
        parsedArguments.append(argv[iArg + 1])

        iAdditional = iArg + 2
        if len(argv) > iAdditional and argv[iAdditional] in mapsByCategory:
            categoryRestriction = argv[iAdditional]
            parsedArguments.append(argv[iAdditional])
            iAdditional += 1
        
        if len(argv) > iAdditional and argv[iAdditional] in gamemodes:
            gamemodeRestriction = argv[iAdditional]
            parsedArguments.append(argv[iAdditional])
            iAdditional += 1


        if collectionEvent == 'golden_bloon':
            allAvailablePlaythroughs = filterAllAvailablePlaythroughs(allAvailablePlaythroughs, getMonkeyKnowledgeStatus(), handlePlaythroughValidation, categoryRestriction, gamemodeRestriction, requiredFlags=['gB'])
            customPrint(f'Mode: playing games with golden bloons using special playthroughs' + (f' on {gamemodeRestriction}' if gamemodeRestriction else '') + (f' in {categoryRestriction} category' if categoryRestriction else '') + '!')
        else:
            allAvailablePlaythroughs = filterAllAvailablePlaythroughs(allAvailablePlaythroughs, getMonkeyKnowledgeStatus(), handlePlaythroughValidation, categoryRestriction, gamemodeRestriction)
            customPrint(f'Mode: playing games with increased {collectionEvent} collection event rewards' + (f' on {gamemodeRestriction}' if gamemodeRestriction else '') + (f' in {categoryRestriction} category' if categoryRestriction else '') + '!')
        allAvailablePlaythroughsList = allPlaythroughsToList(allAvailablePlaythroughs)

        originalObjectives.append({'type': State.MANAGE_OBJECTIVES})
        mode = Mode.CHASE_REWARDS
        usesAllAvailablePlaythroughsList = True
    # py replay.py achievements [achievement]
    # plays all achievement related playthroughs
    # if achievement is provided it just plays plays until said achievement is unlocked
    # userconfig.json can be used to specify which achievements have already been unlocked or to document progress(e. g. games won only using primary monkeys)
    # refer to userconfig.example.json for an example
    elif argv[iArg] == 'achievements':
        pass
    # py replay.py missing [category]
    # plays all playthroughs with missing medals
    # if category is not provided from easiest category to hardest
    # if category is provided in said category
    # requires userconfig.json to specify which medals have already been earned
    # unlocking of maps has do be done manually
    elif argv[iArg] == 'missing':
        pass
    # py replay.py xp [int n=1]
    # plays one of the n most efficient(in terms of xp/hour) playthroughs
    # with -r: plays indefinitely
    elif argv[iArg] == 'xp':
        allAvailablePlaythroughsList = sortPlaythroughsByXPGain(allAvailablePlaythroughsList)

        if len(argv) > iArg + 1 and argv[iArg + 1].isdigit():
            allAvailablePlaythroughsList = allAvailablePlaythroughsList[:int(argv[iArg + 1])]
            parsedArguments.append(argv[iArg + 1])
        else:
            allAvailablePlaythroughsList = allAvailablePlaythroughsList[:1]
        
        originalObjectives.append({'type': State.MANAGE_OBJECTIVES})
        mode = Mode.XP_FARMING
        valueUnit = 'XP/h'
        usesAllAvailablePlaythroughsList = True
    # py replay.py mm [int n=1]
    # plays one of the n most efficient(in terms of mm/hour) playthroughs
    # with -r: plays indefinitely
    elif argv[iArg] == 'mm' or argv[iArg] == 'monkey_money':
        allAvailablePlaythroughsList = sortPlaythroughsByMonkeyMoneyGain(allAvailablePlaythroughsList)

        if len(argv) > iArg + 1 and argv[iArg + 1].isdigit():
            allAvailablePlaythroughsList = allAvailablePlaythroughsList[:int(argv[iArg + 1])]
            parsedArguments.append(argv[iArg + 1])
        else:
            allAvailablePlaythroughsList = allAvailablePlaythroughsList[:1]
        
        originalObjectives.append({'type': State.MANAGE_OBJECTIVES})
        mode = Mode.MM_FARMING
        valueUnit = 'MM/h'
        usesAllAvailablePlaythroughsList = True
    # py replay.py validate file <filename>
    # or
    # py replay.py validate all [category]
    elif argv[iArg] == 'validate':
        
        if len(argv) <= iArg + 1:
            customPrint('requested validation but arguments missing!')
            return

        parsedArguments.append(argv[iArg + 1])

        if getMonkeyKnowledgeStatus():
            customPrint('Mode validate only works with monkey knowledge disabled!')
            return

        if argv[iArg + 1] == 'file':
            if len(argv) <= iArg + 2:
                customPrint('no filename provided!')
                return

            if not parseBTD6InstructionFileName(argv[iArg + 2]):
                customPrint('"' + str(argv[iArg + 2]) + '" can\'t be recognized as a playthrough filename! exiting!')
                return
            elif str(argv[iArg + 1]).count('/') or str(argv[iArg + 2]).count('\\') and exists(argv[iArg + 2]):
                filename = argv[iArg + 1]
            elif exists('own_playthroughs/' + argv[iArg + 2]):
                filename = 'own_playthroughs/' + argv[iArg + 2]
            elif exists('playthroughs/' + argv[iArg + 2]):
                filename = 'playthroughs/' + argv[iArg + 2]
            elif exists('unvalidated_playthroughs/' + argv[iArg + 2]):
                filename = 'unvalidated_playthroughs/' + argv[iArg + 2]
            else:
                customPrint('requested playthrough ' + str(argv[iArg + 2]) + ' not found! exiting!')
                return

            parsedArguments.append(argv[iArg + 2])
            
            fileConfig = parseBTD6InstructionFileName(filename)
            allAvailablePlaythroughsList = [{'filename': filename, 'fileConfig': fileConfig, 'gamemode': fileConfig['gamemode'], 'isOriginalGamemode': True}]
        elif argv[iArg + 1] == 'all':
            iAdditional = iArg + 2

            if len(argv) > iAdditional and argv[iAdditional] in mapsByCategory:
                categoryRestriction = argv[iAdditional]
                parsedArguments.append(argv[iAdditional])
                iAdditional += 1

            customPrint('Mode: validating all playthroughs' + (' in ' + categoryRestriction + ' category' if categoryRestriction else '') + '!')

