syntax: add 'std' feature

In effect, this adds support for no_std by depending on only core and
alloc. There is still currently some benefit to enabling std support,
namely, getting the 'std::error::Error' trait impls for the various
error types. (Although, it seems like the 'Error' trait is going to get
moved to 'core' finally.) Otherwise, the only 'std' things we use are in
tests for tweaking stack sizes.

This is the first step in an effort to make 'regex' itself work without
depending on 'std'. 'regex' itself will be more precarious since it uses
things like HashMap and Mutex that we'll need to find a way around.
Getting around HashMap is easy (just use BTreeMap), but figuring out how
to synchronize the threadpool will be interesting.

Ref #476, Ref #477

Author: BurntSushi

Cargo.toml

@@ -14,7 +14,8 @@ finite automata and guarantees linear time matching on all inputs.
 categories = ["text-processing"]
 autotests = false
 exclude = ["/scripts/*", "/.github/*"]
-edition = "2018"
+edition = "2021"
+resolver = "2"
 
 [workspace]
 members = [

regex-syntax/Cargo.toml

@@ -13,7 +13,8 @@ edition = "2021"
 # Features are documented in the "Crate features" section of the crate docs:
 # https://docs.rs/regex-syntax/*/#crate-features
 [features]
-default = ["unicode"]
+default = ["std", "unicode"]
+std = []
 
 unicode = [
   "unicode-age",

regex-syntax/src/ast/mod.rs

@@ -2,8 +2,9 @@
 Defines an abstract syntax for regular expressions.
 */
 
-use std::cmp::Ordering;
-use std::fmt;
+use core::cmp::Ordering;
+
+use alloc::{boxed::Box, string::String, vec, vec::Vec};
 
 pub use crate::ast::visitor::{visit, Visitor};
 
@@ -174,23 +175,24 @@ pub enum ErrorKind {
     UnsupportedLookAround,
 }
 
+#[cfg(feature = "std")]
 impl std::error::Error for Error {}
 
-impl fmt::Display for Error {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl core::fmt::Display for Error {
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
         crate::error::Formatter::from(self).fmt(f)
     }
 }
 
-impl fmt::Display for ErrorKind {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl core::fmt::Display for ErrorKind {
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
         use self::ErrorKind::*;
         match *self {
             CaptureLimitExceeded => write!(
                 f,
                 "exceeded the maximum number of \
                  capturing groups ({})",
-                ::std::u32::MAX
+                u32::MAX
             ),
             ClassEscapeInvalid => {
                 write!(f, "invalid escape sequence found in character class")
@@ -283,8 +285,8 @@ pub struct Span {
     pub end: Position,
 }
 
-impl fmt::Debug for Span {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl core::fmt::Debug for Span {
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
         write!(f, "Span({:?}, {:?})", self.start, self.end)
     }
 }
@@ -316,8 +318,8 @@ pub struct Position {
     pub column: usize,
 }
 
-impl fmt::Debug for Position {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl core::fmt::Debug for Position {
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
         write!(
             f,
             "Position(o: {:?}, l: {:?}, c: {:?})",
@@ -497,8 +499,8 @@ impl Ast {
 ///
 /// This implementation uses constant stack space and heap space proportional
 /// to the size of the `Ast`.
-impl fmt::Display for Ast {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl core::fmt::Display for Ast {
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
         use crate::ast::print::Printer;
         Printer::new().print(self, f)
     }
@@ -1315,7 +1317,7 @@ pub enum Flag {
 /// space but heap space proportional to the depth of the `Ast`.
 impl Drop for Ast {
     fn drop(&mut self) {
-        use std::mem;
+        use core::mem;
 
         match *self {
             Ast::Empty(_)
@@ -1365,7 +1367,7 @@ impl Drop for Ast {
 /// stack space but heap space proportional to the depth of the `ClassSet`.
 impl Drop for ClassSet {
     fn drop(&mut self) {
-        use std::mem;
+        use core::mem;
 
