def __init__(self,
parser=None,
save=None,
sent=None,
requester=None,
login=None,
url_filter=None,
proxy=None,
queue_parse_size=-1,
queue_save_size=-1,
queue_proxy_size=-1):
self._cust_parser = parser #Parser解析器实例，解析器的子类或无
self._cust_saver = save #saver实例，saver的子类或无
self._cust_request = requester #url实例，requester的子类
self._cust_proxy = proxy # 代理服务器实例
self._cust_sent = sent #sent实例，senter的子类
self._cust_login = login #login实例
self._url_filter = url_filter #默认值：无，也可以是UrlFilter（） # default: None, also can be UrlFilter()
self._thread_REQUEST_list = [] #获取url线程列表，在start_working（）中定义长度
self._thread_parser = None #解析器线程，如果没有解析器实例，则为None
self._thread_saver = None #saver线程，如果没有saver实例，则为None
self._thread_senter = None #senter线程，如果没有sent实例，则为None
self._thread_requester = None #requester线程
self._thread_proxy = None #proxy线程
self._thread_stop_flag = False #默认值：False，线程的停止标志 # default: False, stop flag of threads
self._login_flag = False
self._cookie = {}
self._queue_deal = queue.PriorityQueue(-1)
self._queue_url = queue.PriorityQueue(-1)
self._queue_parse = queue.PriorityQueue(queue_parse_size)
self._queue_save = queue.PriorityQueue(queue_save_size)
self._queue_proxy = queue.PriorityQueue(queue_proxy_size)
self._stop_spider = False
self._lock = threading.Lock() #_number_dict需要的锁
self._number_dict = {
TypeEnum.URL_DEAL_RUN: 0, #通过urler去重处理完成的url计数
TypeEnum.URL_DEAL_NOT: 0, # 通过urler去重处理完成的url计数
TypeEnum.URL_DEAL_SUCC: 0, # 通过urler去重处理完成的url计数
TypeEnum.URL_DEAL_FAIL: 0, # 通过urler去重处理完成的url计数
TypeEnum.URL_REQUEST_RUN: 0, #正在请求的获取任务的计数
TypeEnum.URL_REQUEST_NOT: 0, #尚未请求的URL的计数
TypeEnum.URL_REQUEST_SUCC: 0, #已成功请求的URL的计数
TypeEnum.URL_REQUEST_FAIL: 0, #请求失败的url计数
TypeEnum.HTM_PARSE_RUN: 0, #正在运行的分析任务的计数
TypeEnum.HTM_PARSE_NOT: 0, #尚未分析的URL的计数
TypeEnum.HTM_PARSE_SUCC: 0, #已成功分析的URL计数
TypeEnum.HTM_PARSE_FAIL: 0, #已分析的URL计数失败
TypeEnum.ITEM_SAVE_RUN: 0, #正在运行的保存任务的计数
TypeEnum.ITEM_SAVE_NOT: 0, #尚未保存的URL的计数
TypeEnum.ITEM_SAVE_SUCC: 0, #已成功保存的URL的计数
TypeEnum.ITEM_SAVE_FAIL: 0, #已保存的URL计数失败
TypeEnum.PROXY_LIFE: 0, # 可用的代理数
TypeEnum.PROXY_FAIL: 0, # 不可用的代理数
}
print("线程池已初始化")
# logging.warning("线程池已初始化")
self._thread_monitor = MonitorThread("monitor", self) #监视线程
self._thread_monitor.setDaemon(True)
self._thread_monitor.start()
return
def initiate_queues(self, s, t):
Q = [queue.PriorityQueue(), queue.PriorityQueue()]
Q[0].put((0, s))
Q[1].put((0, t))
return Q
# 优先级队列 import queue q = queue.PriorityQueue() # 每个元素都是元组 # 数字越小优先级越高 # 同优先级先进先出 q.put((1, "work")) q.put((2, "eat")) q.put((-1, "drink")) print(q.get()) print(q.get()) print(q.get()) # queue.LifoQueue 后进先出队列， 类似堆栈 # q.deque 双向队列
lista = []
while queue:
lista.append(queue.get())
if (queue.empty()):
return lista
t = 1
while (t > 0):
enter = input().split()
n_no = int(enter[0])
n_aresta = int(enter[1])
queue = Q.PriorityQueue()
for i in range(n_aresta):
enter = input().split()
queue.put((int(enter[2]), int(enter[0]), int(enter[1])))
graph = {}
lista = []
lista = convert(queue)
current = lista[0]
graph[current[1]] = [[current[0], current[2]]]
graph[current[2]] = [[current[0], current[1]]]
qtd = 1
val = current[0]
def a_star_search(start_point, goal_point, step, robot_radius, clearance,
ANG_THRESHOLD, DIST_THRESHOLD):
x_node = []
y_node = []
x_parent = []
y_parent = []
start_point = discretize(start_point[0], start_point[1], start_point[2],
ANG_THRESHOLD, DIST_THRESHOLD)
goal_point = discretize(goal_point[0], goal_point[1], goal_point[2],
ANG_THRESHOLD, DIST_THRESHOLD)
visited = np.zeros([
int(300 / DIST_THRESHOLD),
int(200 / DIST_THRESHOLD),
int(360 / ANG_THRESHOLD)
])
start = (0, start_point, None)
goal = (0, goal_point, None)
nodes_explored = queue.PriorityQueue()
path_dict = {}
nodes_explored.put(start)
while True:
current_node = list(nodes_explored.get())
current_node[0] = current_node[0] - cost_to_go(current_node[1],
goal[1])
if (current_node[2] != None):
str1 = str(current_node[1][0])