            allAvailablePlaythroughs = filterAllAvailablePlaythroughs(allAvailablePlaythroughs, True, ValidatedPlaythroughs.EXCLUDE_VALIDATED if handlePlaythroughValidation == ValidatedPlaythroughs.INCLUDE_ALL else ValidatedPlaythroughs.INCLUDE_ALL, categoryRestriction, gamemodeRestriction, onlyOriginalGamemodes=True)
            allAvailablePlaythroughsList = allPlaythroughsToList(allAvailablePlaythroughs)

        originalObjectives.append({'type': State.MANAGE_OBJECTIVES})
        usesAllAvailablePlaythroughsList = True
        mode = Mode.VALIDATE_PLAYTHROUGHS
    # py replay.py costs [+heros]
    # determines the base cost and cost of each upgrade for each monkey as well as the base cost for each hero if '+heros' is specified
    elif argv[iArg] == 'costs':
        if getMonkeyKnowledgeStatus():
            customPrint('Mode validate costs only works with monkey knowledge disabled!')
            return

        includeHeros = False

        if len(argv) >= iArg + 2 and argv[iArg + 1] == '+heros':
            includeHeros = True
            parsedArguments.append(argv[iArg + 1])

        customPrint('Mode: validating monkey costs' + (' including heros' if includeHeros else '') + '!')

        allTestPositions = json.load(open('test_positions.json'))
        if getResolutionString() in allTestPositions:
            testPositions = allTestPositions[getResolutionString()]
        else:
            testPositions = json.loads(convertPositionsInString(json.dumps(allTestPositions['2560x1440']), (2560, 1440), pyautogui.size()))

        selectedMap = None
        for mapname in testPositions:
            if getAvailableSandbox(mapname, ['medium_sandbox']):
                selectedMap = mapname
                break
        
        if selectedMap is None:
            customPrint('This mode requires access to medium sandbox for one of the maps in "test_positions.json"!')
            return

        costs = {'monkeys': {}}

        baseMapConfig = {'category': maps[selectedMap]['category'], 'map': selectedMap, 'page': maps[selectedMap]['page'], 'pos': maps[selectedMap]['pos'], 'difficulty': 'medium', 'gamemode': 'medium_sandbox', 'steps': [], 'extrainstructions': 1, 'filename': None}

        monkeySteps = []
        monkeySteps.append({'action': 'click', 'pos': imageAreas['click']['gamemode_deflation_message_confirmation'], 'cost': 0})
        pos = testPositions[selectedMap]
        pos['any'] = pos['land']
        for monkeyType in towers['monkeys']:
            costs['monkeys'][monkeyType] = {'base': 0, 'upgrades': np.zeros((3, 5))}
            for iPath in range(0, 3):
                monkeySteps.append({'action': 'place', 'type': monkeyType, 'name': f"{monkeyType}{iPath}", 'key': keybinds['monkeys'][monkeyType], 'pos': pos[towers['monkeys'][monkeyType]['class']], 'cost': 1, 'extra': {'group': 'monkeys', 'type': monkeyType}})
                for iUpgrade in range(1, 6):
                    monkeySteps.append({'action': 'upgrade', 'name': f"{monkeyType}{iPath}", 'key': keybinds['path'][str(iPath)], 'pos': pos[towers['monkeys'][monkeyType]['class']], 'path': iPath, 'cost': 1, 'extra': {'group': 'monkeys', 'type': monkeyType, 'upgrade': (iPath, iUpgrade)}})
                    if upgradeRequiresConfirmation({'type': monkeyType, 'upgrades': [(iUpgrade if iTmp == iPath else 0) for iTmp in range(0, 3)]}, iPath):
                        monkeySteps.append({'action': 'click', 'name': f"{monkeyType}{iPath}", 'pos': imageAreas['click']['paragon_message_confirmation'], 'cost': 0})
                monkeySteps.append({'action': 'sell', 'name': f"{monkeyType}{iPath}", 'key': keybinds['others']['sell'], 'pos': pos[towers['monkeys'][monkeyType]['class']], 'cost': -1})
        
        monkeyMapConfig = copy.deepcopy(baseMapConfig)
        monkeyMapConfig['steps'] = monkeySteps
        
        originalObjectives.append({'type': State.GOTO_HOME})
        originalObjectives.append({'type': State.GOTO_INGAME, 'mapConfig': monkeyMapConfig})
        originalObjectives.append({'type': State.INGAME, 'mapConfig': monkeyMapConfig})

        if includeHeros:
            costs['heros'] = {}
            
            for hero in towers['heros']:
                costs['heros'][hero] = {'base' : 0}
                heroMapConfig = copy.deepcopy(baseMapConfig)
                heroMapConfig['hero'] = hero
                heroMapConfig['steps'] = [{'action': 'click', 'pos': imageAreas['click']['gamemode_deflation_message_confirmation'], 'cost': 0}, {'action': 'place', 'type': 'hero', 'name': 'hero0', 'key': keybinds['monkeys']['hero'], 'pos': pos[towers['heros'][hero]['class']], 'cost': 1, 'extra': {'group': 'heros', 'type': hero}}]
                originalObjectives.append({'type': State.GOTO_HOME})
                originalObjectives.append({'type': State.SELECT_HERO, 'mapConfig': heroMapConfig})
                originalObjectives.append({'type': State.GOTO_HOME})
                originalObjectives.append({'type': State.GOTO_INGAME, 'mapConfig': heroMapConfig})
                originalObjectives.append({'type': State.INGAME, 'mapConfig': heroMapConfig})

        originalObjectives.append({'type': State.MANAGE_OBJECTIVES})
        usesAllAvailablePlaythroughsList = False
        mode = Mode.VALIDATE_COSTS

    if mode == Mode.ERROR:
        customPrint('invalid arguments! exiting!')
        return

    if not mode.name in supportedModes:
        customPrint('mode not supported due to missing images!')
        return

    parsedArguments.append(argv[0])
    parsedArguments.append(argv[1])

