Specifying a format when reading a datalink should override the format for the target file, not the datalink file

[radeusgd]

Currently, Data.read "./my_csv.datalink" Plain_Text will override the file format and read the datalink definition itself as plain text.

As discussed with @jdunkerley, this is not the desired behaviour. The datalinks should have special handling and instead always resolve as a datalink. If a format is specified, it should override the format that was specified inside of the datalink.

Note: this means that the user cannot easily read datalink definition. To do so, they first need to rename the file to change the extension. Moving/copying the datalink will still copy the target contents, not the datalink description.

• Change the behaviour for datalinks, overriding the selected format in read.
• Add a test case for overriding the format.
• If a datalink is 'format-less' (e.g. Postgres Table), only accept default format (Auto_Detect). Format overrides will result in an error as they are meaningless in such context.

[enso-bot]

Radosław Waśko reports a new STANDUP for yesterday (2024-03-14):

Progress: Some further debugging, then passed over to Jaroslav who found a fix for the issue that was blocking my PR. Implemented an initial draft of datalink-as-symlinks, but still WIP. It should be finished by 2024-03-20.

Next Day: Next day I will be working on the #8590 task. Serialization is waiting on #9361 so I will try to implement other parts needed for the PR (common IrToTruffle::processName abstraction).

[enso-bot]

Radosław Waśko reports a new STANDUP for yesterday (2024-03-18):

Progress: Getting the pending PRs through CI - fixing some tests. Adding datalink checks to primitive file operations. It should be finished by 2024-03-20.

Next Day: Next day I will be working on the same task. Finish the refactor, check that datalinks still work. Add tests for raw operations and datalink copy/move/delete.

[enso-bot]

Radosław Waśko reports a new STANDUP for yesterday (2024-03-19):

Progress: Added remaining implementation, updated existing tests - got into a state that existing tests seem to be passing again. It should be finished by 2024-03-20.

Next Day: Next day I will be working on the same task. Add more tests for raw operations (No_Follow) and datalink copy/move/delete (on all backends), datalink from S3 (not to S3).

[enso-bot]

Radosław Waśko reports a new 🔴 DELAY for today (2024-03-21):

Summary: There is 2 days delay in implementation of the Specifying a format when reading a datalink should override the format for the target file, not the datalink file (#9324) task.
It will cause 0 days delay for the delivery of this weekly plan.

Delay Cause: Unexpected complexity - only recently we realized we want to handle datalinks in Data.fetch as well - more tests and handling were needed.

[enso-bot]

Radosław Waśko reports a new STANDUP for yesterday (2024-03-20):

Progress: Added more tests for copying scenarios, raw config reading etc. Adding tests for Data.fetch and others. It should be finished by 2024-03-22.

Next Day: Next day I will be working on the same task. Some final touches before making the PR ready. Work on Types.

[enso-bot]

Radosław Waśko reports a new STANDUP for yesterday (2024-03-21):

Progress: Adding some more edge case tests. PR in review. Fixed the IllegalStateBug in Types work - I realized I had a misconception about how BindingsMap works - I have now fixed my logic but it needs a bit of a refactor as the code got too ugly for my liking. It should be finished by 2024-03-22.

Next Day: Next day I will be working on the same task. Get the PR merged. Slight refactor for types. Pick a next task.