str2 = str(current_node[1][1])
str3 = str(current_node[1][2])
str4 = str(current_node[2][0])
str5 = str(current_node[2][1])
str6 = str(current_node[2][2])
str_node = str1 + ',' + str2 + ',' + str3
str_parent = str4 + ',' + str5 + ',' + str6
path_dict[str_node] = str_parent
else:
str1 = str(current_node[1][0])
str2 = str(current_node[1][1])
str3 = str(current_node[1][2])
str4 = str(current_node[2])
str_node = str1 + ',' + str2 + ',' + str3
str_parent = str4
path_dict[str_node] = str_parent
actions = action_step(current_node, step, ANG_THRESHOLD)
for new_node in actions:
angle = new_node[2] + current_node[1][2]
if angle < 0:
angle = angle + 360
if angle > 360:
angle = angle - 360
if angle == 360:
angle = 0
node = (current_node[1][0] + new_node[0],
current_node[1][1] + new_node[1], angle)
node = discretize(node[0], node[1], node[2], ANG_THRESHOLD,
DIST_THRESHOLD)
node_cost = current_node[0] + new_node[3] + cost_to_go(
node, goal[1])
node_parent = current_node[1]
if check_obstacle(node[0], node[1], robot_radius,
clearance) == False:
if (visited[int(node[0] / DIST_THRESHOLD)][int(
node[1] / DIST_THRESHOLD)][int(node[2] /
ANG_THRESHOLD)] == 0):
visited[int(node[0] / DIST_THRESHOLD)][int(
node[1] / DIST_THRESHOLD)][int(node[2] /
ANG_THRESHOLD)] = 1
x_node.append(node[0])
y_node.append(node[1])
x_parent.append(node_parent[0])
y_parent.append(node_parent[1])
new_node = (node_cost, node, node_parent)
nodes_explored.put(new_node)
if check_goal(current_node[1][0], current_node[1][1], goal_point[0],
goal_point[1]):
print("GOAL REACHED!!!")
print("Plotting path...")
path = []
str_p1 = str(current_node[2][0])
str_p2 = str(current_node[2][1])
str_p3 = str(current_node[2][2])
parent = str_p1 + ',' + str_p2 + ',' + str_p3
while parent != "None":
temp = path_dict.get(parent)
if (parent[1] == '.' and parent[5] == '.'):
par_1 = float(parent[0]) + float(parent[2]) / 10
par_2 = float(parent[4]) + float(parent[6]) / 10
if (parent[2] == '.' and parent[7] == '.'):
par_1 = float(parent[0] +
parent[1]) + float(parent[3]) / 10
par_2 = float(parent[5] +
parent[6]) + float(parent[8]) / 10
if (parent[1] == '.' and parent[6] == '.'):
par_1 = float(parent[0]) + float(parent[2]) / 10
par_2 = float(parent[4] +
parent[5]) + float(parent[7]) / 10
if (parent[2] == '.' and parent[6] == '.'):
par_1 = float(parent[0] +
parent[1]) + float(parent[3]) / 10
par_2 = float(parent[5]) + float(parent[7]) / 10
if (parent[3] == '.' and parent[9] == '.'):
par_1 = float(parent[0] + parent[1] +
parent[2]) + float(parent[4]) / 10
par_2 = float(parent[6] + parent[7] +
parent[8]) + float(parent[10]) / 10
if (parent[3] == '.' and parent[7] == '.'):
par_1 = float(parent[0] + parent[1] +
parent[2]) + float(parent[4]) / 10
par_2 = float(parent[6]) + float(parent[8]) / 10
if (parent[3] == '.' and parent[8] == '.'):
par_1 = float(parent[0] + parent[1] +
parent[2]) + float(parent[4]) / 10
par_2 = float(parent[6] +
parent[7]) + float(parent[9]) / 10
if (parent[1] == '.' and parent[7] == '.'):
par_1 = float(parent[0]) + float(parent[2]) / 10
par_2 = float(parent[4] + parent[5] +
parent[6]) + float(parent[8]) / 10
if (parent[2] == '.' and parent[8] == '.'):
par_1 = float(parent[0] +
parent[1]) + float(parent[3]) / 10
par_2 = float(parent[5] + parent[6] +
parent[7]) + float(parent[9]) / 10
path.append((par_1, par_2))
parent = temp
if ((par_1, par_2) == (start_point[0], start_point[1])):
break
path.append((start_point[0], start_point[1]))
return path, x_node, y_node, x_parent, y_parent
def uniformCostOrAStar(sleeptime=0, astar=True, tracepath=False):
start = datetime.datetime.now() # Starts a timer
resetColors(
) # Recolors the previous shown pathway (if there is one) back to white
board.resetVisitorFlags() # Resets the visitor flags back to False
# Both A* and UCS use a priorityqueue as agenda, that's also why I have combined both in one function
agenda = queue.PriorityQueue()
previous_pathway = []