    unparsedArguments = []
    parsedArgumentsTmp = np.array(parsedArguments)
    for arg in sys.argv:
        if len(np.where(parsedArgumentsTmp == arg)[0]):
            parsedArgumentsTmp = np.delete(parsedArgumentsTmp, np.where(parsedArgumentsTmp == arg)[0])
        else:
            unparsedArguments.append(arg)

    if len(unparsedArguments):
        customPrint('unrecognized arguments:')
        customPrint(unparsedArguments)
        customPrint('exiting!')
        return

    if listAvailablePlaythroughs:
        if usesAllAvailablePlaythroughsList:
            customPrint(str(len(allAvailablePlaythroughsList)) + ' playthroughs found:')
            for playthrough in allAvailablePlaythroughsList:
                customPrint(playthrough['filename'] + ': ' + playthrough['fileConfig']['map'] + ' - ' + playthrough['gamemode'] + (' with ' + str(playthrough['value']) + (' ' + valueUnit if len(valueUnit) else '') if 'value' in playthrough else ''))
        else:
            customPrint('Mode doesn\'t qualify for listing all available playthroughs')
        return

    if usesAllAvailablePlaythroughsList and len(allAvailablePlaythroughsList) == 0:
        customPrint('no playthroughs matching requirements found!')

    keyboard.add_hotkey('ctrl+space', setExitAfterGame)

    objectives = copy.deepcopy(originalObjectives)
        
    state = objectives[0]['type']
    lastStateTransitionSuccessful = True
    objectiveFailed = False
    mapConfig = objectives[0]['mapConfig'] if 'mapConfig' in objectives[0] else None

    gamesPlayed = 0

    lastIterationBalance = -1
    lastIterationRound = -1
    lastIterationScreenshotAreas = []
    lastIterationCost = 0
    iterationBalances = []
    thisIterationAction = None
    lastIterationAction = None

    fast = True

    validationResult = None

    playthroughLog = {}

    lastHeroSelected = None

    increasedRewardsPlaythrough = None

    lastPlaythrough = None
    lastPlaythroughStats = {}

    lastScreen = Screen.UNKNOWN
    lastState = State.UNDEFINED

    unknownScreenHasWaited = False

    segmentCoordinates = None
    
    while True:
        screenshot = np.array(pyautogui.screenshot())[:, :, ::-1].copy()

        screen = Screen.UNKNOWN
        activeWindow = ahk.get_active_window()
        if not activeWindow or not isBTD6Window(activeWindow.title):
            screen = Screen.BTD6_UNFOCUSED
        else:
            bestMatchDiff = None
            for screenCfg in [
                (Screen.STARTMENU, comparisonImages["screens"]["startmenu"], imageAreas["compare"]["screens"]["startmenu"]),
                (Screen.MAP_SELECTION, comparisonImages["screens"]["map_selection"], imageAreas["compare"]["screens"]["map_selection"]),
                (Screen.DIFFICULTY_SELECTION, comparisonImages["screens"]["difficulty_selection"], imageAreas["compare"]["screens"]["difficulty_selection"]),
                (Screen.GAMEMODE_SELECTION, comparisonImages["screens"]["gamemode_selection"], imageAreas["compare"]["screens"]["gamemode_selection"]),
                (Screen.HERO_SELECTION, comparisonImages["screens"]["hero_selection"], imageAreas["compare"]["screens"]["hero_selection"]),
                (Screen.INGAME, comparisonImages["screens"]["ingame"], imageAreas["compare"]["screens"]["ingame"]),
                (Screen.INGAME_PAUSED, comparisonImages["screens"]["ingame_paused"], imageAreas["compare"]["screens"]["ingame_paused"]),
                (Screen.VICTORY_SUMMARY, comparisonImages["screens"]["victory_summary"], imageAreas["compare"]["screens"]["victory_summary"]),
                (Screen.VICTORY, comparisonImages["screens"]["victory"], imageAreas["compare"]["screens"]["victory"]),
                (Screen.DEFEAT, comparisonImages["screens"]["defeat"], imageAreas["compare"]["screens"]["defeat"]),
                (Screen.OVERWRITE_SAVE, comparisonImages["screens"]["overwrite_save"], imageAreas["compare"]["screens"]["overwrite_save"]),
                (Screen.LEVELUP, comparisonImages["screens"]["levelup"], imageAreas["compare"]["screens"]["levelup"]),
                (Screen.APOPALYPSE_HINT, comparisonImages["screens"]["apopalypse_hint"], imageAreas["compare"]["screens"]["apopalypse_hint"]),
                (Screen.ROUND_100_INSTA, comparisonImages["screens"]["round_100_insta"], imageAreas["compare"]["screens"]["round_100_insta"]),
                (Screen.COLLECTION_CLAIM_CHEST, comparisonImages["screens"]["collection_claim_chest"], imageAreas["compare"]["screens"]["collection_claim_chest"]),
            ]:
                diff = cv2.matchTemplate(cutImage(screenshot, screenCfg[2]), cutImage(screenCfg[1], screenCfg[2]), cv2.TM_SQDIFF_NORMED)[0][0]
                if diff < 0.05 and (bestMatchDiff is None or diff < bestMatchDiff):
                    bestMatchDiff = diff
                    screen = screenCfg[0]

        if screen != lastScreen:
            customPrint("screen " + screen.name + "!")

        if screen == Screen.BTD6_UNFOCUSED:
            pass
        # don't do anything when ctrl is pressed: useful for alt + tab / sending SIGINT(ctrl + c) to the script
        elif keyboard.is_pressed('ctrl'):
            pass
        elif state == State.MANAGE_OBJECTIVES:
            customPrint("entered objective management!")
            
