Go to file
2016-11-01 01:08:38 -07:00
.github Fix Travis not pushing gh-pages for tags 2016-08-17 15:25:21 +02:00
clippy_lints Small fixes for get_unwrap lint 2016-11-01 01:08:38 -07:00
mini-macro Rustup to ea0dc92972 2016-07-01 17:44:17 +02:00
src Fix small nits on the help message 2016-10-25 15:09:56 +02:00
tests implement 'Re-implementing PartialEq::ne' lint 2016-10-30 15:00:59 -04:00
util fix dogfood.sh 2016-10-30 14:25:06 -04:00
.editorconfig Add EditorConfig 2016-07-15 13:08:21 +02:00
.gitignore Make Python utils more idiomatic, use better names, fix -c mode of update_wiki. 2016-08-07 08:09:26 +02:00
.remarkrc.json use .remarkrc.json for travis 2016-05-06 16:13:05 +02:00
.travis.yml Move set -e of travis down 2016-10-01 00:38:01 +02:00
Cargo.toml Bump to 0.0.96 2016-10-22 16:00:17 +02:00
CHANGELOG.md Run update_lints.py 2016-10-31 23:43:19 -07:00
CONTRIBUTING.md Make the lint docstrings more consistent. 2016-08-07 08:06:19 +02:00
README.md Run update_lints.py 2016-10-31 23:43:19 -07:00
rustfmt.toml Rustfmt config 2016-01-04 09:55:29 +05:30

rust-clippy

Build Status Clippy Linting Result Current Version License: MPL-2.0

A collection of lints to catch common mistakes and improve your Rust code.

Table of contents:

Usage

As a general rule clippy will only work with the latest Rust nightly for now.

As a Compiler Plugin

Since stable Rust is backwards compatible, you should be able to compile your stable programs with nightly Rust with clippy plugged in to circumvent this.

Add in your Cargo.toml:

[dependencies]
clippy = "*"

You then need to add #![feature(plugin)] and #![plugin(clippy)] to the top of your crate entry point (main.rs or lib.rs).

Sample main.rs:

#![feature(plugin)]

#![plugin(clippy)]


fn main(){
    let x = Some(1u8);
    match x {
        Some(y) => println!("{:?}", y),
        _ => ()
    }
}

Produces this warning:

src/main.rs:8:5: 11:6 warning: you seem to be trying to use match for destructuring a single type. Consider using `if let`, #[warn(single_match)] on by default
src/main.rs:8     match x {
src/main.rs:9         Some(y) => println!("{:?}", y),
src/main.rs:10         _ => ()
src/main.rs:11     }
src/main.rs:8:5: 11:6 help: Try
if let Some(y) = x { println!("{:?}", y) }

As a cargo subcommand (cargo clippy)

An alternate way to use clippy is by installing clippy through cargo as a cargo subcommand.

cargo install clippy

Now you can run clippy by invoking cargo clippy, or rustup run nightly cargo clippy directly from a directory that is usually compiled with stable.

In case you are not using rustup, you need to set the environment flag SYSROOT during installation so clippy knows where to find librustc and similar crates.

SYSROOT=/path/to/rustc/sysroot cargo install clippy

Running clippy from the command line without installing

To have cargo compile your crate with clippy without needing #![plugin(clippy)] in your code, you can use:

cargo rustc -- -L /path/to/clippy_so -Z extra-plugins=clippy

Note: Be sure that clippy was compiled with the same version of rustc that cargo invokes here!

Optional dependency

If you want to make clippy an optional dependency, you can do the following:

In your Cargo.toml:

[dependencies]
clippy = {version = "*", optional = true}

[features]
default = []

And, in your main.rs or lib.rs:

#![cfg_attr(feature="clippy", feature(plugin))]

#![cfg_attr(feature="clippy", plugin(clippy))]

Then build by enabling the feature: cargo build --features "clippy"

Instead of adding the cfg_attr attributes you can also run clippy on demand: cargo rustc --features clippy -- -Z no-trans -Z extra-plugins=clippy (the -Z no trans, while not neccessary, will stop the compilation process after typechecking (and lints) have completed, which can significantly reduce the runtime).

Configuration

Some lints can be configured in a clippy.toml file. It contains basic variable = value mapping eg.

blacklisted-names = ["toto", "tata", "titi"]
cyclomatic-complexity-threshold = 30

See the wiki for more information about which lints can be configured and the meaning of the variables.

You can also specify the path to the configuration file with:

#![plugin(clippy(conf_file="path/to/clippy's/configuration"))]

To deactivate the “for further information visit wiki-link” message you can define the CLIPPY_DISABLE_WIKI_LINKS environment variable.

