Improve smart case to only consider literal uppercase chars (#103)

Cargo.lock

@@ -9,6 +9,7 @@ dependencies = [
  "ignore 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "num_cpus 1.7.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "regex 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "regex-syntax 0.4.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "tempdir 0.3.5 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 

Cargo.toml

@@ -25,6 +25,7 @@ ansi_term = "0.9"
 clap = "2.26.0"
 atty = "0.2"
 regex = "0.2"
+regex-syntax = "0.4"
 ignore = "0.2"
 num_cpus = "1.6.2"
 

src/internal.rs

@@ -4,6 +4,7 @@ use std::io::Write;
 
 use lscolors::LsColors;
 use walk::FileType;
+use regex_syntax::{Expr, ExprBuilder};
 
 /// Root directory
 #[cfg(unix)]
@@ -78,3 +79,28 @@ pub fn error(message: &str) -> ! {
     writeln!(&mut ::std::io::stderr(), "{}", message).expect("Failed writing to stderr");
     process::exit(1);
 }
+
+/// Determine if a regex pattern contains a literal uppercase character.
+pub fn pattern_has_uppercase_char(pattern: &str) -> bool {
+    ExprBuilder::new()
+        .parse(pattern)
+        .map(|expr| expr_has_uppercase_char(&expr))
+        .unwrap_or(false)
+}
+
+/// Determine if a regex expression contains a literal uppercase character.
+fn expr_has_uppercase_char(expr: &Expr) -> bool {
+    match *expr {
+        Expr::Literal { ref chars, .. } => chars.iter().any(|c| c.is_uppercase()),
+        Expr::Class(ref ranges) => {
+            ranges.iter().any(|r| {
+                r.start.is_uppercase() || r.end.is_uppercase()
+            })
+        }
+        Expr::Group { ref e, .. } => expr_has_uppercase_char(e),
+        Expr::Repeat { ref e, .. } => expr_has_uppercase_char(e),
+        Expr::Concat(ref es) => es.iter().any(expr_has_uppercase_char),
+        Expr::Alternate(ref es) => es.iter().any(expr_has_uppercase_char),
+        _ => false,
+    }
+}

src/main.rs

@@ -3,6 +3,7 @@ extern crate clap;
 extern crate ansi_term;
 extern crate atty;
 extern crate regex;
+extern crate regex_syntax;
 extern crate ignore;
 extern crate num_cpus;
 
@@ -22,7 +23,7 @@ use std::time;
 use atty::Stream;
 use regex::RegexBuilder;
 
-use internal::{error, FdOptions, PathDisplay, ROOT_DIR};
+use internal::{error, pattern_has_uppercase_char, FdOptions, PathDisplay, ROOT_DIR};
 use lscolors::LsColors;
 use walk::FileType;
 
@@ -65,11 +66,8 @@ fn main() {
 
     // The search will be case-sensitive if the command line flag is set or
     // if the pattern has an uppercase character (smart case).
-    let case_sensitive = if !matches.is_present("ignore-case") {
-        matches.is_present("case-sensitive") || pattern.chars().any(char::is_uppercase)
-    } else {
-        false
-    };
+    let case_sensitive = !matches.is_present("ignore-case") &&
+        (matches.is_present("case-sensitive") || pattern_has_uppercase_char(pattern));
 
     let colored_output = match matches.value_of("color") {
         Some("always") => true,

tests/tests.rs

@@ -115,6 +115,14 @@ fn test_smart_case() {
     te.assert_output(&["C.Foo"], "one/two/C.Foo2");
 
     te.assert_output(&["Foo"], "one/two/C.Foo2");
+
+    // Only literal uppercase chars should trigger case sensitivity.
+    te.assert_output(
+        &["\\Ac"],
+        "one/two/c.foo
+        one/two/C.Foo2",
+    );
+    te.assert_output(&["\\AC"], "one/two/C.Foo2");
 }
 
 /// Case sensitivity (--case-sensitive)