Add a blocklist for URLs. (#4515)

* Add a blocklist for URLs.

* Fix SQL format

* Make clippy happy.

* Use regex for URL matching.

* Escape regex chars in URLs.

* Use post for modification.

* Make URL block regex static and remove API routes.

* Add date fields to table and use transaction.

* Use Cache for blocklist.

* Rename check_links + move list to parameters of process_markdown.

* SQL format.

* Format, again.

* Remove println.

* Add API test.

* Set a shorter lifetime for regex in debug mode.

* Add missing macro.

* Update lemmy-js-client

* Update api_test/pnpm-lock.yaml

* Don't break other tests

* Use different URL for test

---------

Co-authored-by: Dessalines <dessalines@users.noreply.github.com>
Co-authored-by: Nutomic <me@nutomic.com>

api_tests/package.json

@@ -27,7 +27,7 @@
     "eslint": "^8.57.0",
     "eslint-plugin-prettier": "^5.0.1",
     "jest": "^29.5.0",
-    "lemmy-js-client": "0.19.4-alpha.6",
+    "lemmy-js-client": "0.19.4-alpha.8",
     "prettier": "^3.2.5",
     "ts-jest": "^29.1.0",
     "typescript": "^5.3.3"

api_tests/pnpm-lock.yaml

@@ -30,8 +30,8 @@ devDependencies:
     specifier: ^29.5.0
     version: 29.7.0(@types/node@20.11.22)
   lemmy-js-client:
-    specifier: 0.19.4-alpha.6
+    specifier: 0.19.4-alpha.8
-    version: 0.19.4-alpha.6
+    version: 0.19.4-alpha.8
   prettier:
     specifier: ^3.2.5
     version: 3.2.5
@@ -2390,8 +2390,8 @@ packages:
     engines: {node: '>=6'}
     dev: true
 
-  /lemmy-js-client@0.19.4-alpha.6:
+  /lemmy-js-client@0.19.4-alpha.8:
-    resolution: {integrity: sha512-x4htMlpoZ7hzrhrIk82aompVxbpu2ZDWtmWNGraM0+27nUCDf6gYxJH5nb5R/o39BQe5KSHq6zoBdliBwAY40w==}
+    resolution: {integrity: sha512-8vjqUYVOhyUTcmG9FvPLjrWziVwNa2/Zi+kSflTrajJsK0V+5DclJ5dhdVMUQ4DEA70gb0OuNMDlipPG2FoS5A==}
     dependencies:
       cross-fetch: 4.0.0
       form-data: 4.0.0

api_tests/src/post.spec.ts

@@ -18,6 +18,7 @@ import {
   resolveBetaCommunity,
   createComment,
   deletePost,
+  delay,
   removePost,
   getPost,
   unfollowRemotes,
@@ -710,3 +711,25 @@ test("Fetch post via redirect", async () => {
   expect(gammaPost.post?.post.ap_id).toBe(alphaPost.post_view.post.ap_id);
   await unfollowRemotes(alpha);
 });
+
+test("Block post that contains banned URL", async () => {
+  let editSiteForm: EditSite = {
+    blocked_urls: ["https://evil.com/"],
+  };
+
+  await epsilon.editSite(editSiteForm);
+
+  await delay(500);
+
+  if (!betaCommunity) {
+    throw "Missing beta community";
+  }
+
+  expect(
+    createPost(epsilon, betaCommunity.community.id, "https://evil.com"),
+  ).rejects.toStrictEqual(Error("blocked_url"));
+
+  // Later tests need this to be empty
+  editSiteForm.blocked_urls = [];
+  await epsilon.editSite(editSiteForm);
+});

crates/api/src/local_user/save_settings.rs

@@ -3,6 +3,7 @@ use lemmy_api_common::{
   context::LemmyContext,
   person::SaveUserSettings,
   utils::{
+    get_url_blocklist,
     local_site_to_slur_regex,
     process_markdown_opt,
     proxy_image_link_opt_api,
@@ -35,7 +36,10 @@ pub async fn save_user_settings(
   let site_view = SiteView::read_local(&mut context.pool()).await?;
 
   let slur_regex = local_site_to_slur_regex(&site_view.local_site);
-  let bio = diesel_option_overwrite(process_markdown_opt(&data.bio, &slur_regex, &context).await?);
+  let url_blocklist = get_url_blocklist(&context).await?;
+  let bio = diesel_option_overwrite(
+    process_markdown_opt(&data.bio, &slur_regex, &url_blocklist, &context).await?,
+  );
 
   let avatar = proxy_image_link_opt_api(&data.avatar, &context).await?;
   let banner = proxy_image_link_opt_api(&data.banner, &context).await?;

crates/api/src/site/leave_admin.rs

@@ -4,6 +4,7 @@ use lemmy_db_schema::{
   source::{
     actor_language::SiteLanguage,
     language::Language,
+    local_site_url_blocklist::LocalSiteUrlBlocklist,
     local_user::{LocalUser, LocalUserUpdateForm},
     moderator::{ModAdd, ModAddForm},
     tagline::Tagline,
@@ -62,6 +63,7 @@ pub async fn leave_admin(
   let taglines = Tagline::get_all(&mut context.pool(), site_view.local_site.id).await?;
   let custom_emojis =
     CustomEmojiView::get_all(&mut context.pool(), site_view.local_site.id).await?;
+  let blocked_urls = LocalSiteUrlBlocklist::get_all(&mut context.pool()).await?;
 
   Ok(Json(GetSiteResponse {
     site_view,
@@ -72,5 +74,6 @@ pub async fn leave_admin(
     discussion_languages,
     taglines,
     custom_emojis,
+    blocked_urls,
   }))
 }

crates/api_common/Cargo.toml

@@ -59,6 +59,8 @@ uuid = { workspace = true, optional = true }
 tokio = { workspace = true, optional = true }
 reqwest = { workspace = true, optional = true }
 ts-rs = { workspace = true, optional = true }
+moka.workspace = true
+anyhow.workspace = true
 once_cell = { workspace = true, optional = true }
 actix-web = { workspace = true, optional = true }
 enum-map = { workspace = true }

crates/api_common/src/site.rs

@@ -6,6 +6,7 @@ use lemmy_db_schema::{
     federation_queue_state::FederationQueueState,
     instance::Instance,
     language::Language,
+    local_site_url_blocklist::LocalSiteUrlBlocklist,
     tagline::Tagline,
   },
   ListingType,
@@ -268,6 +269,8 @@ pub struct EditSite {
   pub allowed_instances: Option<Vec<String>>,
   /// A list of blocked instances.
   pub blocked_instances: Option<Vec<String>>,
+  /// A list of blocked URLs
+  pub blocked_urls: Option<Vec<String>>,
   /// A list of taglines shown at the top of the front page.
   pub taglines: Option<Vec<String>>,
   pub registration_mode: Option<RegistrationMode>,
@@ -305,6 +308,7 @@ pub struct GetSiteResponse {
   pub taglines: Vec<Tagline>,
   /// A list of custom emojis your site supports.
   pub custom_emojis: Vec<CustomEmojiView>,
+  pub blocked_urls: Vec<LocalSiteUrlBlocklist>,
 }
 