Allowing/denying lints

You can add options to allow/warn/deny:

  • the whole set of Warn lints using the clippy lint group (#![deny(clippy)])

  • all lints using both the clippy and clippy_pedantic lint groups (#![deny(clippy)], #![deny(clippy_pedantic)]). Note that clippy_pedantic contains some very aggressive lints prone to false positives.

  • only some lints (#![deny(single_match, box_vec)], etc)

  • allow/warn/deny can be limited to a single function or module using #[allow(...)], etc

Note: deny produces errors instead of warnings.

For convenience, cargo clippy automatically defines a cargo-clippy features. This lets you set lints level and compile with or without clippy transparently:

#[cfg_attr(feature = "cargo-clippy", allow(needless_lifetimes))]

clippy-service is a rust web initiative providing rust-clippy as a web service.

Both projects are independent and maintained by different people (even if some clippy-service's contributions are authored by some rust-clippy members).

You can check out this great service at clippy.bashy.io.

Lints

There are 177 lints included in this crate:

name default triggers on
absurd_extreme_comparisons warn a comparison with a maximum or minimum value that is always true or false
almost_swapped warn foo = bar; bar = foo sequence
approx_constant warn the approximate of a known float constant (in std::fXX::consts)
assign_op_pattern warn assigning the result of an operation on a variable to that same variable
assign_ops allow any compound assignment operation
bad_bit_mask warn expressions of the form _ & mask == select that will only ever return true or false
blacklisted_name warn usage of a blacklisted/placeholder name
block_in_if_condition_expr warn braces that can be eliminated in conditions, e.g if { true } ...
block_in_if_condition_stmt warn complex blocks in conditions, e.g. if { let x = true; x } ...
bool_comparison warn comparing a variable to a boolean, e.g. if x == true
box_vec warn usage of Box<Vec<T>>, vector elements are already on the heap
boxed_local warn using Box<T> where unnecessary
builtin_type_shadow warn shadowing a builtin type
cast_possible_truncation allow casts that may cause truncation of the value, e.g x as u8 where x: u32, or x as i32 where x: f32
cast_possible_wrap allow casts that may cause wrapping around the value, e.g x as i32 where x: u32 and x > i32::MAX
cast_precision_loss allow casts that cause loss of precision, e.g x as f32 where x: u64
cast_sign_loss allow casts from signed types to unsigned types, e.g x as u32 where x: i32
char_lit_as_u8 warn casting a character literal to u8
chars_next_cmp warn using .chars().next() to check if a string starts with a char
clone_double_ref warn using clone on &&T
clone_on_copy warn using clone on a Copy type
cmp_nan deny comparisons to NAN, which will always return false, probably not intended
cmp_null warn comparing a pointer to a null pointer, suggesting to use .is_null() instead.
cmp_owned warn creating owned instances for comparing with others, e.g. x == "foo".to_string()
collapsible_if warn ifs that can be collapsed (e.g. if x { if y { ... } } and else { if x { ... } })
crosspointer_transmute warn transmutes that have to or from types that are a pointer to the other
cyclomatic_complexity warn functions that should be split up into multiple functions
deprecated_semver warn use of #[deprecated(since = "x")] where x is not semver
derive_hash_xor_eq warn deriving Hash but implementing PartialEq explicitly
diverging_sub_expression warn whether an expression contains a diverging sub expression
doc_markdown warn presence of _, :: or camel-case outside backticks in documentation
double_neg warn --x, which is a double negation of x and not a pre-decrement as in C/C++
drop_ref warn calls to std::mem::drop with a reference instead of an owned value
duplicate_underscore_argument warn function arguments having names which only differ by an underscore
empty_loop warn empty loop {}, which should block or sleep
enum_clike_unportable_variant warn C-like enums that are repr(isize/usize) and have values that don't fit into an i32
enum_glob_use allow use items that import all variants of an enum
enum_variant_names warn enums where all variants share a prefix/postfix
eq_op warn equal operands on both sides of a comparison or bitwise combination (e.g. x == x)
eval_order_dependence warn whether a variable read occurs before a write depends on sub-expression evaluation order
expl_impl_clone_on_copy warn implementing Clone explicitly on Copy types
explicit_counter_loop warn for-looping with an explicit counter when _.enumerate() would do
explicit_into_iter_loop warn for-looping over _.into_iter() when _ would do
explicit_iter_loop warn for-looping over _.iter() or _.iter_mut() when &_ or &mut _ would do
extend_from_slice warn .extend_from_slice(_) is a faster way to extend a Vec by a slice
filter_map allow using combinations of filter, map, filter_map and flat_map which can usually be written as a single method call
filter_next warn using filter(p).next(), which is more succinctly expressed as .find(p)
float_arithmetic allow any floating-point arithmetic statement
float_cmp warn using == or != on float values instead of comparing difference with an epsilon
for_kv_map warn looping on a map using iter when keys or values would do
