Shortcuts on this page
r m x toggle line displays
j k next/prev highlighted chunk
0 (zero) top of page
1 (one) first highlighted chunk
1import os
2import re
5_default_seps = os.sep + str(os.altsep) * bool(os.altsep)
8class Translator:
9 """
10 >>> Translator('xyz')
11 Traceback (most recent call last):
12 ...
13 AssertionError: Invalid separators
15 >>> Translator('')
16 Traceback (most recent call last):
17 ...
18 AssertionError: Invalid separators
19 """
21 seps: str
23 def __init__(self, seps: str = _default_seps):
24 assert seps and set(seps) <= set(_default_seps), "Invalid separators"
25 self.seps = seps
27 def translate(self, pattern):
28 """
29 Given a glob pattern, produce a regex that matches it.
30 """
31 return self.extend(self.translate_core(pattern))
33 def extend(self, pattern):
34 r"""
35 Extend regex for pattern-wide concerns.
37 Apply '(?s:)' to create a non-matching group that
38 matches newlines (valid on Unix).
40 Append '\Z' to imply fullmatch even when match is used.
41 """
42 return rf'(?s:{pattern})\Z'
44 def translate_core(self, pattern):
45 r"""
46 Given a glob pattern, produce a regex that matches it.
48 >>> t = Translator()
49 >>> t.translate_core('*.txt').replace('\\\\', '')
50 '[^/]*\\.txt'
51 >>> t.translate_core('a?txt')
52 'a[^/]txt'
53 >>> t.translate_core('**/*').replace('\\\\', '')
54 '.*/[^/][^/]*'
55 """
56 self.restrict_rglob(pattern)
57 return ''.join(map(self.replace, separate(self.star_not_empty(pattern))))
59 def replace(self, match):
60 """
61 Perform the replacements for a match from :func:`separate`.
62 """
63 return match.group('set') or (
64 re.escape(match.group(0))
65 .replace('\\*\\*', r'.*')
66 .replace('\\*', rf'[^{re.escape(self.seps)}]*')
67 .replace('\\?', r'[^/]')
68 )
70 def restrict_rglob(self, pattern):
71 """
72 Raise ValueError if ** appears in anything but a full path segment.
74 >>> Translator().translate('**foo')
75 Traceback (most recent call last):
76 ...
77 ValueError: ** must appear alone in a path segment
78 """
79 seps_pattern = rf'[{re.escape(self.seps)}]+'
80 segments = re.split(seps_pattern, pattern)
81 if any('**' in segment and segment != '**' for segment in segments):
82 raise ValueError("** must appear alone in a path segment")
84 def star_not_empty(self, pattern):
85 """
86 Ensure that * will not match an empty segment.
87 """
89 def handle_segment(match):
90 segment = match.group(0)
91 return '?*' if segment == '*' else segment
93 not_seps_pattern = rf'[^{re.escape(self.seps)}]+'
94 return re.sub(not_seps_pattern, handle_segment, pattern)
97def separate(pattern):
98 """
99 Separate out character sets to avoid translating their contents.
101 >>> [m.group(0) for m in separate('*.txt')]
102 ['*.txt']
103 >>> [m.group(0) for m in separate('a[?]txt')]
104 ['a', '[?]', 'txt']
105 """
106 return re.finditer(r'([^\[]+)|(?P<set>[\[].*?[\]])|([\[][^\]]*$)', pattern)