actions/typos
1
0
Fork 0
mirror of https://github.com/crate-ci/typos.git synced 2024-11-24 18:10:56 -05:00
Code Issues Projects Releases Packages Wiki Activity Actions

refactor: Split out typos-dict

Browse source
This commit is contained in:
Ed Page 2019-08-08 10:24:50 -05:00
parent 164ee9cb84
commit 1bdd1c928a
17 changed files with 173 additions and 127 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

36
Cargo.lock generated
View file

@ -78,6 +78,7 @@ dependencies = [
"lazy_static 1.3.0 (registry+https://github.com/rust-lang/crates.io-index)",
"memchr 2.1.3 (registry+https://github.com/rust-lang/crates.io-index)",
"regex-automata 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)",
"serde 1.0.85 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
@ -158,16 +159,19 @@ dependencies = [
[[package]]
name = "csv"
version = "1.0.5"
version = "1.1.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
dependencies = [
"csv-core 0.1.5 (registry+https://github.com/rust-lang/crates.io-index)",
"bstr 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
"csv-core 0.1.6 (registry+https://github.com/rust-lang/crates.io-index)",
"itoa 0.4.3 (registry+https://github.com/rust-lang/crates.io-index)",
"ryu 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
"serde 1.0.85 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
name = "csv-core"
version = "0.1.5"
version = "0.1.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
dependencies = [
"memchr 2.1.3 (registry+https://github.com/rust-lang/crates.io-index)",
@ -697,6 +701,11 @@ name = "ryu"
version = "0.2.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
[[package]]
name = "ryu"
version = "1.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
[[package]]
name = "same-file"
version = "1.0.4"
@ -880,14 +889,11 @@ name = "typos"
version = "0.1.0"
dependencies = [
"bstr 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
"csv 1.0.5 (registry+https://github.com/rust-lang/crates.io-index)",
"derive_more 0.15.0 (registry+https://github.com/rust-lang/crates.io-index)",
"failure 0.1.5 (registry+https://github.com/rust-lang/crates.io-index)",
"itertools 0.8.0 (registry+https://github.com/rust-lang/crates.io-index)",
"lazy_static 1.3.0 (registry+https://github.com/rust-lang/crates.io-index)",
"log 0.4.6 (registry+https://github.com/rust-lang/crates.io-index)",
"phf 0.7.24 (registry+https://github.com/rust-lang/crates.io-index)",
"phf_codegen 0.7.24 (registry+https://github.com/rust-lang/crates.io-index)",
"regex 1.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
"serde 1.0.85 (registry+https://github.com/rust-lang/crates.io-index)",
"serde_json 1.0.36 (registry+https://github.com/rust-lang/crates.io-index)",
@ -910,6 +916,19 @@ dependencies = [
"structopt 0.2.14 (registry+https://github.com/rust-lang/crates.io-index)",
"toml 0.4.10 (registry+https://github.com/rust-lang/crates.io-index)",
"typos 0.1.0",
"typos-dict 0.1.0",
]
[[package]]
name = "typos-dict"
version = "0.1.0"
dependencies = [
