Show More
@@ -1,505 +1,489 b'' | |||||
1 | # pycompat.py - portability shim for python 3 |
|
1 | # pycompat.py - portability shim for python 3 | |
2 | # |
|
2 | # | |
3 | # This software may be used and distributed according to the terms of the |
|
3 | # This software may be used and distributed according to the terms of the | |
4 | # GNU General Public License version 2 or any later version. |
|
4 | # GNU General Public License version 2 or any later version. | |
5 |
|
5 | |||
6 | """Mercurial portability shim for python 3. |
|
6 | """Mercurial portability shim for python 3. | |
7 |
|
7 | |||
8 | This contains aliases to hide python version-specific details from the core. |
|
8 | This contains aliases to hide python version-specific details from the core. | |
9 | """ |
|
9 | """ | |
10 |
|
10 | |||
11 |
|
11 | |||
12 | import builtins |
|
12 | import builtins | |
13 | import codecs |
|
13 | import codecs | |
14 | import concurrent.futures as futures |
|
14 | import concurrent.futures as futures | |
15 | import functools |
|
|||
16 | import getopt |
|
15 | import getopt | |
17 | import http.client as httplib |
|
16 | import http.client as httplib | |
18 | import http.cookiejar as cookielib |
|
17 | import http.cookiejar as cookielib | |
19 | import inspect |
|
18 | import inspect | |
20 | import io |
|
19 | import io | |
21 | import json |
|
20 | import json | |
22 | import os |
|
21 | import os | |
23 | import queue |
|
22 | import queue | |
24 | import shlex |
|
23 | import shlex | |
25 | import socketserver |
|
24 | import socketserver | |
26 | import struct |
|
25 | import struct | |
27 | import sys |
|
26 | import sys | |
28 | import tempfile |
|
27 | import tempfile | |
29 | import xmlrpc.client as xmlrpclib |
|
28 | import xmlrpc.client as xmlrpclib | |
30 |
|
29 | |||
31 | from typing import ( |
|
30 | from typing import ( | |
32 | Any, |
|
31 | Any, | |
33 | AnyStr, |
|
32 | AnyStr, | |
34 | BinaryIO, |
|
33 | BinaryIO, | |
35 | Callable, |
|
34 | Callable, | |
36 | Dict, |
|
35 | Dict, | |
37 | Iterable, |
|
36 | Iterable, | |
38 | Iterator, |
|
37 | Iterator, | |
39 | List, |
|
38 | List, | |
40 | Mapping, |
|
39 | Mapping, | |
41 | NoReturn, |
|
40 | NoReturn, | |
42 | Optional, |
|
41 | Optional, | |
43 | Sequence, |
|
42 | Sequence, | |
44 | Tuple, |
|
43 | Tuple, | |
45 | Type, |
|
44 | Type, | |
46 | TypeVar, |
|
45 | TypeVar, | |
47 | cast, |
|
46 | cast, | |
48 | overload, |
|
47 | overload, | |
49 | ) |
|
48 | ) | |
50 |
|
49 | |||
51 | ispy3 = sys.version_info[0] >= 3 |
|
50 | ispy3 = sys.version_info[0] >= 3 | |
52 | ispypy = '__pypy__' in sys.builtin_module_names |
|
51 | ispypy = '__pypy__' in sys.builtin_module_names | |
53 | TYPE_CHECKING = False |
|
52 | TYPE_CHECKING = False | |
54 |
|
53 | |||
55 | if not globals(): # hide this from non-pytype users |
|
54 | if not globals(): # hide this from non-pytype users | |
56 | import typing |
|
55 | import typing | |
57 |
|
56 | |||
58 | TYPE_CHECKING = typing.TYPE_CHECKING |
|
57 | TYPE_CHECKING = typing.TYPE_CHECKING | |
59 |
|
58 | |||
60 | _GetOptResult = Tuple[List[Tuple[bytes, bytes]], List[bytes]] |
|
59 | _GetOptResult = Tuple[List[Tuple[bytes, bytes]], List[bytes]] | |
61 | _T0 = TypeVar('_T0') |
|
60 | _T0 = TypeVar('_T0') | |
62 | _T1 = TypeVar('_T1') |
|
61 | _T1 = TypeVar('_T1') | |
63 | _S = TypeVar('_S') |
|
62 | _S = TypeVar('_S') | |
64 | _Tbytestr = TypeVar('_Tbytestr', bound='bytestr') |
|
63 | _Tbytestr = TypeVar('_Tbytestr', bound='bytestr') | |
65 |
|
64 | |||
66 |
|
65 | |||
67 | def future_set_exception_info(f, exc_info): |
|
66 | def future_set_exception_info(f, exc_info): | |
68 | f.set_exception(exc_info[0]) |
|
67 | f.set_exception(exc_info[0]) | |
69 |
|
68 | |||
70 |
|
69 | |||
71 | FileNotFoundError = builtins.FileNotFoundError |
|
70 | FileNotFoundError = builtins.FileNotFoundError | |
72 |
|
71 | |||
73 |
|
72 | |||
74 | def identity(a: _T0) -> _T0: |
|
73 | def identity(a: _T0) -> _T0: | |
75 | return a |
|
74 | return a | |
76 |
|
75 | |||
77 |
|
76 | |||
78 | def _rapply(f, xs): |
|
77 | def _rapply(f, xs): | |
79 | if xs is None: |
|
78 | if xs is None: | |
80 | # assume None means non-value of optional data |
|
79 | # assume None means non-value of optional data | |
81 | return xs |
|
80 | return xs | |
