From c0734731cb1b4500091d0cf1ba12f45ee68f2886 Mon Sep 17 00:00:00 2001
From: yggverse <yggverse@project>
Date: Wed, 7 Jan 2026 19:02:00 +0200
Subject: [PATCH] disallow nullable title/description values for the content
 table, implement `contents_by_channel_item_id_source_id`, return last insert
 id for `insert_content`, fix `content_id` data type, implement initial
 content version save on crawl

---
 crates/crawler/src/main.rs | 52 ++++++++++++++++++++++++++++----------
 crates/mysql/src/lib.rs    | 26 ++++++++++++++++---
 2 files changed, 60 insertions(+), 18 deletions(-)

diff --git a/crates/crawler/src/main.rs b/crates/crawler/src/main.rs
index 7cdfc0b..fe86502 100644
--- a/crates/crawler/src/main.rs
+++ b/crates/crawler/src/main.rs
@@ -139,37 +139,61 @@ fn crawl(db: &mut Mysql, channel_config: &config::Channel) -> Result<()> {
             )?,
         };
 
-        // @TODO preload remote content
+        // preload remote content
 
         let title = match channel_config.content_title_selector {
             Some(ref selector) => match scrape(&link, selector) {
-                Ok(value) => value,
+                Ok(value) => match value {
+                    Some(title) => title,
+                    None => {
+                        warn!("Could not scrape `title` selector in `{channel_url}`");
+                        continue;
+                    }
+                },
                 Err(e) => {
                     warn!("Could not update `title` selector in `{channel_url}`: `{e}`");
                     continue;
                 }
             },
-            None => None,
+            None => match channel_item.title {
+                Some(ref title) => title.clone(),
+                None => {
+                    warn!(
+                        "Could not assign `title` from channel item for content in `{channel_url}`"
+                    );
+                    continue;
+                }
+            },
         };
-
         let description = match channel_config.content_description_selector {
             Some(ref selector) => match scrape(&link, selector) {
-                Ok(value) => value,
+                Ok(value) => match value {
+                    Some(description) => description,
+                    None => {
+                        warn!("Could not scrape `description` selector in `{channel_url}`");
+                        continue;
+                    }
+                },
                 Err(e) => {
                     warn!("Could not update `description` selector in `{channel_url}`: `{e}`");
                     continue;
                 }
             },
-            None => None,
+            None => match channel_item.description {
+                Some(ref description) => description.clone(),
+                None => {
+                    warn!(
+                        "Could not assign `description` from channel item for content in `{channel_url}`"
+                    );
+                    continue;
+                }
+            },
         };
-
-        if title.is_none() && description.is_none() {
-            continue;
-        }
-
-        // @TODO insert content record
-
-        println!("{:?}", description)
+        assert!(
+            db.contents_by_channel_item_id_source_id(channel_item_id, None, Some(1))?
+                .is_empty()
+        );
+        let _content_id = db.insert_content(channel_item_id, None, title, description)?;
     }
     Ok(())
 }
diff --git a/crates/mysql/src/lib.rs b/crates/mysql/src/lib.rs
index 92ffb27..4219d50 100644
--- a/crates/mysql/src/lib.rs
+++ b/crates/mysql/src/lib.rs
@@ -93,17 +93,34 @@ impl Mysql {
         Ok(self.connection.last_insert_id())
     }
 
+    pub fn contents_by_channel_item_id_source_id(
+        &mut self,
+        channel_item_id: u64,
+        source_id: Option<u64>,
+        limit: Option<usize>,
+    ) -> Result<Vec<Content>, Error> {
+        self.connection.exec_map(
+            format!(
+                "SELECT `content_id`, `channel_item_id`, `source_id`, `title`, `description` FROM `content` WHERE `channel_item_id` = ? AND `source_id` = ? LIMIT {}",
+                limit.unwrap_or(DEFAULT_LIMIT)
+            ),
+            (channel_item_id, source_id),
+            |(content_id, channel_item_id,source_id, title, description)| Content { content_id, channel_item_id, source_id, title, description },
+        )
+    }
+
     pub fn insert_content(
         &mut self,
         channel_item_id: u64,
         source_id: Option<u64>,
-        title: &str,
-        description: &str,
-    ) -> Result<(), Error> {
+        title: String,
+        description: String,
+    ) -> Result<u64, Error> {
         self.connection.exec_drop(
             "INSERT INTO `content` SET `channel_item_id` = ?, `source_id` = ?, `title` = ?, `description` = ?",
             (channel_item_id, source_id, title, description ),
-        )
+        )?;
+        Ok(self.connection.last_insert_id())
     }
 }
 
@@ -126,6 +143,7 @@ pub struct ChannelItem {
 
 #[derive(Debug, PartialEq, Eq)]
 pub struct Content {
+    pub content_id: u64,
     pub channel_item_id: u64,
     /// None if the original `title` and `description` values
     /// parsed from the channel item on crawl