 #[skip_serializing_none]

crates/api_common/src/utils.rs

@@ -17,6 +17,7 @@ use lemmy_db_schema::{
     instance_block::InstanceBlock,
     local_site::LocalSite,
     local_site_rate_limit::LocalSiteRateLimit,
+    local_site_url_blocklist::LocalSiteUrlBlocklist,
     password_reset_request::PasswordResetRequest,
     person::{Person, PersonUpdateForm},
     person_block::PersonBlock,
@@ -38,18 +39,24 @@ use lemmy_utils::{
   rate_limit::{ActionType, BucketConfig},
   settings::structs::{PictrsImageMode, Settings},
   utils::{
-    markdown::markdown_rewrite_image_links,
+    markdown::{markdown_check_for_blocked_urls, markdown_rewrite_image_links},
     slurs::{build_slur_regex, remove_slurs},
   },
 };
-use regex::Regex;
+use moka::future::Cache;
+use once_cell::sync::Lazy;
+use regex::{escape, Regex, RegexSet};
 use rosetta_i18n::{Language, LanguageId};
-use std::collections::HashSet;
+use std::{collections::HashSet, time::Duration};
 use tracing::warn;
 use url::{ParseError, Url};
 use urlencoding::encode;
 
 pub static AUTH_COOKIE_NAME: &str = "jwt";
+#[cfg(debug_assertions)]
+static URL_BLOCKLIST_RECHECK_DELAY: Duration = Duration::from_millis(500);
+#[cfg(not(debug_assertions))]
+static URL_BLOCKLIST_RECHECK_DELAY: Duration = Duration::from_secs(60);
 
 #[tracing::instrument(skip_all)]
 pub async fn is_mod_or_admin(
@@ -516,6 +523,47 @@ pub fn local_site_opt_to_sensitive(local_site: &Option<LocalSite>) -> bool {
     .unwrap_or(false)
 }
 
+pub async fn get_url_blocklist(context: &LemmyContext) -> LemmyResult<RegexSet> {
+  static URL_BLOCKLIST: Lazy<Cache<(), RegexSet>> = Lazy::new(|| {
+    Cache::builder()
+      .max_capacity(1)
+      .time_to_live(URL_BLOCKLIST_RECHECK_DELAY)
+      .build()
+  });
+
+  Ok(
+    URL_BLOCKLIST
+      .try_get_with::<_, LemmyError>((), async {
+        let urls = LocalSiteUrlBlocklist::get_all(&mut context.pool()).await?;
+
+        let regexes = urls.iter().map(|url| {
+          let url = &url.url;
+          let parsed = Url::parse(url).expect("Coundln't parse URL.");
+          if url.ends_with('/') {
+            format!(
+              "({}://)?{}{}?",
+              parsed.scheme(),
+              escape(parsed.domain().expect("No domain.")),
+              escape(parsed.path())
+            )
+          } else {
+            format!(
+              "({}://)?{}{}",
+              parsed.scheme(),
+              escape(parsed.domain().expect("No domain.")),
+              escape(parsed.path())
+            )
+          }
+        });
+
+        let set = RegexSet::new(regexes)?;
+        Ok(set)
+      })
+      .await
+      .map_err(|e| anyhow::anyhow!("Failed to build URL blocklist due to `{}`", e))?,
+  )
+}
+
 pub async fn send_application_approved_email(
   user: &LocalUserView,
   settings: &Settings,
@@ -867,9 +915,13 @@ fn limit_expire_time(expires: DateTime<Utc>) -> LemmyResult<Option<DateTime<Utc>
 pub async fn process_markdown(
   text: &str,
   slur_regex: &Option<Regex>,
+  url_blocklist: &RegexSet,
   context: &LemmyContext,
 ) -> LemmyResult<String> {
   let text = remove_slurs(text, slur_regex);
+
+  markdown_check_for_blocked_urls(&text, url_blocklist)?;
+
   if context.settings().pictrs_config()?.image_mode() == PictrsImageMode::ProxyAllImages {
     let (text, links) = markdown_rewrite_image_links(text);
     RemoteImage::create(&mut context.pool(), links).await?;
@@ -882,10 +934,13 @@ pub async fn process_markdown(
 pub async fn process_markdown_opt(
   text: &Option<String>,
   slur_regex: &Option<Regex>,
+  url_blocklist: &RegexSet,
   context: &LemmyContext,
 ) -> LemmyResult<Option<String>> {
   match text {
-    Some(t) => process_markdown(t, slur_regex, context).await.map(Some),
+    Some(t) => process_markdown(t, slur_regex, url_blocklist, context)
+      .await
+      .map(Some),
     None => Ok(None),
   }
 }

crates/api_crud/src/comment/create.rs

@@ -10,6 +10,7 @@ use lemmy_api_common::{
     check_post_deleted_or_removed,
     generate_local_apub_endpoint,
     get_post,
+    get_url_blocklist,
     is_mod_or_admin,
     local_site_to_slur_regex,
     process_markdown,
@@ -44,7 +45,8 @@ pub async fn create_comment(
   let local_site = LocalSite::read(&mut context.pool()).await?;
 
   let slur_regex = local_site_to_slur_regex(&local_site);
-  let content = process_markdown(&data.content, &slur_regex, &context).await?;
+  let url_blocklist = get_url_blocklist(&context).await?;
+  let content = process_markdown(&data.content, &slur_regex, &url_blocklist, &context).await?;
   is_valid_body_field(&Some(content.clone()), false)?;
 
