feat: add fancy-regex

2020-11-24 23:52:30 +08:00
parent 9eecc77ac0
commit 113911b6ac
4 changed files with 202 additions and 0 deletions
--- a/__misc/fancy-regex/src/main.rs
+++ b/__misc/fancy-regex/src/main.rs
@@ -0,0 +1,104 @@
+use fancy_regex::internal::{analyze, compile, run_trace, Insn, Prog};
+use fancy_regex::*;
+use std::env;
+use std::str::FromStr;
+
+fn main() {
+    let mut args = env::args().skip(1);
+    if let Some(cmd) = args.next() {
+        if cmd == "parse" {
+            if let Some(re) = args.next() {
+                let e = Expr::parse_tree(&re);
+                println!("{:#?}", e);
+            }
+        } else if cmd == "analyze" {
+            if let Some(re) = args.next() {
+                let tree = Expr::parse_tree(&re).unwrap();
+                let a = analyze(&tree);
+                println!("{:#?}", a);
+            }
+        } else if cmd == "compile" {
+            if let Some(re) = args.next() {
+                let r = Regex::new(&re).unwrap();
+                r.debug_print();
+            }
+        } else if cmd == "run" {
+            let re = args.next().expect("expected regexp argument");
+            let r = Regex::new(&re).unwrap();
+            let text = args.next().expect("expected text argument");
+            let mut pos = 0;
+            if let Some(pos_str) = args.next() {
+                pos = usize::from_str(&pos_str).unwrap();
+            }
+            if let Some(caps) = r.captures_from_pos(&text, pos).unwrap() {
+                print!("captures:");
+                for i in 0..caps.len() {
+                    print!(" {}:", i);
+                    if let Some(m) = caps.get(i) {
+                        print!("[{}..{}] \"{}\"", m.start(), m.end(), m.as_str());
+                    } else {
+                        print!("_");
+                    }
+                }
+                println!("");
+                for cap in caps.iter() {
+                    println!("iterate {:?}", cap);
+                }
+            } else {
+                println!("no match");
+            }
+        } else if cmd == "trace" {
+            if let Some(re) = args.next() {
+                let prog = prog(&re);
+                if let Some(s) = args.next() {
+                    run_trace(&prog, &s, 0).unwrap();
+                }
+            }
+        } else if cmd == "trace-inner" {
+            if let Some(re) = args.next() {
+                let tree = Expr::parse_tree(&re).unwrap();
+                let a = analyze(&tree).unwrap();
+                let p = compile(&a).unwrap();
+                if let Some(s) = args.next() {
+                    run_trace(&p, &s, 0).unwrap();
+                }
+            }
+        } else if cmd == "graph" {
+            let re = args.next().expect("expected regexp argument");
+            graph(&re);
+        } else {
+            println!("commands: parse|analyze|compile|graph <expr>, run|trace|trace-inner <expr> <input>");
+        }
+    }
+}
+
+fn graph(re: &str) {
+    let prog = prog(re);
+    println!("digraph G {{");
+    for (i, insn) in prog.body.iter().enumerate() {
+        let label = format!("{:?}", insn)
+            .replace(r#"\"#, r#"\\"#)
+            .replace(r#"""#, r#"\""#);
+        println!(r#"{:3} [label="{}: {}"];"#, i, i, label);
+        match *insn {
+            Insn::Split(a, b) => {
+                println!("{:3} -> {};", i, a);
+                println!("{:3} -> {};", i, b);
+            }
+            Insn::Jmp(target) => {
+                println!("{:3} -> {};", i, target);
+            }
+            Insn::End => {}
+            _ => {
+                println!("{:3} -> {};", i, i + 1);
+            }
+        }
+    }
+    println!("}}");
+}
+
+fn prog(re: &str) -> Prog {
+    let tree = Expr::parse_tree(re).expect("Expected parsing regex to work");
+    let result = analyze(&tree).expect("Expected analyze to succeed");
+    compile(&result).expect("Expected compile to succeed")
+}