            if exitAfterGame:
                state = State.EXIT
                continue
            
            if mode == Mode.VALIDATE_PLAYTHROUGHS:
                if validationResult != None:
                    customPrint('validation result: playthrough ' + lastPlaythrough['filename'] + ' is ' + ('valid' if validationResult else 'invalid') + '!')
                    updatePlaythroughValidationStatus(lastPlaythrough['filename'], validationResult)
                if len(allAvailablePlaythroughsList):
                    playthrough = allAvailablePlaythroughsList.pop(0)
                    customPrint('validation playthrough chosen: ' + playthrough['fileConfig']['map'] + ' on ' + playthrough['gamemode'] + ' (' + playthrough['filename'] + ')')
                    
                    gamemode = getAvailableSandbox(playthrough['fileConfig']['map'])
                    if gamemode:
                        mapConfig = parseBTD6InstructionsFile(playthrough['filename'], gamemode=gamemode)
                        objectives = []
                        objectives.append({'type': State.GOTO_HOME})
                        if 'hero' in mapConfig and lastHeroSelected != mapConfig['hero']:
                            objectives.append({'type': State.SELECT_HERO, 'mapConfig': mapConfig})
                            objectives.append({'type': State.GOTO_HOME})
                        objectives.append({'type': State.GOTO_INGAME, 'mapConfig': mapConfig})
                        objectives.append({'type': State.INGAME, 'mapConfig': mapConfig})
                        objectives.append({'type': State.MANAGE_OBJECTIVES})

                        validationResult = True
                        lastPlaythrough = playthrough
                    else:
                        customPrint('missing sandbox access for ' + playthrough['fileConfig']['map'])
                        objectives = []
                        objectives.append({'type': State.MANAGE_OBJECTIVES})
                else:
                    objectives = []
                    objectives.append({'type': State.EXIT})
            elif mode == Mode.VALIDATE_COSTS:
                oldTowers = copy.deepcopy(towers)
                changes = 0
                for monkeyType in costs['monkeys']:
                    if costs['monkeys'][monkeyType]['base'] and costs['monkeys'][monkeyType]['base'] != oldTowers['monkeys'][monkeyType]['base']:
                        print(f"{monkeyType} base cost: {oldTowers['monkeys'][monkeyType]['base']} -> {int(costs['monkeys'][monkeyType]['base'])}")
                        towers['monkeys'][monkeyType]['base'] = int(costs['monkeys'][monkeyType]['base'])
                        changes += 1
                    for iPath in range(0, 3):
                        for iUpgrade in range(0, 5):
                            if costs['monkeys'][monkeyType]['upgrades'][iPath][iUpgrade] and costs['monkeys'][monkeyType]['upgrades'][iPath][iUpgrade] != oldTowers['monkeys'][monkeyType]['upgrades'][iPath][iUpgrade]:
                                print(f"{monkeyType} path {iPath + 1} upgrade {iUpgrade + 1} cost: {oldTowers['monkeys'][monkeyType]['upgrades'][iPath][iUpgrade]} -> {int(costs['monkeys'][monkeyType]['upgrades'][iPath][iUpgrade])}")
                                towers['monkeys'][monkeyType]['upgrades'][iPath][iUpgrade] = int(costs['monkeys'][monkeyType]['upgrades'][iPath][iUpgrade])
                                changes += 1
                if 'heros' in costs:
                    for hero in costs['heros']:
                        if costs['heros'][hero]['base'] and costs['heros'][hero]['base'] != oldTowers['heros'][hero]['base']:
                            print(f"hero {hero} base cost: {oldTowers['heros'][hero]['base']} -> {int(costs['heros'][hero]['base'])}")
                            towers['heros'][hero]['base'] = int(costs['heros'][hero]['base'])
                            changes += 1

                if changes:
                    print(f"updating \"towers.json\" with {changes} changes!")
                    fp = open('towers_backup.json', "w")
                    fp.write(json.dumps(oldTowers, indent=4))
                    fp.close()
                    fp = open('towers.json', "w")
                    fp.write(json.dumps(towers, indent=4))
                    fp.close()
                else:
                    print(f"no price changes in comparison to \"towers.json\" detected!")
                
                return
            elif repeatObjectives or gamesPlayed == 0:
                if mode == Mode.SINGLE_MAP:
                    objectives = copy.deepcopy(originalObjectives)
                elif mode == Mode.RANDOM_MAP or mode == Mode.XP_FARMING or mode == Mode.MM_FARMING:
                    objectives = []
                    playthrough = random.choice(allAvailablePlaythroughsList)
                    customPrint('random playthrough chosen: ' + playthrough['fileConfig']['map'] + ' on ' + playthrough['gamemode'] + ' (' + playthrough['filename'] + ')')
                    mapConfig = parseBTD6InstructionsFile(playthrough['filename'], gamemode=playthrough['gamemode'])
                    
                    objectives.append({'type': State.GOTO_HOME})
                    if 'hero' in mapConfig and lastHeroSelected != mapConfig['hero']:
                        objectives.append({'type': State.SELECT_HERO, 'mapConfig': mapConfig})
                        objectives.append({'type': State.GOTO_HOME})
                    objectives.append({'type': State.GOTO_INGAME, 'mapConfig': mapConfig})
                    objectives.append({'type': State.INGAME, 'mapConfig': mapConfig})
                    objectives.append({'type': State.MANAGE_OBJECTIVES})
                    lastPlaythrough = playthrough
                elif mode == Mode.CHASE_REWARDS:
                    objectives = []
                    if increasedRewardsPlaythrough:
                        playthrough = increasedRewardsPlaythrough
                        customPrint('highest reward playthrough chosen: ' + playthrough['fileConfig']['map'] + ' on ' + playthrough['gamemode'] + ' (' + playthrough['filename'] + ')')
                        mapConfig = parseBTD6InstructionsFile(playthrough['filename'], gamemode=playthrough['gamemode'])

