mirror of
https://github.com/InternLM/InternBootcamp.git
synced 2026-04-19 12:58:04 +00:00
324 lines
No EOL
13 KiB
Python
Executable file
324 lines
No EOL
13 KiB
Python
Executable file
import sys
|
||
|
||
from libs.wordladder.solving.puzzle import Puzzle
|
||
from libs.wordladder.solving.solver import Solver
|
||
from libs.wordladder.words.dictionary import Dictionary
|
||
from libs.wordladder.words.word import Word
|
||
from libs.wordladder.interactive import Interactive
|
||
from time import perf_counter
|
||
import json
|
||
|
||
|
||
def solver_main():
|
||
if len(sys.argv) > 2:
|
||
solve(sys.argv[1:])
|
||
else:
|
||
Interactive(sys.argv[1:]).run()
|
||
|
||
def solver_main2(param_data):
|
||
if len(sys.argv) > 0:
|
||
return solve2(sys.argv[1:],param_data)
|
||
else:
|
||
Interactive(sys.argv[1:]).run()
|
||
|
||
def solve(args):
|
||
first = args[0]
|
||
second = args[1]
|
||
if len(second) != len(first):
|
||
print('Start word \'%s\' and end word \'%s\' not the same length!' % (first, second))
|
||
sys.exit(-1)
|
||
start = perf_counter()
|
||
dictionary = Dictionary(len(first))
|
||
took = (perf_counter() - start) * 1000
|
||
print('Took %.2fms to load dictionary' % took)
|
||
start_word: Word = dictionary[first]
|
||
if start_word is None:
|
||
print('Start word \'%s\' not in dictionary' % first)
|
||
sys.exit(-1)
|
||
elif start_word.is_island:
|
||
print('Start word \'%s\' is an island word' % first)
|
||
sys.exit(-1)
|
||
end_word: Word = dictionary[second]
|
||
if end_word is None:
|
||
print('End word \'%s\' not in dictionary' % second)
|
||
sys.exit(-1)
|
||
elif end_word.is_island:
|
||
print('End word \'%s\' is an island word' % second)
|
||
sys.exit(-1)
|
||
|
||
puzzle: Puzzle = Puzzle(start_word, end_word)
|
||
max_ladder_length: int = -1
|
||
if len(args) > 2:
|
||
try:
|
||
max_ladder_length = int(args[2])
|
||
if max_ladder_length < 1:
|
||
raise ValueError
|
||
except ValueError:
|
||
print('Max ladder length arg must be an integer (greater than zero)')
|
||
sys.exit(-1)
|
||
else:
|
||
start = perf_counter()
|
||
min_ladder = puzzle.calculate_minimum_ladder_length()
|
||
took = (perf_counter() - start) * 1000
|
||
if min_ladder is None:
|
||
print('Cannot solve \'%s\' to \'%s\'' % (first, second))
|
||
sys.exit(-1)
|
||
max_ladder_length = min_ladder
|
||
print('Took %.2fms to determine minimum ladder length of %d' % (took, max_ladder_length))
|
||
|
||
solver = Solver(puzzle)
|
||
start = perf_counter()
|
||
solutions = solver.solve(max_ladder_length)
|
||
took = (perf_counter() - start) * 1000
|
||
if len(solutions) == 0:
|
||
print('Cannot solve \'%s\' to \'%s\' in ladder length %d (took %.2fms)' % (first, second, max_ladder_length, took))
|
||
sys.exit(-1)
|
||
slen = len(solutions)
|
||
print('Took %.2fms to find %d solutions (explored %d solutions)' % (took, slen, solver.explored_count))
|
||
solutions.sort()
|
||
for i in range(slen):
|
||
print('%d/%d %s' % (i + 1, slen, solutions[i]))
|
||
|
||
