core: filter all messages with links in text when link content filter is used (#6591)

* core: change msg_content_tag stored as blob to text

* track if items have links and use has_link when searching for links (MCLink_ MsgContentTag is passed)

* fix test (broken in master)

* merge migrations, update api and schema

* minor fixes

* tests

* update plans

* update migrations
This commit is contained in:
Evgeny
2026-01-23 12:31:42 +00:00
committed by GitHub
parent d08202ee2d
commit 5453d8875c
21 changed files with 355 additions and 82 deletions
@@ -25,6 +25,7 @@ import Simplex.Chat.Store.Postgres.Migrations.M20251117_member_relations_vector
import Simplex.Chat.Store.Postgres.Migrations.M20251128_migrate_member_relations
import Simplex.Chat.Store.Postgres.Migrations.M20251230_strict_tables
import Simplex.Chat.Store.Postgres.Migrations.M20260108_chat_indices
import Simplex.Chat.Store.Postgres.Migrations.M20260122_has_link
import Simplex.Messaging.Agent.Store.Shared (Migration (..))
schemaMigrations :: [(String, Text, Maybe Text)]
@@ -49,7 +50,8 @@ schemaMigrations =
("20251117_member_relations_vector", m20251117_member_relations_vector, Just down_m20251117_member_relations_vector),
("20251128_migrate_member_relations", m20251128_migrate_member_relations, Just down_m20251128_migrate_member_relations),
("20251230_strict_tables", m20251230_strict_tables, Just down_m20251230_strict_tables),
("20260108_chat_indices", m20260108_chat_indices, Just down_m20260108_chat_indices)
("20260108_chat_indices", m20260108_chat_indices, Just down_m20260108_chat_indices),
("20260122_has_link", m20260122_has_link, Just down_m20260122_has_link)
]
-- | The list of migrations in ascending order by date
@@ -0,0 +1,32 @@
{-# LANGUAGE OverloadedStrings #-}
{-# LANGUAGE QuasiQuotes #-}
module Simplex.Chat.Store.Postgres.Migrations.M20260122_has_link where
import Data.Text (Text)
import Text.RawString.QQ (r)
m20260122_has_link :: Text
m20260122_has_link =
[r|
ALTER TABLE chat_items ADD COLUMN has_link SMALLINT NOT NULL DEFAULT 0;
UPDATE chat_items SET msg_content_tag = 'text' WHERE msg_content_tag = 'liveText';
UPDATE chat_items SET has_link = 1
WHERE msg_content_tag = 'link' OR item_text LIKE '%https://%';
CREATE INDEX idx_chat_items_groups_has_link_item_ts ON chat_items(user_id, group_id, has_link, item_ts);
CREATE INDEX idx_chat_items_contacts_has_link_created_at ON chat_items(user_id, contact_id, has_link, created_at);
CREATE INDEX idx_chat_items_note_folder_has_link_created_at ON chat_items(user_id, note_folder_id, has_link, created_at);
|]
down_m20260122_has_link :: Text
down_m20260122_has_link =
[r|
DROP INDEX idx_chat_items_note_folder_has_link_created_at;
DROP INDEX idx_chat_items_contacts_has_link_created_at;
DROP INDEX idx_chat_items_groups_has_link_item_ts;
ALTER TABLE chat_items DROP COLUMN has_link;
|]
@@ -342,7 +342,8 @@ CREATE TABLE test_chat_schema.chat_items (
user_mention smallint DEFAULT 0 NOT NULL,
group_scope_tag text,
group_scope_group_member_id bigint,
show_group_as_sender smallint DEFAULT 0 NOT NULL
show_group_as_sender smallint DEFAULT 0 NOT NULL,
has_link smallint DEFAULT 0 NOT NULL
);
@@ -1813,6 +1814,10 @@ CREATE INDEX idx_chat_items_contacts_created_at ON test_chat_schema.chat_items U
CREATE INDEX idx_chat_items_contacts_has_link_created_at ON test_chat_schema.chat_items USING btree (user_id, contact_id, has_link, created_at);
CREATE INDEX idx_chat_items_contacts_msg_content_tag_created_at ON test_chat_schema.chat_items USING btree (user_id, contact_id, msg_content_tag, created_at);
@@ -1873,6 +1878,10 @@ CREATE INDEX idx_chat_items_groups ON test_chat_schema.chat_items USING btree (u
CREATE INDEX idx_chat_items_groups_has_link_item_ts ON test_chat_schema.chat_items USING btree (user_id, group_id, has_link, item_ts);
CREATE INDEX idx_chat_items_groups_history ON test_chat_schema.chat_items USING btree (user_id, group_id, include_in_history, item_deleted, item_ts, chat_item_id);
@@ -1901,6 +1910,10 @@ CREATE INDEX idx_chat_items_item_status ON test_chat_schema.chat_items USING btr
CREATE INDEX idx_chat_items_note_folder_has_link_created_at ON test_chat_schema.chat_items USING btree (user_id, note_folder_id, has_link, created_at);
CREATE INDEX idx_chat_items_note_folder_msg_content_tag_created_at ON test_chat_schema.chat_items USING btree (user_id, note_folder_id, msg_content_tag, created_at);