From 568d83f029e46d61e4f0d4df687db9712abbedcc Mon Sep 17 00:00:00 2001
From: Bill Thiede <git@xinu.tv>
Date: Sat, 3 Feb 2024 11:10:51 -0800
Subject: [PATCH] linkify URLs in plaintext emails.

---
 Cargo.lock            | 10 ++++++++++
 server/Cargo.toml     |  1 +
 server/src/graphql.rs |  9 ++++++---
 server/src/lib.rs     | 19 +++++++++++++++++--
 4 files changed, 34 insertions(+), 5 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 7bfffdb..f3b5381 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1747,6 +1747,15 @@ version = "0.2.152"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "13e3bf6590cbc649f4d1a3eefc9d5d6eb746f5200ffb04e5e142700b8faa56e7"
 
+[[package]]
+name = "linkify"
+version = "0.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f1dfa36d52c581e9ec783a7ce2a5e0143da6237be5811a0b3153fedfdbe9f780"
+dependencies = [
+ "memchr",
+]
+
 [[package]]
 name = "linux-raw-sys"
 version = "0.4.13"
@@ -3228,6 +3237,7 @@ dependencies = [
  "async-graphql-rocket",
  "css-inline",
  "glog",
+ "linkify",
  "log 0.4.20",
  "lol_html",
  "mailparse",
diff --git a/server/Cargo.toml b/server/Cargo.toml
index da88b79..01afb28 100644
--- a/server/Cargo.toml
+++ b/server/Cargo.toml
@@ -27,6 +27,7 @@ lol_html = "1.2.0"
 css-inline = "0.13.0"
 anyhow = "1.0.79"
 maplit = "1.0.2"
+linkify = "0.10.0"
 
 [dependencies.rocket_contrib]
 version = "0.4.11"
diff --git a/server/src/graphql.rs b/server/src/graphql.rs
index 55a4e14..c5a7af9 100644
--- a/server/src/graphql.rs
+++ b/server/src/graphql.rs
@@ -16,7 +16,7 @@ use memmap::MmapOptions;
 use notmuch::Notmuch;
 use rocket::time::Instant;
 
-use crate::sanitize_html;
+use crate::{linkify_html, sanitize_html};
 
 pub struct QueryRoot;
 
@@ -347,8 +347,11 @@ impl QueryRoot {
                 .get_first_value("date")
                 .and_then(|d| mailparse::dateparse(&d).ok());
             let body = match extract_body(&m)? {
-                Body::PlainText(PlainText { text, content_tree }) => Body::PlainText(PlainText {
-                    text,
+                Body::PlainText(PlainText { text, content_tree }) => Body::Html(Html {
+                    html: format!(
+                        r#"<p class="view-part-text-plain">{}</p>"#,
+                        linkify_html(&text)
+                    ),
                     content_tree: if debug_content_tree {
                         render_content_type_tree(&m)
                     } else {
diff --git a/server/src/lib.rs b/server/src/lib.rs
index 9ec4903..139bef7 100644
--- a/server/src/lib.rs
+++ b/server/src/lib.rs
@@ -3,6 +3,7 @@ pub mod graphql;
 pub mod nm;
 
 use css_inline::{CSSInliner, InlineError, InlineOptions};
+use linkify::{LinkFinder, LinkKind};
 use log::error;
 use lol_html::{element, errors::RewritingError, rewrite_str, RewriteStrSettings};
 use maplit::{hashmap, hashset};
@@ -16,6 +17,22 @@ pub enum SanitizeError {
     InlineError(#[from] InlineError),
 }
 
+pub fn linkify_html(text: &str) -> String {
+    let finder = LinkFinder::new();
+    let mut parts = Vec::new();
+    for span in finder.spans(text) {
+        // TODO(wathiede): use Cow<str>?
+        match span.kind() {
+            // Text as-is
+            None => parts.push(span.as_str().to_string()),
+            // Wrap in anchor tag
+            Some(LinkKind::Url) => parts.push(format!(r#"<a href="{0}">{0}</a>"#, span.as_str())),
+            _ => todo!("unhandled kind: {:?}", span.kind().unwrap()),
+        }
+    }
+    parts.join("")
+}
+
 pub fn sanitize_html(html: &str) -> Result<String, SanitizeError> {
     let element_content_handlers = vec![
         // Open links in new tab
@@ -204,8 +221,6 @@ pub fn sanitize_html(html: &str) -> Result<String, SanitizeError> {
     //let clean_html = inlined_html;
 
     Ok(rewrite_str(
-        // TODO(wathiede): replace ammonia with more lol-html rules.
-        // &ammonia::clean(&html),
         &clean_html,
         RewriteStrSettings {
             element_content_handlers,