From b484a0834ef76ba9aa02cc34da0516ac278506d2 Mon Sep 17 00:00:00 2001
From: jesopo <github@lolnerd.net>
Date: Tue, 21 Mar 2023 21:50:02 +0000
Subject: [PATCH] first code commit

---
 .gitignore     |  2 ++
 Cargo.toml     |  8 +++++
 src/lib.rs     | 88 ++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/basic.rs | 21 ++++++++++++
 4 files changed, 119 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 Cargo.toml
 create mode 100644 src/lib.rs
 create mode 100644 tests/basic.rs

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..4fffb2f
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+/target
+/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
new file mode 100644
index 0000000..5157cae
--- /dev/null
+++ b/Cargo.toml
@@ -0,0 +1,8 @@
+[package]
+name = "irctokens"
+version = "0.1.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
diff --git a/src/lib.rs b/src/lib.rs
new file mode 100644
index 0000000..ee8aa4a
--- /dev/null
+++ b/src/lib.rs
@@ -0,0 +1,88 @@
+use std::collections::{HashMap, VecDeque};
+
+pub struct Line {
+    // tags are promised to be utf8 encoded
+    pub tags: Option<HashMap<String, Option<String>>>,
+    pub source: Option<Vec<u8>>,
+    // commands are promised to be ascii encoded
+    pub command: String,
+    pub args: Vec<Vec<u8>>,
+}
+
+#[derive(Debug)]
+pub enum Error {
+    Empty,
+    MissingCommand,
+    CommandDecode,
+    TagKeyDecode,
+    TagValueDecode,
+}
+
+trait TakeWord<'a> {
+    fn take_word(&mut self, sep: u8) -> &'a [u8];
+}
+
+impl<'a> TakeWord<'a> for &'a [u8] {
+    fn take_word(&mut self, sep: u8) -> &'a [u8] {
+        if let Some(i) = self.iter().position(|c| c == &sep) {
+            let word = &self[..i];
+            *self = &self[i + 1..];
+            word
+        } else {
+            let word = &self[..];
+            *self = &self[self.len()..];
+            word
+        }
+    }
+}
+
+pub fn tokenise(mut line: &[u8]) -> Result<Line, Error> {
+    let tags = match line.first() {
+        Some(b'@') => {
+            let mut tags = &line.take_word(b' ')[1..];
+            let mut tags_map = HashMap::new();
+
+            while !tags.is_empty() {
+                let mut keyvalue = tags.take_word(b';');
+                let tag = keyvalue.take_word(b'=');
+                tags_map.insert(
+                    String::from_utf8(tag.to_vec()).map_err(|_| Error::TagKeyDecode)?,
+                    match keyvalue {
+                        b"" | b"=" => None,
+                        _ => Some(
+                            String::from_utf8(keyvalue.to_vec())
+                                .map_err(|_| Error::TagValueDecode)?,
+                        ),
+                    },
+                );
+            }
+
+            Some(tags_map)
+        }
+        _ => None,
+    };
+
+    let source = match line.first() {
+        Some(b':') => Some(line.take_word(b' ')[1..].to_vec()),
+        _ => None,
+    };
+
+    let mut args = VecDeque::<Vec<u8>>::new();
+    while !line.is_empty() {
+        if line[0] == b':' {
+            args.push_back(line[1..].to_vec());
+            line = &[];
+        } else {
+            args.push_back(line.take_word(b' ').to_vec());
+        }
+    }
+
+    let command = args.pop_front().ok_or(Error::MissingCommand)?;
+
+    Ok(Line {
+        tags,
+        source,
+        command: String::from_utf8(command).map_err(|_| Error::CommandDecode)?,
+        args: args.into(),
+    })
+}
diff --git a/tests/basic.rs b/tests/basic.rs
new file mode 100644
index 0000000..acfd5fb
--- /dev/null
+++ b/tests/basic.rs
@@ -0,0 +1,21 @@
+use irctokens::tokenise;
+
+#[test]
+fn basic() {
+    let line =
+        tokenise(b"@tag1=tag1value;tag2=;tag3 :source COMMAND arg1 arg2 :arg3 with space").unwrap();
+
+    assert_eq!(line.source, Some(b"source".to_vec()));
+    assert_eq!(&line.command, "COMMAND");
+
+    assert_eq!(line.args.len(), 3);
+    assert_eq!(line.args[0], b"arg1");
+    assert_eq!(line.args[1], b"arg2");
+    assert_eq!(line.args[2], b"arg3 with space");
+
+    let tags = line.tags.unwrap();
+    assert_eq!(tags.len(), 3);
+    assert_eq!(tags["tag1"], Some("tag1value".to_string()));
+    assert_eq!(tags["tag2"], None);
+    assert_eq!(tags["tag3"], None);
+}