         match *self {
             ClassSet::Item(ref item) => match *item {

regex-syntax/src/ast/parse.rs

@@ -2,17 +2,26 @@
 This module provides a regular expression parser.
 */
 
-use std::borrow::Borrow;
-use std::cell::{Cell, RefCell};
-use std::mem;
-use std::result;
-
-use crate::ast::{self, Ast, Position, Span};
-use crate::either::Either;
-
-use crate::is_meta_character;
-
-type Result<T> = result::Result<T, ast::Error>;
+use core::{
+    borrow::Borrow,
+    cell::{Cell, RefCell},
+    mem,
+};
+
+use alloc::{
+    boxed::Box,
+    string::{String, ToString},
+    vec,
+    vec::Vec,
+};
+
+use crate::{
+    ast::{self, Ast, Position, Span},
+    either::Either,
+    is_meta_character,
+};
+
+type Result<T> = core::result::Result<T, ast::Error>;
 
 /// A primitive is an expression with no sub-expressions. This includes
 /// literals, assertions and non-set character classes. This representation
@@ -1533,9 +1542,6 @@ impl<'s, P: Borrow<Parser>> ParserI<'s, P> {
     /// Assuming the preconditions are met, this routine can never fail.
     #[inline(never)]
     fn parse_octal(&self) -> ast::Literal {
-        use std::char;
-        use std::u32;
-
         assert!(self.parser().octal);
         assert!('0' <= self.char() && self.char() <= '7');
         let start = self.pos();
@@ -1600,9 +1606,6 @@ impl<'s, P: Borrow<Parser>> ParserI<'s, P> {
         &self,
         kind: ast::HexLiteralKind,
     ) -> Result<ast::Literal> {
-        use std::char;
-        use std::u32;
-
         let mut scratch = self.parser().scratch.borrow_mut();
         scratch.clear();
 
@@ -1646,9 +1649,6 @@ impl<'s, P: Borrow<Parser>> ParserI<'s, P> {
         &self,
         kind: ast::HexLiteralKind,
     ) -> Result<ast::Literal> {
-        use std::char;
-        use std::u32;
-
         let mut scratch = self.parser().scratch.borrow_mut();
         scratch.clear();
 
@@ -2146,7 +2146,7 @@ impl<'p, 's, P: Borrow<Parser>> NestLimiter<'p, 's, P> {
         let new = self.depth.checked_add(1).ok_or_else(|| {
             self.p.error(
                 span.clone(),
-                ast::ErrorKind::NestLimitExceeded(::std::u32::MAX),
+                ast::ErrorKind::NestLimitExceeded(u32::MAX),
             )
         })?;
         let limit = self.p.parser().nest_limit;
@@ -2297,11 +2297,14 @@ fn specialize_err<T>(
 
 #[cfg(test)]
 mod tests {
-    use std::ops::Range;
+    use core::ops::Range;
+
+    use alloc::format;
 
-    use super::{Parser, ParserBuilder, ParserI, Primitive};
     use crate::ast::{self, Ast, Position, Span};
 
+    use super::*;
+
     // Our own assert_eq, which has slightly better formatting (but honestly
     // still kind of crappy).
     macro_rules! assert_eq {
@@ -4272,7 +4275,7 @@ bar
                 Ok(Primitive::Literal(ast::Literal {
                     span: span(0..pat.len()),
                     kind: ast::LiteralKind::Octal,
-                    c: ::std::char::from_u32(i).unwrap(),
+                    c: char::from_u32(i).unwrap(),
                 }))
             );
         }
@@ -4347,7 +4350,7 @@ bar
                 Ok(Primitive::Literal(ast::Literal {
                     span: span(0..pat.len()),
                     kind: ast::LiteralKind::HexFixed(ast::HexLiteralKind::X),
-                    c: ::std::char::from_u32(i).unwrap(),
+                    c: char::from_u32(i).unwrap(),
                 }))
             );
         }
@@ -4378,7 +4381,7 @@ bar
     #[test]
     fn parse_hex_four() {
         for i in 0..65536 {
-            let c = match ::std::char::from_u32(i) {
+            let c = match char::from_u32(i) {
                 None => continue,
                 Some(c) => c,
             };
@@ -4442,7 +4445,7 @@ bar
     #[test]
     fn parse_hex_eight() {
         for i in 0..65536 {
-            let c = match ::std::char::from_u32(i) {
+            let c = match char::from_u32(i) {
                 None => continue,
                 Some(c) => c,
             };

