mirror of
https://github.com/rust-lang/rust-analyzer.git
synced 2025-09-28 21:05:02 +00:00
Finish implementing char validation
This commit is contained in:
parent
d1b242262a
commit
c56db92d1f
3 changed files with 93 additions and 8 deletions
|
@ -219,7 +219,7 @@ mod tests {
|
||||||
|
|
||||||
#[test]
|
#[test]
|
||||||
fn test_unicode_escapes() {
|
fn test_unicode_escapes() {
|
||||||
let unicode_escapes = &[r"{DEAD}", "{BEEF}", "{FF}", ""];
|
let unicode_escapes = &[r"{DEAD}", "{BEEF}", "{FF}", "{}", ""];
|
||||||
for escape in unicode_escapes {
|
for escape in unicode_escapes {
|
||||||
let escape_sequence = format!(r"'\u{}'", escape);
|
let escape_sequence = format!(r"'\u{}'", escape);
|
||||||
let component = closed_char_component(&escape_sequence);
|
let component = closed_char_component(&escape_sequence);
|
||||||
|
|
|
@ -1,3 +1,5 @@
|
||||||
|
use std::u32;
|
||||||
|
|
||||||
use crate::{
|
use crate::{
|
||||||
algo::visit::{visitor_ctx, VisitorCtx},
|
algo::visit::{visitor_ctx, VisitorCtx},
|
||||||
ast::{self, AstNode},
|
ast::{self, AstNode},
|
||||||
|
@ -42,15 +44,82 @@ fn validate_char(node: ast::Char, errors: &mut Vec<SyntaxError>) {
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
AsciiCodeEscape => {
|
AsciiCodeEscape => {
|
||||||
// TODO:
|
// An AsciiCodeEscape has 4 chars, example: `\xDD`
|
||||||
// * First digit is octal
|
if text.len() < 4 {
|
||||||
// * Second digit is hex
|
errors.push(SyntaxError::new(TooShortAsciiCodeEscape, range));
|
||||||
|
} else {
|
||||||
|
assert!(text.chars().count() == 4, "AsciiCodeEscape cannot be longer than 4 chars");
|
||||||
|
|
||||||
|
match u8::from_str_radix(&text[2..], 16) {
|
||||||
|
Ok(code) if code < 128 => { /* Escape code is valid */ },
|
||||||
|
Ok(_) => errors.push(SyntaxError::new(AsciiCodeEscapeOutOfRange, range)),
|
||||||
|
Err(_) => errors.push(SyntaxError::new(MalformedAsciiCodeEscape, range)),
|
||||||
|
}
|
||||||
|
|
||||||
|
}
|
||||||
}
|
}
|
||||||
UnicodeEscape => {
|
UnicodeEscape => {
|
||||||
// TODO:
|
assert!(&text[..2] == "\\u", "UnicodeEscape always starts with \\u");
|
||||||
// * Only hex digits or underscores allowed
|
|
||||||
// * Max 6 chars
|
if text.len() == 2 {
|
||||||
// * Within allowed range (must be at most 10FFFF)
|
// No starting `{`
|
||||||
|
errors.push(SyntaxError::new(MalformedUnicodeEscape, range));
|
||||||
|
return;
|
||||||
|
}
|
||||||
|
|
||||||
|
if text.len() == 3 {
|
||||||
|
// Only starting `{`
|
||||||
|
errors.push(SyntaxError::new(UnclosedUnicodeEscape, range));
|
||||||
|
return;
|
||||||
|
}
|
||||||
|
|
||||||
|
let mut code = String::new();
|
||||||
|
let mut closed = false;
|
||||||
|
for c in text[3..].chars() {
|
||||||
|
assert!(!closed, "no characters after escape is closed");
|
||||||
|
|
||||||
|
if c.is_digit(16) {
|
||||||
|
code.push(c);
|
||||||
|
} else if c == '_' {
|
||||||
|
// Reject leading _
|
||||||
|
if code.len() == 0 {
|
||||||
|
errors.push(SyntaxError::new(MalformedUnicodeEscape, range));
|
||||||
|
return;
|
||||||
|
}
|
||||||
|
} else if c == '}' {
|
||||||
|