def solve2(args,param_data):
|
||
solutions_cnt = 0
|
||
current_ladder = 1
|
||
solutions_rounds = 0
|
||
|
||
first = param_data["start_word"]
|
||
second = param_data["end_word"]
|
||
|
||
if len(second) != len(first):
|
||
print('Start word \'%s\' and end word \'%s\' not the same length!' % (first, second))
|
||
sys.exit(-1)
|
||
start = perf_counter()
|
||
dictionary = Dictionary(len(first))
|
||
took = (perf_counter() - start) * 1000
|
||
print('Took %.2fms to load dictionary' % took)
|
||
start_word: Word = dictionary[first]
|
||
if start_word is None:
|
||
print('Start word \'%s\' not in dictionary' % first)
|
||
return save_to_file(param_data,None,code = 2)
|
||
sys.exit(-1)
|
||
elif start_word.is_island:
|
||
print('Start word \'%s\' is an island word' % first)
|
||
return save_to_file(param_data,None,code = 3)
|
||
sys.exit(-1)
|
||
end_word: Word = dictionary[second]
|
||
if end_word is None:
|
||
print('End word \'%s\' not in dictionary' % second)
|
||
return save_to_file(param_data,None,code = 4)
|
||
sys.exit(-1)
|
||
elif end_word.is_island:
|
||
print('End word \'%s\' is an island word' % second)
|
||
return save_to_file(param_data,None,code = 5)
|
||
sys.exit(-1)
|
||
|
||
puzzle: Puzzle = Puzzle(start_word, end_word)
|
||
|
||
max_ladder_length: int = -1
|
||
while(solutions_cnt<param_data["solutions"] ):
|
||
|
||
if len(args) > 2:
|
||
try:
|
||
max_ladder_length = current_ladder
|
||
if max_ladder_length < 1:
|
||
raise ValueError
|
||
except ValueError:
|
||
print('Max ladder length arg must be an integer (greater than zero)')
|
||
sys.exit(-1)
|
||
else:
|
||
start = perf_counter()
|
||
min_ladder = puzzle.calculate_minimum_ladder_length()
|
||
took = (perf_counter() - start) * 1000
|
||
if min_ladder is None:
|
||
print('Cannot solve \'%s\' to \'%s\'' % (first, second))
|
||
return save_to_file(param_data,solutions="", code = 6)
|
||
sys.exit(-1)
|
||
max_ladder_length = min_ladder + solutions_rounds
|
||
# print('Took %.2fms to determine minimum ladder length of %d' % (took, max_ladder_length))
|
||
|
||
if( max_ladder_length > param_data["max_ladder"] or max_ladder_length > 20):
|
||
print("max_ladder 超出限额")
|
||
break
|
||
|
||
solver = Solver(puzzle)
|
||
start = perf_counter()
|
||
# print("-- max_ladder_length (start solve):" + str(max_ladder_length))
|
||
solutions = solver.solve(max_ladder_length)
|
||
took = (perf_counter() - start) * 1000
|
||
if len(solutions) == 0:
|
||
print('Cannot solve \'%s\' to \'%s\' in ladder length %d (took %.2fms)' % (first, second, max_ladder_length, took))
|
||
sys.exit(-1)
|
||
slen = len(solutions)
|
||
solutions_rounds += 1
|
||
if slen < param_data["solutions"]:
|
||
print('Took %.2fms to find %d solutions (explored %d solutions)' % (took, slen, solver.explored_count))
|
||
print("not enough,pass to next ladder")
|
||
current_ladder += 1
|
||
max_ladder_length += 1
|
||
continue
|
||
else:
|
||
solutions_cnt += slen
|
||
print('Took %.2fms to find %d solutions (explored %d solutions)' % (took, slen, solver.explored_count))