"csv 1.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
"log 0.4.6 (registry+https://github.com/rust-lang/crates.io-index)",
"phf 0.7.24 (registry+https://github.com/rust-lang/crates.io-index)",
"phf_codegen 0.7.24 (registry+https://github.com/rust-lang/crates.io-index)",
"typos 0.1.0",
"unicase 1.4.2 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
@ -1033,8 +1052,8 @@ dependencies = [
"checksum cloudabi 0.0.3 (registry+https://github.com/rust-lang/crates.io-index)" = "ddfc5b9aa5d4507acaf872de71051dfd0e309860e88966e1051e462a077aac4f"
"checksum crossbeam-channel 0.3.6 (registry+https://github.com/rust-lang/crates.io-index)" = "137bc235f622ffaa0428e3854e24acb53291fc0b3ff6fb2cb75a8be6fb02f06b"
"checksum crossbeam-utils 0.6.3 (registry+https://github.com/rust-lang/crates.io-index)" = "41ee4864f4797060e52044376f7d107429ce1fb43460021b126424b7180ee21a"
"checksum csv 1.0.5 (registry+https://github.com/rust-lang/crates.io-index)" = "9fd1c44c58078cfbeaf11fbb3eac9ae5534c23004ed770cc4bfb48e658ae4f04"
"checksum csv-core 0.1.5 (registry+https://github.com/rust-lang/crates.io-index)" = "fa5cdef62f37e6ffe7d1f07a381bc0db32b7a3ff1cac0de56cb0d81e71f53d65"
"checksum csv 1.1.1 (registry+https://github.com/rust-lang/crates.io-index)" = "37519ccdfd73a75821cac9319d4fce15a81b9fcf75f951df5b9988aa3a0af87d"
"checksum csv-core 0.1.6 (registry+https://github.com/rust-lang/crates.io-index)" = "9b5cadb6b25c77aeff80ba701712494213f4a8418fcda2ee11b6560c3ad0bf4c"
"checksum derive_more 0.15.0 (registry+https://github.com/rust-lang/crates.io-index)" = "7a141330240c921ec6d074a3e188a7c7ef95668bb95e7d44fa0e5778ec2a7afe"
"checksum difference 2.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "524cbf6897b527295dff137cec09ecf3a05f4fddffd7dfcd1585403449e74198"
"checksum either 1.5.0 (registry+https://github.com/rust-lang/crates.io-index)" = "3be565ca5c557d7f59e7cfcf1844f9e3033650c929c6566f511e8005f205c1d0"
@ -1097,6 +1116,7 @@ dependencies = [
"checksum rustc-demangle 0.1.13 (registry+https://github.com/rust-lang/crates.io-index)" = "adacaae16d02b6ec37fdc7acfcddf365978de76d1983d3ee22afc260e1ca9619"
"checksum rustc_version 0.2.3 (registry+https://github.com/rust-lang/crates.io-index)" = "138e3e0acb6c9fb258b19b67cb8abd63c00679d2851805ea151465464fe9030a"
"checksum ryu 0.2.7 (registry+https://github.com/rust-lang/crates.io-index)" = "eb9e9b8cde282a9fe6a42dd4681319bfb63f121b8a8ee9439c6f4107e58a46f7"
"checksum ryu 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "c92464b447c0ee8c4fb3824ecc8383b81717b9f1e74ba2e72540aef7b9f82997"
"checksum same-file 1.0.4 (registry+https://github.com/rust-lang/crates.io-index)" = "8f20c4be53a8a1ff4c1f1b2bd14570d2f634628709752f0702ecdd2b3f9a5267"
"checksum scopeguard 0.3.3 (registry+https://github.com/rust-lang/crates.io-index)" = "94258f53601af11e6a49f722422f6e3425c52b06245a5cf9bc09908b174f5e27"
"checksum semver 0.9.0 (registry+https://github.com/rust-lang/crates.io-index)" = "1d7eb9ef2c18661902cc47e535f9bc51b78acd254da71d375c2f6720d9a40403"

3
Cargo.toml
View file

@ -1,5 +1,5 @@
[workspace]
members = ["typos"]
members = ["typos", "typos-dict"]
[package]
name = "typos-cli"
@ -23,6 +23,7 @@ iterate_unstable = []
[dependencies]
typos = { version = "0.1", path = "typos" }
typos-dict = { version = "0.1", path = "typos-dict" }
failure = "0.1"
structopt = "0.2"
clap = "2"

6
benches/corrections.rs
View file

@ -4,12 +4,12 @@ extern crate test;
#[bench]
fn load_corrections(b: &mut test::Bencher) {
b.iter(|| typos::BuiltIn::new());
b.iter(|| typos_dict::BuiltIn::new());
}
#[bench]
fn correct_word_hit(b: &mut test::Bencher) {
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let input = typos::tokens::Word::new("successs", 0).unwrap();
assert_eq!(
corrections.correct_word(input),
@ -20,7 +20,7 @@ fn correct_word_hit(b: &mut test::Bencher) {
#[bench]
fn correct_word_miss(b: &mut test::Bencher) {
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let input = typos::tokens::Word::new("success", 0).unwrap();
assert_eq!(corrections.correct_word(input), None);
b.iter(|| corrections.correct_word(input));

2
benches/data.rs
View file

@ -28,4 +28,4 @@ fn main() {
}
";
pub const CORPUS: &str = include_str!("../typos/assets/words.csv");
pub const CORPUS: &str = include_str!("../typos-dict/assets/words.csv");

12
benches/file.rs
View file

@ -12,7 +12,7 @@ fn process_empty(b: &mut test::Bencher) {
let sample_path = temp.child("sample");
sample_path.write_str(data::EMPTY).unwrap();
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let parser = typos::tokens::Parser::new();
let checks = typos::checks::CheckSettings::new().build(&corrections, &parser);
b.iter(|| checks.check_file(sample_path.path(), true, typos::report::print_silent));
@ -26,7 +26,7 @@ fn process_no_tokens(b: &mut test::Bencher) {
let sample_path = temp.child("sample");
sample_path.write_str(data::NO_TOKENS).unwrap();
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let parser = typos::tokens::Parser::new();
let checks = typos::checks::CheckSettings::new().build(&corrections, &parser);
b.iter(|| checks.check_file(sample_path.path(), true, typos::report::print_silent));
@ -40,7 +40,7 @@ fn process_single_token(b: &mut test::Bencher) {
let sample_path = temp.child("sample");
sample_path.write_str(data::SINGLE_TOKEN).unwrap();
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let parser = typos::tokens::Parser::new();
let checks = typos::checks::CheckSettings::new().build(&corrections, &parser);
b.iter(|| checks.check_file(sample_path.path(), true, typos::report::print_silent));
@ -54,7 +54,7 @@ fn process_sherlock(b: &mut test::Bencher) {
let sample_path = temp.child("sample");
sample_path.write_str(data::SHERLOCK).unwrap();
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let parser = typos::tokens::Parser::new();
let checks = typos::checks::CheckSettings::new().build(&corrections, &parser);
b.iter(|| checks.check_file(sample_path.path(), true, typos::report::print_silent));
@ -68,7 +68,7 @@ fn process_code(b: &mut test::Bencher) {
let sample_path = temp.child("sample");
sample_path.write_str(data::CODE).unwrap();
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let parser = typos::tokens::Parser::new();
let checks = typos::checks::CheckSettings::new().build(&corrections, &parser);
b.iter(|| checks.check_file(sample_path.path(), true, typos::report::print_silent));
@ -82,7 +82,7 @@ fn process_corpus(b: &mut test::Bencher) {
let sample_path = temp.child("sample");
sample_path.write_str(data::CORPUS).unwrap();
let corrections = typos::BuiltIn::new();
let corrections = typos_dict::BuiltIn::new();
let parser = typos::tokens::Parser::new();
let checks = typos::checks::CheckSettings::new().build(&corrections, &parser);
b.iter(|| checks.check_file(sample_path.path(), true, typos::report::print_silent));

2
src/main.rs
View file

@ -310,7 +310,7 @@ fn run() -> Result<i32, failure::Error> {
config.default.update(&args.overrides);
let config = config;
let dictionary = typos::BuiltIn::new();
let dictionary = typos_dict::BuiltIn::new();
let parser = typos::tokens::ParserBuilder::new()
.ignore_hex(config.default.ignore_hex())

31
typos-dict/Cargo.toml Normal file
View file

@ -0,0 +1,31 @@
[package]
name = "typos-dict"
version = "0.1.0"
authors = ["Ed Page <eopage@gmail.com>"]
description = "Source Code Spelling Correction"
repository = "https://github.com/epage/typos"
documentation = "https://docs.rs/typos-dict"
readme = "README.md"
categories = ["development-tools", "text-processing"]
keywords = ["development", "spelling"]
license = "MIT"
edition = "2018"
[badges]
travis-ci = { repository = "epage/typos" }
appveyor = { repository = "epage/typos" }
[features]
# Support quickly iterating
iterate_unstable = []
[dependencies]
typos = { version = "0.1", path = "../typos" }
phf = { version = "0.7", features = ["unicase"] }
unicase = "1.1"
log = "0.4"
[build-dependencies]
phf_codegen = "0.7"
csv = "1.0"
unicase = "1.1"

0
typos/assets/main.go → typos-dict/assets/main.go
View file

0
typos/assets/words.csv → typos-dict/assets/words.csv
View file

Can't render this file because it is too large.

0
typos/assets/words.go → typos-dict/assets/words.go
View file

0
typos/build.rs → typos-dict/build.rs
View file

97
typos-dict/src/dict.rs Normal file
View file

@ -0,0 +1,97 @@
use std::borrow::Cow;
use unicase::UniCase;
use typos::tokens::Case;
#[derive(Default)]
pub struct BuiltIn {}
impl BuiltIn {
pub fn new() -> Self {
Self {}
}
pub fn correct_ident<'s, 'w>(
&'s self,
_ident: typos::tokens::Identifier<'w>,
) -> Option<Cow<'s, str>> {
None
}
pub fn correct_word<'s, 'w>(&'s self, word: typos::tokens::Word<'w>) -> Option<Cow<'s, str>> {
map_lookup(&crate::dict_codegen::WORD_DICTIONARY, word.token())
.map(|s| case_correct(s, word.case()))
}
}
impl typos::Dictionary for BuiltIn {
fn correct_ident<'s, 'w>(
&'s self,
ident: typos::tokens::Identifier<'w>,
) -> Option<Cow<'s, str>> {
BuiltIn::correct_ident(self, ident)
}
fn correct_word<'s, 'w>(&'s self, word: typos::tokens::Word<'w>) -> Option<Cow<'s, str>> {
BuiltIn::correct_word(self, word)
}
}
fn map_lookup(
map: &'static phf::Map<UniCase<&'static str>, &'static str>,
key: &str,
) -> Option<&'static str> {
// This transmute should be safe as `get` will not store the reference with
// the expanded lifetime. This is due to `Borrow` being overly strict and
// can't have an impl for `&'static str` to `Borrow<&'a str>`.
//
//
// See https://github.com/rust-lang/rust/issues/28853#issuecomment-158735548
unsafe {
let key = ::std::mem::transmute::<_, &'static str>(key);
map.get(&UniCase(key)).cloned()
}
}
fn case_correct(correction: &str, case: Case) -> Cow<'_, str> {
match case {
Case::Lower | Case::None => correction.into(),
Case::Title => {
let mut title = String::with_capacity(correction.as_bytes().len());
let mut char_indices = correction.char_indices();
if let Some((_, c)) = char_indices.next() {
title.extend(c.to_uppercase());
if let Some((i, _)) = char_indices.next() {
title.push_str(&correction[i..]);
}
}
title.into()
}
Case::Scream => correction
.chars()
.flat_map(|c| c.to_uppercase())
.collect::<String>()
.into(),
}
}
#[cfg(test)]
mod test {
use super::*;
#[test]
fn test_case_correct() {
let cases = [
("foo", Case::Lower, "foo"),
("foo", Case::None, "foo"),
("foo", Case::Title, "Foo"),
("foo", Case::Scream, "FOO"),
("fOo", Case::None, "fOo"),
];
for (correction, case, expected) in cases.iter() {
let actual = case_correct(correction, *case);
assert_eq!(*expected, actual);
}
}
}

0
typos/src/dict_codegen.rs → typos-dict/src/dict_codegen.rs
View file

4
typos-dict/src/lib.rs Normal file
View file

@ -0,0 +1,4 @@
mod dict;
mod dict_codegen;
pub use crate::dict::*;

10
typos/Cargo.toml
View file

@ -15,13 +15,8 @@ edition = "2018"
travis-ci = { repository = "epage/typos" }
appveyor = { repository = "epage/typos" }
[features]
# Support quickly iterating
iterate_unstable = []
[dependencies]
failure = "0.1"
phf = { version = "0.7", features = ["unicase"] }
regex = "1.0"
lazy_static = "1.2.0"
serde = { version = "1.0", features = ["derive"] }
@ -32,8 +27,3 @@ bstr = "0.2"
log = "0.4"
unicode-segmentation = "1.3.0"
derive_more = "0.15.0"
[build-dependencies]
phf_codegen = "0.7"
csv = "1.0"
unicase = "1.1"

96
typos/src/dict.rs
View file

@ -1,9 +1,5 @@
use std::borrow::Cow;
use unicase::UniCase;
use crate::tokens::Case;
pub trait Dictionary {
fn correct_ident<'s, 'w>(
&'s self,
@ -12,95 +8,3 @@ pub trait Dictionary {
fn correct_word<'s, 'w>(&'s self, word: crate::tokens::Word<'w>) -> Option<Cow<'s, str>>;
}
#[derive(Default)]
pub struct BuiltIn {}
impl BuiltIn {
pub fn new() -> Self {
Self {}
}
pub fn correct_ident<'s, 'w>(
&'s self,
_ident: crate::tokens::Identifier<'w>,
) -> Option<Cow<'s, str>> {
None
}
pub fn correct_word<'s, 'w>(&'s self, word: crate::tokens::Word<'w>) -> Option<Cow<'s, str>> {
map_lookup(&crate::dict_codegen::WORD_DICTIONARY, word.token())
.map(|s| case_correct(s, word.case()))
}
}
impl Dictionary for BuiltIn {
fn correct_ident<'s, 'w>(
&'s self,
ident: crate::tokens::Identifier<'w>,
) -> Option<Cow<'s, str>> {
BuiltIn::correct_ident(self, ident)
}
fn correct_word<'s, 'w>(&'s self, word: crate::tokens::Word<'w>) -> Option<Cow<'s, str>> {
BuiltIn::correct_word(self, word)
}
}
fn map_lookup(
map: &'static phf::Map<UniCase<&'static str>, &'static str>,
key: &str,
) -> Option<&'static str> {
// This transmute should be safe as `get` will not store the reference with
// the expanded lifetime. This is due to `Borrow` being overly strict and
// can't have an impl for `&'static str` to `Borrow<&'a str>`.
//
//
// See https://github.com/rust-lang/rust/issues/28853#issuecomment-158735548
unsafe {
let key = ::std::mem::transmute::<_, &'static str>(key);
map.get(&UniCase(key)).cloned()
}
}
fn case_correct(correction: &str, case: Case) -> Cow<'_, str> {
match case {
Case::Lower | Case::None => correction.into(),
Case::Title => {
let mut title = String::with_capacity(correction.as_bytes().len());
let mut char_indices = correction.char_indices();
if let Some((_, c)) = char_indices.next() {
title.extend(c.to_uppercase());
if let Some((i, _)) = char_indices.next() {
title.push_str(&correction[i..]);
}
}
title.into()
}
Case::Scream => correction
.chars()
.flat_map(|c| c.to_uppercase())
.collect::<String>()
.into(),
}
}
#[cfg(test)]
mod test {
use super::*;
#[test]
fn test_case_correct() {
let cases = [
("foo", Case::Lower, "foo"),
("foo", Case::None, "foo"),
("foo", Case::Title, "Foo"),
("foo", Case::Scream, "FOO"),
("fOo", Case::None, "fOo"),
];
for (correction, case, expected) in cases.iter() {
let actual = case_correct(correction, *case);
assert_eq!(*expected, actual);
}
}
}

1
typos/src/lib.rs
View file

@ -1,5 +1,4 @@
mod dict;
mod dict_codegen;
pub mod checks;
pub mod report;