82 | if isinstance(xs, (list, set, tuple)): |
|
81 | if isinstance(xs, (list, set, tuple)): | |
83 | return type(xs)(_rapply(f, x) for x in xs) |
|
82 | return type(xs)(_rapply(f, x) for x in xs) | |
84 | if isinstance(xs, dict): |
|
83 | if isinstance(xs, dict): | |
85 | return type(xs)((_rapply(f, k), _rapply(f, v)) for k, v in xs.items()) |
|
84 | return type(xs)((_rapply(f, k), _rapply(f, v)) for k, v in xs.items()) | |
86 | return f(xs) |
|
85 | return f(xs) | |
87 |
|
86 | |||
88 |
|
87 | |||
89 | def rapply(f, xs): |
|
88 | def rapply(f, xs): | |
90 | """Apply function recursively to every item preserving the data structure |
|
89 | """Apply function recursively to every item preserving the data structure | |
91 |
|
90 | |||
92 | >>> def f(x): |
|
91 | >>> def f(x): | |
93 | ... return 'f(%s)' % x |
|
92 | ... return 'f(%s)' % x | |
94 | >>> rapply(f, None) is None |
|
93 | >>> rapply(f, None) is None | |
95 | True |
|
94 | True | |
96 | >>> rapply(f, 'a') |
|
95 | >>> rapply(f, 'a') | |
97 | 'f(a)' |
|
96 | 'f(a)' | |
98 | >>> rapply(f, {'a'}) == {'f(a)'} |
|
97 | >>> rapply(f, {'a'}) == {'f(a)'} | |
99 | True |
|
98 | True | |
100 | >>> rapply(f, ['a', 'b', None, {'c': 'd'}, []]) |
|
99 | >>> rapply(f, ['a', 'b', None, {'c': 'd'}, []]) | |
101 | ['f(a)', 'f(b)', None, {'f(c)': 'f(d)'}, []] |
|
100 | ['f(a)', 'f(b)', None, {'f(c)': 'f(d)'}, []] | |
102 |
|
101 | |||
103 | >>> xs = [object()] |
|
102 | >>> xs = [object()] | |
104 | >>> rapply(identity, xs) is xs |
|
103 | >>> rapply(identity, xs) is xs | |
105 | True |
|
104 | True | |
106 | """ |
|
105 | """ | |
107 | if f is identity: |
|
106 | if f is identity: | |
108 | # fast path mainly for py2 |
|
107 | # fast path mainly for py2 | |
109 | return xs |
|
108 | return xs | |
110 | return _rapply(f, xs) |
|
109 | return _rapply(f, xs) | |
111 |
|
110 | |||
112 |
|
111 | |||
113 | if os.name == r'nt': |
|
112 | if os.name == r'nt': | |
114 | # MBCS (or ANSI) filesystem encoding must be used as before. |
|
113 | # MBCS (or ANSI) filesystem encoding must be used as before. | |
115 | # Otherwise non-ASCII filenames in existing repositories would be |
|
114 | # Otherwise non-ASCII filenames in existing repositories would be | |
116 | # corrupted. |
|
115 | # corrupted. | |
117 | # This must be set once prior to any fsencode/fsdecode calls. |
|
116 | # This must be set once prior to any fsencode/fsdecode calls. | |
118 | sys._enablelegacywindowsfsencoding() # pytype: disable=module-attr |
|
117 | sys._enablelegacywindowsfsencoding() # pytype: disable=module-attr | |
119 |
|
118 | |||
120 | fsencode = os.fsencode |
|
119 | fsencode = os.fsencode | |
121 | fsdecode = os.fsdecode |
|
120 | fsdecode = os.fsdecode | |
122 | oscurdir: bytes = os.curdir.encode('ascii') |
|
121 | oscurdir: bytes = os.curdir.encode('ascii') | |
123 | oslinesep: bytes = os.linesep.encode('ascii') |
|
122 | oslinesep: bytes = os.linesep.encode('ascii') | |
124 | osname: bytes = os.name.encode('ascii') |
|
123 | osname: bytes = os.name.encode('ascii') | |
125 | ospathsep: bytes = os.pathsep.encode('ascii') |
|
124 | ospathsep: bytes = os.pathsep.encode('ascii') | |
126 | ospardir: bytes = os.pardir.encode('ascii') |
|
125 | ospardir: bytes = os.pardir.encode('ascii') | |
127 | ossep: bytes = os.sep.encode('ascii') |
|
126 | ossep: bytes = os.sep.encode('ascii') | |
128 | osaltsep: Optional[bytes] = os.altsep.encode('ascii') if os.altsep else None |
|
127 | osaltsep: Optional[bytes] = os.altsep.encode('ascii') if os.altsep else None | |
129 | osdevnull: bytes = os.devnull.encode('ascii') |
|
128 | osdevnull: bytes = os.devnull.encode('ascii') | |
130 |
|
129 | |||
131 | sysplatform: bytes = sys.platform.encode('ascii') |
|
130 | sysplatform: bytes = sys.platform.encode('ascii') | |
132 | sysexecutable: bytes = os.fsencode(sys.executable) if sys.executable else b'' |
|
131 | sysexecutable: bytes = os.fsencode(sys.executable) if sys.executable else b'' | |
133 |
|
132 | |||
134 |
|
133 | |||
135 | if TYPE_CHECKING: |
|
134 | if TYPE_CHECKING: | |
136 |
|
135 | |||
137 | @overload |
|
136 | @overload | |
138 | def maplist(f: Callable[[_T0], _S], arg: Iterable[_T0]) -> List[_S]: |
|
137 | def maplist(f: Callable[[_T0], _S], arg: Iterable[_T0]) -> List[_S]: | |
139 | ... |
|
138 | ... | |
140 |
|
139 | |||