# If the algorithm is A* then we will incorporate a heuristic, we will subtract the board.start value since this
# will be added later on. If the algorithm is UCS then we will not incorporate the heuristic. An item in the agenda
# will look like this (value, [travelled_pathway]) where value is either heuristic + travelled distance (A*) or just
# travelled distance in case of UCS.
if astar:
agenda.put(
(calculateHeuristic(board.start, board.destination) -
board.board[board.start[0]][board.start[1]], [board.start]))
alg = "Algorithm A*"
else:
agenda.put(
[-board.board[board.start[0]][board.start[1]], [board.start]])
alg = "Uniform Cost Search"
while agenda.qsize() > 0:
pointer = agenda.get() # Gets the first item from the priorityqueue
pathway, current_point = pointer[1], pointer[1][-1]
shortest_distance_to_current = board.visited[current_point[0]][
current_point[1]]
travelled_distance = pointer[0] + board.board[current_point[0]][
current_point[1]]
if astar:
# In the A* algorithm the heuristic is incorporated in the first value of the agenda
travelled_distance -= calculateHeuristic(current_point,
board.destination)
if str(shortest_distance_to_current).isdigit(
) and travelled_distance >= shortest_distance_to_current:
# If this point has been visited yet and the current travelled distance is more than the previous
# shortest distance to this point then we will not continue with this pathway
continue
# Else we will update the shortest distance to this point to the current travelled distance
board.visited[current_point[0]][current_point[1]] = travelled_distance
# If the tracepath is selected on the GUI then we will update the pathway each iteration
if tracepath:
time.sleep(sleeptime)
# Delete the part of the previous pathway which isn't present in the current pathway
unshowCalculation(
[coord for coord in previous_pathway if coord not in pathway])
# Show the part of the current pathway which isn't active yet
showCalculation(
[coord for coord in pathway if coord not in previous_pathway])
if current_point == board.destination: # If we have reached the destination
if not tracepath: # Show the pathway if tracepath was disabled, otherwise it is already shown
showCalculation(pathway)
messagebox.config(
state='normal') # Allow us to make a change in the messagebox
messagebox.insert(
'end',
f'Algorithm: {alg}\nTotal Distance: {travelled_distance}\n'
f'Elapsed time: {datetime.datetime.now() - start}\n'
) # update messagebox
messagebox.config(
state='disabled'
) # Close access to messagebox again (prevents user input/disrupt)
return
# If we are not at the destination yet then we will add all possible and viable neighbors to the agenda
# Viable neighbors are those which are on the board, which are not obstacles, which are not in the pathway
# yet (prevents cycles) and which have not been visited yet or have only been visited by a longer path
for point in filter(
lambda point: not str(board.board[point[0]][point[1]]).isalpha(
) and point not in pathway and
(not board.visited[point[0]][point[1]] or travelled_distance >
board.visited[point[0]][point[1]]),
findNeighbors(current_point, board.size)):
if astar:
agenda.put((calculateHeuristic(point, board.destination) +
travelled_distance, pathway + [point]))
else:
agenda.put((travelled_distance, pathway + [point]))
previous_pathway = pathway
# This will only execute when no pathway has been found
unshowCalculation(
previous_pathway[1:]) # Decolor the last pathway of the calculation
messagebox.config(
state='normal') # Allow us to make a change in the messagebox
messagebox.insert('end', f'Algorithm: {alg} \nNo pathway found!\n')
messagebox.config(
state='disabled'
) # Close access to messagebox again (prevents user input/disrupt)
return
def Solve_ThiningAstar(maze, start, goal, dim, prob):
print("Thining A star")
maze_org = copy.deepcopy(maze)
assert (prob > 0.02)
blocked_nodes = int(dim * dim * prob)
new_blocked_nodes = int(
dim * dim *
(prob * 0.5)) # reduce number of blockages by 50% of current value(q)
blockage_to_remove = blocked_nodes - new_blocked_nodes
# Simplified maze
while blockage_to_remove > 0:
row = random.randint(0, dim - 1)
column = random.randint(0, dim - 1)
while maze[row][column] == 0:
row = random.randint(0, dim - 1)
column = random.randint(0, dim - 1)
maze[row][column] = 0
blockage_to_remove -= 1
## Print the new maze
#print("Simplified maze")
#for row in range(0,dim):
# for column in range(0,dim):
# print("%d "%maze[row][column] ,end="")
# print()
total_eu = math.sqrt(2 * ((dim - 1)**2))
assert (maze[0][0] == 0) and (maze[dim - 1][dim - 1] == 0)
Path_queue = queue.PriorityQueue()
Path_queue.put((total_eu, [start]))
Visited = set()
Done_find = False
path = []
explored_nodes = []
