Add a Markdown parser to convert Discord formatting to Minecraft (#4)

Add a Markdown parser to convert Discord formatting to Minecraft

Signed-off-by: Evan Maddock <[email protected]>

CHANGELOG.md

@@ -1,14 +1,20 @@
 # Changelog
+
 All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 ## [Unreleased]
 
+### Added
+
+- Add a converter to show Discord formatting in Minecraft
+
 ## [v2.2.0] - 2021-01-23
 
 ### Changed
+
 - Update dependencies
   - Tokio runtime updated to 1.0
   - Serenity updated to 0.10
@@ -20,40 +26,48 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## [v2.1.1] - 2020-12-11
 
 ### Fixed
+
 - Fix full username mentions with discriminator not being parsed correctly
 
 ## [v2.1.0] - 2020-12-03
 
 ### Added
+
 - Add ability to mention roles and channels from Minecraft
 - Add an optional webserver implementation to listen for messages from other machines
 
 ### Fixed
+
 - Log files being moved (such as maybe during log rotation) should no longer break the bot, if that was happening
 - Fix mentions from Minecraft with spaces not creating a mention
 
 ### Changed
+
 - Print nicer-looking error messages
 
 ## [v2.0.1] - 2020-11-22
 
 ### Fixed
+
 - Fix a bad value in the default configuration
 
 ### Changed
+
 - Eliminated a call to the Discord REST API when messages are received from Minecraft
 - Replace ugly Discord mentions with names in messages to Minecraft
 - Escape double quote characters in messages to Minecraft
 
 ## [v2.0.0] - 2020-11-8
 
 ### Added
+
 - Add more customization options for chat formatting in Minecraft
 
 ### Changed
+
 - Improve experience when a user sends an attachment in Discord
 
-[Unreleased]: https://github.com/EbonJaeger/dolphin-rs/compare/v2.2.0...master
+[unreleased]: https://github.com/EbonJaeger/dolphin-rs/compare/v2.2.0...master
 [v2.2.0]: https://github.com/EbonJaeger/dolphin-rs/compare/v2.1.1...v2.2.0
 [v2.1.1]: https://github.com/EbonJaeger/dolphin-rs/compare/v2.1.0...v2.1.1
 [v2.1.0]: https://github.com/EbonJaeger/dolphin-rs/compare/v2.0.1...v2.1.0

Cargo.toml

@@ -12,8 +12,9 @@ confy = "0.4.0"
 err-derive = "0.3"
 lazy_static = "1.4.0"
 linemux = {git = "https://github.com/jmagnuson/linemux"}
+pipeline = "0.5"
 rcon = "0.4"
-regex = "1.4.2"
+fancy-regex = "0.4"
 serde_json = "1.0"
 tracing = "0.1.21"
 tracing-subscriber = "0.2.15"

src/discord.rs

@@ -1,9 +1,10 @@
 use crate::config::RootConfig;
 use crate::errors::DolphinError;
 use crate::listener::{Listener, LogTailer, Webserver};
+use crate::markdown;
 use crate::minecraft::{MinecraftMessage, Source};
+use fancy_regex::Regex;
 use rcon::Connection;
-use regex::Regex;
 use serenity::{
     async_trait,
     model::{
@@ -15,12 +16,9 @@ use serenity::{
     prelude::*,
     utils::parse_channel,
 };
-use std::{
-    str::Split,
-    sync::{
-        atomic::{AtomicBool, AtomicU64, Ordering},
-        Arc,
-    },
+use std::sync::{
+    atomic::{AtomicBool, AtomicU64, Ordering},
+    Arc,
 };
 use tokio::sync::mpsc;
 use tracing::{debug, error, info, warn};
@@ -191,8 +189,17 @@ impl EventHandler for Handler {
         let content = self.sanitize_message(&ctx, &msg).await;
 
         // Send a separate message for each line
-        let lines = content.split("\n");
-        let lines = truncate_lines(lines);
+        let lines = content.split('\n');
+
+        // Parse and convert any Markdown
+        let mut marked = Vec::new();
+        lines.for_each(|line| {
+            let blocks = markdown::parse(&line);
+            debug!("event_handler:message: parsed plocks: {:?}", blocks);
+            marked.push(markdown::to_minecraft_format(&blocks));
+        });
+
+        let lines = truncate_lines(marked);
         let mut lines = self.apply_line_template(lines);
 
         // Add attachement message if an attachment is present
@@ -286,7 +293,7 @@ impl EventHandler for Handler {
 /// by default 100. If a line is over the limit, it will be split at that
 /// number of chacacters, and a new line inserted into the line Vector.
 ///
-fn truncate_lines<'a>(lines: Split<'a, &'a str>) -> Vec<String> {
+fn truncate_lines(lines: Vec<String>) -> Vec<String> {
     let mut truncated: Vec<String> = Vec::new();
 
     for mut line in lines {
@@ -303,8 +310,8 @@ fn truncate_lines<'a>(lines: Split<'a, &'a str>) -> Vec<String> {
 
