rpms
/
rust-grep
20
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update to 0.1.9

Browse Source
epel9
Josh Stone 7 years ago
parent fc7616ce77
commit d77b3dcd43
6 changed files with 7 additions and 809 deletions
Show Stats Download Patch File Download Diff File

1
.gitignore vendored
Unescape View File

@ -1,2 +1,3 @@
/grep-0.1.7.crate /grep-0.1.7.crate
/grep-0.1.8.crate /grep-0.1.8.crate
/grep-0.1.9.crate

460
0001-grep-upgrade-to-regex-syntax-0.5.patch
Unescape View File

@ -1,460 +0,0 @@
From cd08707c7c82058559bd5557efb3c1d0379dbf1d Mon Sep 17 00:00:00 2001
From: Andrew Gallant <jamslam@gmail.com>
Date: Tue, 13 Mar 2018 20:38:50 -0400
Subject: [PATCH 1/2] grep: upgrade to regex-syntax 0.5
This update brings with it many bug fixes:
* Better error messages are printed overall. We also include
explicit call out for unsupported features like backreferences
and look-around.
* Regexes like `\s*{` no longer emit incomprehensible errors.
* Unicode escape sequences, such as `\u{..}` are now supported.
For the most part, this upgrade was done in a straight-forward way. We
resist the urge to refactor the `grep` crate, in anticipation of it
being rewritten anyway.
Note that we removed the `--fixed-strings` suggestion whenever a regex
syntax error occurs. In practice, I've found that it results in a lot of
false positives, and I believe that its use is not as paramount now that
regex parse errors are much more readable.
Closes #268, Closes #395, Closes #702, Closes #853
---
src/literals.rs | 119 +++++++++++++++++++++++-----------------------
src/nonl.rs | 85 ++++++++++++++++++---------------
src/search.rs | 19 ++++----
src/word_boundary.rs | 31 ++++++------
4 files changed, 130 insertions(+), 124 deletions(-)
diff --git a/src/literals.rs b/src/literals.rs
index eebeac4c7249..3e1c385bcda9 100644
--- a/src/literals.rs
+++ b/src/literals.rs
@@ -10,10 +10,8 @@ principled.
use std::cmp;
use regex::bytes::RegexBuilder;
-use syntax::{
- Expr, Literals, Lit,
- ByteClass, ByteRange, CharClass, ClassRange, Repeater,
-};
+use syntax::hir::{self, Hir, HirKind};
+use syntax::hir::literal::{Literal, Literals};
#[derive(Clone, Debug)]
pub struct LiteralSets {
@@ -23,12 +21,12 @@ pub struct LiteralSets {
}
impl LiteralSets {
- pub fn create(expr: &Expr) -> Self {
+ pub fn create(expr: &Hir) -> Self {
let mut required = Literals::empty();
union_required(expr, &mut required);
LiteralSets {
- prefixes: expr.prefixes(),
- suffixes: expr.suffixes(),
+ prefixes: Literals::prefixes(expr),
+ suffixes: Literals::suffixes(expr),
required: required,
}
}
@@ -93,60 +91,52 @@ impl LiteralSets {
}
}
-fn union_required(expr: &Expr, lits: &mut Literals) {
- use syntax::Expr::*;
- match *expr {
- Literal { ref chars, casei: false } => {
- let s: String = chars.iter().cloned().collect();
- lits.cross_add(s.as_bytes());
+fn union_required(expr: &Hir, lits: &mut Literals) {
+ match *expr.kind() {
+ HirKind::Literal(hir::Literal::Unicode(c)) => {
+ let mut buf = [0u8; 4];
+ lits.cross_add(c.encode_utf8(&mut buf).as_bytes());
}
- Literal { ref chars, casei: true } => {
- for &c in chars {
- let cls = CharClass::new(vec![
- ClassRange { start: c, end: c },
- ]).case_fold();
- if !lits.add_char_class(&cls) {
- lits.cut();
- return;
- }
+ HirKind::Literal(hir::Literal::Byte(b)) => {