                        objectives.append({'type': State.GOTO_HOME})
                        if 'hero' in mapConfig and lastHeroSelected != mapConfig['hero']:
                            objectives.append({'type': State.SELECT_HERO, 'mapConfig': mapConfig})
                            objectives.append({'type': State.GOTO_HOME})
                        objectives.append({'type': State.GOTO_INGAME, 'mapConfig': mapConfig})
                        objectives.append({'type': State.INGAME, 'mapConfig': mapConfig})
                        objectives.append({'type': State.MANAGE_OBJECTIVES})
                        increasedRewardsPlaythrough = None
                        lastPlaythrough = playthrough
                    else:
                        objectives.append({'type': State.GOTO_HOME})
                        objectives.append({'type': State.FIND_HARDEST_INCREASED_REWARDS_MAP})
                        objectives.append({'type': State.MANAGE_OBJECTIVES})
                else:
                    objectives = copy.deepcopy(originalObjectives)
            else:
                objectives = []
                objectives.append({'type': State.EXIT})

            state = objectives[0]['type']
            lastStateTransitionSuccessful = True
            objectiveFailed = False
        elif state == State.UNDEFINED:
            customPrint("entered state management!")
            if exitAfterGame:
                state = State.EXIT
            if objectiveFailed:
                customPrint("objective failed on step " + objectives[0]['type'].name + "(screen " + lastScreen.name + ")!")
                if repeatObjectives:
                    state = State.MANAGE_OBJECTIVES
                else:
                    state = State.EXIT
            elif not lastStateTransitionSuccessful:
                state = objectives[0]['type']
                if 'mapConfig' in objectives[0]:
                    mapConfig = objectives[0]['mapConfig']
                lastStateTransitionSuccessful = True
            elif lastStateTransitionSuccessful and len(objectives):
                objectives.pop(0)
                state = objectives[0]['type']
                if 'mapConfig' in objectives[0]:
                    mapConfig = objectives[0]['mapConfig']
            else:
                state = State.EXIT
        elif state == State.IDLE:
            pass
        elif state == State.EXIT:
            customPrint("goal EXIT! exiting!")
            return
        elif state == State.GOTO_HOME:
            if screen == Screen.STARTMENU:
                customPrint("goal GOTO_HOME fullfilled!")
                state = State.UNDEFINED
            elif screen == Screen.UNKNOWN:
                if lastScreen == Screen.UNKNOWN and unknownScreenHasWaited:
                    unknownScreenHasWaited = False
                    sendKey('{Esc}')
                else:
                    unknownScreenHasWaited = True
                    time.sleep(2)
            elif screen == Screen.INGAME:
                sendKey('{Esc}')
            elif screen == Screen.INGAME_PAUSED:
                pyautogui.click(imageAreas["click"]["screen_ingame_paused_button_home"])
            elif screen == Screen.HERO_SELECTION:
                sendKey('{Esc}')
            elif screen == Screen.GAMEMODE_SELECTION:
                sendKey('{Esc}')
            elif screen == Screen.DIFFICULTY_SELECTION:
                sendKey('{Esc}')
            elif screen == Screen.MAP_SELECTION:
                sendKey('{Esc}')
            elif screen == Screen.DEFEAT:
                result = cv2.matchTemplate(screenshot, locateImages['button_home'], cv2.TM_SQDIFF_NORMED)
                pyautogui.click(cv2.minMaxLoc(result)[2])
            elif screen == Screen.VICTORY_SUMMARY:
                pyautogui.click(imageAreas["click"]["screen_victory_summary_button_next"])
            elif screen == Screen.VICTORY:
                pyautogui.click(imageAreas["click"]["screen_victory_button_home"])
            elif screen == Screen.OVERWRITE_SAVE:
                sendKey('{Esc}')
            elif screen == Screen.LEVELUP:
                pyautogui.click(100, 100)
                time.sleep(menuChangeDelay)
                pyautogui.click(100, 100)
            elif screen == Screen.ROUND_100_INSTA:
                pyautogui.click(100, 100)
                time.sleep(menuChangeDelay)
            elif screen == Screen.COLLECTION_CLAIM_CHEST:
                pyautogui.click(imageAreas["click"]["collection_claim_chest"])
                time.sleep(menuChangeDelay * 2)
                while True:
                    newScreenshot = np.array(pyautogui.screenshot())[:, :, ::-1].copy()
                    result = [cv2.minMaxLoc(cv2.matchTemplate(newScreenshot, locateImages['unknown_insta'], cv2.TM_SQDIFF_NORMED, mask=locateImages['unknown_insta_mask']))[i] for i in [0,2]]
                    if result[0] < 0.01:
                        pyautogui.click(result[1])
                        time.sleep(menuChangeDelay)
                        pyautogui.click(result[1])
                        time.sleep(menuChangeDelay)
                    else:
                        break
                pyautogui.click(round(resolution[0] / 2), round(resolution[1] / 2))
                time.sleep(menuChangeDelay)
                sendKey('{Esc}')
            elif screen == Screen.APOPALYPSE_HINT:
                pyautogui.click(imageAreas["click"]["gamemode_apopalypse_message_confirmation"])
        elif state == State.GOTO_INGAME:
            if screen == Screen.STARTMENU:
                pyautogui.click(imageAreas["click"]["screen_startmenu_button_play"])
                time.sleep(menuChangeDelay)
                if mapConfig['category'] == 'beginner':
                    pyautogui.click(imageAreas["click"]["map_categories"]['advanced'])
                    time.sleep(menuChangeDelay)
                    pyautogui.click(imageAreas["click"]["map_categories"][mapConfig['category']])
                    time.sleep(menuChangeDelay)
                else:
                    pyautogui.click(imageAreas["click"]["map_categories"]['beginner'])
                    time.sleep(menuChangeDelay)
                    pyautogui.click(imageAreas["click"]["map_categories"][mapConfig['category']])
                    time.sleep(menuChangeDelay)
                tmpClicks = mapConfig['page']
                while tmpClicks > 0:
                    pyautogui.click(imageAreas["click"]["map_categories"][mapConfig['category']])
                    tmpClicks -= 1
                    time.sleep(menuChangeDelay)
                pyautogui.click(imageAreas["click"]["map_positions"][mapConfig['pos']])
                time.sleep(menuChangeDelay)
                pyautogui.click(imageAreas["click"]["gamedifficulty_positions"][mapConfig['difficulty']])
                time.sleep(menuChangeDelay)
                pyautogui.click(getGamemodePosition(mapConfig['gamemode']))
            elif screen == Screen.OVERWRITE_SAVE:
                pyautogui.click(imageAreas["click"]["screen_overwrite_save_button_ok"])
            elif screen == Screen.APOPALYPSE_HINT:
                pyautogui.click(imageAreas["click"]["gamemode_apopalypse_message_confirmation"])
            elif screen == Screen.INGAME:
                customPrint("goal GOTO_INGAME fullfilled!")
                customPrint("game: " + mapConfig['map'] + ' - ' + mapConfig['difficulty'])
                segmentCoordinates = getIngameOcrSegments(mapConfig)
                iterationBalances = []
                if logStats:
                    lastPlaythroughStats = {'gamemode': mapConfig['gamemode'], 'time': [], 'result': PlaythroughResult.UNDEFINED}
                    lastPlaythroughStats['time'].append(('start', time.time()))
                lastIterationBalance = -1
                lastIterationCost = 0
                state = State.UNDEFINED
            elif screen == Screen.UNKNOWN:
                pass
            else:
                customPrint("task GOTO_INGAME, but not in startmenu!")
                state = State.GOTO_HOME
                lastStateTransitionSuccessful = False
        elif state == State.SELECT_HERO:
            if screen == Screen.STARTMENU:
                pyautogui.click(imageAreas["click"]["screen_startmenu_button_hero_selection"])
                time.sleep(menuChangeDelay)
                pyautogui.click(imageAreas["click"]["hero_positions"][mapConfig['hero']])
                time.sleep(menuChangeDelay)
                pyautogui.click(imageAreas["click"]["screen_hero_selection_select_hero"])
                customPrint("goal SELECT_HERO " + mapConfig['hero'] + " fullfilled!")
                lastHeroSelected = mapConfig['hero']
                state = State.UNDEFINED
            elif screen == Screen.UNKNOWN:
                pass
            else:
                customPrint("task SELECT_HERO, but not in startmenu!")
                state = State.GOTO_HOME
                lastStateTransitionSuccessful = False
        elif state == State.FIND_HARDEST_INCREASED_REWARDS_MAP:
            if screen == Screen.STARTMENU:
                pyautogui.click(imageAreas["click"]["screen_startmenu_button_play"])
                time.sleep(menuChangeDelay)

