From df7b01b5c18d7027c711a20ae88621591278b698 Mon Sep 17 00:00:00 2001
From: froge <froge@git.repo.cafe>
Date: Sun, 8 Oct 2023 03:51:36 +1000
Subject: [PATCH] Fix issues #12 and #13

---
 Cargo.toml    |  2 +-
 src/embeds.rs | 42 +++++++++++++++++++++++++++---------------
 src/lib.rs    |  3 +++
 3 files changed, 31 insertions(+), 16 deletions(-)

diff --git a/Cargo.toml b/Cargo.toml
index d95666d..d7bfc9e 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "frogbot"
-version = "0.1.0"
+version = "0.1.1"
 edition = "2021"
 
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
diff --git a/src/embeds.rs b/src/embeds.rs
index 9f76a51..e85df46 100644
--- a/src/embeds.rs
+++ b/src/embeds.rs
@@ -15,6 +15,7 @@ use regex::Regex;
 use scraper::{Html, Selector};
 
 /// Represents an Embed in the chat
+#[derive(Default)]
 pub struct Embed {
     /// The title of the embed
     pub title: String,
@@ -30,7 +31,7 @@ impl Embed {
 }
 
 /// Scrapes the HTML of a webpage and generates an [`Embed`] with the scraped information.
-pub fn parse_metadata(page: &str) -> Embed {
+pub fn parse_metadata(page: &str) -> Option<Embed> {
     let doc_body = Html::parse_document(page);
 
     // Selectors used to get metadata are defined here
@@ -44,19 +45,23 @@ pub fn parse_metadata(page: &str) -> Embed {
     let mut meta_title = String::default();
     let mut meta_description = String::default();
 
-    if let Some(title) = title {
-        meta_title = title.text().collect();
-    } else {
-        warn!("Failed to parse title HTML");
+    match (title, desc) {
+        // If both title and description aren't found return None
+        (None, None) => {
+            warn!("Couldn't parse any metadata for URL");
+            return None;
+        },
+        // Otherwise set the title/description to whatever we find
+        (Some(title), Some(desc)) => {
+            meta_title = title.text().collect();
+            meta_description = desc.value().attr("content").unwrap().to_string();
+        }
+        // Handle logging of parse failures
+        (Some(_), None) => warn!("Failed to parse description HTML"),
+        (None, Some(_)) => warn!("Failed to parse title HTML"),
     }
 
-    if let Some(desc) = desc {
-        meta_description = desc.value().attr("content").unwrap().to_string();
-    } else {
-        warn!("Failed to parse description HTML");
-    }
-
-    Embed::new(meta_title, meta_description)
+    Some(Embed::new(meta_title, meta_description))
 }
 
 /// Check if the message has any urls in it and get them if it does
@@ -119,15 +124,22 @@ pub async fn embed_handler(event: OriginalSyncRoomMessageEvent, room: Room, clie
 
         let urls = get_urls_from_message(&text_content.body);
 
-        let reqwest_client = reqwest::Client::builder().user_agent("Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36").build().unwrap();
+        let reqwest_client = reqwest::Client::builder().user_agent("Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/118.0.0.0 Safari/537.36").build().unwrap();
 
-        for url in urls {
+        for mut url in urls {
             if let Ok(req) = reqwest_client.get(url).send().await {
                 if let Ok(res) = req.text().await {
                     // beware, dirty HTML parsing code
-                    let embed = parse_metadata(&res);
+                    let metadata = parse_metadata(&res);
+
+                    // If we didn't get any metadata set URL to nothing so it won't get repeated
+                    // With no other embed data in the bot's embed message
+                    if metadata.is_none() {
+                        url = "";
+                    }
 
                     // Build our message reply
+                    let embed = metadata.unwrap_or(Embed::new("No metadata found".to_string(), "".to_string()));
                     let bot_reply = RoomMessageEventContent::text_html(
                         &embed.title,
                         format!(
diff --git a/src/lib.rs b/src/lib.rs
index 829cec6..ae93518 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -141,6 +141,9 @@ pub async fn run(config: Config) -> anyhow::Result<()> {
         .await
         .expect("frogbot couldn't log into it's account.");
 
+    // Set the bot account's display name according to config
+    client.account().set_display_name(Some(&config.display_name)).await?;
+
     warn!("Logged in successfully!");
     warn!(
         "server: '{}', username: '{}', display name: '{}'",