+ lits.cross_add(&[b]);
+ }
+ HirKind::Class(hir::Class::Unicode(ref cls)) => {
+ if count_unicode_class(cls) >= 5 || !lits.add_char_class(cls) {
+ lits.cut();
+ }
+ }
+ HirKind::Class(hir::Class::Bytes(ref cls)) => {
+ if count_byte_class(cls) >= 5 || !lits.add_byte_class(cls) {
+ lits.cut();
}
}
- LiteralBytes { ref bytes, casei: false } => {
- lits.cross_add(bytes);
+ HirKind::Group(hir::Group { ref hir, .. }) => {
+ union_required(&**hir, lits);
}
- LiteralBytes { ref bytes, casei: true } => {
- for &b in bytes {
- let cls = ByteClass::new(vec![
- ByteRange { start: b, end: b },
- ]).case_fold();
- if !lits.add_byte_class(&cls) {
+ HirKind::Repetition(ref x) => {
+ match x.kind {
+ hir::RepetitionKind::ZeroOrOne => lits.cut(),
+ hir::RepetitionKind::ZeroOrMore => lits.cut(),
+ hir::RepetitionKind::OneOrMore => {
+ union_required(&x.hir, lits);
lits.cut();
- return;
+ }
+ hir::RepetitionKind::Range(ref rng) => {
+ let (min, max) = match *rng {
+ hir::RepetitionRange::Exactly(m) => (m, Some(m)),
+ hir::RepetitionRange::AtLeast(m) => (m, None),
+ hir::RepetitionRange::Bounded(m, n) => (m, Some(n)),
+ };
+ repeat_range_literals(
+ &x.hir, min, max, x.greedy, lits, union_required);
}
}
}
- Class(_) => {
- lits.cut();
- }
- ClassBytes(_) => {
- lits.cut();
+ HirKind::Concat(ref es) if es.is_empty() => {}
+ HirKind::Concat(ref es) if es.len() == 1 => {
+ union_required(&es[0], lits)
}
- Group { ref e, .. } => {
- union_required(&**e, lits);
- }
- Repeat { r: Repeater::ZeroOrOne, .. } => lits.cut(),
- Repeat { r: Repeater::ZeroOrMore, .. } => lits.cut(),
- Repeat { ref e, r: Repeater::OneOrMore, .. } => {
- union_required(&**e, lits);
- lits.cut();
- }
- Repeat { ref e, r: Repeater::Range { min, max }, greedy } => {
- repeat_range_literals(
- &**e, min, max, greedy, lits, union_required);
- }
- Concat(ref es) if es.is_empty() => {}
- Concat(ref es) if es.len() == 1 => union_required(&es[0], lits),
- Concat(ref es) => {
+ HirKind::Concat(ref es) => {
for e in es {
let mut lits2 = lits.to_empty();
union_required(e, &mut lits2);
@@ -157,7 +147,6 @@ fn union_required(expr: &Expr, lits: &mut Literals) {
if lits2.contains_empty() {
lits.cut();
}
- // if !lits.union(lits2) {
if !lits.cross_product(&lits2) {
// If this expression couldn't yield any literal that
// could be extended, then we need to quit. Since we're
@@ -167,15 +156,15 @@ fn union_required(expr: &Expr, lits: &mut Literals) {
}
}
}
- Alternate(ref es) => {
+ HirKind::Alternation(ref es) => {
alternate_literals(es, lits, union_required);
}
_ => lits.cut(),
}
}
-fn repeat_range_literals<F: FnMut(&Expr, &mut Literals)>(
- e: &Expr,
+fn repeat_range_literals<F: FnMut(&Hir, &mut Literals)>(
+ e: &Hir,
min: u32,
max: Option<u32>,
_greedy: bool,
@@ -204,8 +193,8 @@ fn repeat_range_literals<F: FnMut(&Expr, &mut Literals)>(
}
}
-fn alternate_literals<F: FnMut(&Expr, &mut Literals)>(
- es: &[Expr],
+fn alternate_literals<F: FnMut(&Hir, &mut Literals)>(
+ es: &[Hir],
lits: &mut Literals,
mut f: F,
) {
@@ -234,11 +223,21 @@ fn alternate_literals<F: FnMut(&Expr, &mut Literals)>(
}
lits.cut();
if !lcs.is_empty() {
- lits.add(Lit::empty());
- lits.add(Lit::new(lcs.to_vec()));
+ lits.add(Literal::empty());
+ lits.add(Literal::new(lcs.to_vec()));
}
}
+/// Return the number of characters in the given class.