# Pick the first entry from the priority queue
# Get its non-blocked children by running A* on the thinned maze.
# Add the children to the priority queue based on the heuristic distance
# If goal is found or queue is empty, exit the loop and return the path
while not Path_queue.empty() and not Done_find:
path_full = Path_queue.get()
path = path_full[-1]
node = path[-1]
explored_nodes.append(node)
if node == goal:
break
elif node not in Visited:
for node_child in GetChildNodeforParentWithThining(
maze_org, maze, node, dim, (dim * dim) - 1, "Manhattan",
Visited):
if node_child[0] not in Visited:
path_1 = list(path)
path_1.append(node_child[0])
Path_queue.put((node_child[1], path_1))
if node_child[0] == goal:
Done_find = True
path = path_1
break
Visited.add(node)
return path, maze_org, explored_nodes
import queue q1 = queue.Queue(5) # 创建一个先进先出的队列 q2 = queue.LifoQueue(5) # 创建一个后进先出的队列 q3 = queue.PriorityQueue(5) # 优先级队列 # put() : 输入 q1.put(1) # get() : 获取 q1.get() # qsize : 获取当前队列的长度,不是百分之百的准确 print(q1.qsize()) # empty : 判断队列是否为空,返回的是布尔值 print(q1.empty()) # full : 判断队列是否为满,返回为布尔值 print(q1.full())
def search(model, vocab, prob_sequence, sequence, post_stress, state, session, \
temp, dictMeters, fsaLine, dictWordTransitions, dictPartSpeechTags, breadth, wordPool):
def beamSearchOneLevel(model, vocab, prob_sequence, sequence, post_stress, state, session, \
temp, dictMeters, fsaLine, dictWordTransitions, dictPartSpeechTags, breadth, wordPool):
def decayRepeat(word,sequence, scale):
safe_repeat_words = []
#safe_repeat_words = set(["with,the,of,in,i"])
score_adjust = 0
decr = -scale
for w in range(len(sequence)):
if(word==sequence[w] and word not in safe_repeat_words):
score_adjust += decr
decr += scale/10 #decreases penalty as the words keep getting further from the new word
return score_adjust
def partsOfSpeechFilter(word1,word2,dictPartSpeechTags,dictPossiblePartsSpeech):
okay_tags = set(["RB","RBR","RBS"]) #THESE ARE THE ADVERBS
tag1 = dictPartSpeechTags[word1]
tag2 = dictPartSpeechTags[word2]
#if(tag1==tag2 and tag1 not in okay_tags):
# return True
if(tag1 not in dictPossiblePartsSpeech[tag2]):
return True
else:
return False
if(post_stress==0):
return("begin_line")
ret = []
scale = .02 #scale is the significant magnitude required to affect the score of bad/good things
dist, state = model.compute_fx(session, vocab, prob_sequence, sequence, state, temp)
for pred_stress in list(fsaLine[post_stress].prevs):
word_set = set([])
for word in dictWordTransitions[(pred_stress,post_stress)]:
#PREVENTS REPEAT ADJACENT WORDS OR PROBLEM-TAGGED WORDS
if(word == sequence[0]):
continue
if(partsOfSpeechFilter(word,sequence[0],dictPartSpeechTags,dictPossiblePartsSpeech)):
continue
#FACTORS IN SCORE ADJUSTMENTS
score_adjust = decayRepeat(word, sequence, 100*scale) #repeats
score_adjust += scale*len(word)/50 #length word
if(word in wordPool):
score_adjust += scale
#CALCULATES ACTUAL SCORE
key = np.array([[vocab[word]]])
new_prob = dist[key]
score_tuple = (new_prob, state)
score_tup = (score_tuple[0]+score_adjust,score_tuple[1]) #NOTE SHOULD SCORE_ADJUST BE ADDED HERE OR JUST IN THE ITEM LINE?
item = (score_tup[0],(score_tup, [word]+sequence, pred_stress))
if(item[0]==[[-float("inf")]]):
continue
ret+=[item]
return ret
masterPQ = Q.PriorityQueue()
checkList = []
checkSet = set([])
score_tuple = (prob_sequence, state)
first = (score_tuple[0],(score_tuple, sequence, post_stress))
masterPQ.put(first)#initial case
set_explored = set([])
while(not masterPQ.empty()):
depthPQ = Q.PriorityQueue()
while(not masterPQ.empty()):
try:
next_search = masterPQ.get()
except:
continue
possible_branches = beamSearchOneLevel(model, vocab, next_search[1][0][0], next_search[1][1], next_search[1][2],\
next_search[1][0][1], session, temp, dictMeters, fsaLine, dictWordTransitions,\
dictPartSpeechTags, breadth, wordPool)
if(possible_branches == "begin_line"):
checkList+=[next_search]
continue
for branch in possible_branches:
if(branch == []):
continue
test = tuple(branch[1][1]) #need to make sure each phrase is being checked uniquely (want it to be checked once in possible branches then never again)
if(test in set_explored):
continue
set_explored.add(test)
depthPQ.put(branch)
try:
if(depthPQ.qsize()>breadth):
depthPQ.get()
except:
pass
masterPQ = depthPQ
return checkList
def download_zip(self,
destination,
progress_callback=None,
parallel_downloads=16,
parallel_chunks_size=10):
"""Download this task's assets archive to a directory.