regex-syntax/src/ast/print.rs

@@ -2,10 +2,13 @@
 This module provides a regular expression printer for `Ast`.
 */
 
-use std::fmt;
+use core::fmt;
 
-use crate::ast::visitor::{self, Visitor};
-use crate::ast::{self, Ast};
+use crate::ast::{
+    self,
+    visitor::{self, Visitor},
+    Ast,
+};
 
 /// A builder for constructing a printer.
 ///
@@ -395,9 +398,12 @@ impl<W: fmt::Write> Writer<W> {
 
 #[cfg(test)]
 mod tests {
-    use super::Printer;
+    use alloc::string::String;
+
     use crate::ast::parse::ParserBuilder;
 
+    use super::*;
+
     fn roundtrip(given: &str) {
         roundtrip_with(|b| b, given);
     }

regex-syntax/src/ast/visitor.rs

@@ -1,4 +1,4 @@
-use std::fmt;
+use alloc::{vec, vec::Vec};
 
 use crate::ast::{self, Ast};
 
@@ -475,8 +475,8 @@ impl<'a> ClassInduct<'a> {
     }
 }
 
-impl<'a> fmt::Debug for ClassFrame<'a> {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl<'a> core::fmt::Debug for ClassFrame<'a> {
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
         let x = match *self {
             ClassFrame::Union { .. } => "Union",
             ClassFrame::Binary { .. } => "Binary",
@@ -487,8 +487,8 @@ impl<'a> fmt::Debug for ClassFrame<'a> {
     }
 }
 
-impl<'a> fmt::Debug for ClassInduct<'a> {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl<'a> core::fmt::Debug for ClassInduct<'a> {
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
         let x = match *self {
             ClassInduct::Item(it) => match *it {
                 ast::ClassSetItem::Empty(_) => "Item(Empty)",

regex-syntax/src/error.rs

@@ -1,9 +1,13 @@
-use std::cmp;
-use std::fmt;
-use std::result;
+use core::{cmp, fmt, result};
 
-use crate::ast;
-use crate::hir;
+use alloc::{
+    format,
+    string::{String, ToString},
+    vec,
+    vec::Vec,
+};
+
+use crate::{ast, hir};
 
 /// A type alias for dealing with errors returned by this crate.
 pub type Result<T> = result::Result<T, Error>;
@@ -35,6 +39,7 @@ impl From<hir::Error> for Error {
     }
 }
 
+#[cfg(feature = "std")]
 impl std::error::Error for Error {}
 
 impl fmt::Display for Error {
@@ -266,11 +271,13 @@ impl<'p> Spans<'p> {
 }
 
 fn repeat_char(c: char, count: usize) -> String {
-    ::std::iter::repeat(c).take(count).collect()
+    core::iter::repeat(c).take(count).collect()
 }
 
 #[cfg(test)]
 mod tests {
+    use alloc::string::ToString;
+
     use crate::ast::parse::Parser;
 
     fn assert_panic_message(pattern: &str, expected_msg: &str) {

regex-syntax/src/hir/interval.rs

@@ -1,8 +1,6 @@
-use std::char;
-use std::cmp;
-use std::fmt::Debug;
-use std::slice;
-use std::u8;
+use core::{char, cmp, fmt::Debug, slice};
+
+use alloc::vec::Vec;
 
 use crate::unicode;