   // Check for a community ban

crates/api_crud/src/comment/update.rs

@@ -5,7 +5,12 @@ use lemmy_api_common::{
   comment::{CommentResponse, EditComment},
   context::LemmyContext,
   send_activity::{ActivityChannel, SendActivityData},
-  utils::{check_community_user_action, local_site_to_slur_regex, process_markdown_opt},
+  utils::{
+    check_community_user_action,
+    get_url_blocklist,
+    local_site_to_slur_regex,
+    process_markdown_opt,
+  },
 };
 use lemmy_db_schema::{
   source::{
@@ -54,7 +59,8 @@ pub async fn update_comment(
   .await?;
 
   let slur_regex = local_site_to_slur_regex(&local_site);
-  let content = process_markdown_opt(&data.content, &slur_regex, &context).await?;
+  let url_blocklist = get_url_blocklist(&context).await?;
+  let content = process_markdown_opt(&data.content, &slur_regex, &url_blocklist, &context).await?;
   is_valid_body_field(&content, false)?;
 
   let comment_id = data.comment_id;

crates/api_crud/src/community/create.rs

@@ -9,6 +9,7 @@ use lemmy_api_common::{
     generate_inbox_url,
     generate_local_apub_endpoint,
     generate_shared_inbox_url,
+    get_url_blocklist,
     is_admin,
     local_site_to_slur_regex,
     process_markdown_opt,
@@ -53,9 +54,11 @@ pub async fn create_community(
   }
 
   let slur_regex = local_site_to_slur_regex(&local_site);
+  let url_blocklist = get_url_blocklist(&context).await?;
   check_slurs(&data.name, &slur_regex)?;
   check_slurs(&data.title, &slur_regex)?;
-  let description = process_markdown_opt(&data.description, &slur_regex, &context).await?;
+  let description =
+    process_markdown_opt(&data.description, &slur_regex, &url_blocklist, &context).await?;
   let icon = proxy_image_link_api(&data.icon, &context).await?;
   let banner = proxy_image_link_api(&data.banner, &context).await?;
 

crates/api_crud/src/community/update.rs

@@ -7,6 +7,7 @@ use lemmy_api_common::{
   send_activity::{ActivityChannel, SendActivityData},
   utils::{
     check_community_mod_action,
+    get_url_blocklist,
     local_site_to_slur_regex,
     process_markdown_opt,
     proxy_image_link_opt_api,
@@ -36,8 +37,10 @@ pub async fn update_community(
   let local_site = LocalSite::read(&mut context.pool()).await?;
 
   let slur_regex = local_site_to_slur_regex(&local_site);
+  let url_blocklist = get_url_blocklist(&context).await?;
   check_slurs_opt(&data.title, &slur_regex)?;
-  let description = process_markdown_opt(&data.description, &slur_regex, &context).await?;
+  let description =
+    process_markdown_opt(&data.description, &slur_regex, &url_blocklist, &context).await?;
   is_valid_body_field(&data.description, false)?;
 
   let description = diesel_option_overwrite(description);

crates/api_crud/src/post/create.rs

@@ -9,6 +9,7 @@ use lemmy_api_common::{
   utils::{
     check_community_user_action,
     generate_local_apub_endpoint,
+    get_url_blocklist,
     honeypot_check,
     local_site_to_slur_regex,
     mark_post_as_read,
@@ -38,6 +39,7 @@ use lemmy_utils::{
     validation::{
       check_url_scheme,
       clean_url_params,
+      is_url_blocked,
       is_valid_alt_text_field,
       is_valid_body_field,
       is_valid_post_title,
@@ -60,8 +62,9 @@ pub async fn create_post(
 
   let slur_regex = local_site_to_slur_regex(&local_site);
   check_slurs(&data.name, &slur_regex)?;
+  let url_blocklist = get_url_blocklist(&context).await?;
 
-  let body = process_markdown_opt(&data.body, &slur_regex, &context).await?;
+  let body = process_markdown_opt(&data.body, &slur_regex, &url_blocklist, &context).await?;
   let data_url = data.url.as_ref();
   let url = data_url.map(clean_url_params); // TODO no good way to handle a "clear"
   let custom_thumbnail = data.custom_thumbnail.as_ref().map(clean_url_params);
@@ -69,6 +72,7 @@ pub async fn create_post(
   is_valid_post_title(&data.name)?;
   is_valid_body_field(&body, true)?;
   is_valid_alt_text_field(&data.alt_text)?;
+  is_url_blocked(&url, &url_blocklist)?;
   check_url_scheme(&url)?;
   check_url_scheme(&custom_thumbnail)?;
 

crates/api_crud/src/post/update.rs

@@ -8,6 +8,7 @@ use lemmy_api_common::{
   send_activity::{ActivityChannel, SendActivityData},
   utils::{
     check_community_user_action,
+    get_url_blocklist,
     local_site_to_slur_regex,
     process_markdown_opt,
     proxy_image_link_opt_apub,
@@ -30,6 +31,7 @@ use lemmy_utils::{
     validation::{
       check_url_scheme,
       clean_url_params,
+      is_url_blocked,
       is_valid_alt_text_field,
       is_valid_body_field,
       is_valid_post_title,
@@ -51,9 +53,11 @@ pub async fn update_post(
   let url = data.url.as_ref().map(clean_url_params);
   let custom_thumbnail = data.custom_thumbnail.as_ref().map(clean_url_params);
 
+  let url_blocklist = get_url_blocklist(&context).await?;
+
   let slur_regex = local_site_to_slur_regex(&local_site);
   check_slurs_opt(&data.name, &slur_regex)?;
-  let body = process_markdown_opt(&data.body, &slur_regex, &context).await?;
+  let body = process_markdown_opt(&data.body, &slur_regex, &url_blocklist, &context).await?;
 
   if let Some(name) = &data.name {
     is_valid_post_title(name)?;
@@ -61,6 +65,7 @@ pub async fn update_post(
 
   is_valid_body_field(&body, true)?;
   is_valid_alt_text_field(&data.alt_text)?;
+  is_url_blocked(&url, &url_blocklist)?;
   check_url_scheme(&url)?;
   check_url_scheme(&custom_thumbnail)?;
 

crates/api_crud/src/private_message/create.rs

@@ -8,6 +8,7 @@ use lemmy_api_common::{
     check_person_block,
     generate_local_apub_endpoint,
     get_interface_language,
+    get_url_blocklist,
     local_site_to_slur_regex,
     process_markdown,
     send_email_to_user,
@@ -36,7 +37,8 @@ pub async fn create_private_message(
   let local_site = LocalSite::read(&mut context.pool()).await?;
 
   let slur_regex = local_site_to_slur_regex(&local_site);
-  let content = process_markdown(&data.content, &slur_regex, &context).await?;
+  let url_blocklist = get_url_blocklist(&context).await?;
+  let content = process_markdown(&data.content, &slur_regex, &url_blocklist, &context).await?;
   is_valid_body_field(&Some(content.clone()), false)?;
 
   check_person_block(

crates/api_crud/src/private_message/update.rs

@@ -4,7 +4,7 @@ use lemmy_api_common::{
   context::LemmyContext,
   private_message::{EditPrivateMessage, PrivateMessageResponse},
   send_activity::{ActivityChannel, SendActivityData},
-  utils::{local_site_to_slur_regex, process_markdown},
+  utils::{get_url_blocklist, local_site_to_slur_regex, process_markdown},
 };
 use lemmy_db_schema::{
   source::{
@@ -37,7 +37,8 @@ pub async fn update_private_message(
 