141 | @overload |
|
140 | @overload | |
142 | def maplist( |
|
141 | def maplist( | |
143 | f: Callable[[_T0, _T1], _S], arg1: Iterable[_T0], arg2: Iterable[_T1] |
|
142 | f: Callable[[_T0, _T1], _S], arg1: Iterable[_T0], arg2: Iterable[_T1] | |
144 | ) -> List[_S]: |
|
143 | ) -> List[_S]: | |
145 | ... |
|
144 | ... | |
146 |
|
145 | |||
147 |
|
146 | |||
148 | def maplist(f, *args): |
|
147 | def maplist(f, *args): | |
149 | return list(map(f, *args)) |
|
148 | return list(map(f, *args)) | |
150 |
|
149 | |||
151 |
|
150 | |||
152 | def rangelist(*args) -> List[int]: |
|
151 | def rangelist(*args) -> List[int]: | |
153 | return list(range(*args)) |
|
152 | return list(range(*args)) | |
154 |
|
153 | |||
155 |
|
154 | |||
156 | def ziplist(*args): |
|
155 | def ziplist(*args): | |
157 | return list(zip(*args)) |
|
156 | return list(zip(*args)) | |
158 |
|
157 | |||
159 |
|
158 | |||
160 | rawinput = input |
|
159 | rawinput = input | |
161 | getargspec = inspect.getfullargspec |
|
160 | getargspec = inspect.getfullargspec | |
162 |
|
161 | |||
163 | long = int |
|
162 | long = int | |
164 |
|
163 | |||
165 | if builtins.getattr(sys, 'argv', None) is not None: |
|
164 | if builtins.getattr(sys, 'argv', None) is not None: | |
166 | # On POSIX, the char** argv array is converted to Python str using |
|
165 | # On POSIX, the char** argv array is converted to Python str using | |
167 | # Py_DecodeLocale(). The inverse of this is Py_EncodeLocale(), which |
|
166 | # Py_DecodeLocale(). The inverse of this is Py_EncodeLocale(), which | |
168 | # isn't directly callable from Python code. In practice, os.fsencode() |
|
167 | # isn't directly callable from Python code. In practice, os.fsencode() | |
169 | # can be used instead (this is recommended by Python's documentation |
|
168 | # can be used instead (this is recommended by Python's documentation | |
170 | # for sys.argv). |
|
169 | # for sys.argv). | |
171 | # |
|
170 | # | |
172 | # On Windows, the wchar_t **argv is passed into the interpreter as-is. |
|
171 | # On Windows, the wchar_t **argv is passed into the interpreter as-is. | |
173 | # Like POSIX, we need to emulate what Py_EncodeLocale() would do. But |
|
172 | # Like POSIX, we need to emulate what Py_EncodeLocale() would do. But | |
174 | # there's an additional wrinkle. What we really want to access is the |
|
173 | # there's an additional wrinkle. What we really want to access is the | |
175 | # ANSI codepage representation of the arguments, as this is what |
|
174 | # ANSI codepage representation of the arguments, as this is what | |
176 | # `int main()` would receive if Python 3 didn't define `int wmain()` |
|
175 | # `int main()` would receive if Python 3 didn't define `int wmain()` | |
177 | # (this is how Python 2 worked). To get that, we encode with the mbcs |
|
176 | # (this is how Python 2 worked). To get that, we encode with the mbcs | |
178 | # encoding, which will pass CP_ACP to the underlying Windows API to |
|
177 | # encoding, which will pass CP_ACP to the underlying Windows API to | |
179 | # produce bytes. |
|
178 | # produce bytes. | |
180 | sysargv: List[bytes] = [] |
|
179 | sysargv: List[bytes] = [] | |
181 | if os.name == r'nt': |
|
180 | if os.name == r'nt': | |
182 | sysargv = [a.encode("mbcs", "ignore") for a in sys.argv] |
|
181 | sysargv = [a.encode("mbcs", "ignore") for a in sys.argv] | |
183 | else: |
|
182 | else: | |
184 | sysargv = [fsencode(a) for a in sys.argv] |
|
183 | sysargv = [fsencode(a) for a in sys.argv] | |
185 |
|
184 | |||
186 | bytechr = struct.Struct('>B').pack |
|
185 | bytechr = struct.Struct('>B').pack | |
187 | byterepr = b'%r'.__mod__ |
|
186 | byterepr = b'%r'.__mod__ | |
188 |
|
187 | |||
189 |
|
188 | |||
190 | class bytestr(bytes): |
|
189 | class bytestr(bytes): | |
191 | """A bytes which mostly acts as a Python 2 str |
|
190 | """A bytes which mostly acts as a Python 2 str | |
192 |
|
191 | |||
193 | >>> bytestr(), bytestr(bytearray(b'foo')), bytestr(u'ascii'), bytestr(1) |
|
192 | >>> bytestr(), bytestr(bytearray(b'foo')), bytestr(u'ascii'), bytestr(1) | |
194 | ('', 'foo', 'ascii', '1') |
|
193 | ('', 'foo', 'ascii', '1') | |
195 | >>> s = bytestr(b'foo') |
|
194 | >>> s = bytestr(b'foo') | |
196 | >>> assert s is bytestr(s) |
|
195 | >>> assert s is bytestr(s) | |