|
||
solutions.sort()
|
||
for i in range(slen):
|
||
print('%d/%d %s' % (i + 1, slen, solutions[i]))
|
||
return save_to_file(param_data,solutions, code = 0)
|
||
sys.exit(-1)
|
||
|
||
return save_to_file(param_data,None,code = 1)
|
||
|
||
|
||
def save_to_file2(param_data, solutions , code):
|
||
# 获取前n个元素
|
||
# txt输出，已弃用
|
||
is_nlp = param_data["is_nlp"]
|
||
|
||
filepath = param_data["output_filepath"]
|
||
n = param_data["solutions"]
|
||
|
||
with open(filepath, 'a', encoding='utf-8') as file:
|
||
if code == 0:
|
||
if is_nlp :
|
||
title = "在一次Word Ladder中,当start word是" + param_data["start_word"] + "，end word是" + param_data["end_word"] + "时，至少可以得出以下" + str (n) + "条ladder路径："
|
||
else:
|
||
title = param_data["start_word"] + " " + param_data["end_word"]
|
||
file.write(title + '\n')
|
||
for i in range(n):
|
||
file.write(str(solutions[i]) + '\n')
|
||
elif code == 1:
|
||
if is_nlp :
|
||
title = "在一次Word Ladder中,当start word是" + param_data["start_word"] + "，end word是" + param_data["end_word"] + "时，没能解出" + str (n) + "条ladder路径"
|
||
else:
|
||
title = param_data["start_word"] + " " + param_data["end_word"]
|
||
file.write(title + '\n')
|
||
elif code == 2:
|
||
if is_nlp :
|
||
title = "在一次Word Ladder中,当start word是" + param_data["start_word"] + "，end word是" + param_data["end_word"] + "时，无法求解，因为" + param_data["start_word"] + "不在字典中"
|
||
else:
|
||
title = param_data["start_word"] + " " + param_data["end_word"]
|
||
file.write(title + '\n')
|
||
elif code == 3:
|
||
if is_nlp :
|
||
title = "在一次Word Ladder中,当start word是" + param_data["start_word"] + "，end word是" + param_data["end_word"] + "时，无法求解，因为" + param_data["start_word"] + "是一个island word"
|
||
else:
|
||
title = param_data["start_word"] + " " + param_data["end_word"]
|
||
file.write(title + '\n')
|
||
elif code == 4:
|
||
if is_nlp :
|
||
title = "在一次Word Ladder中,当start word是" + param_data["start_word"] + "，end word是" + param_data["end_word"] + "时，无法求解，因为" + param_data["end_word"] + "不在字典中"
|
||
else:
|
||
title = param_data["start_word"] + " " + param_data["end_word"]
|
||
file.write(title + '\n')
|
||
elif code == 5:
|
||
if is_nlp :
|
||
title = "在一次Word Ladder中,当start word是" + param_data["start_word"] + "，end word是" + param_data["end_word"] + "时，无法求解，因为" + param_data["end_word"] + "是一个island word"
|
||
else:
|
||
title = param_data["start_word"] + " " + param_data["end_word"]
|
||
file.write(title + '\n')
|
||
elif code == 6:
|
||
if is_nlp :
|
||
title = "在一次Word Ladder中,当start word是" + param_data["start_word"] + "，end word是" + param_data["end_word"] + "时，无法求解，因为没有路径可以从start word到end word"
|
||
else:
|
||
title = param_data["start_word"] + " " + param_data["end_word"]
|
||
file.write(title + '\n')
|
||
else:
|
||
if is_nlp :
|
||
title = "【出现未知异常，跳过此条训练数据1】"
|
||
else:
|
||
title = "error"
|
||
file.write(title + '\n')
|
||
|
||
def save_to_file(param_data, solutions, code):
|
||
is_nlp = param_data["is_nlp"]
|
||