             // Shorten the line for the next iteration
             line = match line.get(MAX_LINE_LENGTH..) {
-                Some(sub) => sub,
-                None => "",
+                Some(sub) => sub.to_string(),
+                None => String::new(),
             };
         }
     }
@@ -482,22 +489,20 @@ fn split_webhook_url(url: &str) -> Option<(u64, &str)> {
             Regex::new(r"^https://discord.com/api/webhooks/(?P<id>.*)/(?P<token>.*)$").unwrap();
     }
 
-    let captures = match WEBHOOK_REGEX.captures(&url) {
-        Some(captures) => captures,
-        None => return None,
-    };
+    let mut ret = None;
 
-    if captures.len() != 3 {
-        return None;
-    }
+    if let Ok(Some(captures)) = WEBHOOK_REGEX.captures(&url) {
+        if captures.len() != 3 {
+            return None;
+        }
 
-    let id = captures.name("id").unwrap().as_str();
-    let id = match id.parse::<u64>() {
-        Ok(num) => num,
-        Err(_) => return None,
-    };
+        let id = captures.name("id").unwrap().as_str();
+        if let Ok(id) = id.parse::<u64>() {
+            ret = Some((id, captures.name("token").unwrap().as_str()));
+        }
+    }
 
-    Some((id, captures.name("token").unwrap().as_str()))
+    ret
 }
 
 #[cfg(test)]
@@ -508,12 +513,11 @@ mod tests {
     #[test]
     fn split_long_line() {
         // Given
-        let input = String::from("01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789");
-        let split = input.split("\n");
+        let input = vec!(String::from("01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789"));
         let expected = vec!("0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789", "0123456789");
 
         // When
-        let result = truncate_lines(split);
+        let result = truncate_lines(input);
 
         // Then
         assert_eq!(result, expected);
@@ -522,12 +526,11 @@ mod tests {
     #[test]
     fn no_split_line() {
         // Given
-        let input = String::from("0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789");
-        let split = input.split("\n");
+        let input = vec!(String::from("0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789"));
         let expected = vec!("0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789");
 
         // When
-        let result = truncate_lines(split);
+        let result = truncate_lines(input);
 
         // Then
         assert_eq!(result, expected);

src/main.rs

@@ -3,12 +3,15 @@ mod config;
 mod discord;
 mod errors;
 mod listener;
+mod markdown;
 mod minecraft;
 
 #[macro_use]
 extern crate clap;
 #[macro_use]
 extern crate lazy_static;
+#[macro_use]
+extern crate pipeline;
 
 use clap::{App, Arg};
 use commands::{general::*, hooks::after, minecraft::*};

src/markdown/README.md

@@ -0,0 +1,19 @@
+This is a stripped down version of the Rust crate `markdown` written by Johann Hofmann, licensed under the Apache 2.0 license. You can find the original work [here](https://github.com/johannhof/markdown.rs). The vast majority of the credit here goes to him.
+
+# Differences
+
+Since this application only deals with Discord's flavor of Markdown and Minecraft formatting, there is a lot of the original crate that isn't needed here. On top of that, there are a couple of formatting types that Discord uses that aren't present in the upstream library.
+
+Only the following elements are implemented:
+
+- Blockquotes
+- Emphasis
+- Strikethrough
+- Strong
+- Underline
+
+# License
+
+All work **except** the strikethrough and underline parsers, and the Minecraft format conversion code is © Johann Hofmann.
+
+The license for the original work can be found [here](https://github.com/johannhof/markdown.rs/blob/master/LICENSE-APACHE). I really make no claims on top of that.

src/markdown/block/blockquote.rs

@@ -0,0 +1,41 @@
+use super::Block;
+use super::Block::Blockquote;
+use crate::markdown::block::parse_blocks;
+
+pub fn parse_blockquote(line: &str) -> Option<Block> {
+    if line.is_empty() || !line.starts_with("> ") {
+        return None;
+    }
+
+    let mut content = String::new();
+
+    // Push the content of the quote after the opening `>`
+    content.push_str(&line[2..line.len()]);
+
+    Some(Blockquote(parse_blocks(&content)))
+}
+
+#[cfg(test)]
+mod tests {
+    use super::parse_blockquote;
+    use super::Block::Blockquote;
+
+    #[test]
+    fn finds_blockquote() {
+        match parse_blockquote("> quote") {
+            Some(Blockquote(_)) => (),
+            _ => panic!(),
+        }
+    }
+
+    #[test]
+    fn no_false_positives() {
+        assert_eq!(parse_blockquote(">shouldn't parse"), None);
+        assert_eq!(parse_blockquote("shouldn't > parse"), None);
+    }
+
+    #[test]
+    fn no_early_matching() {
+        assert_eq!(parse_blockquote("first > quote > another blah"), None);
+    }
+}