surrealpatch/lib/src/doc/insert.rs

use crate::ctx::Context;
use crate::dbs::Options;
use crate::dbs::Statement;
use crate::dbs::Transaction;
use crate::doc::Document;
use crate::err::Error;
use crate::sql::value::Value;

impl<'a> Document<'a> {
	pub async fn insert(
		&mut self,
		_ctx: &Context<'_>,
		_opt: &Options,
		_txn: &Transaction,
		_stm: &Statement<'_>,
	) -> Result<Value, Error> {
		todo!()
	}
}
Don’t clone variables when processing sub-contexts Closes SUR-53 When creating a new context for subqueries or statement clauses, we used to have to clone any variables/values, and freeze the context, so that it could be used across threads and async boundaries. Now with the new executor pattern for parallel queries, we can pass references instead, improving performance by removing unnecessary cloning of values. 2022-05-14 12:35:08 +00:00			`use crate::ctx::Context;`
Add initial iterator blueprint implementation 2022-02-06 01:14:56 +00:00			`use crate::dbs::Options;`
			`use crate::dbs::Statement;`
Only pass transaction when processing queries Instead of passing the executor instance, we only need to pass the current transaction which is to be used for processing and running any queries. 2022-02-15 01:00:30 +00:00			`use crate::dbs::Transaction;`
Add initial iterator blueprint implementation 2022-02-06 01:14:56 +00:00			`use crate::doc::Document;`
			`use crate::err::Error;`
			`use crate::sql::value::Value;`
Add document package blueprint 2022-01-13 07:00:50 +00:00
Use copy-on-write semantics for document values Instead of blindly cloning document values (even if we don't need to mutate them), we now use copy-on-write semantics to only create a copy if we need to update a document value. This improves performance, especially when selecting documents. 2022-02-13 19:03:00 +00:00			`impl<'a> Document<'a> {`
Add initial iterator blueprint implementation 2022-02-06 01:14:56 +00:00			`pub async fn insert(`
			`&mut self,`
Don’t clone variables when processing sub-contexts Closes SUR-53 When creating a new context for subqueries or statement clauses, we used to have to clone any variables/values, and freeze the context, so that it could be used across threads and async boundaries. Now with the new executor pattern for parallel queries, we can pass references instead, improving performance by removing unnecessary cloning of values. 2022-05-14 12:35:08 +00:00			`_ctx: &Context<'_>,`
Don't store selected NS and DB on executor 2022-02-06 21:06:52 +00:00			`_opt: &Options,`
Switch to asynchronous owned mutex locks in underlying kv storage 2022-02-15 03:33:16 +00:00			`_txn: &Transaction,`
Don’t use static futures for parallel query execution Instead of using static futures when processing a query in parallel, we now use a new executor model, which allows us to process futures which make use of references. As a result, we can remove the need to store each statement in an Arc. 2022-05-13 20:46:56 +00:00			`_stm: &Statement<'_>,`
Add initial iterator blueprint implementation 2022-02-06 01:14:56 +00:00			`) -> Result<Value, Error> {`
			`todo!()`
			`}`
			`}`