feat: implement batch operation for data import

This commit is contained in:
qaz741wsd856 2025-11-27 17:21:21 +08:00
parent 77a604c600
commit fd7fc3c202
2 changed files with 60 additions and 9 deletions

View file

@ -2,7 +2,7 @@ use axum::{extract::State, Json};
use chrono::Utc;
use std::sync::Arc;
use uuid::Uuid;
use worker::{query, Env};
use worker::{query, D1PreparedStatement, Env};
use crate::auth::Claims;
use crate::db;
@ -11,6 +11,38 @@ use crate::models::cipher::{Cipher, CipherData};
use crate::models::folder::Folder;
use crate::models::import::ImportRequest;
/// Get the batch size from environment variable IMPORT_BATCH_SIZE.
/// Defaults to 30 if not set or invalid.
fn get_batch_size(env: &Env) -> usize {
env.var("IMPORT_BATCH_SIZE")
.ok()
.and_then(|v| v.to_string().parse::<usize>().ok())
.unwrap_or(30)
}
/// Execute statements in batches. If batch_size is 0, execute all in one batch.
async fn execute_in_batches(
db: &worker::D1Database,
statements: Vec<D1PreparedStatement>,
batch_size: usize,
) -> Result<(), AppError> {
if statements.is_empty() {
return Ok(());
}
if batch_size == 0 {
// Execute all statements in one batch
db.batch(statements).await?;
} else {
// Execute in chunks of batch_size
for chunk in statements.chunks(batch_size) {
db.batch(chunk.to_vec()).await?;
}
}
Ok(())
}
#[worker::send]
pub async fn import_data(
claims: Claims,
@ -20,6 +52,10 @@ pub async fn import_data(
let db = db::get_db(&env)?;
let now = Utc::now();
let now = now.format("%Y-%m-%dT%H:%M:%S%.3fZ").to_string();
let batch_size = get_batch_size(&env);
// Prepare all folder insert statements
let mut folder_statements: Vec<D1PreparedStatement> = Vec::with_capacity(payload.folders.len());
for import_folder in &payload.folders {
let folder = Folder {
@ -30,7 +66,7 @@ pub async fn import_data(
updated_at: now.clone(),
};
query!(
let stmt = query!(
&db,
"INSERT OR IGNORE INTO folders (id, user_id, name, created_at, updated_at) VALUES (?1, ?2, ?3, ?4, ?5)",
folder.id,
@ -39,11 +75,15 @@ pub async fn import_data(
folder.created_at,
folder.updated_at
)
.map_err(|_| AppError::Database)?
.run()
.await?;
.map_err(|_| AppError::Database)?;
folder_statements.push(stmt);
}
// Execute folder inserts in batches
execute_in_batches(&db, folder_statements, batch_size).await?;
// Process folder relationships
for relationship in payload.folder_relationships {
if let Some(cipher) = payload.ciphers.get_mut(relationship.key) {
if let Some(folder) = payload.folders.get(relationship.value) {
@ -52,6 +92,9 @@ pub async fn import_data(
}
}
// Prepare all cipher insert statements
let mut cipher_statements: Vec<D1PreparedStatement> = Vec::with_capacity(payload.ciphers.len());
for import_cipher in payload.ciphers {
if import_cipher.encrypted_for != claims.sub {
return Err(AppError::BadRequest("Cipher encrypted for wrong user".to_string()));
@ -91,7 +134,7 @@ pub async fn import_data(
let data = serde_json::to_string(&cipher.data).map_err(|_| AppError::Internal)?;
query!(
let stmt = query!(
&db,
"INSERT OR IGNORE INTO ciphers (id, user_id, organization_id, type, data, favorite, folder_id, created_at, updated_at)
VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9)",
@ -104,10 +147,13 @@ pub async fn import_data(
cipher.folder_id,
cipher.created_at,
cipher.updated_at,
).map_err(|_|AppError::Database)?
.run()
.await?;
).map_err(|_| AppError::Database)?;
cipher_statements.push(stmt);
}
// Execute cipher inserts in batches
execute_in_batches(&db, cipher_statements, batch_size).await?;
Ok(Json(()))
}

View file

@ -5,6 +5,11 @@ compatibility_date = "2025-09-19"
[build]
command = "cargo install -q worker-build && worker-build --release"
[vars]
# Optional: Set the batch size for imports. Defaults to 30 if not set.
# Set to 0 means no batching (all records imported in a single batch).
# IMPORT_BATCH_SIZE = "30"
[[d1_databases]]
binding = "vault1"
database_name = "vault1"