Commit graph

4313 commits

Author SHA1 Message Date
Manish Goregaokar
6df102cdab Merge pull request #238 from birkenfeld/to_string_fix
methods: suggest correct replacement for `to_string()` (fixes #232)
2015-08-26 02:43:19 +05:30
llogiq
56cdd58ad2 Merge branch 'master' of https://github.com/Manishearth/rust-clippy into shadow 2015-08-25 20:09:50 +02:00
Georg Brandl
6fa34cca29 methods: suggest correct replacement for to_string() (fixes #232) 2015-08-25 18:39:21 +02:00
Georg Brandl
ffed5b0b23 loops: use a whitelist for the "x.iter() -> &x" lint (fixes #236) 2015-08-25 18:28:05 +02:00
llogiq
5225feceaa shadowing detection 2015-08-25 13:40:26 +02:00
Georg Brandl
64cd1fc6ba eq_op: cut back to expressions that are guaranteed side effect free
fixes #229
2015-08-25 12:34:37 +02:00
Georg Brandl
56e8db476c new lint: inherent methods that should be trait impls (fixes #218) 2015-08-25 11:10:42 +02:00
llogiq
209e6981a3 shadowing detection 2015-08-24 16:59:02 +02:00
R.Chavignat
e80f2470b7 Merge branch 'master' into cast_iusize_improvements 2015-08-23 00:08:16 +02:00
R.Chavignat
79ef13592e Completed the implementation of *size handling.
Added some more cases to the test, and implemented a new lint,
cast_possible_wrap, triggered when casting from an unsigned type to a
signed type of the same size.
2015-08-22 23:49:03 +02:00
R.Chavignat
807dab943b Updated test case for cast lints.
Also improved readability and reworded the messages.
2015-08-22 22:06:58 +02:00
Georg Brandl
7580da306e matches: special message for this case
match &e { &Pat1 => {}, &Pat2 => {}, ... }

(inspired by dogfood fixes)
2015-08-22 14:34:39 +02:00
Georg Brandl
017dac2301 new lint: using &Ref patterns instead of matching on *expr (fixes #187) 2015-08-22 14:34:39 +02:00
Georg Brandl
630bb76f96 new lint: type complexity (fixes #93)
Still very naive, but it's a start.
2015-08-22 08:57:11 +02:00
R.Chavignat
f1255d5f5d Casts : work in progress handling *size separately 2015-08-22 02:46:22 +02:00
Georg Brandl
707e95f2e5 types: use middle::ty types instead of ast types
This gets rid of the match_ty_unwrap function.
2015-08-21 19:21:19 +02:00
Georg Brandl
b2df15d65a ptr_arg improvements (fixes #214)
* do not trigger on mutable references
* use "real" type from ty, not AST type
2015-08-21 19:21:18 +02:00
llogiq
6ffe1ed337 Merge pull request #212 from Manishearth/eq_op_const
Eq op const
2015-08-21 12:30:53 +02:00
llogiq
a22b3cdcee const folding for eq_op 2015-08-21 12:26:03 +02:00
R.Chavignat
ad0bc66402 Added support for isize/usize in the CastPass lint pass.
Extracted the match that determines an integer types's size in a
utility function and implemented support for usize/isize.
Added a needed feature to the crate root.
Added some tests to cover those cases, and a test I previously forgot.
Silenced two errors signaled by dogfood.sh in unicode.rs.
2015-08-21 03:03:37 +02:00
R.Chavignat
dbc9b7f46e Reworked the error messages for more heplfulness.
Renamed the cast_possible_overflow lint to cast_possible_truncation,
and updated the error message, readme and crate root accordingly.
Added some more information to the message for the cast_precision_loss
lint.
Updated the test case to reflect changes.
2015-08-20 22:44:40 +02:00
R.Chavignat
993239d33a Initial implementation of lossy cast lints.
Introduces 3 lints :
  cast_possible_overflow
  cast_precision_loss
  cast_sign_loss
Add a compile-test test case.
Fix errors spotted by dogfood script.
