DELETE implementation #475
Conversation
- currently doesn't work because in delete_record `cell_idx` is being calculated wrong.
| page.set_dirty(); | ||
| self.pager.add_dirty(page.get().id); | ||
|
|
||
| return_if_io!(self.balance_after_delete()); |
There was a problem hiding this comment.
I would need more time to reviews this. As far as I know sqlite deletes a cell from a page and the there is a loop that calls balance. balance checks wether there is an overflow (what we do now) or underflow. I want to refactor the balance because it should be as close to sqlite as possible so we should stick to that instead of having different types of balancing.
| ); | ||
| let start_offset = program.offset(); | ||
|
|
||
| let table_name = &qualified_table_name.name.0; |
There was a problem hiding this comment.
need to use normalize_ident() for this to make sure e.g. both "users" and [users] match the table users
| let table_name = &qualified_table_name.name.0; | ||
| let table = match schema.get_table(table_name) { | ||
| Some(t) => t, | ||
| None => crate::bail_corrupt_error!("No such table: {}", table_name), |
There was a problem hiding this comment.
this should be bail_parse_error!
| let mut expr = *where_expr.clone(); | ||
| bind_column_references(&mut expr, &referenced_tables)?; | ||
|
|
||
| // Handle WHERE id = value case |
There was a problem hiding this comment.
you should be able to use translate_condition_expr() for this
| returning, | ||
| .. | ||
| } => { | ||
| assert!(returning.is_none()); // Don't handle RETURNING yet |
There was a problem hiding this comment.
bail_parse_error!("DELETE ... RETURNING is not supported yet")
| .. | ||
| } => { | ||
| assert!(returning.is_none()); // Don't handle RETURNING yet | ||
| let boxed_where = where_clause.map(Box::new); |
|
Is this PR being actively worked on? 🙏 |
|
@seonWKim, yes I'm working on it. I'm working on making the PR more comprehensive. I was away for the past two weeks. |
|
@krishvishal PR for delete planning has been added to main. Maybe you can make use of it. Btw, is there any docs or code you are referring to when implementing DELETE on btree.rs? I also want to take a look |
core/storage/btree.rsdelete_recordcell_idxis being calculated wrong.I'm still studying the sqlite source code and some other material. So this implementation can be wrong.