+fn count_unicode_class(cls: &hir::ClassUnicode) -> u32 {
+ cls.iter().map(|r| 1 + (r.end() as u32 - r.start() as u32)).sum()
+}
+
+/// Return the number of bytes in the given class.
+fn count_byte_class(cls: &hir::ClassBytes) -> u32 {
+ cls.iter().map(|r| 1 + (r.end() as u32 - r.start() as u32)).sum()
+}
+
/// Converts an arbitrary sequence of bytes to a literal suitable for building
/// a regular expression.
fn bytes_to_regex(bs: &[u8]) -> String {
diff --git a/src/nonl.rs b/src/nonl.rs
index 361b0b003eb7..3beb5f61ce57 100644
--- a/src/nonl.rs
+++ b/src/nonl.rs
@@ -1,4 +1,4 @@
-use syntax::Expr;
+use syntax::hir::{self, Hir, HirKind};
use {Error, Result};
@@ -9,59 +9,66 @@ use {Error, Result};
///
/// If `byte` is not an ASCII character (i.e., greater than `0x7F`), then this
/// function panics.
-pub fn remove(expr: Expr, byte: u8) -> Result<Expr> {
- // TODO(burntsushi): There is a bug in this routine where only `\n` is
- // handled correctly. Namely, `AnyChar` and `AnyByte` need to be translated
- // to proper character classes instead of the special `AnyCharNoNL` and
- // `AnyByteNoNL` classes.
- use syntax::Expr::*;
+pub fn remove(expr: Hir, byte: u8) -> Result<Hir> {
assert!(byte <= 0x7F);
let chr = byte as char;
assert!(chr.len_utf8() == 1);
- Ok(match expr {
- Literal { chars, casei } => {
- if chars.iter().position(|&c| c == chr).is_some() {
+ Ok(match expr.into_kind() {
+ HirKind::Empty => Hir::empty(),
+ HirKind::Literal(hir::Literal::Unicode(c)) => {
+ if c == chr {
return Err(Error::LiteralNotAllowed(chr));
}
- Literal { chars: chars, casei: casei }
+ Hir::literal(hir::Literal::Unicode(c))
}
- LiteralBytes { bytes, casei } => {
- if bytes.iter().position(|&b| b == byte).is_some() {
+ HirKind::Literal(hir::Literal::Byte(b)) => {
+ if b as char == chr {
return Err(Error::LiteralNotAllowed(chr));
}
- LiteralBytes { bytes: bytes, casei: casei }
+ Hir::literal(hir::Literal::Byte(b))
}
- AnyChar => AnyCharNoNL,
- AnyByte => AnyByteNoNL,
- Class(mut cls) => {
- cls.remove(chr);
- Class(cls)
- }
- ClassBytes(mut cls) => {
- cls.remove(byte);
- ClassBytes(cls)
- }
- Group { e, i, name } => {
- Group {
- e: Box::new(remove(*e, byte)?),
- i: i,
- name: name,
+ HirKind::Class(hir::Class::Unicode(mut cls)) => {
+ let remove = hir::ClassUnicode::new(Some(
+ hir::ClassUnicodeRange::new(chr, chr),
+ ));
+ cls.difference(&remove);
+ if cls.iter().next().is_none() {
+ return Err(Error::LiteralNotAllowed(chr));
}
+ Hir::class(hir::Class::Unicode(cls))
}
- Repeat { e, r, greedy } => {
- Repeat {
- e: Box::new(remove(*e, byte)?),
- r: r,
- greedy: greedy,
+ HirKind::Class(hir::Class::Bytes(mut cls)) => {
+ let remove = hir::ClassBytes::new(Some(
+ hir::ClassBytesRange::new(byte, byte),
+ ));
+ cls.difference(&remove);
+ if cls.iter().next().is_none() {
+ return Err(Error::LiteralNotAllowed(chr));
}
+ Hir::class(hir::Class::Bytes(cls))
+ }
+ HirKind::Anchor(x) => Hir::anchor(x),
+ HirKind::WordBoundary(x) => Hir::word_boundary(x),
+ HirKind::Repetition(mut x) => {
+ x.hir = Box::new(remove(*x.hir, byte)?);
+ Hir::repetition(x)
+ }
+ HirKind::Group(mut x) => {
+ x.hir = Box::new(remove(*x.hir, byte)?);
+ Hir::group(x)
}
- Concat(exprs) => {
- Concat(exprs.into_iter().map(|e| remove(e, byte)).collect::<Result<Vec<Expr>>>()?)
