Replies: 3 comments
|
From the current source, the CSV node does not expose arbitrary DuckDB The CSV UI currently maps to a fixed set of properties: path, header, delimiter, quote character, encoding, skipped lines, null sentinel, plus schema/date-format handling. The engine then builds For your specific case, the closest current workaround is to use a DuckDB source with Advanced custom SQL and call DuckDB directly, for example: SELECT *
FROM read_csv_auto('/path/to/folder/*.csv', filename = true);If the UI requires a DuckDB database path for that source, point it at a small/empty Longer term, I think this would fit well as either:
That would cover this without forcing the whole node into custom SQL. |
|
Nice one guys! I will review and implement soon! |
|
Done in v0.5.0. Please download - https://github.com/slothflowlabs/duckle/releases/tag/v0.5.0 and let me know! |
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
DuckDB supports quite a number of parameters for loading the supported sources. Do you currently or plan to have the ability to manually control setting additional parameter options that are not presented in your UI (Basic Tab)? For example, I want to read all CSV files from a folder (wildcard in filename), and have the filename=True option so I can additionally extract data from the filename. Alternatively, allow me execute the read using my own custom SQL?
All reactions