   // Doing the update
   let slur_regex = local_site_to_slur_regex(&local_site);
-  let content = process_markdown(&data.content, &slur_regex, &context).await?;
+  let url_blocklist = get_url_blocklist(&context).await?;
+  let content = process_markdown(&data.content, &slur_regex, &url_blocklist, &context).await?;
   is_valid_body_field(&Some(content.clone()), false)?;
 
   let private_message_id = data.private_message_id;

crates/api_crud/src/site/create.rs

@@ -6,6 +6,7 @@ use lemmy_api_common::{
   site::{CreateSite, SiteResponse},
   utils::{
     generate_shared_inbox_url,
+    get_url_blocklist,
     is_admin,
     local_site_rate_limit_to_rate_limit_config,
     local_site_to_slur_regex,
@@ -58,7 +59,8 @@ pub async fn create_site(
   let keypair = generate_actor_keypair()?;
 
   let slur_regex = local_site_to_slur_regex(&local_site);
-  let sidebar = process_markdown_opt(&data.sidebar, &slur_regex, &context).await?;
+  let url_blocklist = get_url_blocklist(&context).await?;
+  let sidebar = process_markdown_opt(&data.sidebar, &slur_regex, &url_blocklist, &context).await?;
   let icon = proxy_image_link_opt_api(&data.icon, &context).await?;
   let banner = proxy_image_link_opt_api(&data.banner, &context).await?;
 

crates/api_crud/src/site/read.rs

@@ -6,6 +6,7 @@ use lemmy_api_common::{
 use lemmy_db_schema::source::{
   actor_language::{LocalUserLanguage, SiteLanguage},
   language::Language,
+  local_site_url_blocklist::LocalSiteUrlBlocklist,
   tagline::Tagline,
 };
 use lemmy_db_views::structs::{CustomEmojiView, LocalUserView, SiteView};
@@ -47,6 +48,7 @@ pub async fn get_site(
       let taglines = Tagline::get_all(&mut context.pool(), site_view.local_site.id).await?;
       let custom_emojis =
         CustomEmojiView::get_all(&mut context.pool(), site_view.local_site.id).await?;
+      let blocked_urls = LocalSiteUrlBlocklist::get_all(&mut context.pool()).await?;
       Ok(GetSiteResponse {
         site_view,
         admins,
@@ -56,6 +58,7 @@ pub async fn get_site(
         discussion_languages,
         taglines,
         custom_emojis,
+        blocked_urls,
       })
     })
     .await

crates/api_crud/src/site/update.rs

@@ -4,6 +4,7 @@ use lemmy_api_common::{
   context::LemmyContext,
   site::{EditSite, SiteResponse},
   utils::{
+    get_url_blocklist,
     is_admin,
     local_site_rate_limit_to_rate_limit_config,
     local_site_to_slur_regex,
@@ -18,6 +19,7 @@ use lemmy_db_schema::{
     federation_blocklist::FederationBlockList,
     local_site::{LocalSite, LocalSiteUpdateForm},
     local_site_rate_limit::{LocalSiteRateLimit, LocalSiteRateLimitUpdateForm},
+    local_site_url_blocklist::LocalSiteUrlBlocklist,
     local_user::LocalUser,
     site::{Site, SiteUpdateForm},
     tagline::Tagline,
@@ -34,6 +36,7 @@ use lemmy_utils::{
     validation::{
       build_and_check_regex,
       check_site_visibility_valid,
+      check_urls_are_valid,
       is_valid_body_field,
       site_description_length_check,
       site_name_length_check,
@@ -61,7 +64,8 @@ pub async fn update_site(
   }
 
   let slur_regex = local_site_to_slur_regex(&local_site);
-  let sidebar = process_markdown_opt(&data.sidebar, &slur_regex, &context).await?;
+  let url_blocklist = get_url_blocklist(&context).await?;
+  let sidebar = process_markdown_opt(&data.sidebar, &slur_regex, &url_blocklist, &context).await?;
   let icon = proxy_image_link_opt_api(&data.icon, &context).await?;
   let banner = proxy_image_link_opt_api(&data.banner, &context).await?;
 
@@ -137,6 +141,11 @@ pub async fn update_site(
   let blocked = data.blocked_instances.clone();
   FederationBlockList::replace(&mut context.pool(), blocked).await?;
 
+  if let Some(url_blocklist) = data.blocked_urls.clone() {
+    let parsed_urls = check_urls_are_valid(&url_blocklist)?;
+    LocalSiteUrlBlocklist::replace(&mut context.pool(), parsed_urls).await?;
+  }
+
   // TODO can't think of a better way to do this.
   // If the server suddenly requires email verification, or required applications, no old users
   // will be able to log in. It really only wants this to be a requirement for NEW signups.
@@ -578,6 +587,7 @@ mod tests {
       captcha_difficulty: None,
       allowed_instances: None,
       blocked_instances: None,
+      blocked_urls: None,
       taglines: None,
       registration_mode: site_registration_mode,
       reports_email_admins: None,

crates/apub/src/objects/comment.rs

@@ -18,7 +18,7 @@ use activitypub_federation::{
 use chrono::{DateTime, Utc};
 use lemmy_api_common::{
   context::LemmyContext,
-  utils::{is_mod_or_admin, local_site_opt_to_slur_regex, process_markdown},
+  utils::{get_url_blocklist, is_mod_or_admin, local_site_opt_to_slur_regex, process_markdown},
 };
 use lemmy_db_schema::{
   source::{
@@ -165,7 +165,8 @@ impl Object for ApubComment {
 
     let local_site = LocalSite::read(&mut context.pool()).await.ok();
     let slur_regex = &local_site_opt_to_slur_regex(&local_site);
-    let content = process_markdown(&content, slur_regex, context).await?;
+    let url_blocklist = get_url_blocklist(context).await?;
+    let content = process_markdown(&content, slur_regex, &url_blocklist, context).await?;
     let language_id =
       LanguageTag::to_language_id_single(note.language, &mut context.pool()).await?;
 

crates/apub/src/objects/community.rs

@@ -21,6 +21,7 @@ use lemmy_api_common::{
     generate_featured_url,
     generate_moderators_url,
     generate_outbox_url,
+    get_url_blocklist,
     local_site_opt_to_slur_regex,
     process_markdown_opt,
     proxy_image_link_opt_apub,
@@ -141,8 +142,10 @@ impl Object for ApubCommunity {
 
     let local_site = LocalSite::read(&mut context.pool()).await.ok();
     let slur_regex = &local_site_opt_to_slur_regex(&local_site);
+    let url_blocklist = get_url_blocklist(context).await?;
     let description = read_from_string_or_source_opt(&group.summary, &None, &group.source);
-    let description = process_markdown_opt(&description, slur_regex, context).await?;
+    let description =
+      process_markdown_opt(&description, slur_regex, &url_blocklist, context).await?;
     let icon = proxy_image_link_opt_apub(group.icon.map(|i| i.url), context).await?;
     let banner = proxy_image_link_opt_apub(group.image.map(|i| i.url), context).await?;
 

