Furst/src/route.rs

use std::env;
use std::fmt::Write as _;
use std::io;
use std::os::unix::process::CommandExt;
use std::path::Path;
use std::process::Command;

use crate::config::{Config, Rule};
use crate::url::Target;

/// A command furst is prepared to run, and where it came from.
struct Candidate {
    label: String,
    argv: Vec<String>,
}

/// Never returns `Ok`: on success this process has been replaced by the handler.
pub fn dispatch(raw: &str) -> Result<String, String> {
    let cfg = Config::load()?;
    let target = Target::parse(raw);
    let candidates = candidates(&cfg, &target);

    if candidates.is_empty() {
        return Err(format!("no rule matched {raw}, and no default is set"));
    }

    let mut skipped: Vec<String> = Vec::new();
    for c in &candidates {
        // exec replaces this process, so it only returns on failure.
        let err = Command::new(&c.argv[0]).args(&c.argv[1..]).exec();
        if err.kind() == io::ErrorKind::NotFound {
            skipped.push(c.argv[0].clone());
            continue;
        }
        return Err(format!("{}: {err}", c.argv[0]));
    }
    Err(format!("nothing to run for {raw} (not installed: {})", skipped.join(", ")))
}

pub fn explain(raw: &str) -> Result<String, String> {
    let cfg = Config::load()?;
    let target = Target::parse(raw);
    let mut o = String::new();

    let dash = |s: &str| if s.is_empty() { "-".to_string() } else { s.to_string() };
    let _ = writeln!(o, "scheme  {}", dash(&target.scheme));
    let _ = writeln!(o, "host    {}", dash(&target.host));
    let _ = writeln!(o, "path    {}\n", dash(&target.path));

    let candidates = candidates(&cfg, &target);
    if candidates.is_empty() {
        let _ = writeln!(o, "no rule matched, and no default is set");
        return Ok(o);
    }

    // The first candidate that is actually installed is the one that runs.
    let winner = candidates.iter().position(|c| which(&c.argv[0]));
    for (i, c) in candidates.iter().enumerate() {
        let arrow = if Some(i) == winner { "->" } else { "  " };
        let note = if which(&c.argv[0]) { "" } else { "   (not installed)" };
        let _ = writeln!(o, "{arrow} [{}] {}{}", c.label, quote(&c.argv), note);
    }
    if winner.is_none() {
        let _ = writeln!(o, "\nnone of these are installed");
    }
    Ok(o)
}

pub fn list() -> Result<String, String> {
    let cfg = Config::load()?;
    let mut o = String::new();
    for r in &cfg.rules {
        let _ = writeln!(o, "{}", r.name);
        for (k, v) in [
            ("schemes", &r.schemes),
            ("hosts", &r.hosts),
            ("paths", &r.paths),
            ("contains", &r.contains),
        ] {
            if !v.is_empty() {
                let _ = writeln!(o, "  {k:<9}{}", v.join(", "));
            }
        }
        let note = if which(&r.run[0]) { "" } else { "   (not installed)" };
        let _ = writeln!(o, "  {:<9}{}{}", "run", quote(&r.run), note);
    }
    if !cfg.default.is_empty() {
        let _ = writeln!(o, "default\n  {:<9}{}", "run", quote(&cfg.default));
    }
    Ok(o)
}

fn candidates(cfg: &Config, t: &Target) -> Vec<Candidate> {
    let mut out: Vec<Candidate> = cfg
        .rules
        .iter()
        .filter(|r| matches(r, t))
        .map(|r| Candidate { label: r.name.clone(), argv: build(&r.run, r.terminal, t) })
        .collect();
    if !cfg.default.is_empty() {
        out.push(Candidate { label: "default".to_string(), argv: build(&cfg.default, false, t) });
    }
    out
}

/// A rule matches when every criterion it *states* is satisfied. A criterion is
/// satisfied by any one of its patterns. A rule that states nothing matches all.
fn matches(r: &Rule, t: &Target) -> bool {
    let ok = |pats: &Vec<String>, f: &dyn Fn(&str) -> bool| pats.is_empty() || pats.iter().any(|p| f(p));
    ok(&r.schemes, &|p| p.eq_ignore_ascii_case(&t.scheme))
        && ok(&r.hosts, &|p| host_matches(p, &t.host))
        && ok(&r.paths, &|p| glob(&p.to_ascii_lowercase(), &t.path.to_ascii_lowercase()))
        && ok(&r.contains, &|p| t.raw.contains(p))
}

fn host_matches(pat: &str, host: &str) -> bool {
    if pat == "*" {
        return true;
    }
    if let Some(exact) = pat.strip_prefix('=') {
        return exact.eq_ignore_ascii_case(host);
    }
    let base = pat.strip_prefix("*.").unwrap_or(pat).to_ascii_lowercase();
    // Domain match: the apex plus every subdomain under it.
    host == base || host.ends_with(&format!(".{base}"))
}