filepath = param_data["output_filepath"]
|
||
start_word = param_data["start_word"]
|
||
end_word = param_data["end_word"]
|
||
n = param_data["solutions"]
|
||
|
||
# 定义基础描述
|
||
base_title = f"{start_word} {end_word}" if not is_nlp else f"在一次Word Ladder中,当start word是{start_word}，end word是{end_word}时，"
|
||
|
||
# 处理不同的code
|
||
if code == 0:
|
||
describe = f"{base_title}至少可以得出以下{n}条ladder路径：" if is_nlp else base_title
|
||
solutions_data = [str(solutions[i]) for i in range(n)] if solutions else []
|
||
describe = parse_word_ladder(solutions_data,start_word=start_word,end_word=end_word,n=n)
|
||
elif code == 1:
|
||
describe = f"{base_title}没能解出{n}条ladder路径" if is_nlp else base_title
|
||
solutions_data = "此题目无解"
|
||
elif code == 2:
|
||
describe = f"{base_title}无法求解，因为{start_word}不在字典中" if is_nlp else base_title
|
||
solutions_data = "此题目无解"
|
||
elif code == 3:
|
||
describe = f"{base_title}无法求解，因为{start_word}是一个island word" if is_nlp else base_title
|
||
solutions_data = "此题目无解"
|
||
elif code == 4:
|
||
describe = f"{base_title}无法求解，因为{end_word}不在字典中" if is_nlp else base_title
|
||
solutions_data = "此题目无解"
|
||
elif code == 5:
|
||
describe = f"{base_title}无法求解，因为{end_word}是一个island word" if is_nlp else base_title
|
||
solutions_data = "此题目无解"
|
||
elif code == 6:
|
||
describe = f"{base_title}无法求解，因为没有路径可以从start word到end word" if is_nlp else base_title
|
||
solutions_data = "此题目无解"
|
||
else:
|
||
describe = "【出现未知异常，跳过此条训练数据2】" if is_nlp else "error"
|
||
|
||
solutions_data = "此题目无解"
|
||
|
||
|
||
|
||
return solutions_data,describe
|
||
|
||
# 准备要写入的数据
|
||
record = {
|
||
"start_word": start_word,
|
||
"end_word": end_word,
|
||
"describe": describe,
|
||
"solutions": solutions_data
|
||
}
|
||
|
||
|
||
|
||
# 写入文件
|
||
with open(filepath, 'a', encoding='utf-8') as file:
|
||
file.write(json.dumps(record, ensure_ascii=False) + '\n')
|
||
|
||
def parse_word_ladder(data, start_word, end_word, n):
|
||
"""
|
||
解析 Word Ladder 数据为自然语言格式。
|
||
|
||
参数：
|
||
data (list of str): 包含每条路径的字符串列表。
|
||
start_word (str): 初始单词。
|
||
end_word (str): 结束单词。
|
||
n (int): 数据中路径的数量。
|
||
|
||
返回：
|
||
str: 格式化后的自然语言文本。
|
||
"""
|
||
nl_output = []
|
||
nl_output.append(f"在一次Word Ladder中，从{start_word}到{end_word}至少可以得出以下{n}条ladder路径：")
|
||
|
||
for i, path in enumerate(data[:n], start=1):
|
||
words = path.strip('[]').split(',')
|
||
words = [word.strip() for word in words]
|
||
changes = []
|
||
|
||
# 解析路径中的每步变化
|
||
for j in range(1, len(words)):
|
||
prev_word, curr_word = words[j - 1], words[j]
|
||
for k, (prev_char, curr_char) in enumerate(zip(prev_word, curr_word)):
|
||
if prev_char != curr_char:
|
||
changes.append(f"将第{k + 1}个字母由{prev_char}更换成{curr_char}")
|
||
|
||
# 构建路径描述
|
||
path_description = f"第{i}条：{words[0]}" + ''.join([f"，{change}，得到{words[j]}" for j, change in enumerate(changes, start=1)])
|
||
nl_output.append(path_description + '；')
|
||
|
||
return '\n'.join(nl_output)
|
||
|
||
# 完整的原solver
|
||
if __name__ == '__main__':
|
||
solver_main() |