2015-08-20 01:04:06 +02:00
llogiq
150840667e Merge pull request #204 from Manishearth/bit-mask
Extend and correct the ineffective_bit_mask check
2015-08-19 13:25:43 +02:00
llogiq
973d5e5c6b Mul and Div for integers 2015-08-19 11:58:59 +02:00
llogiq
98d24b5b56 fixed #203 and #197 2015-08-19 09:39:03 +02:00
Georg Brandl
8f4499f3ae new lint: comparing unit types (fixes #201) 2015-08-19 08:11:00 +02:00
Georg Brandl
7f52239cab consts: convert to using a struct with state
Struct has the context reference (as an Option) and the
needed_resolution flag.
2015-08-17 19:58:21 +02:00
llogiq
a2ee637be6 added test and fixed negativity check in Partial{Eq, Ord} impl 2015-08-17 16:24:57 +02:00
llogiq
9f134f8e95 added PartialEq/PartialOrd to ConstantVariant, used to implement comparing binops 2015-08-17 13:18:14 +02:00
llogiq
0e67c0134f make float_cmp check for zero 2015-08-17 12:08:45 +02:00
llogiq
e354fdc3e8 added regression test for #189 2015-08-17 12:08:45 +02:00
llogiq
fb715ce45d fix 189, fixed a few warnings, ==/!= for consts, refactored consts test 2015-08-17 12:08:45 +02:00
Manish Goregaokar
4f1fcd4d5b Merge pull request #184 from Manishearth/identity_op
Identity op now uses const folding, no longer follows const bindings
2015-08-17 14:35:45 +05:30
Georg Brandl
e9a41e2374 new lint: lint when iterating over any Iterator::next() result (fixes #182) 2015-08-17 07:28:40 +02:00
llogiq
759b45a46d made is_negative(..) public (+doctest), fixed identity_op and precedence 2015-08-16 23:09:56 +02:00
llogiq
c4b07ead17 Merge pull request #179 from nweston/step-by-zero
New lint: Range::step_by(0) (fixes #95)
2015-08-16 20:12:52 +02:00
Nathan Weston
23a38c4170 New lint: Range::step_by(0) (fixes #95)
Uses type information so it can detect non-literal ranges as well
(Range or RangeFrom -- the other range types don't have step_by).
2015-08-16 12:58:54 -04:00
llogiq
e1438e7010 copied over cmp_owned fix from master 2015-08-16 16:13:44 +02:00
llogiq
03c7d7074d With working test now 2015-08-16 15:56:09 +02:00
Manish Goregaokar
164907ece2 restrict toplevel_ref_arg to only functions (fixes #170) 2015-08-16 17:24:03 +05:30
Manish Goregaokar
f494f14aa6 Merge pull request #177 from birkenfeld/if_let_mini_fix
misc: fix check for unit body in "match -> if let" lint (fixes #172)
2015-08-15 14:12:34 +05:30
Georg Brandl
542bf8d50e misc: fix check for unit body in "match -> if let" lint (fixes #172) 2015-08-15 09:56:15 +02:00
Georg Brandl
b299433de3 lifetimes: fix case with one unnamed and one static ref (fixes #171) 2015-08-15 09:36:07 +02:00
llogiq
f23af0cfd5 changed const to consts to avoid keyword, added test, fixed a lot of bugs 2015-08-14 17:14:54 +02:00
llogiq
5d99ebec72 fixed false positives (at the cost of some false negatives) 2015-08-14 14:37:12 +02:00
llogiq
ffda91a8c7 removed String::from_str(..) to fix build with 1.4.0-nightly/2015-08-14 2015-08-14 14:26:57 +02:00
llogiq
dece5a6cb5 added empty line test 2015-08-13 23:18:34 +05:30
llogiq
f4b5d21533 added a few unit tests to trim_multiline 2015-08-13 23:18:34 +05:30
Manish Goregaokar
847070e19d Merge pull request #151 from birkenfeld/needless_bool_msg
spelling fix, rework needless_bool with snippet (fixes #150)
2015-08-13 23:17:18 +05:30
Manish Goregaokar