+ HirKind::Concat(xs) => {
+ let xs = xs.into_iter()
+ .map(|e| remove(e, byte))
+ .collect::<Result<Vec<Hir>>>()?;
+ Hir::concat(xs)
}
- Alternate(exprs) => {
- Alternate(exprs.into_iter().map(|e| remove(e, byte)).collect::<Result<Vec<Expr>>>()?)
+ HirKind::Alternation(xs) => {
+ let xs = xs.into_iter()
+ .map(|e| remove(e, byte))
+ .collect::<Result<Vec<Hir>>>()?;
+ Hir::alternation(xs)
}
- e => e,
})
}
diff --git a/src/search.rs b/src/search.rs
index 8d056796ac14..1d5d7e29cccf 100644
--- a/src/search.rs
+++ b/src/search.rs
@@ -1,10 +1,10 @@
use memchr::{memchr, memrchr};
use regex::bytes::{Regex, RegexBuilder};
-use syntax;
use literals::LiteralSets;
use nonl;
-use syntax::Expr;
+use syntax::ParserBuilder;
+use syntax::hir::Hir;
use word_boundary::strip_unicode_word_boundaries;
use Result;
@@ -166,7 +166,7 @@ impl GrepBuilder {
/// Creates a new regex from the given expression with the current
/// configuration.
- fn regex(&self, expr: &Expr) -> Result<Regex> {
+ fn regex(&self, expr: &Hir) -> Result<Regex> {
let mut builder = RegexBuilder::new(&expr.to_string());
builder.unicode(true);
self.regex_build(builder)
@@ -184,15 +184,16 @@ impl GrepBuilder {
/// Parses the underlying pattern and ensures the pattern can never match
/// the line terminator.
- fn parse(&self) -> Result<syntax::Expr> {
- let expr =
- syntax::ExprBuilder::new()
- .allow_bytes(true)
- .unicode(true)
+ fn parse(&self) -> Result<Hir> {
+ let expr = ParserBuilder::new()
+ .allow_invalid_utf8(true)
.case_insensitive(self.is_case_insensitive()?)
+ .multi_line(true)
+ .build()
.parse(&self.pattern)?;
+ debug!("original regex HIR pattern:\n{}", expr);
let expr = nonl::remove(expr, self.opts.line_terminator)?;
- debug!("regex ast:\n{:#?}", expr);
+ debug!("transformed regex HIR pattern:\n{}", expr);
Ok(expr)
}
diff --git a/src/word_boundary.rs b/src/word_boundary.rs
index 6df5c6574933..8e6b86d12df8 100644
--- a/src/word_boundary.rs
+++ b/src/word_boundary.rs
@@ -1,4 +1,4 @@
-use syntax::Expr;
+use syntax::hir::{self, Hir, HirKind};
/// Strips Unicode word boundaries from the given expression.
///
@@ -8,7 +8,7 @@ use syntax::Expr;
/// false negatives.
///
/// If no word boundaries could be stripped, then None is returned.