crates/apub/src/objects/instance.rs

@@ -19,7 +19,12 @@ use activitypub_federation::{
 use chrono::{DateTime, Utc};
 use lemmy_api_common::{
   context::LemmyContext,
-  utils::{local_site_opt_to_slur_regex, process_markdown_opt, proxy_image_link_opt_apub},
+  utils::{
+    get_url_blocklist,
+    local_site_opt_to_slur_regex,
+    process_markdown_opt,
+    proxy_image_link_opt_apub,
+  },
 };
 use lemmy_db_schema::{
   newtypes::InstanceId,
@@ -138,8 +143,9 @@ impl Object for ApubSite {
 
     let local_site = LocalSite::read(&mut context.pool()).await.ok();
     let slur_regex = &local_site_opt_to_slur_regex(&local_site);
+    let url_blocklist = get_url_blocklist(context).await?;
     let sidebar = read_from_string_or_source_opt(&apub.content, &None, &apub.source);
-    let sidebar = process_markdown_opt(&sidebar, slur_regex, context).await?;
+    let sidebar = process_markdown_opt(&sidebar, slur_regex, &url_blocklist, context).await?;
     let icon = proxy_image_link_opt_apub(apub.icon.map(|i| i.url), context).await?;
     let banner = proxy_image_link_opt_apub(apub.image.map(|i| i.url), context).await?;
 

crates/apub/src/objects/person.rs

@@ -22,6 +22,7 @@ use lemmy_api_common::{
   context::LemmyContext,
   utils::{
     generate_outbox_url,
+    get_url_blocklist,
     local_site_opt_to_slur_regex,
     process_markdown_opt,
     proxy_image_link_opt_apub,
@@ -152,8 +153,9 @@ impl Object for ApubPerson {
 
     let local_site = LocalSite::read(&mut context.pool()).await.ok();
     let slur_regex = &local_site_opt_to_slur_regex(&local_site);
+    let url_blocklist = get_url_blocklist(context).await?;
     let bio = read_from_string_or_source_opt(&person.summary, &None, &person.source);
-    let bio = process_markdown_opt(&bio, slur_regex, context).await?;
+    let bio = process_markdown_opt(&bio, slur_regex, &url_blocklist, context).await?;
     let avatar = proxy_image_link_opt_apub(person.icon.map(|i| i.url), context).await?;
     let banner = proxy_image_link_opt_apub(person.image.map(|i| i.url), context).await?;
 

crates/apub/src/objects/post.rs

@@ -26,6 +26,7 @@ use lemmy_api_common::{
   context::LemmyContext,
   request::fetch_link_metadata_opt,
   utils::{
+    get_url_blocklist,
     local_site_opt_to_sensitive,
     local_site_opt_to_slur_regex,
     process_markdown_opt,
@@ -246,9 +247,10 @@ impl Object for ApubPost {
       let thumbnail_url = proxy_image_link_opt_apub(thumbnail_url, context).await?;
 
       let slur_regex = &local_site_opt_to_slur_regex(&local_site);
+      let url_blocklist = get_url_blocklist(context).await?;
 
       let body = read_from_string_or_source_opt(&page.content, &page.media_type, &page.source);
-      let body = process_markdown_opt(&body, slur_regex, context).await?;
+      let body = process_markdown_opt(&body, slur_regex, &url_blocklist, context).await?;
       let language_id =
         LanguageTag::to_language_id_single(page.language, &mut context.pool()).await?;
 

crates/apub/src/objects/private_message.rs

@@ -14,7 +14,7 @@ use activitypub_federation::{
 use chrono::{DateTime, Utc};
 use lemmy_api_common::{
   context::LemmyContext,
-  utils::{check_person_block, local_site_opt_to_slur_regex, process_markdown},
+  utils::{check_person_block, get_url_blocklist, local_site_opt_to_slur_regex, process_markdown},
 };
 use lemmy_db_schema::{
   source::{
@@ -127,8 +127,9 @@ impl Object for ApubPrivateMessage {
 
     let local_site = LocalSite::read(&mut context.pool()).await.ok();
     let slur_regex = &local_site_opt_to_slur_regex(&local_site);
+    let url_blocklist = get_url_blocklist(context).await?;
     let content = read_from_string_or_source(&note.content, &None, &note.source);
-    let content = process_markdown(&content, slur_regex, context).await?;
+    let content = process_markdown(&content, slur_regex, &url_blocklist, context).await?;
 
     let form = PrivateMessageInsertForm {
       creator_id: creator.id,

crates/db_schema/src/impls/local_site_url_blocklist.rs

@@ -0,0 +1,49 @@
+use crate::{
+  schema::local_site_url_blocklist,
+  source::local_site_url_blocklist::{LocalSiteUrlBlocklist, LocalSiteUrlBlocklistForm},
+  utils::{get_conn, DbPool},
+};
+use diesel::{dsl::insert_into, result::Error};
+use diesel_async::{AsyncPgConnection, RunQueryDsl};
+
+impl LocalSiteUrlBlocklist {
+  pub async fn replace(pool: &mut DbPool<'_>, url_blocklist: Vec<String>) -> Result<(), Error> {
+    let conn = &mut get_conn(pool).await?;
+
+    conn
+      .build_transaction()
+      .run(|conn| {
+        Box::pin(async move {
+          use crate::schema::local_site_url_blocklist::dsl::local_site_url_blocklist;
+
+          Self::clear(conn).await?;
+
+          let forms = url_blocklist
+            .into_iter()
+            .map(|url| LocalSiteUrlBlocklistForm { url, updated: None })
+            .collect::<Vec<_>>();
+
+          insert_into(local_site_url_blocklist)
+            .values(forms)
+            .execute(conn)
+            .await?;
+
+          Ok(())
+        }) as _
+      })
+      .await
+  }
+
+  async fn clear(conn: &mut AsyncPgConnection) -> Result<usize, Error> {
+    diesel::delete(local_site_url_blocklist::table)
+      .execute(conn)
+      .await
+  }
+
+  pub async fn get_all(pool: &mut DbPool<'_>) -> Result<Vec<Self>, Error> {
+    let conn = &mut get_conn(pool).await?;
+    local_site_url_blocklist::table
+      .get_results::<Self>(conn)
+      .await
+  }
+}

crates/db_schema/src/impls/mod.rs

@@ -17,6 +17,7 @@ pub mod instance_block;
 pub mod language;
 pub mod local_site;
 pub mod local_site_rate_limit;
+pub mod local_site_url_blocklist;
 pub mod local_user;
 pub mod local_user_vote_display_mode;
 pub mod login_token;

