restore the default SIGPIPE behavior as a temporary workaround

See BurntSushi#200.

Cargo.toml

@@ -38,6 +38,7 @@ env_logger = { version = "0.4", default-features = false }
 grep = { version = "0.1.5", path = "grep" }
 ignore = { version = "0.2.2", path = "ignore" }
 lazy_static = "0.2"
+libc = "0.2"
 log = "0.3"
 memchr = "1"
 memmap = "0.5"

src/main.rs

@@ -54,6 +54,7 @@ mod worker;
 pub type Result<T> = result::Result<T, Box<Error + Send + Sync>>;
 
 fn main() {
+    reset_sigpipe();
     match Args::parse().map(Arc::new).and_then(run) {
         Ok(0) => process::exit(1),
         Ok(_) => process::exit(0),
@@ -329,3 +330,22 @@ fn eprint_nothing_searched() {
                applied a filter you didn't expect. \
                Try running again with --debug.");
 }
+
+// The Rust standard library suppresses the default SIGPIPE behavior, so that
+// writing to a closed pipe doesn't kill the process. The goal is to instead
+// handle errors through the normal result mechanism. Ripgrep needs some
+// refactoring before it will be able to do that, however, so we re-enable the
+// standard SIGPIPE behavior as a workaround. See
+// https://github.com/BurntSushi/ripgrep/issues/200.
+#[cfg(unix)]
+fn reset_sigpipe() {
+    extern crate libc;
+    unsafe {
+        libc::signal(libc::SIGPIPE, libc::SIG_DFL);
+    }
+}
+
+#[cfg(not(unix))]
+fn reset_sigpipe() {
+    // no-op
+}