-pub fn strip_unicode_word_boundaries(expr: &Expr) -> Option<Expr> {
+pub fn strip_unicode_word_boundaries(expr: &Hir) -> Option<Hir> {
// The real reason we do this is because Unicode word boundaries are the
// one thing that Rust's regex DFA engine can't handle. When it sees a
// Unicode word boundary among non-ASCII text, it falls back to one of the
@@ -16,23 +16,24 @@ pub fn strip_unicode_word_boundaries(expr: &Expr) -> Option<Expr> {
// a regex to find candidate matches without a Unicode word boundary. We'll
// only then use the full (and slower) regex to confirm a candidate as a
// match or not during search.
- use syntax::Expr::*;
-
- match *expr {
- Concat(ref es) if !es.is_empty() => {
+ //
+ // It looks like we only check the outer edges for `\b`? I guess this is
+ // an attempt to optimize for the `-w/--word-regexp` flag? ---AG
+ match *expr.kind() {
+ HirKind::Concat(ref es) if !es.is_empty() => {
let first = is_unicode_word_boundary(&es[0]);
let last = is_unicode_word_boundary(es.last().unwrap());
// Be careful not to strip word boundaries if there are no other
// expressions to match.
match (first, last) {
(true, false) if es.len() > 1 => {
- Some(Concat(es[1..].to_vec()))
+ Some(Hir::concat(es[1..].to_vec()))
}
(false, true) if es.len() > 1 => {
- Some(Concat(es[..es.len() - 1].to_vec()))
+ Some(Hir::concat(es[..es.len() - 1].to_vec()))
}
(true, true) if es.len() > 2 => {
- Some(Concat(es[1..es.len() - 1].to_vec()))
+ Some(Hir::concat(es[1..es.len() - 1].to_vec()))
}
_ => None,
}
@@ -42,13 +43,11 @@ pub fn strip_unicode_word_boundaries(expr: &Expr) -> Option<Expr> {
}
/// Returns true if the given expression is a Unicode word boundary.
-fn is_unicode_word_boundary(expr: &Expr) -> bool {
- use syntax::Expr::*;
-
- match *expr {
- WordBoundary => true,
- NotWordBoundary => true,
- Group { ref e, .. } => is_unicode_word_boundary(e),
+fn is_unicode_word_boundary(expr: &Hir) -> bool {
+ match *expr.kind() {
+ HirKind::WordBoundary(hir::WordBoundary::Unicode) => true,
+ HirKind::WordBoundary(hir::WordBoundary::UnicodeNegate) => true,
+ HirKind::Group(ref x) => is_unicode_word_boundary(&x.hir),
_ => false,
}
}
--
2.14.3

328
0002-grep-add-perfect-smart-case-detection.patch
Unescape View File

@ -1,328 +0,0 @@
From 42b8132d0ad1918c1c0dc677015d87c12819fa26 Mon Sep 17 00:00:00 2001
From: Andrew Gallant <jamslam@gmail.com>
Date: Tue, 13 Mar 2018 21:43:23 -0400
Subject: [PATCH 2/2] grep: add "perfect" smart case detection
This commit removes the previous smart case detection logic and replaces
it with detection based on the regex AST. This particular AST is a faithful
representation of the concrete syntax, which lets us be very precise in
how we handle it.
Closes #851
---
src/lib.rs | 1 +
src/search.rs | 58 +++------------
src/smart_case.rs | 191 +++++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 201 insertions(+), 49 deletions(-)
create mode 100644 src/smart_case.rs
diff --git a/src/lib.rs b/src/lib.rs
index 3b2f0ebd65d5..023cd64ac36a 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -19,6 +19,7 @@ pub use search::{Grep, GrepBuilder, Iter, Match};
mod literals;
mod nonl;
mod search;
+mod smart_case;
mod word_boundary;
/// Result is a convenient type alias that fixes the type of the error to
diff --git a/src/search.rs b/src/search.rs
index 1d5d7e29cccf..49ddf1f875c4 100644
--- a/src/search.rs
+++ b/src/search.rs
@@ -1,10 +1,11 @@
use memchr::{memchr, memrchr};
+use syntax::ParserBuilder;
+use syntax::hir::Hir;
use regex::bytes::{Regex, RegexBuilder};
use literals::LiteralSets;
use nonl;
-use syntax::ParserBuilder;
-use syntax::hir::Hir;
+use smart_case::Cased;
use word_boundary::strip_unicode_word_boundaries;
use Result;
@@ -205,7 +206,11 @@ impl GrepBuilder {
if !self.opts.case_smart {
return Ok(false);
}
- Ok(!has_uppercase_literal(&self.pattern))
+ let cased = match Cased::from_pattern(&self.pattern) {
+ None => return Ok(false),
+ Some(cased) => cased,
+ };
+ Ok(cased.any_literal && !cased.any_uppercase)
}
}
@@ -311,44 +316,15 @@ impl<'b, 's> Iterator for Iter<'b, 's> {
}
}
-/// Determine whether the pattern contains an uppercase character which should
-/// negate the effect of the smart-case option.