3cf5c36296 Address review comments, move to travis 2015-08-13 21:45:50 +05:30
Georg Brandl
49e51fe65a lifetimes: try to fix w.r.t. lifetimes from parent scopes (fixes #162) 2015-08-13 21:45:42 +05:30
Georg Brandl
9578403638 new lint: looping over x.iter() or x.iter_mut() (fixes #157) 2015-08-13 16:31:16 +02:00
Georg Brandl
8a98736f51 spelling fix, rework needless_bool with snippet (fixes #150) 2015-08-13 15:31:16 +02:00
Andre Bogus
f67175b4cd fixed error messages in compile-fail test 2015-08-13 11:38:39 +02:00
llogiq
f9e851e212 pulled strings passes together, added more tests 2015-08-13 11:36:39 +02:00
llogiq
0b08e9e83e added string_add lint and fixed string_add_assign + test 2015-08-13 11:36:39 +02:00
Manish Goregaokar
8dfa02938d Merge branch 'pr-149' 2015-08-13 13:00:29 +05:30
Manish Goregaokar
510ea0c5e5 Merge branch 'pr-140'
Conflicts:
	src/lib.rs
2015-08-13 12:52:14 +05:30
Georg Brandl
7aee04878f tests: use fragment of lint text for error checking
(Did not touch strings.rs, which is fixed by @llogiq's PR)
2015-08-13 08:12:07 +02:00
Georg Brandl
5952a29543 lifetimes test: use explicit message prefix 2015-08-13 07:51:24 +02:00
Georg Brandl
2f7693094f lifetimes lint: include support for lifetimes as generic params 2015-08-13 06:43:25 +02:00
Georg Brandl
b349f9e88d new lint for needless lifetimes (fixes #115) 2015-08-13 06:35:40 +02:00
Georg Brandl
f6090909d3 new lint: using for i in 0..x { .. vec[i] .. } instead of iterator (fixes #3) 2015-08-13 06:34:08 +02:00
Georg Brandl
3044d3d633 unicode: add lint against non-ascii chars in literals (Allow by default), #85 2015-08-12 20:36:35 +02:00
Manish Goregaokar
c58f803511 Merge branch 'pr-138'
Conflicts:
	src/lib.rs
2015-08-12 21:22:01 +05:30
Georg Brandl
4074c1f968 methods: lint against String.to_string (fixes #100) 2015-08-12 17:03:13 +02:00
Georg Brandl
e8fed074cf new lint: warn if let-binding has unit value (fixes #74) 2015-08-12 13:21:07 +02:00
Georg Brandl
b831bd1d1d len_zero: display full suggested expr in message 2015-08-12 10:53:14 +02:00
Georg Brandl
bcd95aec1c all: make style of lint messages consistent
* start first sentence lowercased
* use backticks to delimit code snippets
* use "this is wrong. Consider doing X." consistently
2015-08-12 10:47:09 +02:00
Georg Brandl
f6dc48fe3a new lint for "let x = EXPR; x" at the end of functions (fixes #104) 2015-08-11 22:25:47 +02:00
Georg Brandl
2bcc151888 new lint for Option.unwrap() and Result.unwrap()
The latter is set to Allow by default (fixes #24)
2015-08-11 21:19:11 +02:00
Manish Goregaokar
6bc69fa342 Merge pull request #127 from birkenfeld/better-helptext-if-let
better help text for "match -> if let" lint
2015-08-12 00:44:51 +05:30
Georg Brandl
e318328d63 all: whitespace cleanup
* 4-space indentation
* no trailing whitespace
* no tabs
2015-08-11 20:22:50 +02:00
Manish Goregaokar
2cb26126d3 Merge branch 'pr-78'
Conflicts:
	src/lib.rs
2015-08-11 23:28:06 +05:30
Georg Brandl
cab9905705 better help text for "match -> if let" lint
Implements the suggestion from #87.

Changes span_help_and_lint(), which is only used for this lint,
to use fileline_help() instead of span_help() to avoid printing
the span twice.

Also adds complete suggested new code. I had to distinguish
between blocks, which need no additionals braces, and other exprs.