closed = true;
|
||||||
|
} else {
|
||||||
|
errors.push(SyntaxError::new(MalformedUnicodeEscape, range));
|
||||||
|
return;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
if !closed {
|
||||||
|
errors.push(SyntaxError::new(UnclosedUnicodeEscape, range))
|
||||||
|
}
|
||||||
|
|
||||||
|
if code.len() == 0 {
|
||||||
|
errors.push(SyntaxError::new(EmptyUnicodeEcape, range));
|
||||||
|
return;
|
||||||
|
}
|
||||||
|
|
||||||
|
if code.len() > 6 {
|
||||||
|
errors.push(SyntaxError::new(OverlongUnicodeEscape, range));
|
||||||
|
}
|
||||||
|
|
||||||
|
match u32::from_str_radix(&code, 16) {
|
||||||
|
Ok(code_u32) if code_u32 > 0x10FFFF => {
|
||||||
|
errors.push(SyntaxError::new(UnicodeEscapeOutOfRange, range));
|
||||||
|
}
|
||||||
|
Ok(_) => {
|
||||||
|
// Valid escape code
|
||||||
|
}
|
||||||
|
Err(_) => {
|
||||||
|
errors.push(SyntaxError::new(MalformedUnicodeEscape, range));
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
// FIXME: we really need tests for this
|
||||||
}
|
}
|
||||||
// Code points are always valid
|
// Code points are always valid
|
||||||
CodePoint => (),
|
CodePoint => (),
|
||||||
|
|
|
@ -69,6 +69,14 @@ pub enum SyntaxErrorKind {
|
||||||
LongChar,
|
LongChar,
|
||||||
EmptyAsciiEscape,
|
EmptyAsciiEscape,
|
||||||
InvalidAsciiEscape,
|
InvalidAsciiEscape,
|
||||||
|
TooShortAsciiCodeEscape,
|
||||||
|
AsciiCodeEscapeOutOfRange,
|
||||||
|
MalformedAsciiCodeEscape,
|
||||||
|
UnclosedUnicodeEscape,
|
||||||
|
MalformedUnicodeEscape,
|
||||||
|
EmptyUnicodeEcape,
|
||||||
|
OverlongUnicodeEscape,
|
||||||
|
UnicodeEscapeOutOfRange,
|
||||||
}
|
}
|
||||||
|
|
||||||
#[derive(Debug, Clone, PartialEq, Eq, Hash)]
|
#[derive(Debug, Clone, PartialEq, Eq, Hash)]
|
||||||
|
@ -83,6 +91,14 @@ impl fmt::Display for SyntaxErrorKind {
|
||||||
EmptyChar => write!(f, "Empty char literal"),
|
EmptyChar => write!(f, "Empty char literal"),
|
||||||
UnclosedChar => write!(f, "Unclosed char literal"),
|
UnclosedChar => write!(f, "Unclosed char literal"),
|
||||||
LongChar => write!(f, "Char literal should be one character long"),
|
LongChar => write!(f, "Char literal should be one character long"),
|
||||||
|
TooShortAsciiCodeEscape => write!(f, "Escape sequence should have two digits"),
|
||||||
|
AsciiCodeEscapeOutOfRange => write!(f, "Escape sequence should be between \\x00 and \\x7F"),
|
||||||
|
MalformedAsciiCodeEscape => write!(f, "Escape sequence should be a hexadecimal number"),
|
||||||
|
UnclosedUnicodeEscape => write!(f, "Missing `}}`"),
|
||||||
|
MalformedUnicodeEscape => write!(f, "Malformed unicode escape sequence"),
|
||||||
|
EmptyUnicodeEcape => write!(f, "Empty unicode escape sequence"),
|
||||||
|
OverlongUnicodeEscape => write!(f, "Unicode escape sequence should have at most 6 digits"),
|
||||||
|
UnicodeEscapeOutOfRange => write!(f, "Unicode escape code should be at most 0x10FFFF"),
|
||||||
ParseError(msg) => write!(f, "{}", msg.0),
|
ParseError(msg) => write!(f, "{}", msg.0),
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
Loading…
Add table
Add a link
Reference in a new issue