-///
-/// Ideally we would be able to check the AST in order to correctly handle
-/// things like '\p{Ll}' and '\p{Lu}' (which should be treated as explicitly
-/// cased), but we don't currently have that option. For now, our 'good enough'
-/// solution is to simply perform a semi-naïve scan of the input pattern and
-/// ignore all characters following a '\'. The ExprBuilder will handle any
-/// actual errors, and this at least lets us support the most common cases,
-/// like 'foo\w' and 'foo\S', in an intuitive manner.
-fn has_uppercase_literal(pattern: &str) -> bool {
- let mut chars = pattern.chars();
- while let Some(c) = chars.next() {
- if c == '\\' {
- chars.next();
- } else if c.is_uppercase() {
- return true;
- }
- }
- false
-}
-
#[cfg(test)]
mod tests {
- #![allow(unused_imports)]
-
use memchr::{memchr, memrchr};
use regex::bytes::Regex;
- use super::{GrepBuilder, Match, has_uppercase_literal};
+ use super::{GrepBuilder, Match};
static SHERLOCK: &'static [u8] = include_bytes!("./data/sherlock.txt");
- #[allow(dead_code)]
- fn s(bytes: &[u8]) -> String {
- String::from_utf8(bytes.to_vec()).unwrap()
- }
-
fn find_lines(pat: &str, haystack: &[u8]) -> Vec<Match> {
let re = Regex::new(pat).unwrap();
let mut lines = vec![];
@@ -377,20 +353,4 @@ mod tests {
assert_eq!(expected.len(), got.len());
assert_eq!(expected, got);
}
-
- #[test]
- fn pattern_case() {
- assert_eq!(has_uppercase_literal(&"".to_string()), false);
- assert_eq!(has_uppercase_literal(&"foo".to_string()), false);
- assert_eq!(has_uppercase_literal(&"Foo".to_string()), true);
- assert_eq!(has_uppercase_literal(&"foO".to_string()), true);
- assert_eq!(has_uppercase_literal(&"foo\\\\".to_string()), false);
- assert_eq!(has_uppercase_literal(&"foo\\w".to_string()), false);
- assert_eq!(has_uppercase_literal(&"foo\\S".to_string()), false);
- assert_eq!(has_uppercase_literal(&"foo\\p{Ll}".to_string()), true);
- assert_eq!(has_uppercase_literal(&"foo[a-z]".to_string()), false);
- assert_eq!(has_uppercase_literal(&"foo[A-Z]".to_string()), true);
- assert_eq!(has_uppercase_literal(&"foo[\\S\\t]".to_string()), false);
- assert_eq!(has_uppercase_literal(&"foo\\\\S".to_string()), true);
- }
}
diff --git a/src/smart_case.rs b/src/smart_case.rs
new file mode 100644
index 000000000000..1379b32620bc
--- /dev/null
+++ b/src/smart_case.rs
@@ -0,0 +1,191 @@
+use syntax::ast::{self, Ast};
+use syntax::ast::parse::Parser;
+
+/// The results of analyzing a regex for cased literals.
+#[derive(Clone, Debug, Default)]
+pub struct Cased {
+ /// True if and only if a literal uppercase character occurs in the regex.
