From 4efbb952043ccde0408dfe5535ae98261e87d8c9 Mon Sep 17 00:00:00 2001
From: Joshua Coles <josh@coles.to>
Date: Sat, 27 Jul 2024 20:28:39 +0100
Subject: [PATCH] SOA for batch inserts using UNNEST as per docs

---
 Cargo.lock       | 21 +++++++++++++++++++
 Cargo.toml       |  1 +
 src/main.rs      | 54 +++++++++++++++++++++++++-----------------------
 src/toggl/mod.rs |  5 +++--
 4 files changed, 53 insertions(+), 28 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index c61002b..ed0b505 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1912,6 +1912,26 @@ version = "1.13.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3c5e1a9a646d36c3599cd173a41282daf47c44583ad367b8e6837255952e5c67"
 
+[[package]]
+name = "soa-rs"
+version = "0.6.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5c76685dff09a60d416b82a23e192e14cd47d147858b9aca2ce5ca05877acd93"
+dependencies = [
+ "soa-rs-derive",
+]
+
+[[package]]
+name = "soa-rs-derive"
+version = "0.6.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a4281fd25ca7e6cc6e84cadaf296ffabac12719e830e1af160cd243dca831577"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.71",
+]
+
 [[package]]
 name = "socket2"
 version = "0.5.7"
@@ -2343,6 +2363,7 @@ dependencies = [
  "serde_json",
  "serde_json_path_to_error",
  "serde_with",
+ "soa-rs",
  "sqlx",
  "thiserror",
  "tokio",
diff --git a/Cargo.toml b/Cargo.toml
index e803cc4..66ef1d4 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -25,3 +25,4 @@ futures = "0.3.30"
 tracing = "0.1.40"
 tracing-subscriber = "0.3.18"
 itertools = "0.13.0"
+soa-rs = "0.6.1"
diff --git a/src/main.rs b/src/main.rs
index fcdccd5..09cb69c 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -1,8 +1,9 @@
 use chrono::{DateTime, DurationRound, NaiveDate, TimeDelta, Utc};
 use sqlx::{Connection, PgConnection, Postgres, QueryBuilder};
 use toggl::TogglApi;
-use crate::toggl::types::{TimeEntry, TogglReportFilters};
+use crate::toggl::types::{Tag, TimeEntry, TogglReportFilters};
 use itertools::Itertools;
+use soa_rs::Soa;
 
 mod toggl;
 mod sensitive;
@@ -256,31 +257,32 @@ impl Worker {
 
     async fn update_tags(&mut self) -> Result<(), AppError> {
         let tags = self.toggl_api.get_tags().await?;
+        let tags: Soa<Tag> = Soa::from(tags.as_slice());
 
-        for tag in tags {
-            sqlx::query!(
-                r#"
-                INSERT INTO tags (id, name, workspace_id, creator_id, updated_at, deleted_at, permissions)
-                VALUES ($1, $2, $3, $4, $5, $6, $7)
-                ON CONFLICT (id) DO UPDATE SET
-                    name = excluded.name,
-                    workspace_id = excluded.workspace_id,
-                    creator_id = excluded.creator_id,
-                    updated_at = excluded.updated_at,
-                    deleted_at = excluded.deleted_at,
-                    permissions = excluded.permissions
-                "#,
-                tag.id as i64,
-                tag.name,
-                tag.workspace_id as i64,
-                tag.creator_id as i64,
-                tag.updated_at,
-                tag.deleted_at,
-                tag.permissions,
-            )
-                .execute(&mut self.db)
-                .await?;
-        }
+        sqlx::query!(
+            r#"
+            INSERT INTO tags (id, name, workspace_id, creator_id, updated_at, deleted_at, permissions)
+            SELECT * FROM UNNEST($1::bigint[], $2::text[], $3::bigint[], $4::bigint[], $5::timestamptz[], $6::timestamptz[], $7::text[])
+            ON CONFLICT (id) DO UPDATE SET
+                name = excluded.name,
+                workspace_id = excluded.workspace_id,
+                creator_id = excluded.creator_id,
+                updated_at = excluded.updated_at,
+                deleted_at = excluded.deleted_at,
+                permissions = excluded.permissions
+            "#,
+            &tags.id().iter().map(|id| *id as i64).collect::<Vec<_>>()[..],
+            tags.name(),
+            &tags.workspace_id().iter().map(|id| *id as i64).collect::<Vec<_>>()[..],
+            &tags.creator_id().iter().map(|id| *id as i64).collect::<Vec<_>>()[..],
+            tags.updated_at(),
+            // Nullable fields fail to type check with UNNEST batch inserts so we silence the
+            // errors using `: _`.
+            tags.deleted_at(): _,
+            tags.permissions(): _,
+        )
+            .execute(&mut self.db)
+            .await?;
 
         Ok(())
     }
@@ -349,7 +351,7 @@ async fn main() {
         toggl_api: api,
     };
 
-    worker.update_clients()
+    worker.update_tags()
         .await
         .unwrap();
 
diff --git a/src/toggl/mod.rs b/src/toggl/mod.rs
index 6e3820b..e8f7794 100644
--- a/src/toggl/mod.rs
+++ b/src/toggl/mod.rs
@@ -349,7 +349,7 @@ pub mod types {
         }
     }
 
-    #[derive(Serialize, Deserialize, Debug, Clone)]
+    #[derive(Soars, Serialize, Deserialize, Debug, Clone)]
     pub struct TrackingClient {
         /// The unique identifier for the client.
         pub id: i64,
@@ -378,7 +378,7 @@ pub mod types {
         pub permissions: Option<String>,
     }
 
-    #[derive(Serialize, Deserialize, Debug, Clone)]
+    #[derive(Soars, Serialize, Deserialize, Debug, Clone)]
     pub struct Tag {
         pub id: u64,
         pub name: String,
@@ -493,6 +493,7 @@ pub mod types {
     }
 
     use std::fmt;
+    use soa_rs::Soars;
 
     impl fmt::Debug for TogglReportFilters {
         fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {