mikros/rust
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Optimize literal, doc comment lint as well, extract function.

Browse Source
This commit is contained in:
Hans Kratz 2021-11-04 23:31:42 +01:00
parent a5b25a2cfa
commit 7885233df0
5 changed files with 54 additions and 49 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
compiler/rustc_ast/src/lib.rs
View File

@ -16,6 +16,7 @@
#![feature(nll)]
#![feature(min_specialization)]
#![recursion_limit = "256"]
#![feature(slice_internals)]
#[macro_use]
extern crate rustc_macros;
@ -25,6 +26,7 @@ pub mod util {
pub mod comments;
pub mod literal;
pub mod parser;
pub mod unicode;
}
pub mod ast;

44
compiler/rustc_ast/src/util/unicode.rs Normal file
View File

@ -0,0 +1,44 @@
pub const TEXT_FLOW_CONTROL_CHARS: &[char] = &[
'\u{202A}', '\u{202B}', '\u{202D}', '\u{202E}', '\u{2066}', '\u{2067}', '\u{2068}', '\u{202C}',
'\u{2069}',
];
#[inline]
pub fn contains_text_flow_control_chars(s: &str) -> bool {
// Char - UTF-8
// U+202A - E2 80 AA
// U+202B - E2 80 AB
// U+202C - E2 80 AC
// U+202D - E2 80 AD
// U+202E - E2 80 AE
// U+2066 - E2 81 A6
// U+2067 - E2 81 A7
// U+2068 - E2 81 A8
// U+2069 - E2 81 A9
let mut bytes = s.as_bytes();
loop {
match core::slice::memchr::memchr(0xE2, &bytes) {
Some(idx) => {
// bytes are valid UTF-8 -> E2 must be followed by two bytes
let ch = &bytes[idx..idx + 3];
match ch[1] {
0x80 => {
if (0xAA..=0xAE).contains(&ch[2]) {
break true;
}
}
0x81 => {
if (0xA6..=0xA9).contains(&ch[2]) {
break true;
}
}
_ => {}
}
bytes = &bytes[idx + 3..];
}
None => {
break false;
}
}
}
}

4
compiler/rustc_lint/src/context.rs
View File

@ -16,9 +16,9 @@
use self::TargetLint::*;
use crate::hidden_unicode_codepoints::UNICODE_TEXT_FLOW_CHARS;
use crate::levels::{is_known_lint_tool, LintLevelsBuilder};
use crate::passes::{EarlyLintPassObject, LateLintPassObject};
use ast::util::unicode::TEXT_FLOW_CONTROL_CHARS;
use rustc_ast as ast;
use rustc_data_structures::fx::FxHashMap;
use rustc_data_structures::sync;
@ -602,7 +602,7 @@ fn lookup_with_diagnostics(
let spans: Vec<_> = content
.char_indices()
.filter_map(|(i, c)| {
UNICODE_TEXT_FLOW_CHARS.contains(&c).then(|| {
TEXT_FLOW_CONTROL_CHARS.contains(&c).then(|| {
let lo = span.lo() + BytePos(2 + i as u32);
(c, span.with_lo(lo).with_hi(lo + BytePos(c.len_utf8() as u32)))
})

12
compiler/rustc_lint/src/hidden_unicode_codepoints.rs
View File

@ -1,4 +1,5 @@
use crate::{EarlyContext, EarlyLintPass, LintContext};
use ast::util::unicode::{contains_text_flow_control_chars, TEXT_FLOW_CONTROL_CHARS};
use rustc_ast as ast;
use rustc_errors::{Applicability, SuggestionStyle};
use rustc_span::{BytePos, Span, Symbol};
@ -37,11 +38,6 @@
declare_lint_pass!(HiddenUnicodeCodepoints => [TEXT_DIRECTION_CODEPOINT_IN_LITERAL]);
crate const UNICODE_TEXT_FLOW_CHARS: &[char] = &[
'\u{202A}', '\u{202B}', '\u{202D}', '\u{202E}', '\u{2066}', '\u{2067}', '\u{2068}', '\u{202C}',
'\u{2069}',
];
impl HiddenUnicodeCodepoints {
fn lint_text_direction_codepoint(
&self,
@ -57,7 +53,7 @@ fn lint_text_direction_codepoint(
.as_str()
.char_indices()
.filter_map(|(i, c)| {
UNICODE_TEXT_FLOW_CHARS.contains(&c).then(|| {
TEXT_FLOW_CONTROL_CHARS.contains(&c).then(|| {
let lo = span.lo() + BytePos(i as u32 + padding);
(c, span.with_lo(lo).with_hi(lo + BytePos(c.len_utf8() as u32)))
})
@ -131,7 +127,7 @@ fn lint_text_direction_codepoint(
impl EarlyLintPass for HiddenUnicodeCodepoints {
fn check_attribute(&mut self, cx: &EarlyContext<'_>, attr: &ast::Attribute) {
if let ast::AttrKind::DocComment(_, comment) = attr.kind {
if comment.as_str().contains(UNICODE_TEXT_FLOW_CHARS) {
if contains_text_flow_control_chars(&comment.as_str()) {
self.lint_text_direction_codepoint(cx, comment, attr.span, 0, false, "doc comment");
}
}
@ -142,7 +138,7 @@ fn check_expr(&mut self, cx: &EarlyContext<'_>, expr: &ast::Expr) {
let (text, span, padding) = match &expr.kind {
ast::ExprKind::Lit(ast::Lit { token, kind, span }) => {
let text = token.symbol;
if !text.as_str().contains(UNICODE_TEXT_FLOW_CHARS) {
if !contains_text_flow_control_chars(&text.as_str()) {
return;
}
let padding = match kind {

41
compiler/rustc_parse/src/lexer/mod.rs
View File

@ -1,6 +1,7 @@
use rustc_ast::ast::{self, AttrStyle};
use rustc_ast::token::{self, CommentKind, Token, TokenKind};
use rustc_ast::tokenstream::{Spacing, TokenStream};
use rustc_ast::util::unicode::contains_text_flow_control_chars;
use rustc_errors::{error_code, Applicability, DiagnosticBuilder, FatalError, PResult};
use rustc_lexer::unescape::{self, Mode};
use rustc_lexer::{Base, DocStyle, RawStrError};
@ -137,45 +138,7 @@ fn lint_unicode_text_flow(&self, start: BytePos) {
// Opening delimiter of the length 2 is not included into the comment text.
let content_start = start + BytePos(2);
let content = self.str_from(content_start);
// Char - UTF-8
// U+202A - E2 80 AA
// U+202B - E2 80 AB
// U+202C - E2 80 AC
// U+202D - E2 80 AD
// U+202E - E2 80 AE
// U+2066 - E2 81 A6
// U+2067 - E2 81 A7
// U+2068 - E2 81 A8
// U+2069 - E2 81 A9
let mut bytes = content.as_bytes();
let contains_flow_control_chars = loop {
match core::slice::memchr::memchr(0xE2, &bytes) {
Some(idx) => {
// bytes are valid UTF-8 -> E2 must be followed by two bytes
let ch = &bytes[idx..idx + 3];
match ch[1] {
0x80 => {
if (0xAA..=0xAE).contains(&ch[2]) {
break true;
}
}
0x81 => {
if (0xA6..=0xA9).contains(&ch[2]) {
break true;
}
}
_ => {}
}
bytes = &bytes[idx + 3..];
}
None => {
break false;
}
}
};
if contains_flow_control_chars {
if contains_text_flow_control_chars(content) {
let span = self.mk_sp(start, self.pos);
self.sess.buffer_lint_with_diagnostic(
&TEXT_DIRECTION_CODEPOINT_IN_COMMENT,