Args:
destination (str): directory where to download assets archive. If the directory does not exist, it will be created.
progress_callback (function): an optional callback with one parameter, the download progress percentage.
parallel_downloads (int): maximum number of parallel downloads if the node supports http range.
parallel_chunks_size (int): size in MB of chunks for parallel downloads
Returns:
str: path to archive file (.zip)
"""
info = self.info()
if info.status != TaskStatus.COMPLETED:
raise NodeResponseError("Cannot download task, task status is " +
str(info.status))
if not os.path.exists(destination):
os.makedirs(destination, exist_ok=True)
try:
download_stream = self.get('/task/{}/download/all.zip'.format(
self.uuid),
stream=True)
headers = download_stream.headers
zip_path = os.path.join(
destination, "{}_{}_all.zip".format(self.uuid,
int(time.time())))
# Keep track of download progress (if possible)
content_length = download_stream.headers.get('content-length')
total_length = int(
content_length) if content_length is not None else None
downloaded = 0
chunk_size = int(parallel_chunks_size * 1024 * 1024)
use_fallback = False
accept_ranges = headers.get('accept-ranges')
# Can we do parallel downloads?
if accept_ranges is not None and accept_ranges.lower(
) == 'bytes' and total_length is not None and total_length > chunk_size and parallel_downloads > 1:
num_chunks = int(math.ceil(total_length / float(chunk_size)))
num_workers = parallel_downloads
class nonloc:
completed_chunks = AtomicCounter(0)
merge_chunks = [False] * num_chunks
error = None
def merge():
current_chunk = 0
with open(zip_path, "wb") as out_file:
while current_chunk < num_chunks and nonloc.error is None:
if nonloc.merge_chunks[current_chunk]:
chunk_file = "%s.part%s" % (zip_path,
current_chunk)
with open(chunk_file, "rb") as fd:
out_file.write(fd.read())
os.unlink(chunk_file)
current_chunk += 1
else:
time.sleep(0.1)
def worker():
while True:
task = q.get()
part_num, bytes_range = task
if bytes_range is None or nonloc.error is not None:
q.task_done()
break
try:
# Download chunk
res = self.get(
'/task/{}/download/all.zip'.format(self.uuid),
stream=True,
headers={'Range': 'bytes=%s-%s' % bytes_range})
if res.status_code == 206:
with open("%s.part%s" % (zip_path, part_num),
'wb') as fd:
for chunk in res.iter_content(4096):
fd.write(chunk)
with nonloc.completed_chunks.lock:
nonloc.completed_chunks.value += 1
if progress_callback is not None:
progress_callback(
100.0 *
nonloc.completed_chunks.value /
num_chunks)
nonloc.merge_chunks[part_num] = True
else:
nonloc.error = RangeNotAvailableError()
except OdmError as e:
time.sleep(5)
q.put((part_num, bytes_range))
except Exception as e:
nonloc.error = e
finally:
q.task_done()
q = queue.PriorityQueue()
threads = []
for i in range(num_workers):
t = threading.Thread(target=worker)
t.start()
threads.append(t)
merge_thread = threading.Thread(target=merge)
merge_thread.start()
range_start = 0
for i in range(num_chunks):
range_end = min(range_start + chunk_size - 1,
total_length - 1)
q.put((i, (range_start, range_end)))
range_start = range_end + 1
# block until all tasks are done
q.join()
# stop workers
for i in range(len(threads)):
q.put((-1, None))
for t in threads:
t.join()
merge_thread.join()
if nonloc.error is not None:
if isinstance(nonloc.error, RangeNotAvailableError):
use_fallback = True
else:
raise nonloc.error
else:
use_fallback = True
if use_fallback:
# Single connection, boring download
with open(zip_path, 'wb') as fd:
for chunk in download_stream.iter_content(4096):
downloaded += len(chunk)
if progress_callback is not None and total_length is not None:
progress_callback(
(100.0 * float(downloaded) / total_length))
fd.write(chunk)
except (requests.exceptions.Timeout,
requests.exceptions.ConnectionError, ReadTimeoutError) as e:
raise NodeConnectionError(e)
return zip_path
import numpy as np
import queue
with open("input") as f:
cave = np.array([[int(c) for c in l] for l in f.read().splitlines()])
h, w = cave.shape
risks = np.full(cave.shape, -1)
frontier = queue.PriorityQueue(cave.size)
frontier.put((0, (0, 0)))
risks[0, 0] = 0
while not frontier.empty():
risk, (y, x) = frontier.get()
if (y, x) == (h - 1, w - 1):
break
for ny, nx in [(y - 1, x), (y + 1, x), (y, x - 1), (y, x + 1)]:
if 0 <= ny < h and 0 <= nx < w:
if risks[ny, nx] == -1:
risks[ny, nx] = risk + cave[ny, nx]
frontier.put((risks[ny, nx], (ny, nx)))
print(risks[h - 1, w - 1])
import random
import queue as q
grid = [[0] * 3 for i in range(3)]
grid_sol = [[1, 2, 3], [4, 5, 6], [7, 8, 0]]
prev_move = None
Q = q.PriorityQueue()
def prettyPrint(grid):
print("====================================")
for arr in grid:
for i in range(3):
print(arr[i], end=" ")
print()
print("=====================================")