197 |
|
196 | |||
198 | __bytes__() should be called if provided: |
|
197 | __bytes__() should be called if provided: | |
199 |
|
198 | |||
200 | >>> class bytesable: |
|
199 | >>> class bytesable: | |
201 | ... def __bytes__(self): |
|
200 | ... def __bytes__(self): | |
202 | ... return b'bytes' |
|
201 | ... return b'bytes' | |
203 | >>> bytestr(bytesable()) |
|
202 | >>> bytestr(bytesable()) | |
204 | 'bytes' |
|
203 | 'bytes' | |
205 |
|
204 | |||
206 | There's no implicit conversion from non-ascii str as its encoding is |
|
205 | There's no implicit conversion from non-ascii str as its encoding is | |
207 | unknown: |
|
206 | unknown: | |
208 |
|
207 | |||
209 | >>> bytestr(chr(0x80)) # doctest: +ELLIPSIS |
|
208 | >>> bytestr(chr(0x80)) # doctest: +ELLIPSIS | |
210 | Traceback (most recent call last): |
|
209 | Traceback (most recent call last): | |
211 | ... |
|
210 | ... | |
212 | UnicodeEncodeError: ... |
|
211 | UnicodeEncodeError: ... | |
213 |
|
212 | |||
214 | Comparison between bytestr and bytes should work: |
|
213 | Comparison between bytestr and bytes should work: | |
215 |
|
214 | |||
216 | >>> assert bytestr(b'foo') == b'foo' |
|
215 | >>> assert bytestr(b'foo') == b'foo' | |
217 | >>> assert b'foo' == bytestr(b'foo') |
|
216 | >>> assert b'foo' == bytestr(b'foo') | |
218 | >>> assert b'f' in bytestr(b'foo') |
|
217 | >>> assert b'f' in bytestr(b'foo') | |
219 | >>> assert bytestr(b'f') in b'foo' |
|
218 | >>> assert bytestr(b'f') in b'foo' | |
220 |
|
219 | |||
221 | Sliced elements should be bytes, not integer: |
|
220 | Sliced elements should be bytes, not integer: | |
222 |
|
221 | |||
223 | >>> s[1], s[:2] |
|
222 | >>> s[1], s[:2] | |
224 | (b'o', b'fo') |
|
223 | (b'o', b'fo') | |
225 | >>> list(s), list(reversed(s)) |
|
224 | >>> list(s), list(reversed(s)) | |
226 | ([b'f', b'o', b'o'], [b'o', b'o', b'f']) |
|
225 | ([b'f', b'o', b'o'], [b'o', b'o', b'f']) | |
227 |
|
226 | |||
228 | As bytestr type isn't propagated across operations, you need to cast |
|
227 | As bytestr type isn't propagated across operations, you need to cast | |
229 | bytes to bytestr explicitly: |
|
228 | bytes to bytestr explicitly: | |
230 |
|
229 | |||
231 | >>> s = bytestr(b'foo').upper() |
|
230 | >>> s = bytestr(b'foo').upper() | |
232 | >>> t = bytestr(s) |
|
231 | >>> t = bytestr(s) | |
233 | >>> s[0], t[0] |
|
232 | >>> s[0], t[0] | |
234 | (70, b'F') |
|
233 | (70, b'F') | |
235 |
|
234 | |||
236 | Be careful to not pass a bytestr object to a function which expects |
|
235 | Be careful to not pass a bytestr object to a function which expects | |
237 | bytearray-like behavior. |
|
236 | bytearray-like behavior. | |
238 |
|
237 | |||
239 | >>> t = bytes(t) # cast to bytes |
|
238 | >>> t = bytes(t) # cast to bytes | |
240 | >>> assert type(t) is bytes |
|
239 | >>> assert type(t) is bytes | |
241 | """ |
|
240 | """ | |
242 |
|
241 | |||
243 | # Trick pytype into not demanding Iterable[int] be passed to __new__(), |
|
242 | # Trick pytype into not demanding Iterable[int] be passed to __new__(), | |
244 | # since the appropriate bytes format is done internally. |
|
243 | # since the appropriate bytes format is done internally. | |
245 | # |
|
244 | # | |
246 | # https://github.com/google/pytype/issues/500 |
|
245 | # https://github.com/google/pytype/issues/500 | |
247 | if TYPE_CHECKING: |
|
246 | if TYPE_CHECKING: | |
248 |
|
247 | |||
249 | def __init__(self, s: object = b'') -> None: |
|
248 | def __init__(self, s: object = b'') -> None: | |
250 | pass |
|
249 | pass | |
251 |
|
250 | |||
252 | def __new__(cls: Type[_Tbytestr], s: object = b'') -> _Tbytestr: |
|
251 | def __new__(cls: Type[_Tbytestr], s: object = b'') -> _Tbytestr: | |
253 | if isinstance(s, bytestr): |
|
252 | if isinstance(s, bytestr): | |
254 | return s |
|
253 | return s | |
255 | if not isinstance( |
|
254 | if not isinstance( | |
256 | s, (bytes, bytearray) |
|
255 | s, (bytes, bytearray) | |
257 | ) and not builtins.hasattr( # hasattr-py3-only |
|
256 | ) and not builtins.hasattr( # hasattr-py3-only | |
258 | s, u'__bytes__' |
|
257 | s, u'__bytes__' | |
259 | ): |
|
258 | ): | |
260 | s = str(s).encode('ascii') |
|
259 | s = str(s).encode('ascii') | |
261 | return bytes.__new__(cls, s) |
|
260 | return bytes.__new__(cls, s) | |
262 |
|
