Re: [PR] feat: Infer partition values from statistics [iceberg-rust]

2025-04-02 Thread via GitHub
jonathanc-n commented on PR #1079: URL: https://github.com/apache/iceberg-rust/pull/1079#issuecomment-2773806669 I think the function name is misleading I will change that. We are passing in the lower and upper bounds computed from the original parquet file read during the `parquet_to_data_

Re: [PR] feat: Infer partition values from statistics [iceberg-rust]

2025-04-02 Thread via GitHub
ZENOTME commented on PR #1079: URL: https://github.com/apache/iceberg-rust/pull/1079#issuecomment-2772641331 > Hi, @jonathanc-n I'm quite confused about this pr, how can you infer partition value from statistics? First of all, statistics are optional, and they are maybe inaccurate. For exam

Re: [PR] feat: Infer partition values from statistics [iceberg-rust]

2025-04-01 Thread via GitHub
ZENOTME commented on code in PR #1079: URL: https://github.com/apache/iceberg-rust/pull/1079#discussion_r2024083268 ## crates/iceberg/src/writer/file_writer/parquet_writer.rs: ## @@ -458,6 +459,54 @@ impl ParquetWriter { Ok(builder) } + +#[allow(dead_code)] +