def whereis(grid, x):
for arr in grid:
try:
j = arr.index(x)
i = grid.index(arr)
except:
continue
return i, j
def getData():
print(
import queue
arr = eval(input())
min_heap = queue.PriorityQueue()
for num in arr:
min_heap.put(num)
pre = 0
while not min_heap.empty():
cur = min_heap.get()
if cur - 1 > 0 and cur != pre + 1:
print(pre + 1)
break
pre = cur
def __init__(self, procs_map):
self.events = queue.PriorityQueue()
self.procs = dict(procs_map)
def clear(self):
try:
import queue
except ImportError:
import Queue as queue
self._queue = queue.PriorityQueue()
#后进先出队列
import queue
q_lifo = queue.LifoQueue()
for i in range(4):
q_lifo.put_nowait(i)
while True:
print(q_lifo.get(timeout=1)) #延迟1秒，如果get不到数据，则抛出异常
'''
#优先队列
import queue
q_PQ = queue.PriorityQueue()
q_PQ.put({1: "LC"}) #按着元组的第一个元素进行对比，小的优先
q_PQ.put({8: "Jack"})
# q_PQ.put({6:"Mark"})
# q_PQ.put({4:"Joe"})
print(q_PQ.queue[0][1])
#
# print(q_PQ.queue[3])
#
# while q_PQ.qsize():
# print(q_PQ.get())
return result
letters = Counter(input().lower())
words = []
for word in open('/usr/share/dict/words').read().split():
word = word.lower()
if not (Counter(word) - letters):
words.append(word)
words.sort()
words.sort(key=len)
#for word in words:
# print(word)
words = [word for word in words if 5 <= len(word) <= 8]
paths = queue.PriorityQueue()
paths.put((0, 0, ['']))
done_paths = []
while len(done_paths) < 50000:
neg_score, dist, path = paths.get_nowait()
if dist > 120:
done_paths.append((-neg_score, dist, path))
continue
for next_word in words:
if next_word in path: continue
next_neg_score = neg_score - score(len(next_word))
next_dist = dist + distance(path[-1], next_word)
paths.put((next_neg_score, next_dist, path + [next_word]))
def greedy_search(
task,
default_parameters,
parameters,
initial_population=3,
limit=10,
report=None,
report_best=None,
):
import itertools
names = [k for k, _ in parameters.items()]
values = [v for _, v in parameters.items()]
all_params = list(itertools.product(*values))
random.shuffle(all_params)
[print(r) for r in all_params]
best = {'eval': None, 'params': None, 'artifact': None}
import queue
open_list = queue.PriorityQueue()
close_list = {}
results = [] # for displaying
def random_configs():
import random
while True:
yield {k: _select(v) for k, v in parameters.items()}
def neighbors(parent):
import random
results = []
for k in names:
for kv in parameters[k]:
if parent[k] is not kv:
other = parent.copy()
other[k] = kv
results.append(other)
return results
def _key(local_parameters):
return tuple(local_parameters[k] for k in names)
def _iter(local_parameters):
artifact, eval = task(merge_hash(default_parameters, local_parameters))
_update_best(artifact, eval, local_parameters, results, best, report,
report_best)
#
close_list[_key(local_parameters)] = eval # tuples are hashable
open_list.put((eval, local_parameters))
try:
for i, local_parameters in zip(range(initial_population),
random_configs()):
_iter(local_parameters)
for i in range(initial_population, limit):
if open_list.empty():
break
_, parent = open_list.get()
children = neighbors(parent)
open_children = []
for c in children:
if _key(c) not in close_list:
open_children.append(c)
_iter(_select(open_children))
finally:
print("Best parameter:\n{}\neval: {}".format(best['params'],
best['eval']))
print(results)
return best['artifact'], best['params'], best['eval']
def __init__(self):
self.q = queue.PriorityQueue()
self.count = 0
data = pickle.load(open('pickles/vectorlists.pickle', 'rb'))
fonts = []
for d in data[1:]:
fonts.append(d[0])
fonts = [f.lower() for f in fonts]
#print(fonts)
transformed = pickle.load(open('pickles/TSNE-trained.pickle', 'rb'))
xs = transformed[:, 0]
ys = transformed[:, 1]
zs = transformed[:, 2]
font = sys.argv[1]
font = font.lower()
idx = fonts.index(font)
distances = Q.PriorityQueue()
for i in range(len(fonts)):
if i != idx:
dist = EuclideanDist(xs[i], ys[i], zs[i], xs[idx], ys[idx], zs[idx])
distances.put(node(i, dist))
for i in range(4):
curr = distances.get()
print(curr.i, fonts[curr.i], curr.dist)
def calculate_shortest_paths(self, start: Vertex) -> dict[Vertex, _Path]:
"""Calculate shortest paths from vertex "start" to all others."""
logger.info("calculating shortest paths from vertex: {}".format(start))
assert start in self._vertices, (
"unexpected vertex: {}".format(start))
shortest_paths: dict[Vertex, _PathRef] = {}
next_path_by_distance: queue.PriorityQueue[_Path] = (
queue.PriorityQueue())
# add all vertices connected to the start vertex to "next" queue
for e in self._vertices_outgoing_edges.get(start, []):
logger.debug("adding initial edge: {}".format(e))
next_path_by_distance.put(e)
# implicit shortest path to starting node is 0
shortest_paths[start] = _PathRef(Edge(start, start, 0))
