From cbaa2c02c7abe907c417874e1af41fda1515ca71 Mon Sep 17 00:00:00 2001 From: Raphael Taylor-Davies <1781103+tustvold@users.noreply.github.com> Date: Sat, 20 Aug 2022 15:47:01 +0100 Subject: [PATCH] More record skipping fuzz tests (#2505) --- parquet/src/arrow/arrow_reader/mod.rs | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/parquet/src/arrow/arrow_reader/mod.rs b/parquet/src/arrow/arrow_reader/mod.rs index b7c2db255b6..bf0fcd2bb09 100644 --- a/parquet/src/arrow/arrow_reader/mod.rs +++ b/parquet/src/arrow/arrow_reader/mod.rs @@ -605,7 +605,6 @@ pub(crate) fn evaluate_predicate( #[cfg(test)] mod tests { - use bytes::Bytes; use std::cmp::min; use std::collections::VecDeque; use std::fmt::Formatter; @@ -614,6 +613,7 @@ mod tests { use std::path::PathBuf; use std::sync::Arc; + use bytes::Bytes; use rand::{thread_rng, Rng, RngCore}; use tempfile::tempfile; @@ -1299,12 +1299,19 @@ mod tests { // Test with nulls and row filter TestOptions::new(2, 256, 93) .with_null_percent(25) + .with_max_data_page_size(10) .with_row_filter(), - // Test with nulls and row filter + // Test with nulls and row filter and small pages TestOptions::new(2, 256, 93) .with_null_percent(25) + .with_max_data_page_size(10) .with_row_selections() .with_row_filter(), + // Test with row selection and no offset index and small pages + TestOptions::new(2, 256, 93) + .with_enabled_statistics(EnabledStatistics::None) + .with_max_data_page_size(10) + .with_row_selections(), ]; all_options.into_iter().for_each(|opts| { @@ -1417,7 +1424,6 @@ mod tests { file.rewind().unwrap(); - // TODO: Should be able to always enable page index (#2434) let options = ArrowReaderOptions::new() .with_page_index(opts.enabled_statistics == EnabledStatistics::Page);