Pylint 抱怨 NLTK 包的特定文件的 R0401 错误代码循环导入,例如
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk -> nltk.internals)
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk.corpus -> nltk.tokenize -> nltk.tokenize.punkt -> nltk.probability)
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk.corpus -> nltk.tokenize -> nltk.tokenize.texttiling)
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk.draw.tree -> nltk.tree)
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk.tree -> nltk.treeprettyprinter)
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk.grammar -> nltk.parse.pchart)
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk.stem -> nltk.stem.porter)
nltk/nltk/ccg/lexicon.py:1: [R0401(cyclic-import), ] Cyclic import (nltk.classify.maxent -> nltk.classify.tadm)
完整列表在 https://github.com/nltk/nltk/issues/2113
但看看进口:
from __future__ import unicode_literals from
import re
from collections import defaultdict
from nltk.ccg.api import PrimitiveCategory, Direction, CCGVar, FunctionalCategory
from nltk.compat import python_2_unicode_compatible
from nltk.internals import deprecated
from nltk.sem.logic import *
但看看nltk.internals
https://github.com/nltk/nltk/blob/develop/nltk/internals.py,没有任何周期性导入指向nltk.ccg.lexicon
:
from __future__ import print_function
import subprocess
import os
import fnmatch
import re
import warnings
import textwrap
import types
import sys
import stat
import locale
# Use the c version of ElementTree, which is faster, if possible:
try:
from xml.etree import cElementTree as ElementTree
except ImportError:
from xml.etree import ElementTree
from six import string_types
from nltk import __file__
from nltk import compat
R0401(cyclic-import)
消息是什么意思?
看看nltk.ccg.lexicon.py
和nltk.internals.py
没有任何循环的进口,那么如何解决这个问题。
不知道为什么 pylint 在nltk/nltk/ccg/lexicon.py
文件中报告这些,但循环导入本身可以在错误消息的右侧看到。
取第一个错误:Cyclic import (nltk -> nltk.internals)
:根__init__.py有一个从nltk.internals
导入,internals.py 有一个从包根导入,这显然是一个循环导入。