Skip to content

Commit c111b21

Browse files
committed
Use official Arrow version
1 parent 4e0589a commit c111b21

5 files changed

Lines changed: 85 additions & 77 deletions

File tree

Cargo.lock

Lines changed: 70 additions & 52 deletions
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

Cargo.toml

Lines changed: 7 additions & 19 deletions
Original file line numberDiff line numberDiff line change
@@ -90,19 +90,12 @@ ahash = { version = "0.8", default-features = false, features = [
9090
"runtime-rng",
9191
] }
9292
apache-avro = { version = "0.20", default-features = false }
93-
arrow = { git = "https://github.com/Coralogix/arrow-rs.git", tag = "v56.0.0-cx.0", features = [
94-
"prettyprint",
95-
"chrono-tz",
96-
] }
97-
arrow-buffer = { git = "https://github.com/Coralogix/arrow-rs.git", tag = "v56.0.0-cx.0", default-features = false }
98-
arrow-flight = { git = "https://github.com/Coralogix/arrow-rs.git", tag = "v56.0.0-cx.0", features = [
99-
"flight-sql-experimental",
100-
] }
101-
arrow-ipc = { git = "https://github.com/Coralogix/arrow-rs.git", tag = "v56.0.0-cx.0", default-features = false, features = [
102-
"lz4",
103-
] }
104-
arrow-ord = { git = "https://github.com/Coralogix/arrow-rs.git", tag = "v56.0.0-cx.0", default-features = false }
105-
arrow-schema = { git = "https://github.com/Coralogix/arrow-rs.git", tag = "v56.0.0-cx.0", default-features = false }
93+
arrow = { version = "56.0.0", features = ["prettyprint", "chrono-tz"] }
94+
arrow-buffer = { version = "56.0.0", default-features = false }
95+
arrow-flight = { version = "56.0.0", features = ["flight-sql-experimental"] }
96+
arrow-ipc = { version = "56.0.0", default-features = false, features = ["lz4"] }
97+
arrow-ord = { version = "56.0.0", default-features = false }
98+
arrow-schema = { version = "56.0.0", default-features = false }
10699
async-trait = "0.1.89"
107100
bigdecimal = "0.4.8"
108101
bytes = "1.10"
@@ -157,12 +150,7 @@ itertools = "0.14"
157150
log = "^0.4"
158151
object_store = { version = ">=0.12.3, <0.13", default-features = false }
159152
parking_lot = "0.12"
160-
parquet = { git = "https://github.com/Coralogix/arrow-rs.git", tag = "v56.0.0-cx.0", default-features = false, features = [
161-
"arrow",
162-
"async",
163-
"object_store",
164-
"encryption",
165-
] }
153+
parquet = { version = "56.0.0", default-features = false, features = ["arrow", "async", "object_store", "encryption"] }
166154
pbjson = { version = "0.7.0" }
167155
pbjson-types = "0.7"
168156
# Should match arrow-flight's version of prost.

datafusion/datasource-parquet/src/metadata.rs

Lines changed: 4 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -39,7 +39,9 @@ use object_store::path::Path;
3939
use object_store::{ObjectMeta, ObjectStore};
4040
use parquet::arrow::arrow_reader::statistics::StatisticsConverter;
4141
use parquet::arrow::parquet_to_arrow_schema;
42-
use parquet::file::metadata::{ParquetMetaData, ParquetMetaDataReader, RowGroupMetaData};
42+
use parquet::file::metadata::{
43+
PageIndexPolicy, ParquetMetaData, ParquetMetaDataReader, RowGroupMetaData,
44+
};
4345
use std::any::Any;
4446
use std::collections::HashMap;
4547
use std::sync::Arc;
@@ -148,7 +150,7 @@ impl<'a> DFParquetMetadata<'a> {
148150

149151
if cache_metadata && file_metadata_cache.is_some() {
150152
// Need to retrieve the entire metadata for the caching to be effective.
151-
reader = reader.with_page_indexes(true);
153+
reader = reader.with_page_index_policy(PageIndexPolicy::Required);
152154
}
153155

154156
let metadata = Arc::new(

datafusion/datasource-parquet/src/opener.rs

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -54,7 +54,7 @@ use log::debug;
5454
use parquet::arrow::arrow_reader::{ArrowReaderMetadata, ArrowReaderOptions};
5555
use parquet::arrow::async_reader::AsyncFileReader;
5656
use parquet::arrow::{ParquetRecordBatchStreamBuilder, ProjectionMask};
57-
use parquet::file::metadata::ParquetMetaDataReader;
57+
use parquet::file::metadata::{PageIndexPolicy, ParquetMetaDataReader};
5858

5959
/// Implements [`FileOpener`] for a parquet file
6060
pub(super) struct ParquetOpener {
@@ -652,8 +652,8 @@ async fn load_page_index<T: AsyncFileReader>(
652652
if missing_column_index || missing_offset_index {
653653
let m = Arc::try_unwrap(Arc::clone(parquet_metadata))
654654
.unwrap_or_else(|e| e.as_ref().clone());
655-
let mut reader =
656-
ParquetMetaDataReader::new_with_metadata(m).with_page_indexes(true);
655+
let mut reader = ParquetMetaDataReader::new_with_metadata(m)
656+
.with_page_index_policy(PageIndexPolicy::Required);
657657
reader.load_page_index(input).await?;
658658
let new_parquet_metadata = reader.finish()?;
659659
let new_arrow_reader =

datafusion/physical-plan/src/spill/mod.rs

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -752,7 +752,7 @@ mod tests {
752752
.unwrap();
753753

754754
let size = get_record_batch_memory_size(&batch);
755-
assert_eq!(size, 8320);
755+
assert_eq!(size, 8208);
756756
}
757757

758758
// ==== Spill manager tests ====

0 commit comments

Comments
 (0)