261 | |||
263 | # The base class uses `int` return in py3, but the point of this class is to |
|
262 | # The base class uses `int` return in py3, but the point of this class is to | |
264 | # behave like py2. |
|
263 | # behave like py2. | |
265 | def __getitem__(self, key) -> bytes: # pytype: disable=signature-mismatch |
|
264 | def __getitem__(self, key) -> bytes: # pytype: disable=signature-mismatch | |
266 | s = bytes.__getitem__(self, key) |
|
265 | s = bytes.__getitem__(self, key) | |
267 | if not isinstance(s, bytes): |
|
266 | if not isinstance(s, bytes): | |
268 | s = bytechr(s) |
|
267 | s = bytechr(s) | |
269 | return s |
|
268 | return s | |
270 |
|
269 | |||
271 | # The base class expects `Iterator[int]` return in py3, but the point of |
|
270 | # The base class expects `Iterator[int]` return in py3, but the point of | |
272 | # this class is to behave like py2. |
|
271 | # this class is to behave like py2. | |
273 | def __iter__(self) -> Iterator[bytes]: # pytype: disable=signature-mismatch |
|
272 | def __iter__(self) -> Iterator[bytes]: # pytype: disable=signature-mismatch | |
274 | return iterbytestr(bytes.__iter__(self)) |
|
273 | return iterbytestr(bytes.__iter__(self)) | |
275 |
|
274 | |||
276 | def __repr__(self) -> str: |
|
275 | def __repr__(self) -> str: | |
277 | return bytes.__repr__(self)[1:] # drop b'' |
|
276 | return bytes.__repr__(self)[1:] # drop b'' | |
278 |
|
277 | |||
279 |
|
278 | |||
280 | def iterbytestr(s: Iterable[int]) -> Iterator[bytes]: |
|
279 | def iterbytestr(s: Iterable[int]) -> Iterator[bytes]: | |
281 | """Iterate bytes as if it were a str object of Python 2""" |
|
280 | """Iterate bytes as if it were a str object of Python 2""" | |
282 | return map(bytechr, s) |
|
281 | return map(bytechr, s) | |
283 |
|
282 | |||
284 |
|
283 | |||
285 | if TYPE_CHECKING: |
|
284 | if TYPE_CHECKING: | |
286 |
|
285 | |||
287 | @overload |
|
286 | @overload | |
288 | def maybebytestr(s: bytes) -> bytestr: |
|
287 | def maybebytestr(s: bytes) -> bytestr: | |
289 | ... |
|
288 | ... | |
290 |
|
289 | |||
291 | @overload |
|
290 | @overload | |
292 | def maybebytestr(s: _T0) -> _T0: |
|
291 | def maybebytestr(s: _T0) -> _T0: | |
293 | ... |
|
292 | ... | |
294 |
|
293 | |||
295 |
|
294 | |||
296 | def maybebytestr(s): |
|
295 | def maybebytestr(s): | |
297 | """Promote bytes to bytestr""" |
|
296 | """Promote bytes to bytestr""" | |
298 | if isinstance(s, bytes): |
|
297 | if isinstance(s, bytes): | |
299 | return bytestr(s) |
|
298 | return bytestr(s) | |
300 | return s |
|
299 | return s | |
301 |
|
300 | |||
302 |
|
301 | |||
303 | def sysbytes(s: AnyStr) -> bytes: |
|
302 | def sysbytes(s: AnyStr) -> bytes: | |
304 | """Convert an internal str (e.g. keyword, __doc__) back to bytes |
|
303 | """Convert an internal str (e.g. keyword, __doc__) back to bytes | |
305 |
|
304 | |||
306 | This never raises UnicodeEncodeError, but only ASCII characters |
|
305 | This never raises UnicodeEncodeError, but only ASCII characters | |
307 | can be round-trip by sysstr(sysbytes(s)). |
|
306 | can be round-trip by sysstr(sysbytes(s)). | |
308 | """ |
|
307 | """ | |
309 | if isinstance(s, bytes): |
|
308 | if isinstance(s, bytes): | |
310 | return s |
|
309 | return s | |
311 | return s.encode('utf-8') |
|
310 | return s.encode('utf-8') | |
312 |
|
311 | |||
313 |
|
312 | |||
314 | def sysstr(s: AnyStr) -> str: |
|
313 | def sysstr(s: AnyStr) -> str: | |
315 | """Return a keyword str to be passed to Python functions such as |
|
314 | """Return a keyword str to be passed to Python functions such as | |
316 | getattr() and str.encode() |
|
315 | getattr() and str.encode() | |
317 |
|
316 | |||
318 | This never raises UnicodeDecodeError. Non-ascii characters are |
|
317 | This never raises UnicodeDecodeError. Non-ascii characters are | |
319 | considered invalid and mapped to arbitrary but unique code points |
|
318 | considered invalid and mapped to arbitrary but unique code points | |
320 | such that 'sysstr(a) != sysstr(b)' for all 'a != b'. |
|
319 | such that 'sysstr(a) != sysstr(b)' for all 'a != b'. | |
321 | """ |
|
320 | """ | |
322 | if isinstance(s, builtins.str): |
|
321 | if isinstance(s, builtins.str): | |
323 | return s |
|
322 | return s | |
324 | return s.decode('latin-1') |
|