for_loop_over_option warn for-looping over an Option, which is more clearly expressed as an if let
for_loop_over_result warn for-looping over a Result, which is more clearly expressed as an if let
get_unwrap warn using .get().unwrap() or .get_mut().unwrap() when using [] would work instead
identity_op warn using identity operations, e.g. x + 0 or y / 1
if_let_redundant_pattern_matching warn use the proper utility function avoiding an if let
if_let_some_result warn usage of ok() in if let Some(pat) statements is unnecessary, match on Ok(pat) instead
if_not_else allow if branches that could be swapped so no negation operation is necessary on the condition
if_same_then_else warn if with the same then and else blocks
ifs_same_cond warn consecutive ifs with the same condition
indexing_slicing allow indexing/slicing usage
ineffective_bit_mask warn expressions where a bit mask will be rendered useless by a comparison, e.g. `(x
inline_always warn use of #[inline(always)]
integer_arithmetic allow any integer arithmetic statement
invalid_regex deny invalid regular expressions
invalid_upcast_comparisons allow a comparison involving an upcast which is always true or false
items_after_statements allow blocks where an item comes after a statement
iter_next_loop warn for-looping over _.next() which is probably not intended
iter_nth warn using .iter().nth() on a standard library type with O(1) element access
iter_skip_next warn using .skip(x).next() on an iterator
len_without_is_empty warn traits or impls with a public len method but no corresponding is_empty method
len_zero warn checking .len() == 0 or .len() > 0 (or similar) when .is_empty() could be used instead
let_and_return warn creating a let-binding and then immediately returning it like let x = expr; x at the end of a block
let_unit_value warn creating a let binding to a value of unit type, which usually can't be used afterwards
linkedlist warn usage of LinkedList, usually a vector is faster, or a more specialized data structure like a VecDeque
logic_bug warn boolean expressions that contain terminals which can be eliminated
manual_swap warn manual swap of two variables
many_single_char_names warn too many single character bindings
map_clone warn using `.map(
map_entry warn use of contains_key followed by insert on a HashMap or BTreeMap
match_bool warn a match on a boolean expression instead of an if..else block
match_overlapping_arm warn a match with overlapping arms
match_ref_pats warn a match or if let with all arms prefixed with & instead of deref-ing the match expression
match_same_arms warn match with identical arm bodies
mem_forget allow mem::forget usage on Drop types, likely to cause memory leaks
min_max warn min(_, max(_, _)) (or vice versa) with bounds clamping the result to a constant
misrefactored_assign_op warn having a variable on both sides of an assign op
missing_docs_in_private_items allow detects missing documentation for public and private members
mixed_case_hex_literals warn hex literals whose letter digits are not consistently upper- or lowercased
module_inception warn modules that have the same name as their parent module
modulo_one warn taking a number modulo 1, which always returns 0
mut_mut allow usage of double-mut refs, e.g. &mut &mut ...
mutex_atomic warn using a mutex where an atomic value could be used instead
mutex_integer allow using a mutex for an integer type
needless_bool warn if-statements with plain booleans in the then- and else-clause, e.g. if p { true } else { false }
needless_borrow warn taking a reference that is going to be automatically dereferenced
needless_lifetimes warn using explicit lifetimes for references in function arguments when elision rules would allow omitting them
needless_range_loop warn for-looping over a range of indices where an iterator over items would do
needless_return warn using a return statement like return expr; where an expression would suffice
needless_update warn using Foo { ..base } when there are no missing fields
neg_multiply warn multiplying integers with -1
new_ret_no_self warn not returning Self in a new method
new_without_default warn fn new() -> Self method without Default implementation
new_without_default_derive warn fn new() -> Self without #[derive]able Default implementation
no_effect warn statements with no effect
non_ascii_literal allow using any literal non-ASCII chars in a string literal instead of using the \\u escape
nonminimal_bool allow boolean expressions that can be written more concisely
nonsensical_open_options warn nonsensical combination of options for opening a file
not_unsafe_ptr_arg_deref warn public functions dereferencing raw pointer arguments but not marked unsafe
ok_expect warn using ok().expect(), which gives worse error messages than calling expect directly on the Result
option_map_unwrap_or allow using Option.map(f).unwrap_or(a), which is more succinctly expressed as map_or(a, f)
option_map_unwrap_or_else allow using Option.map(f).unwrap_or_else(g), which is more succinctly expressed as map_or_else(g, f)
option_unwrap_used allow using Option.unwrap(), which should at least get a better message using expect()
or_fun_call warn using any *or method with a function call, which suggests *or_else
