procmail2notmuch: add debug vs notmuchrc modes

procmail2notmuch/Cargo.toml

@@ -12,3 +12,4 @@ version.workspace = true
 
 [dependencies]
 anyhow = "1.0.69"
+clap = { version = "4.5.37", features = ["derive"] }

procmail2notmuch/src/main.rs

@@ -1,13 +1,15 @@
-use std::{convert::Infallible, io::Write, str::FromStr};
+use std::{collections::HashMap, convert::Infallible, io::Write, str::FromStr};
 
-#[derive(Debug, Default)]
+use clap::{Parser, ValueEnum};
+
+#[derive(Copy, Clone, Debug, Default, PartialEq, Eq, Hash, Ord, PartialOrd)]
 enum MatchType {
     From,
     Sender,
     To,
     Cc,
     Subject,
-    List,
+    ListId,
     DeliveredTo,
     XForwardedTo,
     ReplyTo,
@@ -26,7 +28,7 @@ struct Match {
 #[derive(Debug, Default)]
 struct Rule {
     matches: Vec<Match>,
-    tags: Vec<String>,
+    tags: Option<String>,
 }
 
 fn unescape(s: &str) -> String {
@@ -38,6 +40,10 @@ fn cleanup_match(prefix: &str, s: &str) -> String {
 }
 
 mod matches {
+    // From https://linux.die.net/man/5/procmailrc
+    // If the regular expression contains '^TO_' it will be substituted by '(^((Original-)?(Resent-)?(To|Cc|Bcc)|(X-Envelope |Apparently(-Resent)?)-To):(.*[^-a-zA-Z0-9_.])?)'
+    // If the regular expression contains '^TO' it will be substituted by '(^((Original-)?(Resent-)?(To|Cc|Bcc)|(X-Envelope |Apparently(-Resent)?)-To):(.*[^a-zA-Z])?)', which should catch all destination specifications containing a specific word.
+
     pub const TO: &'static str = "TO";
     pub const CC: &'static str = "Cc";
     pub const TOCC: &'static str = "(TO|Cc)";
@@ -109,7 +115,7 @@ impl FromStr for Match {
                 });
             } else if needle.starts_with(LIST_ID) {
                 return Ok(Match {
-                    match_type: MatchType::List,
+                    match_type: MatchType::ListId,
                     needle: cleanup_match(LIST_ID, needle),
                 });
             } else if needle.starts_with(REPLY_TO) {
@@ -155,7 +161,7 @@ fn notmuch_from_rules<W: Write>(mut w: W, rules: &[Rule]) -> anyhow::Result<()>
     let mut lines = Vec::new();
     for r in rules {
         for m in &r.matches {
-            for t in &r.tags {
+            if let Some(t) = &r.tags {
                 if let MatchType::Unknown = m.match_type {
                     eprintln!("rule has unknown match {:?}", r);
                     continue;
@@ -168,7 +174,7 @@ fn notmuch_from_rules<W: Write>(mut w: W, rules: &[Rule]) -> anyhow::Result<()>
                     MatchType::To => "to:",
                     MatchType::Cc => "to:",
                     MatchType::Subject => "subject:",
-                    MatchType::List => "List-ID:",
+                    MatchType::ListId => "List-ID:",
                     MatchType::Body => "",
                     // TODO(wathiede): these will probably require adding fields to notmuch
                     // index. Handle them later.
@@ -200,11 +206,28 @@ fn notmuch_from_rules<W: Write>(mut w: W, rules: &[Rule]) -> anyhow::Result<()>
     Ok(())
 }
 
+#[derive(Debug, Copy, Clone, PartialEq, Eq, PartialOrd, Ord, ValueEnum)]
+enum Mode {
+    Debug,
+    Notmuchrc,
+}
+
+/// Simple program to greet a person
+#[derive(Parser, Debug)]
+#[command(version, about, long_about = None)]
+struct Args {
+    #[arg(short, long, default_value = "/home/wathiede/dotfiles/procmailrc")]
+    input: String,
+
+    #[arg(value_enum)]
+    mode: Mode,
+}
+
 fn main() -> anyhow::Result<()> {
-    let input = "/home/wathiede/dotfiles/procmailrc";
+    let args = Args::parse();
     let mut rules = Vec::new();
     let mut cur_rule = Rule::default();
-    for l in std::fs::read_to_string(input)?.lines() {
+    for l in std::fs::read_to_string(args.input)?.lines() {
         let l = if let Some(idx) = l.find('#') {
             &l[..idx]
         } else {
@@ -230,7 +253,7 @@ fn main() -> anyhow::Result<()> {
             }
             '.' => {
                 // delivery to folder
-                cur_rule.tags.push(cleanup_match(
+                cur_rule.tags = Some(cleanup_match(
                     "",
                     &l.replace('.', "/")
                         .replace(' ', "")
@@ -240,16 +263,35 @@ fn main() -> anyhow::Result<()> {
                 rules.push(cur_rule);
                 cur_rule = Rule::default();
             }
+            '/' => cur_rule = Rule::default(), // Ex. /dev/null
             '|' => cur_rule = Rule::default(), // external command
             '$' => {
                 // TODO(wathiede): tag messages with no other tag as 'inbox'
-                cur_rule.tags.push(cleanup_match("", "inbox"));
+                cur_rule.tags = Some(cleanup_match("", "inbox"));
                 rules.push(cur_rule);
                 cur_rule = Rule::default();
             } // variable, should only be $DEFAULT in my config
-            _ => panic!("Unhandled first character '{}' {}", first, l),
+            _ => panic!("Unhandled first character '{}'\nLine: {}", first, l),
         }
     }
-    notmuch_from_rules(std::io::stdout(), &rules)?;
+    match args.mode {
+        Mode::Debug => print_rules(&rules),
+        Mode::Notmuchrc => notmuch_from_rules(std::io::stdout(), &rules)?,
+    }
     Ok(())
 }
+
+fn print_rules(rules: &[Rule]) {
+    let mut tally = HashMap::new();
+    for r in rules {
+        for m in &r.matches {
+            *tally.entry(m.match_type).or_insert(0) += 1;
+        }
+    }
+    let mut sorted: Vec<_> = tally.iter().map(|(k, v)| (v, k)).collect();
+    sorted.sort();
+    sorted.reverse();
+    for (v, k) in sorted {
+        println!("{k:?}: {v}");
+    }
+}