323 | return s.decode('latin-1') | |
325 |
|
324 | |||
326 |
|
325 | |||
327 | def strurl(url: AnyStr) -> str: |
|
326 | def strurl(url: AnyStr) -> str: | |
328 | """Converts a bytes url back to str""" |
|
327 | """Converts a bytes url back to str""" | |
329 | if isinstance(url, bytes): |
|
328 | if isinstance(url, bytes): | |
330 | return url.decode('ascii') |
|
329 | return url.decode('ascii') | |
331 | return url |
|
330 | return url | |
332 |
|
331 | |||
333 |
|
332 | |||
334 | def bytesurl(url: AnyStr) -> bytes: |
|
333 | def bytesurl(url: AnyStr) -> bytes: | |
335 | """Converts a str url to bytes by encoding in ascii""" |
|
334 | """Converts a str url to bytes by encoding in ascii""" | |
336 | if isinstance(url, str): |
|
335 | if isinstance(url, str): | |
337 | return url.encode('ascii') |
|
336 | return url.encode('ascii') | |
338 | return url |
|
337 | return url | |
339 |
|
338 | |||
340 |
|
339 | |||
341 | def raisewithtb(exc: BaseException, tb) -> NoReturn: |
|
340 | def raisewithtb(exc: BaseException, tb) -> NoReturn: | |
342 | """Raise exception with the given traceback""" |
|
341 | """Raise exception with the given traceback""" | |
343 | raise exc.with_traceback(tb) |
|
342 | raise exc.with_traceback(tb) | |
344 |
|
343 | |||
345 |
|
344 | |||
346 | def getdoc(obj: object) -> Optional[bytes]: |
|
345 | def getdoc(obj: object) -> Optional[bytes]: | |
347 | """Get docstring as bytes; may be None so gettext() won't confuse it |
|
346 | """Get docstring as bytes; may be None so gettext() won't confuse it | |
348 | with _('')""" |
|
347 | with _('')""" | |
349 | doc = builtins.getattr(obj, '__doc__', None) |
|
348 | doc = builtins.getattr(obj, '__doc__', None) | |
350 | if doc is None: |
|
349 | if doc is None: | |
351 | return doc |
|
350 | return doc | |
352 | return sysbytes(doc) |
|
351 | return sysbytes(doc) | |
353 |
|
352 | |||
354 |
|
353 | |||
355 | def _wrapattrfunc(f): |
|
|||
356 | @functools.wraps(f) |
|
|||
357 | def w(object, name, *args): |
|
|||
358 | if isinstance(name, bytes): |
|
|||
359 | from . import util |
|
|||
360 |
|
||||
361 | msg = b'function "%s" take `str` as argument, not `bytes`' |
|
|||
362 | fname = f.__name__.encode('ascii') |
|
|||
363 | msg %= fname |
|
|||
364 | util.nouideprecwarn(msg, b"6.6", stacklevel=2) |
|
|||
365 | return f(object, sysstr(name), *args) |
|
|||
366 |
|
||||
367 | return w |
|
|||
368 |
|
||||
369 |
|
||||
370 | # these wrappers are automagically imported by hgloader |
|
354 | # these wrappers are automagically imported by hgloader | |
371 |
delattr = |
|
355 | delattr = builtins.delattr | |
372 |
getattr = |
|
356 | getattr = builtins.getattr | |
373 |
hasattr = |
|
357 | hasattr = builtins.hasattr | |
374 |
setattr = |
|
358 | setattr = builtins.setattr | |
375 | xrange = builtins.range |
|
359 | xrange = builtins.range | |
376 | unicode = str |
|
360 | unicode = str | |
377 |
|
361 | |||
378 |
|
362 | |||
379 | def open( |
|
363 | def open( | |
380 | name, |
|
364 | name, | |
381 | mode: AnyStr = b'r', |
|
365 | mode: AnyStr = b'r', | |
382 | buffering: int = -1, |
|
366 | buffering: int = -1, | |
383 | encoding: Optional[str] = None, |
|
367 | encoding: Optional[str] = None, | |
384 | ) -> Any: |
|
368 | ) -> Any: | |
385 | # TODO: assert binary mode, and cast result to BinaryIO? |
|
369 | # TODO: assert binary mode, and cast result to BinaryIO? | |
386 | return builtins.open(name, sysstr(mode), buffering, encoding) |
|
370 | return builtins.open(name, sysstr(mode), buffering, encoding) | |
387 |
|
371 | |||
388 |
|
372 | |||
389 |
safehasattr = |
|
373 | safehasattr = builtins.hasattr | |
390 |
|
374 | |||
391 |
|
375 | |||
392 | def _getoptbwrapper( |
|
376 | def _getoptbwrapper( | |
393 | orig, args: Sequence[bytes], shortlist: bytes, namelist: Sequence[bytes] |
|
377 | orig, args: Sequence[bytes], shortlist: bytes, namelist: Sequence[bytes] | |
394 | ) -> _GetOptResult: |
|
378 | ) -> _GetOptResult: | |
395 | """ |
|
379 | """ | |
396 | Takes bytes arguments, converts them to unicode, pass them to |
|
380 | Takes bytes arguments, converts them to unicode, pass them to | |
397 | getopt.getopt(), convert the returned values back to bytes and then |
|
381 | getopt.getopt(), convert the returned values back to bytes and then | |