# pop the next path. If we've never seen the destination vertex it
# is a shortest path so add a result and add its edges to next
# vertices. If we have seen it check to see if this is an alternate
# shortest path to the vertex, but then add no next vertex. Do this
# until there is no more work (no more items in queue).
while not next_path_by_distance.empty():
next_path = next_path_by_distance.get_nowait()
vertex = next_path.get_destination_vertex()
logger.debug("popped next path: {}. shortest paths: {}".format(
next_path, shortest_paths))
if vertex in shortest_paths:
logger.debug("vertex in shortest paths")
existing_path_ref = shortest_paths[vertex]
existing_path = existing_path_ref.get()
assert existing_path.get_distance() <= next_path.get_distance()
if existing_path.get_distance() == next_path.get_distance():
logger.debug("new alternative path")
# we have an alternate shortest path to this vertex
new_alternatives = _PathAlternatives(
existing_path_ref.copy(), _PathRef(next_path))
logger.debug(
("replacing path with alternatives. "
"Original path: {}. New alternatives: {}").format(
existing_path,
next_path
))
existing_path_ref.set(new_alternatives)
else:
logger.debug("this is a longer path than exists")
# do not add any new paths to search
else:
# a new shortest path!
logger.debug("vertex NOT in shortest paths")
path_ref = _PathRef(next_path)
shortest_paths[vertex] = path_ref
# add all outgoing edges to next paths.
for e in self._vertices_outgoing_edges[vertex]:
new_path = _PathSequence(
path_ref,
e)
logger.debug("adding additional path: {}".format(new_path))
next_path_by_distance.put_nowait(new_path)
# unpack the path refs to return
ret: dict[Vertex, _Path] = {}
for (k, v,) in shortest_paths.items():
ret[k] = v.get()
return ret
# -*- coding: utf-8 -*-
import collections
import threading
import time
import json
import queue
import database
import pymysql
from request import BaseRequest
from urllib import parse
category_map = [
'热血', '格斗', '魔法', '侦探', '竞技', '恐怖', '战国', '魔幻', '冒险', '校园', '搞笑', '少女',
'少男', '科幻', '港产', '其他'
]
priority_queue = queue.PriorityQueue() #存放网址的队列
threads = []
#队列优先级
class Job(object):
def __init__(self, priority, description, url=None):
self.priority = priority
self.description = description
self.url = url
return
def __lt__(self, other):
return self.priority < other.priority
def getPriorityQueue(list):
q = Q.PriorityQueue()
for node in list:
q.put(Ordered_Node(heuristics[node], node))
return q, len(list)
def __init__(self):
SimpleChannel.__init__(self)
self.name=common.random_string(l_min=20,l_max=20)
self._sequence=common.SerializedSequence("channel_"+self.name)
self._queue=queue.PriorityQueue()
import queue
with open("C-small-1-attempt0.in") as f:
with open("output.txt", "w") as g:
t = int(f.readline().strip())
for i in range(1, t + 1):
[n, k] = map(int, f.readline().split())
my_queue = queue.PriorityQueue()
my_queue.put(-n)
for _ in range(k):
tmp = -my_queue.get() - 1
a = tmp // 2 + tmp % 2
b = tmp // 2
my_queue.put(-a)
my_queue.put(-b)
g.write("Case #{}: {} {}\n".format(i, a, b))
import subprocess
from matcher import (wikipedia, database, wikidata_api, netstring, mail,
overpass, space_alert, model)
from time import time, sleep
from datetime import datetime
from matcher.place import Place, PlaceMatcher, bbox_chunk
from matcher.view import app
app.config.from_object('config.default')
database.init_app(app)
re_point = re.compile(r'^Point\(([-E0-9.]+) ([-E0-9.]+)\)$')
active_jobs = {}
task_queue = queue.PriorityQueue()
def wait_for_slot(send_queue):
print('get status')
try:
status = overpass.get_status()
except overpass.OverpassError as e:
r = e.args[0]
body = f'URL: {r.url}\n\nresponse:\n{r.text}'
mail.send_mail('Overpass API unavailable', body)
send_queue.put({'type': 'error', 'error': "Can't access overpass API"})
return False
except requests.exceptions.Timeout:
body = 'Timeout talking to overpass API'
mail.send_mail('Overpass API timeout', body)
def __init__(self, heuristic: 'Heuristic'):
super().__init__()
self.heuristic = heuristic
self.frontier = queue.PriorityQueue()
self.frontier_set = set()
clearbuttons()
def showError(ts):
LOG("Showerror!!! Game Over")
LOG("total numreadfails = " + str(numreadfails))
AddCmd(ts + 0.2, CMD_FLASH_COLOR, [SIMON_ERROR, 0.5, True])
AddCmd(ts + 0.8, CMD_FLASH_COLOR, [SIMON_ERROR, 0.5, True])
AddCmd(ts + 1.4, CMD_FLASH_COLOR, [SIMON_ERROR, 0.5, True])
#################################################################
# our queue and queue functions, definitions, etc
#
cmdq = queue.PriorityQueue(0)
nextCmd = None
# adds a command to the queue sorted by ts, keeping nextCmd as the next one in time sequence
def AddCmd(cmdAt, cmd, data):
global nextCmd
global cmdq
#LOG("added command " + str(cmd) + " " + str(data) + " at " + str(cmdAt))
