tursodatabase/turso - CodSpeed

turso

Blog Docs Changelog

Performance History

Latest Results

core: Implement PRAGMA full_column_names and short_column_names These deprecated SQLite pragmas control how column names appear in result sets. They are needed to pass the upstream select1.test TCL tests (select1-6.1.1 through select1-6.9.16). The column naming rules match SQLite: - full=ON: explicit column refs use REAL_TABLE.COLUMN - short=ON (default), full=OFF: explicit column refs use COLUMN - both OFF: use original expression text (preserving spacing) - SELECT * expansion sets aliases at compile time, bypassing the full/short logic (longNames = full=ON && short=OFF uses TABLE.COLUMN) - AS aliases always take priority

penberg:pragma-column-names

41 minutes ago

Turso 0.6.0-pre.13

main

42 minutes ago

Merge 'Sync add column idempotent replay' from Nikita Sivukhin Make `ALTER TABLE ADD COLUMN` idempotent in the sync engine for both push and pull operations. When both the local client and the remote independently add the same column, sync previously failed with `duplicate column name` errors. Now the operation is safely skipped when the column already exists. ### Pull (local replay) Added `execute_ddl_idempotent()` in the replay generator that parses the DDL statement: if it's an `ALTER TABLE ADD COLUMN` and the column already exists (checked via `PRAGMA table_info`), the statement is skipped. ### Push (HRANA batch) `ALTER TABLE ADD COLUMN` steps in the push batch are tracked by index. The `sql_execute_http` function now accepts a set of step indices whose errors should be ignored, so a duplicate-column error on the remote doesn't abort the entire batch. This is not ideal but we are limited by the hrana capabilities here :( Reviewed-by: Mikaël Francoeur (@LeMikaelF) Closes #6204

main

4 hours ago

Refactor autovacuum pointer map handling Refactor autovacuum pointer map updates for clarity and maintainability.

Ritish134:main

5 hours ago

fix wake panicking on yield completion

pedrocarlo:db-manager-lock

5 hours ago

dont hold database manager lock

pedrocarlo:db-manager-lock

5 hours ago

fix: mark expression-index columns in col_used_mask during UPDATE planning (#6206) When a table has a TEXT PRIMARY KEY and a virtual column with an index, UPDATE panicked because the autoindex was incorrectly considered "covering" -- the col_used_mask only included columns from WHERE/SET expressions, not columns referenced by virtual column expressions in indexes needing update. This caused the table cursor to not be opened, and subsequent reads of non-indexed columns hit an out-of-bounds panic. The fix accumulates all columns referenced by expression-index columns and marks them in col_used_mask, so the covering-index check sees all columns the UPDATE actually needs to read. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

fix-update-gencol-index-panic

7 hours ago

dont hold database manager lock

pedrocarlo:db-manager-lock

8 hours ago

Latest Branches

-1%

core: Implement PRAGMA full_column_names and short_column_names#6200

19 hours ago

1fd54a1

penberg:pragma-column-names

0%

fix(autovacuum): register btree split pages in ptrmap#6193

5 hours ago

362b066

Ritish134:main

0%

Don't hold `DATABASE_MANAGER` lock on db open#6208

5 hours ago

8f77164

pedrocarlo:db-manager-lock

© 2026 CodSpeed Technology

Home Terms Privacy Docs