/// Glob with `*` as the only metacharacter.
fn glob(pat: &str, s: &str) -> bool {
    let parts: Vec<&str> = pat.split('*').collect();
    if parts.len() == 1 {
        return pat == s;
    }
    if !s.starts_with(parts[0]) {
        return false;
    }
    let mut pos = parts[0].len();
    let last = parts.len() - 1;
    for (i, part) in parts.iter().enumerate().skip(1) {
        if i == last {
            return part.is_empty() || (s.len() >= pos + part.len() && s.ends_with(part));
        }
        if part.is_empty() {
            continue;
        }
        match s[pos..].find(part) {
            Some(j) => pos += j + part.len(),
            None => return false,
        }
    }
    true
}

fn build(run: &[String], terminal: bool, t: &Target) -> Vec<String> {
    let mut argv: Vec<String> = run.iter().map(|a| subst(a, t)).collect();
    // Convenience: a rule that never names the URL still gets it, last.
    if !run.iter().any(|a| a.contains("{url}") || a.contains("{url_enc}")) {
        argv.push(t.raw.clone());
    }
    if terminal && let Some(mut term) = terminal_cmd() {
        term.extend(argv);
        return term;
    }
    argv
}

fn subst(arg: &str, t: &Target) -> String {
    arg.replace("{url_enc}", &percent_encode(&t.raw))
        .replace("{url}", &t.raw)
        .replace("{host}", &t.host)
        .replace("{path}", &t.path)
        .replace("{scheme}", &t.scheme)
}

fn percent_encode(s: &str) -> String {
    let mut out = String::with_capacity(s.len());
    for b in s.bytes() {
        match b {
            b'A'..=b'Z' | b'a'..=b'z' | b'0'..=b'9' | b'-' | b'_' | b'.' | b'~' => out.push(b as char),
            _ => out.push_str(&format!("%{b:02X}")),
        }
    }
    out
}

fn terminal_cmd() -> Option<Vec<String>> {
    if let Ok(t) = env::var("TERMINAL")
        && !t.is_empty()
    {
        return Some(vec![t, "-e".to_string()]);
    }
    ["foot", "alacritty", "kitty", "wezterm", "urxvt", "st", "xterm"]
        .iter()
        .find(|t| which(t))
        .map(|t| vec![t.to_string(), "-e".to_string()])
}

pub fn which(cmd: &str) -> bool {
    if cmd.contains('/') {
        return Path::new(cmd).is_file();
    }
    env::var_os("PATH").is_some_and(|paths| env::split_paths(&paths).any(|d| d.join(cmd).is_file()))
}

/// Display-only quoting, so `--explain` output can be pasted into a shell.
fn quote(argv: &[String]) -> String {
    argv.iter()
        .map(|a| {
            if a.is_empty() || a.contains([' ', '\t', '"', '\'', '$', '&', ';', '|', '<', '>', '*', '?']) {
                format!("'{}'", a.replace('\'', r"'\''"))
            } else {
                a.clone()
            }
        })
        .collect::<Vec<_>>()
        .join(" ")
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn host_domain_match_covers_apex_and_subdomains() {
        assert!(host_matches("example.com", "example.com"));
        assert!(host_matches("example.com", "www.example.com"));
        assert!(host_matches("*.example.com", "a.b.example.com"));
        // Must not match a suffix that is not a domain boundary.
        assert!(!host_matches("example.com", "notexample.com"));
        assert!(!host_matches("example.com", "example.com.evil.net"));
    }

    #[test]
    fn exact_host_excludes_subdomains() {
        assert!(host_matches("=example.com", "example.com"));
        assert!(!host_matches("=example.com", "www.example.com"));
    }

    #[test]
    fn globs() {
        assert!(glob("*.pdf", "/doc/a.pdf"));
        assert!(!glob("*.pdf", "/doc/a.pdfx"));
        assert!(glob("/a/*/z*", "/a/bbb/zzz.html"));
        assert!(!glob("/a/*/z*", "/a/bbb/yyy"));
        assert!(glob("*", ""));
        assert!(glob("/exact", "/exact"));
        assert!(!glob("/exact", "/exact/more"));
    }

    #[test]
    fn userinfo_cannot_spoof_the_host() {
        let t = Target::parse("https://meet.google.com@evil.example/x");
        assert_eq!(t.host, "evil.example");
    }

    #[test]
    fn parses_ports_ipv6_and_opaque_schemes() {
        assert_eq!(Target::parse("http://h.test:8080/p?q=1").host, "h.test");
        assert_eq!(Target::parse("http://[::1]:80/p").host, "[::1]");
        assert_eq!(Target::parse("https://H.Test./P").host, "h.test");
        let m = Target::parse("mailto:a@b.test");
        assert_eq!(m.scheme, "mailto");
        assert!(m.host.is_empty());
        // A bare domain is treated as https, the way a browser would.
        assert_eq!(Target::parse("example.com/x").scheme, "https");
    }

    #[test]
    fn url_is_appended_when_the_rule_never_names_it() {
        let t = Target::parse("https://h.test/");
        let argv = build(&["mpv".to_string()], false, &t);
        assert_eq!(argv, vec!["mpv", "https://h.test/"]);
        // ...but not when it does.
        let argv = build(&["mpv".to_string(), "{url}".to_string()], false, &t);
        assert_eq!(argv.len(), 2);
    }

    #[test]
    fn empty_criteria_match_everything() {
        let r = Rule {
            name: "catch-all".into(),
            schemes: vec![], hosts: vec![], paths: vec![], contains: vec![],
            terminal: false, run: vec!["x".into()],
        };
        assert!(matches(&r, &Target::parse("https://anything.test/")));
    }
}