+ ///
+ /// A regex like `\pL` contains no uppercase literals, even though `L`
+ /// is uppercase and the `\pL` class contains uppercase characters.
+ pub any_uppercase: bool,
+ /// True if and only if the regex contains any literal at all. A regex like
+ /// `\pL` has this set to false.
+ pub any_literal: bool,
+}
+
+impl Cased {
+ /// Returns a `Cased` value by doing analysis on the AST of `pattern`.
+ ///
+ /// If `pattern` is not a valid regular expression, then `None` is
+ /// returned.
+ pub fn from_pattern(pattern: &str) -> Option<Cased> {
+ Parser::new()
+ .parse(pattern)
+ .map(|ast| Cased::from_ast(&ast))
+ .ok()
+ }
+
+ fn from_ast(ast: &Ast) -> Cased {
+ let mut cased = Cased::default();
+ cased.from_ast_impl(ast);
+ cased
+ }
+
+ fn from_ast_impl(&mut self, ast: &Ast) {
+ if self.done() {
+ return;
+ }
+ match *ast {
+ Ast::Empty(_)
+ | Ast::Flags(_)
+ | Ast::Dot(_)
+ | Ast::Assertion(_)
+ | Ast::Class(ast::Class::Unicode(_))
+ | Ast::Class(ast::Class::Perl(_)) => {}
+ Ast::Literal(ref x) => {
+ self.from_ast_literal(x);
+ }
+ Ast::Class(ast::Class::Bracketed(ref x)) => {
+ self.from_ast_class_set(&x.kind);
+ }
+ Ast::Repetition(ref x) => {
+ self.from_ast_impl(&x.ast);
+ }
+ Ast::Group(ref x) => {
+ self.from_ast_impl(&x.ast);
+ }
+ Ast::Alternation(ref alt) => {
+ for x in &alt.asts {
+ self.from_ast_impl(x);
+ }
+ }
+ Ast::Concat(ref alt) => {
+ for x in &alt.asts {
+ self.from_ast_impl(x);
+ }
+ }
+ }
+ }
+
+ fn from_ast_class_set(&mut self, ast: &ast::ClassSet) {
+ if self.done() {
+ return;
+ }
+ match *ast {
+ ast::ClassSet::Item(ref item) => {
+ self.from_ast_class_set_item(item);
+ }
+ ast::ClassSet::BinaryOp(ref x) => {
+ self.from_ast_class_set(&x.lhs);
+ self.from_ast_class_set(&x.rhs);
+ }
+ }
+ }
+
+ fn from_ast_class_set_item(&mut self, ast: &ast::ClassSetItem) {
+ if self.done() {
+ return;
+ }
+ match *ast {
+ ast::ClassSetItem::Empty(_)
+ | ast::ClassSetItem::Ascii(_)
+ | ast::ClassSetItem::Unicode(_)
+ | ast::ClassSetItem::Perl(_) => {}
+ ast::ClassSetItem::Literal(ref x) => {
+ self.from_ast_literal(x);
+ }
+ ast::ClassSetItem::Range(ref x) => {
+ self.from_ast_literal(&x.start);
+ self.from_ast_literal(&x.end);
+ }
+ ast::ClassSetItem::Bracketed(ref x) => {
+ self.from_ast_class_set(&x.kind);
+ }
+ ast::ClassSetItem::Union(ref union) => {
+ for x in &union.items {
+ self.from_ast_class_set_item(x);
+ }
+ }
+ }
+ }
+
+ fn from_ast_literal(&mut self, ast: &ast::Literal) {
+ self.any_literal = true;
+ self.any_uppercase = self.any_uppercase || ast.c.is_uppercase();
+ }
+
+ /// Returns true if and only if the attributes can never change no matter
+ /// what other AST it might see.