out_of_bounds_indexing deny out of bounds constant indexing
overflow_check_conditional warn overflow checks inspired by C which are likely to panic
panic_params warn missing parameters in panic! calls
partialeq_ne_impl warn re-implementing PartialEq::ne
precedence warn operations where precedence may be unclear
print_stdout allow printing on stdout
print_with_newline warn using print!() with a format string that ends in a newline
ptr_arg warn fn arguments of the type &Vec<...> or &String, suggesting to use &[...] or &str instead, respectively
range_step_by_zero warn using Range::step_by(0), which produces an infinite iterator
range_zip_with_len warn zipping iterator with a range when enumerate() would do
redundant_closure warn redundant closures, i.e. `
redundant_closure_call warn throwaway closures called in the expression they are defined
redundant_pattern warn using name @ _ in a pattern
regex_macro warn use of regex!(_) instead of Regex::new(_)
result_unwrap_used allow using Result.unwrap(), which might be better handled
reverse_range_loop warn iteration over an empty range, such as 10..0 or 5..5
search_is_some warn using an iterator search followed by is_some(), which is more succinctly expressed as a call to any()
serde_api_misuse warn various things that will negatively affect your serde experience
shadow_reuse allow rebinding a name to an expression that re-uses the original value, e.g. let x = x + 1
shadow_same allow rebinding a name to itself, e.g. let mut x = &mut x
shadow_unrelated allow rebinding a name without even using the original value
should_implement_trait warn defining a method that should be implementing a std trait
similar_names allow similarly named items and bindings
single_char_pattern warn using a single-character str where a char could be used, e.g. _.split("x")
single_match warn a match statement with a single nontrivial arm (i.e, where the other arm is _ => {}) instead of if let
single_match_else allow a match statement with a two arms where the second arm's pattern is a wildcard instead of if let
string_add allow using x + .. where x is a String instead of push_str()
string_add_assign allow using x = x + .. where x is a String instead of push_str()
string_lit_as_bytes warn calling as_bytes on a string literal instead of using a byte string literal
stutter allow type names prefixed/postfixed with their containing module's name
suspicious_assignment_formatting warn suspicious formatting of *=, -= or !=
suspicious_else_formatting warn suspicious formatting of else if
temporary_assignment warn assignments to temporaries
temporary_cstring_as_ptr warn getting the inner pointer of a temporary CString
too_many_arguments warn functions with too many arguments
toplevel_ref_arg warn an entire binding declared as ref, in a function argument or a let statement
transmute_ptr_to_ref warn transmutes from a pointer to a reference type
trivial_regex warn trivial regular expressions
type_complexity warn usage of very complex types that might be better factored into type definitions
unicode_not_nfc allow using a unicode literal not in NFC normal form (see unicode tr15 for further information)
unit_cmp warn comparing unit values
unnecessary_mut_passed warn an argument passed as a mutable reference although the callee only demands an immutable reference
unnecessary_operation warn outer expressions with no effect
unneeded_field_pattern warn struct fields bound to a wildcard instead of using ..
unsafe_removed_from_name warn unsafe removed from API names on import
unseparated_literal_suffix allow literals whose suffix is not separated by an underscore
unused_collect warn collect()ing an iterator without using the result; this is usually better written as a for loop
unused_label warn unused labels
unused_lifetimes warn unused lifetimes in function definitions
use_debug allow use of Debug-based formatting
used_underscore_binding allow using a binding which is prefixed with an underscore
useless_attribute warn use of lint attributes on extern crate items
useless_format warn useless use of format!
useless_let_if_seq warn unidiomatic let mut declaration followed by initialization in if
useless_transmute warn transmutes that have the same to and from types or could be a cast/coercion
useless_vec warn useless vec!
while_let_loop warn loop { if let { ... } else break }, which can be written as a while let loop
while_let_on_iterator warn using a while-let loop instead of a for loop on an iterator
wrong_pub_self_convention allow defining a public method named with an established prefix (like "into_") that takes self with the wrong convention
wrong_self_convention warn defining a method named with an established prefix (like "into_") that takes self with the wrong convention
wrong_transmute warn transmutes that are confusing at best, undefined behaviour at worst and always useless
zero_divided_by_zero warn usage of 0.0 / 0.0 to obtain NaN instead of std::f32::NaN or std::f64::NaN
zero_prefixed_literal warn integer literals starting with 0
zero_width_space deny using a zero-width space in a string literal, which is confusing

More to come, please file an issue if you have ideas!

License

Licensed under MPL. If you're having issues with the license, let me know and I'll try to change it to something more permissive.