crates/db_schema/src/schema.rs

@@ -409,6 +409,15 @@ diesel::table! {
     }
 }
 
+diesel::table! {
+    local_site_url_blocklist (id) {
+        id -> Int4,
+        url -> Text,
+        published -> Timestamptz,
+        updated -> Nullable<Timestamptz>,
+    }
+}
+
 diesel::table! {
     use diesel::sql_types::*;
     use super::sql_types::SortTypeEnum;
@@ -1052,6 +1061,7 @@ diesel::allow_tables_to_appear_in_same_query!(
     local_image,
     local_site,
     local_site_rate_limit,
+    local_site_url_blocklist,
     local_user,
     local_user_language,
     local_user_vote_display_mode,

crates/db_schema/src/source/local_site_url_blocklist.rs

@@ -0,0 +1,28 @@
+#[cfg(feature = "full")]
+use crate::schema::local_site_url_blocklist;
+use chrono::{DateTime, Utc};
+use serde::{Deserialize, Serialize};
+use serde_with::skip_serializing_none;
+#[cfg(feature = "full")]
+use ts_rs::TS;
+
+#[skip_serializing_none]
+#[derive(Clone, PartialEq, Eq, Debug, Serialize, Deserialize)]
+#[cfg_attr(feature = "full", derive(Queryable, Selectable, Identifiable, TS))]
+#[cfg_attr(feature = "full", diesel(table_name = local_site_url_blocklist))]
+#[cfg_attr(feature = "full", diesel(check_for_backend(diesel::pg::Pg)))]
+#[cfg_attr(feature = "full", ts(export))]
+pub struct LocalSiteUrlBlocklist {
+  pub id: i32,
+  pub url: String,
+  pub published: DateTime<Utc>,
+  pub updated: Option<DateTime<Utc>>,
+}
+
+#[derive(Default, Clone)]
+#[cfg_attr(feature = "full", derive(Insertable, AsChangeset))]
+#[cfg_attr(feature = "full", diesel(table_name = local_site_url_blocklist))]
+pub struct LocalSiteUrlBlocklistForm {
+  pub url: String,
+  pub updated: Option<DateTime<Utc>>,
+}

crates/db_schema/src/source/mod.rs

@@ -22,6 +22,7 @@ pub mod instance_block;
 pub mod language;
 pub mod local_site;
 pub mod local_site_rate_limit;
+pub mod local_site_url_blocklist;
 pub mod local_user;
 pub mod local_user_vote_display_mode;
 pub mod login_token;

crates/utils/src/error.rs

@@ -135,6 +135,7 @@ pub enum LemmyErrorType {
   CouldntSetAllRegistrationsAccepted,
   CouldntSetAllEmailVerified,
   Banned,
+  BlockedUrl,
   CouldntGetComments,
   CouldntGetPosts,
   InvalidUrl,

crates/utils/src/utils/markdown/mod.rs

@@ -1,6 +1,7 @@
-use crate::settings::SETTINGS;
+use crate::{error::LemmyResult, settings::SETTINGS, LemmyErrorType};
 use markdown_it::{plugins::cmark::inline::image::Image, MarkdownIt};
 use once_cell::sync::Lazy;
+use regex::RegexSet;
 use url::Url;
 use urlencoding::encode;
 
@@ -98,6 +99,13 @@ pub fn markdown_rewrite_image_links(mut src: String) -> (String, Vec<Url>) {
   (src, links)
 }
 
+pub fn markdown_check_for_blocked_urls(text: &str, blocklist: &RegexSet) -> LemmyResult<()> {
+  if blocklist.is_match(text) {
+    Err(LemmyErrorType::BlockedUrl)?
+  }
+  Ok(())
+}
+
 #[cfg(test)]
 mod tests {
   #![allow(clippy::unwrap_used)]
@@ -109,65 +117,65 @@ mod tests {
   #[test]
   fn test_basic_markdown() {
     let tests: Vec<_> = vec![
-            (
+      (
-                "headings",
+        "headings",
-                "# h1\n## h2\n### h3\n#### h4\n##### h5\n###### h6",
+        "# h1\n## h2\n### h3\n#### h4\n##### h5\n###### h6",
-                "<h1>h1</h1>\n<h2>h2</h2>\n<h3>h3</h3>\n<h4>h4</h4>\n<h5>h5</h5>\n<h6>h6</h6>\n"
+        "<h1>h1</h1>\n<h2>h2</h2>\n<h3>h3</h3>\n<h4>h4</h4>\n<h5>h5</h5>\n<h6>h6</h6>\n"
-            ),
+      ),
-            (
+      (
-                "line breaks",
+        "line breaks",
-                "First\rSecond",
+        "First\rSecond",
-                "<p>First\nSecond</p>\n"),
+        "<p>First\nSecond</p>\n"),
-            (
+      (
-                "emphasis",
+        "emphasis",
-                "__bold__ **bold** *italic* ***bold+italic***",
+        "__bold__ **bold** *italic* ***bold+italic***",
-                "<p><strong>bold</strong> <strong>bold</strong> <em>italic</em> <em><strong>bold+italic</strong></em></p>\n"
+        "<p><strong>bold</strong> <strong>bold</strong> <em>italic</em> <em><strong>bold+italic</strong></em></p>\n"
-            ),
+      ),
-            (
+      (
-                "blockquotes",
+        "blockquotes",
-                "> #### Hello\n > \n > - Hola\n > - 안영 \n>> Goodbye\n",
+        "> #### Hello\n > \n > - Hola\n > - 안영 \n>> Goodbye\n",
-                "<blockquote>\n<h4>Hello</h4>\n<ul>\n<li>Hola</li>\n<li>안영</li>\n</ul>\n<blockquote>\n<p>Goodbye</p>\n</blockquote>\n</blockquote>\n"
+        "<blockquote>\n<h4>Hello</h4>\n<ul>\n<li>Hola</li>\n<li>안영</li>\n</ul>\n<blockquote>\n<p>Goodbye</p>\n</blockquote>\n</blockquote>\n"
-            ),
+      ),
-            (
+      (
-                "lists (ordered, unordered)",
+        "lists (ordered, unordered)",
-                "1. pen\n2. apple\n3. apple pen\n- pen\n- pineapple\n- pineapple pen",
+        "1. pen\n2. apple\n3. apple pen\n- pen\n- pineapple\n- pineapple pen",
-                "<ol>\n<li>pen</li>\n<li>apple</li>\n<li>apple pen</li>\n</ol>\n<ul>\n<li>pen</li>\n<li>pineapple</li>\n<li>pineapple pen</li>\n</ul>\n"
+        "<ol>\n<li>pen</li>\n<li>apple</li>\n<li>apple pen</li>\n</ol>\n<ul>\n<li>pen</li>\n<li>pineapple</li>\n<li>pineapple pen</li>\n</ul>\n"
-            ),
+      ),
-            (
+      (
-                "code and code blocks",
+        "code and code blocks",
-                "this is my amazing `code snippet` and my amazing ```code block```",
+        "this is my amazing `code snippet` and my amazing ```code block```",
-                "<p>this is my amazing <code>code snippet</code> and my amazing <code>code block</code></p>\n"
+        "<p>this is my amazing <code>code snippet</code> and my amazing <code>code block</code></p>\n"
-            ),
+      ),
-            // Links with added nofollow attribute
+      // Links with added nofollow attribute
-            (
+      (
-                "links",
+        "links",
-                "[Lemmy](https://join-lemmy.org/ \"Join Lemmy!\")",
+        "[Lemmy](https://join-lemmy.org/ \"Join Lemmy!\")",
-                "<p><a href=\"https://join-lemmy.org/\" rel=\"nofollow\" title=\"Join Lemmy!\">Lemmy</a></p>\n"
+        "<p><a href=\"https://join-lemmy.org/\" rel=\"nofollow\" title=\"Join Lemmy!\">Lemmy</a></p>\n"
-            ),
+      ),
-            // Remote images with proxy
+      // Remote images with proxy
-            (
+      (
-                "images",
+        "images",
-                "![My linked image](https://example.com/image.png \"image alt text\")",
+        "![My linked image](https://example.com/image.png \"image alt text\")",
-                "<p><img src=\"https://example.com/image.png\" alt=\"My linked image\" title=\"image alt text\" /></p>\n"
+        "<p><img src=\"https://example.com/image.png\" alt=\"My linked image\" title=\"image alt text\" /></p>\n"
-            ),
+      ),
-            // Local images without proxy
+      // Local images without proxy
-            (
+      (
-                "images",
+        "images",
-                "![My linked image](https://lemmy-alpha/image.png \"image alt text\")",
+        "![My linked image](https://lemmy-alpha/image.png \"image alt text\")",
-                "<p><img src=\"https://lemmy-alpha/image.png\" alt=\"My linked image\" title=\"image alt text\" /></p>\n"
+        "<p><img src=\"https://lemmy-alpha/image.png\" alt=\"My linked image\" title=\"image alt text\" /></p>\n"
-            ),
+      ),
-            // Ensure spoiler plugin is added
+      // Ensure spoiler plugin is added
-            (
+      (
-                "basic spoiler",
+        "basic spoiler",
-                "::: spoiler click to see more\nhow spicy!\n:::\n",
+        "::: spoiler click to see more\nhow spicy!\n:::\n",
-                "<details><summary>click to see more</summary><p>how spicy!\n</p></details>\n"
+        "<details><summary>click to see more</summary><p>how spicy!\n</p></details>\n"
-            ),
+      ),
-            (
+      (
-                "escape html special chars",
+        "escape html special chars",
-                "<script>alert('xss');</script> hello &\"",
+        "<script>alert('xss');</script> hello &\"",
-                "<p>&lt;script&gt;alert(‘xss’);&lt;/script&gt; hello &amp;&quot;</p>\n"
+        "<p>&lt;script&gt;alert(‘xss’);&lt;/script&gt; hello &amp;&quot;</p>\n"
-            )
+      )
-        ];
+    ];
 