398 | return them for Python 3 compatibility as getopt.getopt() don't accepts |
|
382 | return them for Python 3 compatibility as getopt.getopt() don't accepts | |
399 | bytes on Python 3. |
|
383 | bytes on Python 3. | |
400 | """ |
|
384 | """ | |
401 | args = [a.decode('latin-1') for a in args] |
|
385 | args = [a.decode('latin-1') for a in args] | |
402 | shortlist = shortlist.decode('latin-1') |
|
386 | shortlist = shortlist.decode('latin-1') | |
403 | namelist = [a.decode('latin-1') for a in namelist] |
|
387 | namelist = [a.decode('latin-1') for a in namelist] | |
404 | opts, args = orig(args, shortlist, namelist) |
|
388 | opts, args = orig(args, shortlist, namelist) | |
405 | opts = [(a[0].encode('latin-1'), a[1].encode('latin-1')) for a in opts] |
|
389 | opts = [(a[0].encode('latin-1'), a[1].encode('latin-1')) for a in opts] | |
406 | args = [a.encode('latin-1') for a in args] |
|
390 | args = [a.encode('latin-1') for a in args] | |
407 | return opts, args |
|
391 | return opts, args | |
408 |
|
392 | |||
409 |
|
393 | |||
410 | def strkwargs(dic: Mapping[bytes, _T0]) -> Dict[str, _T0]: |
|
394 | def strkwargs(dic: Mapping[bytes, _T0]) -> Dict[str, _T0]: | |
411 | """ |
|
395 | """ | |
412 | Converts the keys of a python dictonary to str i.e. unicodes so that |
|
396 | Converts the keys of a python dictonary to str i.e. unicodes so that | |
413 | they can be passed as keyword arguments as dictionaries with bytes keys |
|
397 | they can be passed as keyword arguments as dictionaries with bytes keys | |
414 | can't be passed as keyword arguments to functions on Python 3. |
|
398 | can't be passed as keyword arguments to functions on Python 3. | |
415 | """ |
|
399 | """ | |
416 | dic = {k.decode('latin-1'): v for k, v in dic.items()} |
|
400 | dic = {k.decode('latin-1'): v for k, v in dic.items()} | |
417 | return dic |
|
401 | return dic | |
418 |
|
402 | |||
419 |
|
403 | |||
420 | def byteskwargs(dic: Mapping[str, _T0]) -> Dict[bytes, _T0]: |
|
404 | def byteskwargs(dic: Mapping[str, _T0]) -> Dict[bytes, _T0]: | |
421 | """ |
|
405 | """ | |
422 | Converts keys of python dictionaries to bytes as they were converted to |
|
406 | Converts keys of python dictionaries to bytes as they were converted to | |
423 | str to pass that dictonary as a keyword argument on Python 3. |
|
407 | str to pass that dictonary as a keyword argument on Python 3. | |
424 | """ |
|
408 | """ | |
425 | dic = {k.encode('latin-1'): v for k, v in dic.items()} |
|
409 | dic = {k.encode('latin-1'): v for k, v in dic.items()} | |
426 | return dic |
|
410 | return dic | |
427 |
|
411 | |||
428 |
|
412 | |||
429 | # TODO: handle shlex.shlex(). |
|
413 | # TODO: handle shlex.shlex(). | |
430 | def shlexsplit( |
|
414 | def shlexsplit( | |
431 | s: bytes, comments: bool = False, posix: bool = True |
|
415 | s: bytes, comments: bool = False, posix: bool = True | |
432 | ) -> List[bytes]: |
|
416 | ) -> List[bytes]: | |
433 | """ |
|
417 | """ | |
434 | Takes bytes argument, convert it to str i.e. unicodes, pass that into |
|
418 | Takes bytes argument, convert it to str i.e. unicodes, pass that into | |
435 | shlex.split(), convert the returned value to bytes and return that for |
|
419 | shlex.split(), convert the returned value to bytes and return that for | |
436 | Python 3 compatibility as shelx.split() don't accept bytes on Python 3. |
|
420 | Python 3 compatibility as shelx.split() don't accept bytes on Python 3. | |
437 | """ |
|
421 | """ | |
438 | ret = shlex.split(s.decode('latin-1'), comments, posix) |
|
422 | ret = shlex.split(s.decode('latin-1'), comments, posix) | |
439 | return [a.encode('latin-1') for a in ret] |
|
423 | return [a.encode('latin-1') for a in ret] | |
440 |
|
424 | |||
441 |
|
425 | |||
442 | iteritems = lambda x: x.items() |
|
426 | iteritems = lambda x: x.items() | |
443 | itervalues = lambda x: x.values() |
|
427 | itervalues = lambda x: x.values() | |
444 |
|
428 | |||
445 | json_loads = json.loads |
|
429 | json_loads = json.loads | |
446 |
|
430 | |||
447 | isjython: bool = sysplatform.startswith(b'java') |
|
431 | isjython: bool = sysplatform.startswith(b'java') | |
448 |
|
432 | |||
449 | isdarwin: bool = sysplatform.startswith(b'darwin') |
|
433 | isdarwin: bool = sysplatform.startswith(b'darwin') | |
450 | islinux: bool = sysplatform.startswith(b'linux') |