                if categoryRestriction:
                    pyautogui.click(imageAreas["click"]["map_categories"][('advanced' if categoryRestriction == 'beginner' else 'beginner')])
                    time.sleep(menuChangeDelay)

                    mapname = None
                    for page in range(0, categoryPages[categoryRestriction]):
                        pyautogui.click(imageAreas["click"]["map_categories"][categoryRestriction])
                        if collectionEvent == 'golden_bloon':
                            time.sleep(4)
                        else:
                            time.sleep(menuChangeDelay)
                        newScreenshot = np.array(pyautogui.screenshot())[:, :, ::-1].copy()
                        result = findImageInImage(newScreenshot, locateImages['collection'][collectionEvent])
                        if result[0] < 0.05:
                            mapname = findMapForPxPos(categoryRestriction, page, result[1])
                            break
                    if not mapname:
                        customPrint('no maps with increased rewards found! exiting!')
                        return
                    customPrint('best map: ' + mapname)
                    increasedRewardsPlaythrough = getHighestValuePlaythrough(allAvailablePlaythroughs, mapname, playthroughLog)
                    if not increasedRewardsPlaythrough:
                        customPrint('no playthroughs for map found! exiting!')
                        return
                else:
                    iTmp = 0
                    for category in reversed(list(mapsByCategory.keys())):
                        if iTmp == 0:
                            pyautogui.click(imageAreas["click"]["map_categories"][('advanced' if category == 'beginner' else 'beginner')])
                            time.sleep(menuChangeDelay)

                        mapname = None
                        for page in range(0, categoryPages[category]):
                            pyautogui.click(imageAreas["click"]["map_categories"][category])
                            if collectionEvent == 'golden_bloon':
                                time.sleep(4)
                            else:
                                time.sleep(menuChangeDelay)
                            newScreenshot = np.array(pyautogui.screenshot())[:, :, ::-1].copy()
                            result = findImageInImage(newScreenshot, locateImages['collection'][collectionEvent])
                            if result[0] < 0.05:
                                mapname = findMapForPxPos(category, page, result[1])
                                break
                        if not mapname:
                            customPrint('no maps with increased rewards found! exiting!')
                            return
                        customPrint('best map in ' + category + ': ' + mapname)
                        increasedRewardsPlaythrough = getHighestValuePlaythrough(allAvailablePlaythroughs, mapname, playthroughLog)
                        if increasedRewardsPlaythrough:
                            break
                        else:
                            customPrint('no playthroughs for map found! searching lower map tiers!')
                        iTmp += 1
                    
                    if not increasedRewardsPlaythrough:
                        customPrint('no available playthrough found! exiting!')
                        return
                state = State.UNDEFINED
            elif screen == Screen.UNKNOWN:
                pass
            else:
                customPrint("task FIND_HARDEST_INCREASED_REWARDS_MAP, but not in startmenu!")
                state = State.GOTO_HOME
                lastStateTransitionSuccessful = False
        elif state == State.INGAME:
            if screen == Screen.INGAME_PAUSED:
                if lastScreen != screen and logStats:
                    lastPlaythroughStats['time'].append(('stop', time.time()))
                time.sleep(2)
                if isBTD6Window(ahk.get_active_window().title):
                    sendKey('{Esc}')
            elif screen == Screen.UNKNOWN:
                if lastScreen == Screen.UNKNOWN and unknownScreenHasWaited:
                    unknownScreenHasWaited = False
                    sendKey('{Esc}')
                else:
                    unknownScreenHasWaited = True
                    time.sleep(2)
            elif screen == Screen.LEVELUP:
                pyautogui.click(100, 100)
                time.sleep(menuChangeDelay)
                pyautogui.click(100, 100)
            elif screen == Screen.ROUND_100_INSTA:
                pyautogui.click(100, 100)
                time.sleep(menuChangeDelay)
            elif screen == Screen.VICTORY_SUMMARY:
                if logStats:
                    lastPlaythroughStats['time'].append(('stop', time.time()))
                    lastPlaythroughStats['result'] = PlaythroughResult.WIN
                    updateStatsFile(mapConfig['filename'], lastPlaythroughStats)
                gamesPlayed += 1
                if not mapConfig['filename'] in playthroughLog:
                    playthroughLog[mapConfig['filename']] = {}
                if not mapConfig['gamemode'] in playthroughLog[mapConfig['filename']]:
                    playthroughLog[mapConfig['filename']][mapConfig['gamemode']] = {'attempts': 0, 'wins': 0, 'defeats': 0}
                playthroughLog[mapConfig['filename']][mapConfig['gamemode']]['attempts'] += 1
                playthroughLog[mapConfig['filename']][mapConfig['gamemode']]['wins'] += 1

