Repository navigation
fix!: turn silent fallbacks into clear errors - #1473
Merged
Merged
Conversation
Cursor delete_insert with a unique_key also replaces rows whose key is in the delta. Python loaders type Decimal values as DECIMAL and reject unrecognised types. keep_in_use stops when usage can't be read in full. Databricks freshness observes each source on its own. Project-level target locations resolve like models' locations. Not yet reviewed; make check-ci not yet run.
kvlonge
enabled auto-merge
November 14, 2026 06:11
… ignored connection locations
… keep preserve consistent for loaders
kvlonge
disabled auto-merge
November 14, 2026 08:21
…a check A managed source's loader writes to the source's own schema or the target's loader_schema, so an unused connection schema no longer reports it as an intermediate loader without a schema. Two older tests follow the connection location rules: debug reports no database for a lone named connection, and the Snowflake early-failure case sets no database anywhere, since a lone top-level connection now supplies one.
kvlonge
enabled auto-merge
November 14, 2026 08:50
kvlonge
disabled auto-merge
November 14, 2026 09:07
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
Five paths fell back silently to something weaker than the docs promise: keyed cursor writes duplicated rows, loader decimals lost precision,
keep_in_usearchived without usage, Databricks freshness switched every source to DDL time, and locations fell back to the connection.Changes
delete_insertwith a cursor andunique_keydeletes rows in the window or whose key is in the delta, then inserts. Snowflake stages outside its transaction, BigQuery uses one MERGE, Databricks DELETE then INSERT.Decimalvalues become DECIMAL, written exactly; BigQuery widens to BIGNUMERIC.keep_in_usestops with C503 when usage is missing or partial.[connection]is the default target; with named connections or targets, the target then[defaults], never the connection. S101/S103/S104 andsqb loadname an ignored connection value and the line to add.BREAKING CHANGE: tools no longer read a named connection's location; with
[defaults]set and no target value, loader tables move there. An inline[connection]now places models on Snowflake, BigQuery and Databricks, where 0.223.0 gave S101. With named connections or targets, a connectionschemano longer places models: such projects fail with S104 or S103 naming it. Undeclared loader columns with unsupported, mixed or widening decimal values fail.keep_in_usestops with C503 when usage is incomplete.Verification
make check-cipass.