read from stdin

Cargo.toml

@@ -6,6 +6,12 @@ edition = "2018"
 
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 
+[profile.release]
+opt-level = 3
+lto = "fat"
+codegen-units = 1
+panic = "unwind"
+
 [features]
 perl = ["pcre"]
 

src/main.rs

@@ -1,56 +1,36 @@
-#[cfg(not(feature = "perl"))]
-extern crate regex;
+#![allow(dead_code)]
 
-#[cfg(not(feature = "perl"))]
-use regex::Regex;
+mod re;
+mod text;
 
-#[cfg(not(feature = "perl"))]
-fn main() {
+fn main() -> Result<(), Box<dyn std::error::Error>>
+{
     let args: Vec<String> = std::env::args().collect();
 
     if args.len() < 4 {
 	println!("Usage: {} <str> <regex> <group>", args[0]);
+	println!("Pass `-' as `<str>' to read lines from stdin");
+	std::process::exit(1);
     } else {
-	let re = Regex::new(&args[2]).unwrap();
+	let re = re::Regex::compile(&args[2])?;
 	let text = &args[1];
 	let group: usize = args[3].parse().expect("Invalid group number.");
 
-	match re.captures(&text) {
-
-	    Some(groups) => {
-		if group > groups.len() {
-		    eprintln!("Invalid group number.");
-		} else {
-		    println!("{}", groups.get(group).unwrap().as_str());
-		}
-	    }
-	    None => (),
-	}
-    }
+	if text == "-" {
+	    text::stdin_lines(|text| -> Result<bool, re::Error> {
+		match re.exec(&text)? {
+		    Some(g) if g.len() > group => println!("{}", &g[group]),
+		    _ => (),
 		}
-
-#[cfg(feature = "perl")]
-extern crate pcre;
-
-#[cfg(feature = "perl")]
-use pcre::Pcre;
-
-#[cfg(feature = "perl")]
-fn main () {
-    let args: Vec<String> = std::env::args().collect();
-
-    if args.len() < 4 {
-	println!("Usage: {} <str> <regex> <group>", args[0]);
+		Ok(true)
+	    })?;
 	} else {
-	let mut re = Pcre::compile(&args[2]).expect("Invalid regex.");
-	let text = &args[1];
-	let group: usize = args[3].parse().expect("Invalid group number.");
-
-	match re.exec(&text) {
 
-	    Some(m) if m.string_count() > group => println!("{}", m.group(group)),
+	    match re.exec(&text)? {
+		Some(g) if g.len() > group => println!("{}", &g[group]),
 		_ => (),
 	    }
 	}
     }
-
+    Ok(())
+}

src/re.rs

@@ -0,0 +1,118 @@
+#![allow(unused_imports)]
+
+use std::{
+    error,
+    fmt::{
+	self,
+	Write,
+    },
+    sync::{
+	Arc,
+	Mutex,
+    }
+};
+
+pub type Groups = Vec<String>;
+
+pub struct Regex
+{
+    #[cfg(feature="perl")]
+    internal: Arc<Mutex<pcre::Pcre>>,
+    #[cfg(not(feature = "perl"))]
+    internal: regex::Regex,
+}
+
+#[derive(Debug)]
+pub enum Error
+{
+    Compile(String),
+    Execute,
+    Internal,
+    Unknown,
+}
+
+impl error::Error for Error{}
+impl fmt::Display for Error
+{
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result
+    {
+	write!(f, "regex error: ")?;
+	match self {
+	    Error::Compile(s) => write!(f, "compilation failed: {}", s),
+	    Error::Execute => write!(f, "execution failed"),
+	    Error::Internal => write!(f, "internal"),
+	    _ => write!(f, "unknown"),
+	}
+    }
+}
+
+impl Regex {
+    pub fn compile(string: impl AsRef<str>) -> Result<Self, Error> 
+    {
+	#[cfg(feature = "perl")]
+	return Ok(Self{internal: Arc::new(Mutex::new(pcre::Pcre::compile(string.as_ref())?))});
+	#[cfg(not(feature = "perl"))]
+	return Ok(Self{internal: regex::Regex::new(string.as_ref())?});
+    }
+
+    pub fn exec(&self, string: impl AsRef<str>) -> Result<Option<Groups>, Error>
+    {
+	#[cfg(feature = "perl")]
+	return {
+	    let mut re = self.internal.lock().unwrap();
+	    Ok(match re.exec(string.as_ref()) {
+		Some(m) => {
+		    let len = m.string_count();
+		    let mut output = Vec::with_capacity(len);
+		    for i in 0..len {
+			output.push(m.group(i).to_owned());
+		    }
+		    Some(output)
+		},
+		None => None,
+	    })
+	};
+	#[cfg(not(feature = "perl"))]
+	return {
+	    Ok(match self.internal.captures(string.as_ref()) {
+		Some(m) => {
+		    let mut output = Vec::with_capacity(m.len());
+		    for i in 0..m.len() {
+			let ma = m.get(i).unwrap();
+			let mut op = String::with_capacity(ma.range().len());
+			write!(op, "{}", ma.as_str())?;
+			output.push(op);
+		    }
+		    Some(output)
+		},
+		None => None,
+	    })
+	};
+    }
+}
+
+impl From<fmt::Error> for Error
+{
+    fn from(_er: fmt::Error) -> Self
+    {
+	Self::Internal
+    }
+}
+
+#[cfg(not(feature = "perl"))]
+impl From<regex::Error> for Error
+{
+    fn from(er: regex::Error) -> Self
+    {
+	Self::Compile(format!("{}", er))
+    }
+}
+
+#[cfg(feature = "perl")]
+impl From<pcre::CompilationError> for Error
+{
+    fn from(er: pcre::CompilationError) -> Self
+    {
+	Self::Compile(format!("{}", er))
+    }
+}

src/text.rs

@@ -0,0 +1,39 @@
+use std::{
+    io::{
+	self,
+	Read,
+    },
+    error::Error,
+    marker::{
+	Send, Sync,
+    },
+};
+
+/// Read whole stdin
+pub fn stdin() -> io::Result<String>
+{
+    let mut buffer = String::new();
+    io::stdin().read_to_string(&mut buffer)?;
+
+    Ok(buffer)
+}
+
+/// Read each line with lambda
+pub fn stdin_lines<E, F: FnMut(&str) -> Result<bool, E>>(mut callback: F) -> io::Result<usize>
+    where E: Into<Box<dyn Error + Send + Sync>>
+{
+    let mut lines=0;
+    let mut input = String::new();
+    let stdin = io::stdin();
+    while match stdin.read_line(&mut input) {
+	Ok(0) => return Ok(lines),
+	Err(e) => return Err(e),
+	Ok(v) if input.as_bytes()[v-1] == b'\n' => callback(&input[..v]).or_else(|e| Err(io::Error::new(io::ErrorKind::Other, e)))?,
+	Ok(_) => callback(&input[..]).or_else(|e| Err(io::Error::new(io::ErrorKind::Other, e)))?,
+    } {
+	input.clear();
+	lines+=1;
+    }
+
+    Ok(lines)
+}