|
434 | islinux: bool = sysplatform.startswith(b'linux') | |
451 | isposix: bool = osname == b'posix' |
|
435 | isposix: bool = osname == b'posix' | |
452 | iswindows: bool = osname == b'nt' |
|
436 | iswindows: bool = osname == b'nt' | |
453 |
|
437 | |||
454 |
|
438 | |||
455 | def getoptb( |
|
439 | def getoptb( | |
456 | args: Sequence[bytes], shortlist: bytes, namelist: Sequence[bytes] |
|
440 | args: Sequence[bytes], shortlist: bytes, namelist: Sequence[bytes] | |
457 | ) -> _GetOptResult: |
|
441 | ) -> _GetOptResult: | |
458 | return _getoptbwrapper(getopt.getopt, args, shortlist, namelist) |
|
442 | return _getoptbwrapper(getopt.getopt, args, shortlist, namelist) | |
459 |
|
443 | |||
460 |
|
444 | |||
461 | def gnugetoptb( |
|
445 | def gnugetoptb( | |
462 | args: Sequence[bytes], shortlist: bytes, namelist: Sequence[bytes] |
|
446 | args: Sequence[bytes], shortlist: bytes, namelist: Sequence[bytes] | |
463 | ) -> _GetOptResult: |
|
447 | ) -> _GetOptResult: | |
464 | return _getoptbwrapper(getopt.gnu_getopt, args, shortlist, namelist) |
|
448 | return _getoptbwrapper(getopt.gnu_getopt, args, shortlist, namelist) | |
465 |
|
449 | |||
466 |
|
450 | |||
467 | def mkdtemp( |
|
451 | def mkdtemp( | |
468 | suffix: bytes = b'', prefix: bytes = b'tmp', dir: Optional[bytes] = None |
|
452 | suffix: bytes = b'', prefix: bytes = b'tmp', dir: Optional[bytes] = None | |
469 | ) -> bytes: |
|
453 | ) -> bytes: | |
470 | return tempfile.mkdtemp(suffix, prefix, dir) |
|
454 | return tempfile.mkdtemp(suffix, prefix, dir) | |
471 |
|
455 | |||
472 |
|
456 | |||
473 | # text=True is not supported; use util.from/tonativeeol() instead |
|
457 | # text=True is not supported; use util.from/tonativeeol() instead | |
474 | def mkstemp( |
|
458 | def mkstemp( | |
475 | suffix: bytes = b'', prefix: bytes = b'tmp', dir: Optional[bytes] = None |
|
459 | suffix: bytes = b'', prefix: bytes = b'tmp', dir: Optional[bytes] = None | |
476 | ) -> Tuple[int, bytes]: |
|
460 | ) -> Tuple[int, bytes]: | |
477 | return tempfile.mkstemp(suffix, prefix, dir) |
|
461 | return tempfile.mkstemp(suffix, prefix, dir) | |
478 |
|
462 | |||
479 |
|
463 | |||
480 | # TemporaryFile does not support an "encoding=" argument on python2. |
|
464 | # TemporaryFile does not support an "encoding=" argument on python2. | |
481 | # This wrapper file are always open in byte mode. |
|
465 | # This wrapper file are always open in byte mode. | |
482 | def unnamedtempfile(mode: Optional[bytes] = None, *args, **kwargs) -> BinaryIO: |
|
466 | def unnamedtempfile(mode: Optional[bytes] = None, *args, **kwargs) -> BinaryIO: | |
483 | if mode is None: |
|
467 | if mode is None: | |
484 | mode = 'w+b' |
|
468 | mode = 'w+b' | |
485 | else: |
|
469 | else: | |
486 | mode = sysstr(mode) |
|
470 | mode = sysstr(mode) | |
487 | assert 'b' in mode |
|
471 | assert 'b' in mode | |
488 | return cast(BinaryIO, tempfile.TemporaryFile(mode, *args, **kwargs)) |
|
472 | return cast(BinaryIO, tempfile.TemporaryFile(mode, *args, **kwargs)) | |
489 |
|
473 | |||
490 |
|
474 | |||
491 | # NamedTemporaryFile does not support an "encoding=" argument on python2. |
|
475 | # NamedTemporaryFile does not support an "encoding=" argument on python2. | |
492 | # This wrapper file are always open in byte mode. |
|
476 | # This wrapper file are always open in byte mode. | |
493 | def namedtempfile( |
|
477 | def namedtempfile( | |
494 | mode: bytes = b'w+b', |
|
478 | mode: bytes = b'w+b', | |
495 | bufsize: int = -1, |
|
479 | bufsize: int = -1, | |
496 | suffix: bytes = b'', |
|
480 | suffix: bytes = b'', | |
497 | prefix: bytes = b'tmp', |
|
481 | prefix: bytes = b'tmp', | |
498 | dir: Optional[bytes] = None, |
|
482 | dir: Optional[bytes] = None, | |
499 | delete: bool = True, |
|
483 | delete: bool = True, | |
500 | ): |
|
484 | ): | |
501 | mode = sysstr(mode) |
|
485 | mode = sysstr(mode) | |
502 | assert 'b' in mode |
|
486 | assert 'b' in mode | |
503 | return tempfile.NamedTemporaryFile( |
|
487 | return tempfile.NamedTemporaryFile( | |
504 | mode, bufsize, suffix=suffix, prefix=prefix, dir=dir, delete=delete |
|
488 | mode, bufsize, suffix=suffix, prefix=prefix, dir=dir, delete=delete | |
505 | ) |
|
489 | ) |
General Comments 0
You need to be logged in to leave comments.
Login now