2015-08-11 19:26:51 +02:00
Georg Brandl
0ff476b529 new lint for unneeded return stmts 2015-08-11 19:05:11 +02:00
llogiq
f0eb36c2af added parent method, also changed match-if-let note to help 2015-08-11 15:07:21 +02:00
llogiq
b393752814 New string_add_assign lint (first part of #121), also formatting & refactoring 2015-08-05 15:10:45 +02:00
llogiq
0e5b62c8d8 also included String::from in cmp_owned and fixed deprecation in test 2015-06-11 16:53:23 +02:00
llogiq
23caf3cccc first unicode lint: zero_width_space 2015-06-11 11:35:00 +02:00
llogiq
1f74c4b359 removed false positives from inline_always (issue #84) 2015-06-07 12:03:56 +02:00
llogiq
8563ee60ec fixed issue #88 in bit_mask 2015-06-06 02:27:48 +02:00
llogiq
b44435ef28 extended compile-test.rs to actually observe TESTNAME environment variable and filter tests if available 2015-06-06 02:07:48 +02:00
llogiq
1ee2e4ffe8 Fixed block check, also added macro test to collapsible_if and inline_always 2015-06-01 15:09:17 +02:00
llogiq
7f5891184d merged upstream master 2015-06-01 13:55:55 +02:00
llogiq
cdca2c93c1 now the method lookup actually works (and I understand why! 😄), reduces unnecessary loops, and has a few comments 2015-06-01 12:49:36 +02:00
llogiq
4c76465154 Merge pull request #68 from Manishearth/macro_expn
macro expn detection in mut_mut.rs
2015-06-01 07:44:27 +02:00
llogiq
21cd0c7e70 check for is_empty() method to get rid of false positives 2015-06-01 07:40:33 +02:00
Matthew Hall
423a9666ca Implements #45 - any number mod 1 will be 0 2015-05-31 13:43:31 +01:00
llogiq
77838d6ba7 New lint for issue
#72
2015-05-30 15:10:19 +02:00
Matthew Hall
7e16822925 Add lint for ifs that could be collapsed
"Collapsible" ifs are ones which contain only a then block, and the then
block consists of an if that only has a then block.
2015-05-29 15:41:25 +01:00
llogiq
cd2e621c60 made in_macro distinguish intra-crate and extra-crate macros, as the latter have no working source (note: may fail in the face of compiler plugins doing whatever they like with spans), also one more run-pass test 2015-05-26 13:52:40 +02:00
llogiq
0d651c72ff made macro test even simpler, added a few tests 2015-05-26 01:45:15 +02:00
llogiq
4292dc77a7 new lint: cmp_owned 2015-05-21 14:51:43 +02:00
llogiq
483a546e74 added messages to test error comments 2015-05-20 09:34:02 +02:00
llogiq
e8ca3c6eae new lints len_zero and len_without_is_empty 2015-05-20 08:52:19 +02:00
llogiq
1f8453ab73 mut_mut now more robust (thanks to Manishearth, see issue #9) 2015-05-18 10:41:15 +02:00
llogiq
96bfade4f1 New lint: mut_mut (closes issue #9) 2015-05-18 09:02:24 +02:00
llogiq
edf747ab76 new lint: identity_op, refactored bit_masks a bit 2015-05-15 18:46:43 +02:00
llogiq
6bec4f35df Added 'ineffective bit mask' lint 2015-05-15 14:09:29 +02:00
Manish Goregaokar
2447e1d5be Add eta reduction (fixes #29) 2015-05-10 11:49:08 +05:30
Manish Goregaokar
068e215728 move to compiletest on crates 2015-05-09 15:19:12 +05:30
llogiq
b24433f36d added test for issue #31 2015-05-08 06:01:41 +02:00
llogiq
17bcf0e865 New lint: precedence, see issue #41 2015-05-06 12:59:08 +02:00
llogiq
ac151bb1f0 Added new 'float_cmp' lint (see issue #46) 2015-05-06 10:01:49 +02:00
llogiq
0936e0617a new lint to check for doomed comparisons to NAN 2015-05-04 14:11:15 +02:00
llogiq
2cb84b9d15 New lint: approx_const 2015-05-04 12:01:34 +02:00
llogiq
8d2328d9a5 Added &String matching and renamed to vec_ptr_arg to ptr_arg, also added README section 2015-05-04 08:15:24 +02:00
llogiq
07adeee6e9 Added check for zero bitmask and uncommon directions, wrong comment in needless_bool corrected, added new lint vec_ptr_arg + test 2015-05-04 07:20:34 +02:00
llogiq
53fa76dff9 new lint: needless_bool (TODO: The warnings could give more specific directions) 2015-05-02 00:35:49 +02:00
llogiq
3a9bf24bb3 Added constant lookup (with help from Manish) to bad_bit_mask 2015-04-30 15:17:06 +02:00
llogiq
441b55b328 Added eq_op and bad_bit_mask from the extra_lints project (mostly plain copy, need to refactor to integrate better) 2015-04-30 11:48:43 +02:00
Manish Goregaokar
a5c3102594 Use compiletest 2015-04-14 00:20:44 +05:30