+ fn done(&self) -> bool {
+ self.any_uppercase && self.any_literal
+ }
+}
+
+#[cfg(test)]
+mod tests {
+ use super::*;
+
+ fn cased(pattern: &str) -> Cased {
+ Cased::from_pattern(pattern).unwrap()
+ }
+
+ #[test]
+ fn various() {
+ let x = cased("");
+ assert!(!x.any_uppercase);
+ assert!(!x.any_literal);
+
+ let x = cased("foo");
+ assert!(!x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased("Foo");
+ assert!(x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased("foO");
+ assert!(x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo\\");
+ assert!(!x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo\w");
+ assert!(!x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo\S");
+ assert!(!x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo\p{Ll}");
+ assert!(!x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo[a-z]");
+ assert!(!x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo[A-Z]");
+ assert!(x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo[\S\t]");
+ assert!(!x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"foo\\S");
+ assert!(x.any_uppercase);
+ assert!(x.any_literal);
+
+ let x = cased(r"\p{Ll}");
+ assert!(!x.any_uppercase);
+ assert!(!x.any_literal);
+
+ let x = cased(r"aBc\w");
+ assert!(x.any_uppercase);
+ assert!(x.any_literal);
+ }
+}
--
2.14.3

12
grep-0.1.8-fix-metadata.diff
Unescape View File

@ -1,12 +0,0 @@
--- grep-0.1.8/Cargo.toml 1970-01-01T01:00:00+01:00
+++ grep-0.1.8/Cargo.toml 2018-06-12T09:25:27.056693+02:00
@@ -28,7 +28,7 @@
version = "2"
[dependencies.regex]
-version = "0.2.1"
+version = "1"
[dependencies.regex-syntax]
-version = "0.4.0"
+version = "0.6"

13
rust-grep.spec
Unescape View File

@ -5,19 +5,13 @@
%global crate grep %global crate grep
Name: rust-%{crate} Name: rust-%{crate}
Version: 0.1.8 Version: 0.1.9
Release: 4%{?dist} Release: 1%{?dist}
Summary: Fast line oriented regex searching as a library Summary: Fast line oriented regex searching as a library
License: Unlicense or MIT License: Unlicense or MIT
URL: https://crates.io/crates/grep URL: https://crates.io/crates/grep
Source0: https://crates.io/api/v1/crates/%{crate}/%{version}/download#/%{crate}-%{version}.crate Source0: https://crates.io/api/v1/crates/%{crate}/%{version}/download#/%{crate}-%{version}.crate
# Initial patched metadata
# * bump regex and regex-syntax
Patch0: grep-0.1.8-fix-metadata.diff
# https://github.com/BurntSushi/ripgrep/pull/858/
Patch1: 0001-grep-upgrade-to-regex-syntax-0.5.patch
Patch2: 0002-grep-add-perfect-smart-case-detection.patch
ExclusiveArch: %{rust_arches} ExclusiveArch: %{rust_arches}
@ -62,6 +56,9 @@ which use %{crate} from crates.io.
%{cargo_registry}/%{crate}-%{version}/ %{cargo_registry}/%{crate}-%{version}/
%changelog %changelog
* Sat Aug 04 2018 Josh Stone <jistone@redhat.com> - 0.1.9-1
- Update to 0.1.9
* Sat Jul 14 2018 Fedora Release Engineering <releng@fedoraproject.org> - 0.1.8-4 * Sat Jul 14 2018 Fedora Release Engineering <releng@fedoraproject.org> - 0.1.8-4
- Rebuilt for https://fedoraproject.org/wiki/Fedora_29_Mass_Rebuild - Rebuilt for https://fedoraproject.org/wiki/Fedora_29_Mass_Rebuild

2
sources
Unescape View File

@ -1 +1 @@
SHA512 (grep-0.1.8.crate) = d0a4e9dd51402ed8036aa5d2cfcdb184538d2bf0c81baac75c73e80d05945a63cb08a931d87fa526695af93271cecaed6650a045904ceaa05ed6e06fd23e537b SHA512 (grep-0.1.9.crate) = 1ce3c1c1faf2d001bdafc6e021ec3523e08cfaec611bbaf70492cf0038578c2e738ead0df826229fbc0f5e54ff3f9f9cf928be9223bb53907f12b6f468545ef3

Write Preview
Loading…
Cancel
Save