                if not isContinue:
                    updateMedalStatus(mapConfig['map'], mapConfig['gamemode'])
                
                state = State.UNDEFINED
            elif screen == Screen.DEFEAT:
                if logStats:
                    lastPlaythroughStats['time'].append(('stop', time.time()))
                    lastPlaythroughStats['result'] = PlaythroughResult.DEFEAT
                    updateStatsFile(mapConfig['filename'], lastPlaythroughStats)
                objectiveFailed = True
                gamesPlayed += 1
                if not mapConfig['filename'] in playthroughLog:
                    playthroughLog[mapConfig['filename']] = {}
                if not mapConfig['gamemode'] in playthroughLog[mapConfig['filename']]:
                    playthroughLog[mapConfig['filename']][mapConfig['gamemode']] = {'attempts': 0, 'wins': 0, 'defeats': 0}
                playthroughLog[mapConfig['filename']][mapConfig['gamemode']]['attempts'] += 1
                playthroughLog[mapConfig['filename']][mapConfig['gamemode']]['defeats'] += 1
                
                state = State.UNDEFINED
            elif screen == Screen.INGAME:
                if lastScreen != screen and logStats:
                    lastPlaythroughStats['time'].append(('start', time.time()))

                images = [
                    screenshot[segmentCoordinates[segment][1]:segmentCoordinates[segment][3], segmentCoordinates[segment][0]:segmentCoordinates[segment][2]] for segment in segmentCoordinates
                ]

                currentValues = {}
                thisIterationCost = 0
                thisIterationAction = None
                skippingIteration = False

                try:
                    currentValues['money'] = int(custom_ocr(images[2]))
                    currentValues['round'] = int(custom_ocr(images[3]).split('/')[0])
                except ValueError:
                    currentValues['money'] = -1
                    currentValues['round'] = -1

                
                # to prevent random explosion particles that were recognized as digits from messing up the game
                # still possible: if it habens 2 times in a row
                # potential solution: when placing: check if pixel changed colour(or even is of correct colour) - potentially blocked by particles/projectiles
                # when upgrading: check if corresponding box turned green(for left and right menu)
                # remove obstacle: colour change?

                if len(mapConfig['steps']):
                    if mapConfig['steps'][0]['action'] == 'sell':
                        customPrint('detected money: ' + str(currentValues['money']) + ', required: ' + str(getNextNonSellAction(mapConfig['steps'])['cost'] - sumAdjacentSells(mapConfig['steps'])) + ' (' + str(getNextNonSellAction(mapConfig['steps'])['cost']) + ' - ' + str(sumAdjacentSells(mapConfig['steps'])) + ')' + '          ', end = '', rewriteLine=True)
                    if mapConfig['steps'][0]['action'] == 'await_round':
                        customPrint('detected round: ' + str(currentValues['round']) + ', awaiting: ' + str(mapConfig['steps'][0]['round']) + '          ', end = '', rewriteLine=True)
                    else:
                        customPrint('detected money: ' + str(currentValues['money']) + ', required: ' + str(mapConfig['steps'][0]['cost']) + '          ', end = '', rewriteLine=True)

                if mode == Mode.VALIDATE_PLAYTHROUGHS:
                    if lastIterationBalance != -1 and currentValues['money'] != lastIterationBalance - lastIterationCost:
                        if currentValues['money'] == lastIterationBalance:
                            customPrint('action: ' + str(lastIterationAction) + ' failed!')
                            validationResult = False
                            mapConfig['steps'] = []
                        else:
                            customPrint('# error! expected cost: ' + str(lastIterationCost) + ', detected cost: ' + str(lastIterationBalance - currentValues['money']) + '. Is monkey knowledge disabled?')
                elif mode == Mode.VALIDATE_COSTS:
                    if lastIterationBalance != -1 and lastIterationAction:
                        if lastIterationAction['action'] == 'place':
                            costs[lastIterationAction['extra']['group']][lastIterationAction['extra']['type']]['base'] = int(lastIterationBalance - currentValues['money'])
                        elif lastIterationAction['action'] == 'upgrade':
                            costs[lastIterationAction['extra']['group']][lastIterationAction['extra']['type']]['upgrades'][lastIterationAction['extra']['upgrade'][0]][lastIterationAction['extra']['upgrade'][1] - 1] = int(lastIterationBalance - currentValues['money'])

