Pedro Holanda
Pedro Holanda
When writing a compressed CSV file, the correct extensions will be enforced. For gzip: `.csv.gz` For zstd: `.csv.zst` One potential downside is that users won't be able to write a...
This PR adds: `TIME` `DECIMAL` To types that are directly cast in both the sniffer and the scanner. This means that we don't have to go through `Value` or `Vector`...
Related to duckdb/duckdb#8089
This PR is the continuation of #12167 It adds tests for a variety of situations using the skip operator. Including: * Skipping multiple buffers. * Skipping over big parts of...
When scanning compressed files, we use a different strategy to check for progress. Instead of counting the bytes our CSV scanners have scanned from the file and comparing that to...
This PR introduces an adaptive sniffer that is used for multi-file reading. Instead of always doing a full run of the CSV sniffer on every file, we use the information...
This PR introduces support for generating and consuming query plans with flattening subquery operators (i.e., delim joins and gets). It also now supports full TPC-H roundtrip through DuckDB. It's important...
* Radixsort MSD is giving wrong results after Refinement * Need to reintegrate Consolidation Phase
Hugeint is partially supported in the Arrow integration by defaulting to a `decimal(38,0)`. However, this can cause conversion errors and does not allow for proper round-tripping of this type. To...