     tests.iter().for_each(|&(msg, input, expected)| {
       let result = markdown_to_html(input);
@@ -184,46 +192,46 @@ mod tests {
   fn test_markdown_proxy_images() {
     let tests: Vec<_> =
       vec![
-          (
+        (
-            "remote image proxied",
+          "remote image proxied",
-            "![link](http://example.com/image.jpg)",
+          "![link](http://example.com/image.jpg)",
-            "![link](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage.jpg)",
+          "![link](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage.jpg)",
-          ),
+        ),
-          (
+        (
-            "local image unproxied",
+          "local image unproxied",
-            "![link](http://lemmy-alpha/image.jpg)",
+          "![link](http://lemmy-alpha/image.jpg)",
-            "![link](http://lemmy-alpha/image.jpg)",
+          "![link](http://lemmy-alpha/image.jpg)",
-          ),
+        ),
-          (
+        (
-            "multiple image links",
+          "multiple image links",
-            "![link](http://example.com/image1.jpg) ![link](http://example.com/image2.jpg)",
+          "![link](http://example.com/image1.jpg) ![link](http://example.com/image2.jpg)",
-            "![link](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage1.jpg) ![link](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage2.jpg)",
+          "![link](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage1.jpg) ![link](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage2.jpg)",
-          ),
+        ),
-          (
+        (
-            "empty link handled",
+          "empty link handled",
-            "![image]()",
+          "![image]()",
-            "![image]()"
+          "![image]()"
-          ),
+        ),
-          (
+        (
-            "empty label handled",
+          "empty label handled",
-            "![](http://example.com/image.jpg)",
+          "![](http://example.com/image.jpg)",
-            "![](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage.jpg)"
+          "![](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage.jpg)"
-          ),
+        ),
-         (
+        (
-            "invalid image link removed",
+          "invalid image link removed",
-            "![image](http-not-a-link)",
+          "![image](http-not-a-link)",
-            "![image]()"
+          "![image]()"
-         ),
+        ),
-         (
+        (
-            "label with nested markdown handled",
+          "label with nested markdown handled",
-            "![a *b* c](http://example.com/image.jpg)",
+          "![a *b* c](http://example.com/image.jpg)",
-            "![a *b* c](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage.jpg)"
+          "![a *b* c](https://lemmy-alpha/api/v3/image_proxy?url=http%3A%2F%2Fexample.com%2Fimage.jpg)"
-         ),
+        ),
-          (
+        (
-              "custom emoji support",
+          "custom emoji support",
-              r#"![party-blob](https://www.hexbear.net/pictrs/image/83405746-0620-4728-9358-5f51b040ffee.gif "emoji party-blob")"#,
+          r#"![party-blob](https://www.hexbear.net/pictrs/image/83405746-0620-4728-9358-5f51b040ffee.gif "emoji party-blob")"#,
-              r#"![party-blob](https://lemmy-alpha/api/v3/image_proxy?url=https%3A%2F%2Fwww.hexbear.net%2Fpictrs%2Fimage%2F83405746-0620-4728-9358-5f51b040ffee.gif "emoji party-blob")"#
+          r#"![party-blob](https://lemmy-alpha/api/v3/image_proxy?url=https%3A%2F%2Fwww.hexbear.net%2Fpictrs%2Fimage%2F83405746-0620-4728-9358-5f51b040ffee.gif "emoji party-blob")"#
-              )
+        )
       ];
 
     tests.iter().for_each(|&(msg, input, expected)| {
@@ -237,6 +245,69 @@ mod tests {
     });
   }
 