                if mode == Mode.VALIDATE_PLAYTHROUGHS and len(mapConfig['steps']) and (mapConfig['steps'][0]['action'] == 'await_round' or  mapConfig['steps'][0]['action'] == 'speed'):
                    mapConfig['steps'].pop(0)
                elif currentValues['money'] == -1 or currentValues['round'] == -1 and len(mapConfig['steps']) and mapConfig['steps'][0]['action'] == 'await_round':
                    customPrint('recognition error. money: ' + str(currentValues['money']) + ', round: ' + str(currentValues['round']))
                elif mode != Mode.VALIDATE_COSTS and lastIterationBalance - lastIterationCost > currentValues['money']:
                    customPrint('potential cash recognition error: ' + str(lastIterationBalance) + ' - ' + str(lastIterationCost) + ' -> ' + str(currentValues['money']))
                    # cv2.imwrite('tmp_images/' + time.strftime("%Y-%m-%d_%H-%M-%S") + '_' + str(lastIterationBalance) + '.png', lastIterationScreenshotAreas[2])
                    # cv2.imwrite('tmp_images/' + time.strftime("%Y-%m-%d_%H-%M-%S") + '_' + str(currentValues['money']) + '.png', images[2])
                    skippingIteration = True
                elif mode != Mode.VALIDATE_COSTS and (currentValues['round'] - lastIterationRound > 1 or lastIterationRound > currentValues['round']) and len(mapConfig['steps']) and mapConfig['steps'][0]['action'] == 'await_round':
                    customPrint('potential round recognition error: ' + str(lastIterationRound) + ' -> ' + str(currentValues['round']))
                    skippingIteration = True
                elif len(mapConfig['steps']) and ((mapConfig['steps'][0]['action'] != 'sell' and mapConfig['steps'][0]['action'] != 'await_round' and min(currentValues['money'], lastIterationBalance - lastIterationCost) >= mapConfig['steps'][0]['cost']) 
                or mapConfig['gamemode'] == 'deflation' 
                or mapConfig['steps'][0]['action'] == 'await_round' and currentValues['round'] >= mapConfig['steps'][0]['round']
                or mapConfig['steps'][0]['action'] == 'await_round' and mode == Mode.VALIDATE_PLAYTHROUGHS
                or ((mapConfig['steps'][0]['action'] == 'sell') and min(currentValues['money'], lastIterationBalance - lastIterationCost) + sumAdjacentSells(mapConfig['steps']) >= getNextNonSellAction(mapConfig['steps'])['cost'])):
                    action = mapConfig['steps'].pop(0)
                    thisIterationAction = action
                    if action['action'] != 'sell' and action['action'] != 'await_round':
                        thisIterationCost = action['cost']
                    customPrint('performing action: ' + str(action))
                    if action['action'] == 'place':
                        pyautogui.moveTo(action['pos'])
                        time.sleep(actionDelay)
                        sendKey(action['key'])
                        time.sleep(actionDelay)
                        pyautogui.click()
                    elif action['action'] == 'upgrade' or action['action'] == 'retarget' or action['action'] == 'special':
                        # game hints potentially blocking monkeys
                        pyautogui.click(action['pos'])
                        time.sleep(actionDelay)
                        actionTmp = None
                        while action:
                            if 'to' in action:
                                pyautogui.moveTo(action['to'])
                                time.sleep(smallActionDelay)
                            if action['action'] == 'click':
                                time.sleep(actionDelay)
                                pyautogui.moveTo(action['pos'])
                                pyautogui.click()
                                time.sleep(actionDelay)
                            else:
                                sendKey(action['key'])
                            if 'to' in action and mapConfig['monkeys'][action['name']]['type'] == 'mortar':
                                pyautogui.click()
                            time.sleep(smallActionDelay)
                            actionTmp = action
                            if len(mapConfig['steps']) and 'name' in mapConfig['steps'][0] and mapConfig['steps'][0]['name'] == action['name'] and (mapConfig['steps'][0]['action'] == 'retarget' or mapConfig['steps'][0]['action'] == 'special' or mapConfig['steps'][0]['action'] == 'click'):
                                action = mapConfig['steps'].pop(0)
                                customPrint('+' + action['action'])
                            else:
                                action = None
                        action = actionTmp
                        sendKey('{Esc}')
                    elif action['action'] == 'sell':
                        pyautogui.moveTo(action['pos'])
                        pyautogui.click()
                        time.sleep(actionDelay)
                        sendKey(action['key'])
                    elif action['action'] == 'remove':
                        customPrint('removing obstacle at ' + tupleToStr(action['pos']) + ' for ' + str(action['cost']))
                        pyautogui.moveTo(action['pos'])
                        pyautogui.click()
                        time.sleep(menuChangeDelay)
                        result = cv2.matchTemplate(np.array(pyautogui.screenshot())[:, :, ::-1].copy(), locateImages['remove_obstacle_confirm_button'], cv2.TM_SQDIFF_NORMED)
                        pyautogui.click(cv2.minMaxLoc(result)[2])
                    elif action['action'] == 'click':
                        pyautogui.moveTo(action['pos'])
                        pyautogui.click()
                    elif action['action'] == 'press':
                        sendKey(action['key'])
                    elif action['action'] == 'speed':
                        if action['speed'] == 'fast':
                            fast = True
                        elif action['speed'] == 'slow':
                            fast = False

                elif mode in [Mode.VALIDATE_PLAYTHROUGHS, Mode.VALIDATE_COSTS] and len(mapConfig['steps']) == 0 and lastIterationCost == 0:
                    state = State.UNDEFINED

                if (not doAllStepsBeforeStart and mapConfig['gamemode'] != 'deflation' and not skippingIteration and getNextCostingAction(mapConfig['steps'])['cost'] > min(currentValues['money'], lastIterationBalance - lastIterationCost)) or len(mapConfig['steps']) == 0:
                    bestMatchDiff = None
                    gameState = None
                    for screenCfg in [
                        ('game_playing_fast', comparisonImages['game_state']['game_playing_fast'], imageAreas["compare"]["game_state"]),
                        ('game_playing_slow', comparisonImages['game_state']['game_playing_slow'], imageAreas["compare"]["game_state"]),
                        ('game_paused', comparisonImages['game_state']['game_paused'], imageAreas["compare"]["game_state"]),
                    ]:
                        diff = cv2.matchTemplate(cutImage(screenshot, screenCfg[2]), cutImage(screenCfg[1], screenCfg[2]), cv2.TM_SQDIFF_NORMED)[0][0]
                        if bestMatchDiff is None or diff < bestMatchDiff:
                            bestMatchDiff = diff
                            gameState = screenCfg[0]

                    if gameState == 'game_playing_fast' and not fast:
                        sendKey(keybinds['others']['play'])
                    elif gameState == 'game_playing_slow' and fast:
                        sendKey(keybinds['others']['play'])
                    elif gameState == 'game_paused':
                        sendKey(keybinds['others']['play'])
                    
                lastIterationScreenshotAreas = images
                lastIterationBalance = currentValues['money']
                lastIterationCost = thisIterationCost
                lastIterationAction = thisIterationAction

                lastIterationRound = currentValues['round']

                iterationBalances.append((currentValues['money'], thisIterationCost))
            else:
                customPrint("task INGAME, but not in related screen!")
                state = State.GOTO_HOME
                lastStateTransitionSuccessful = False
        else:
            state = State.UNDEFINED
            lastStateTransitionSuccessful = False

        if state != lastState:
            customPrint("new state " + state.name + "!")

        lastScreen = screen
        lastState = state

        time.sleep(actionDelay if state == State.INGAME else menuChangeDelay)

if __name__ == "__main__":
    main()