if nextCmd == None:
nextCmd = {'ts': cmdAt, 'cmd': cmd, 'data': data}
elif cmdAt < nextCmd['ts']:
cmdq.put((nextCmd['ts'], nextCmd['cmd'], nextCmd['data']))
nextCmd = {'ts': cmdAt, 'cmd': cmd, 'data': data}
else:
cmdq.put((cmdAt, cmd, data))
def ksp(graph, source, target, num_k, weight=None):
graph_copy = graph.copy()
# Shortest path from the source to the target
A = [nx.shortest_path(graph_copy, source, target, weight=weight)]
A_costs = [path_cost(graph_copy, A[0], weight)]
# Initialize the heap to store the potential kth shortest path
B = queue.PriorityQueue()
for k in range(1, num_k):
# The spur node ranges from the first node to the next to last node in the shortest path
try:
for i in range(len(A[k - 1]) - 1):
# Spur node is retrieved from the previous k-shortest path, k - 1
spurNode = A[k - 1][i]
# The sequence of nodes from the source to the spur node of the previous k-shortest path
rootPath = A[k - 1][:i]
# We store the removed edges
removed_edges = []
for path in A:
if len(path) - 1 > i and rootPath == path[:i]:
# Remove the links that are part of the previous shortest paths which share the same root path
edge = (path[i], path[i + 1])
if not graph_copy.has_edge(*edge):
continue
removed_edges.append(
(edge, graph_copy.get_edge_data(*edge)))
graph_copy.remove_edge(*edge)
# Calculate the spur path from the spur node to the sink
try:
spurPath = nx.shortest_path(graph_copy,
spurNode,
target,
weight=weight)
# Entire path is made up of the root path and spur path
totalPath = rootPath + spurPath
totalPathCost = path_cost(graph_copy, totalPath, weight)
# Add the potential k-shortest path to the heap
B.put((totalPathCost, totalPath))
except nx.NetworkXNoPath:
pass
#Add back the edges that were removed from the graph
#for removed_edge in removed_edges:
# print(removed_edge)
# graph_copy.add_edge(
# *removed_edge[0],
# **removed_edge[1]
# )
# Sort the potential k-shortest paths by cost
# B is already sorted
# Add the lowest cost path becomes the k-shortest path.
while True:
try:
cost_, path_ = B.get(False)
if path_ not in A:
A.append(path_)
A_costs.append(cost_)
break
except queue.Empty:
break
except IndexError:
pass
return A, A_costs
def __init__(self,
fetcher,
parser=None,
saver=None,
proxieser=None,
url_filter=None,
max_count=500,
max_count_in_proxies=100):
"""
constructor
"""
self._inst_fetcher = fetcher # fetcher instance, subclass of Fetcher
self._inst_parser = parser # parser instance, subclass of Parser or None
self._inst_saver = saver # saver instance, subclass of Saver or None
self._inst_proxieser = proxieser # proxieser instance, subclass of Proxieser
self._queue_fetch = queue.PriorityQueue(
) # (priority, counter, url, keys, deep, repeat) 优先度，计数器，网址，键，深度，重复次数
self._queue_parse = queue.PriorityQueue(
) # (priority, counter, url, keys, deep, content)
self._queue_save = queue.Queue(
) # (url, keys, item), item can be anything
self._queue_proxies = queue.Queue(
) # {"http": "http://auth@ip:port", "https": "https://auth@ip:port"}
self._thread_fetcher_list = [] # fetcher thread list
self._thread_parser = None # parser thread
self._thread_saver = None # saver thread
self._thread_proxieser = None # proxieser thread
self._thread_stop_flag = False # default: False, stop flag of threads
self._url_filter = url_filter # default: None, also can be UrlFilter()
self._number_dict = {
TPEnum.TASKS_RUNNING: 0, # the count of tasks which are running
TPEnum.URL_FETCH_NOT:
0, # the count of urls which haven't been fetched
TPEnum.URL_FETCH_SUCC:
0, # the count of urls which have been fetched successfully
TPEnum.URL_FETCH_FAIL:
0, # the count of urls which have been fetched failed
TPEnum.URL_FETCH_COUNT:
0, # the count of urls which appeared in self._queue_fetch
TPEnum.HTM_PARSE_NOT:
0, # the count of urls which haven't been parsed
TPEnum.HTM_PARSE_SUCC:
0, # the count of urls which have been parsed successfully
TPEnum.HTM_PARSE_FAIL:
0, # the count of urls which have been parsed failed
TPEnum.ITEM_SAVE_NOT:
0, # the count of urls which haven't been saved
TPEnum.ITEM_SAVE_SUCC:
0, # the count of urls which have been saved successfully
TPEnum.ITEM_SAVE_FAIL:
0, # the count of urls which have been saved failed
TPEnum.PROXIES_LEFT: 0, # the count of proxies which are avaliable
TPEnum.PROXIES_FAIL:
0, # the count of proxies which are unavaliable
}
self._lock = threading.Lock() # the lock which self._number_dict needs
self._max_count = max_count # maximum count of items which in parse queue or save queue
self._max_count_in_proxies = max_count_in_proxies # maximum count of items which in proxies queue
# set monitor thread
self._thread_monitor = MonitorThread("monitor", self) # 创建监控进为守护进程
self._thread_monitor.setDaemon(True)
self._thread_monitor.start()
logging.info("%s has been initialized", self.__class__.__name__)
return
