1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
|
[MASTER]
ignore=
src/python/grpcio/grpc/beta,
src/python/grpcio/grpc/framework,
src/python/grpcio/grpc/framework/common,
src/python/grpcio/grpc/framework/foundation,
src/python/grpcio/grpc/framework/interfaces,
[VARIABLES]
# TODO(https://github.com/PyCQA/pylint/issues/1345): How does the inspection
# not include "unused_" and "ignored_" by default?
dummy-variables-rgx=^ignored_|^unused_
[DESIGN]
# NOTE(nathaniel): Not particularly attached to this value; it just seems to
# be what works for us at the moment (excepting the dead-code-walking Beta
# API).
max-args=6
[MISCELLANEOUS]
# NOTE(nathaniel): We are big fans of "TODO(<issue link>): " and
# "NOTE(<username or issue link>): ". We do not allow "TODO:",
# "TODO(<username>):", "FIXME:", or anything else.
notes=FIXME,XXX
[MESSAGES CONTROL]
disable=
# TODO(https://github.com/PyCQA/pylint/issues/59#issuecomment-283774279):
# Enable cyclic-import after a 1.7-or-later pylint release that
# recognizes our disable=cyclic-import suppressions.
cyclic-import,
# TODO(https://github.com/grpc/grpc/issues/8622): Enable this after the
# Beta API is removed.
duplicate-code,
# TODO(https://github.com/grpc/grpc/issues/261): Doesn't seem to
# understand enum and concurrent.futures; look into this later with the
# latest pylint version.
import-error,
# TODO(https://github.com/grpc/grpc/issues/261): Enable this one.
# Should take a little configuration but not much.
invalid-name,
# TODO(https://github.com/grpc/grpc/issues/261): This doesn't seem to
# work for now? Try with a later pylint?
locally-disabled,
# NOTE(nathaniel): What even is this? *Enabling* an inspection results
# in a warning? How does that encourage more analysis and coverage?
locally-enabled,
# NOTE(nathaniel): We don't write doc strings for most private code
# elements.
missing-docstring,
# NOTE(nathaniel): In numeric comparisons it is better to have the
# lesser (or lesser-or-equal-to) quantity on the left when the
# expression is true than it is to worry about which is an identifier
# and which a literal value.
misplaced-comparison-constant,
# NOTE(nathaniel): Our completely abstract interface classes don't have
# constructors.
no-init,
# TODO(https://github.com/grpc/grpc/issues/261): Doesn't yet play
# nicely with some of our code being implemented in Cython. Maybe in a
# later version?
no-name-in-module,
# TODO(https://github.com/grpc/grpc/issues/261): Suppress these where
# the odd shape of the authentication portion of the API forces them on
# us and enable everywhere else.
protected-access,
# NOTE(nathaniel): Pylint and I will probably never agree on this.
too-few-public-methods,
# NOTE(nathaniel): Pylint and I wil probably never agree on this for
# private classes. For public classes maybe?
too-many-instance-attributes,
# NOTE(nathaniel): Some of our modules have a lot of lines... of
# specification and documentation. Maybe if this were
# lines-of-code-based we would use it.
too-many-lines,
# TODO(https://github.com/grpc/grpc/issues/261): Maybe we could have
# this one if we extracted just a few more helper functions...
too-many-nested-blocks,
# TODO(https://github.com/grpc/grpc/issues/261): Disable unnecessary
# super-init requirement for abstract class implementations for now.
super-init-not-called,
# NOTE(nathaniel): A single statement that always returns program
# control is better than two statements the first of which sometimes
# returns program control and the second of which always returns
# program control. Probably generally, but definitely in the cases of
# if:/else: and for:/else:.
useless-else-on-loop,
no-else-return,
# NOTE(lidiz): Python 3 make object inheritance default, but not PY2
useless-object-inheritance,
|