+  #[test]
+  fn test_url_blocking() {
+    let set = RegexSet::new(vec![r"(https://)?example\.com/?"]).unwrap();
+
+    assert!(
+      markdown_check_for_blocked_urls(&String::from("[](https://example.com)"), &set).is_err()
+    );
+
+    assert!(markdown_check_for_blocked_urls(
+      &String::from("Go to https://example.com to get free Robux"),
+      &set
+    )
+    .is_err());
+
+    assert!(
+      markdown_check_for_blocked_urls(&String::from("[](https://example.blog)"), &set).is_ok()
+    );
+
+    assert!(markdown_check_for_blocked_urls(&String::from("example.com"), &set).is_err());
+
+    assert!(markdown_check_for_blocked_urls(
+      "Odio exercitationem culpa sed sunt
+      et. Sit et similique tempora deserunt doloremque. Cupiditate iusto
+      repellat et quis qui. Cum veritatis facere quasi repellendus sunt
+      eveniet nemo sint. Cumque sit unde est. https://example.com Alias
+      repellendus at quos.",
+      &set
+    )
+    .is_err());
+
+    let set = RegexSet::new(vec![r"(https://)?example\.com/spam\.jpg"]).unwrap();
+    assert!(markdown_check_for_blocked_urls(
+      &String::from("![](https://example.com/spam.jpg)"),
+      &set
+    )
+    .is_err());
+
+    let set = RegexSet::new(vec![
+      r"(https://)?quo\.example\.com/?",
+      r"(https://)?foo\.example\.com/?",
+      r"(https://)?bar\.example\.com/?",
+    ])
+    .unwrap();
+
+    assert!(
+      markdown_check_for_blocked_urls(&String::from("https://baz.example.com"), &set).is_ok()
+    );
+
+    assert!(
+      markdown_check_for_blocked_urls(&String::from("https://bar.example.com"), &set).is_err()
+    );
+
+    let set = RegexSet::new(vec![r"(https://)?example\.com/banned_page"]).unwrap();
+
+    assert!(
+      markdown_check_for_blocked_urls(&String::from("https://example.com/page"), &set).is_ok()
+    );
+
+    let set = RegexSet::new(vec![r"(https://)?ex\.mple\.com/?"]).unwrap();
+
+    assert!(markdown_check_for_blocked_urls("example.com", &set).is_ok());
+  }
+
   #[test]
   fn test_sanitize_html() {
     let sanitized = sanitize_html("<script>alert('xss');</script> hello &\"'");

crates/utils/src/utils/validation.rs

@@ -1,8 +1,8 @@
 use crate::error::{LemmyErrorExt, LemmyErrorType, LemmyResult};
 use itertools::Itertools;
 use once_cell::sync::Lazy;
-use regex::{Regex, RegexBuilder};
+use regex::{Regex, RegexBuilder, RegexSet};
-use url::Url;
+use url::{ParseError, Url};
 
 // From here: https://github.com/vector-im/element-android/blob/develop/matrix-sdk-android/src/main/java/org/matrix/android/sdk/api/MatrixPatterns.kt#L35
 static VALID_MATRIX_ID_REGEX: Lazy<Regex> = Lazy::new(|| {
@@ -299,6 +299,33 @@ pub fn check_url_scheme(url: &Option<Url>) -> LemmyResult<()> {
   }
 }
 
+pub fn is_url_blocked(url: &Option<Url>, blocklist: &RegexSet) -> LemmyResult<()> {
+  if let Some(url) = url {
+    if blocklist.is_match(url.as_str()) {
+      Err(LemmyErrorType::BlockedUrl)?
+    }
+  }
+
+  Ok(())
+}
+
+pub fn check_urls_are_valid(urls: &Vec<String>) -> LemmyResult<Vec<String>> {
+  let mut parsed_urls = vec![];
+  for url in urls {
+    let url = Url::parse(url).or_else(|e| {
+      if e == ParseError::RelativeUrlWithoutBase {
+        Url::parse(&format!("https://{url}"))
+      } else {
+        Err(e)
+      }
+    })?;
+
+    parsed_urls.push(url.to_string());
+  }
+
+  Ok(parsed_urls)
+}
+
 #[cfg(test)]
 mod tests {
   #![allow(clippy::unwrap_used)]
@@ -310,7 +337,9 @@ mod tests {
       build_and_check_regex,
       check_site_visibility_valid,
       check_url_scheme,
+      check_urls_are_valid,
       clean_url_params,
+      is_url_blocked,
       is_valid_actor_name,
       is_valid_bio_field,
       is_valid_display_name,
@@ -550,4 +579,38 @@ mod tests {
     let magnet_link="magnet:?xt=urn:btih:4b390af3891e323778959d5abfff4b726510f14c&dn=Ravel%20Complete%20Piano%20Sheet%20Music%20-%20Public%20Domain&tr=udp%3A%2F%2Fopen.tracker.cl%3A1337%2Fannounce";
     assert!(check_url_scheme(&Some(Url::parse(magnet_link).unwrap())).is_ok());
   }
+
+  #[test]
+  fn test_url_block() {
+    let set = regex::RegexSet::new(vec![
+      r"(https://)?example\.org/page/to/article",
+      r"(https://)?example\.net/?",
+      r"(https://)?example\.com/?",
+    ])
+    .unwrap();
+
+    assert!(is_url_blocked(&Some(Url::parse("https://example.blog").unwrap()), &set).is_ok());
+
+    assert!(is_url_blocked(&Some(Url::parse("https://example.org").unwrap()), &set).is_ok());
+
+    assert!(is_url_blocked(&None, &set).is_ok());
+
+    assert!(is_url_blocked(&Some(Url::parse("https://example.com").unwrap()), &set).is_err());
+  }
+
+  #[test]
+  fn test_url_parsed() {
+    assert_eq!(
+      vec![String::from("https://example.com/")],
+      check_urls_are_valid(&vec![String::from("example.com")]).unwrap()
+    );
+
+    assert!(check_urls_are_valid(&vec![
+      String::from("example.com"),
+      String::from("https://example.blog")
+    ])
+    .is_ok());
+
+    assert!(check_urls_are_valid(&vec![String::from("https://example .com"),]).is_err());
+  }
 }

migrations/2024-03-06-201637_url_blocklist/down.sql

@@ -0,0 +1,3 @@
+-- This file should undo anything in `up.sql`
+DROP TABLE local_site_url_blocklist;
+

migrations/2024-03-06-201637_url_blocklist/up.sql

@@ -0,0 +1,7 @@
+CREATE TABLE local_site_url_blocklist (
+    id serial NOT NULL PRIMARY KEY,
+    url text NOT NULL UNIQUE,
+    published timestamp with time zone NOT NULL DEFAULT now(),
+    updated timestamp with time zone
+);
+