feat: more Results, fewer panics, always have backtraces

Cargo.toml

@@ -131,7 +131,7 @@ uuid = "1.8.0"
 vortex-alp = { version = "0.8.0", path = "./encodings/alp" }
 vortex-array = { version = "0.8.0", path = "./vortex-array" }
 vortex-buffer = { version = "0.8.0", path = "./vortex-buffer" }
-vortex-bytebool = { version = "0.8.0", path = "./encodings/byte-bool" }
+vortex-bytebool = { version = "0.8.0", path = "./encodings/bytebool" }
 vortex-datafusion = { version = "0.8.0", path = "./vortex-datafusion" }
 vortex-datetime-dtype = { version = "0.8.0", path = "./vortex-datetime-dtype" }
 vortex-datetime-parts = { version = "0.8.0", path = "./encodings/datetime-parts" }
@@ -171,11 +171,31 @@ unsafe_op_in_unsafe_fn = "deny"
 
 [workspace.lints.clippy]
 all = { level = "deny", priority = -1 }
+as_ptr_cast_mut = { level = "deny" }
+borrow_as_ptr = { level = "deny" }
+collection_is_never_read = { level = "deny" }
+cognitive_complexity = { level = "deny" }
+debug_assert_with_mut_call = { level = "deny" }
+derive_partial_eq_without_eq = { level = "deny" }
+exit = { level = "deny" }
+expect_fun_call = { level = "deny" }
+expect_used = { level = "deny" }
+equatable_if_let = { level = "deny" }
+fallible_impl_from = { level = "deny" }
+get_unwrap = { level = "deny" }
+host_endian_bytes = { level = "deny" }
 if_then_some_else_none = { level = "deny" }
+inconsistent_struct_constructor = { level = "deny" }
+manual_assert = { level = "deny" }
+manual_is_variant_and = { level = "deny" }
 mem_forget = { level = "deny" }
 or_fun_call = "deny"
 panic_in_result_fn = { level = "deny" }
+panic = { level = "deny" }
 same_name_method = { level = "deny" }
 tests_outside_test_module = { level = "deny" }
+# todo = { level = "deny" }
+# unimplemented = { level = "deny" }
 unwrap_in_result = { level = "deny" }
+unwrap_used = { level = "deny" }
 use_debug = { level = "deny" }

bench-vortex/Cargo.toml

@@ -12,8 +12,13 @@ include = { workspace = true }
 edition = { workspace = true }
 rust-version = { workspace = true }
 
-[lints]
-workspace = true
+[lints.rust]
+warnings = "deny"
+unsafe_op_in_unsafe_fn = "deny"
+
+[lints.clippy]
+all = { level = "deny", priority = -1 }
+or_fun_call = "deny"
 
 [dependencies]
 anyhow = { workspace = true }

bench-vortex/benches/datafusion_benchmark.rs

@@ -1,5 +1,3 @@
-#![allow(clippy::use_debug)]
-
 use std::collections::HashSet;
 use std::sync::Arc;
 
@@ -83,7 +81,7 @@ fn toy_dataset_arrow() -> RecordBatch {
 }
 
 fn toy_dataset_vortex(compress: bool) -> Array {
-    let uncompressed = toy_dataset_arrow().into();
+    let uncompressed = toy_dataset_arrow().try_into().unwrap();
 
     if !compress {
         return uncompressed;

bench-vortex/benches/tpch_benchmark.rs

@@ -1,5 +1,3 @@
-#![allow(clippy::use_debug)]
-
 use bench_vortex::tpch::dbgen::{DBGen, DBGenOptions};
 use bench_vortex::tpch::{load_datasets, run_tpch_query, tpch_queries, Format};
 use criterion::{criterion_group, criterion_main, Criterion};

bench-vortex/src/bin/tpch_benchmark.rs

@@ -1,5 +1,3 @@
-#![allow(clippy::use_debug)]
-
 use std::collections::HashMap;
 use std::process::ExitCode;
 use std::sync;

bench-vortex/src/data_downloads.rs

@@ -45,7 +45,7 @@ pub fn data_vortex_uncompressed(fname_out: &str, downloaded_data: PathBuf) -> Pa
         let array = ChunkedArray::try_new(
             reader
                 .into_iter()
-                .map(|batch_result| Array::from(batch_result.unwrap()))
+                .map(|batch_result| Array::try_from(batch_result.unwrap()).unwrap())
                 .collect(),
             dtype,
         )

bench-vortex/src/lib.rs

@@ -190,7 +190,8 @@ pub fn compress_taxi_data() -> Array {
     let chunks = reader
         .into_iter()
         .map(|batch_result| batch_result.unwrap())
-        .map(Array::from)
+        .map(Array::try_from)
+        .map(Result::unwrap)
         .map(|array| {
             uncompressed_size += array.nbytes();
             compressor.compress(&array).unwrap()
@@ -288,7 +289,7 @@ mod test {
             let struct_arrow: ArrowStructArray = record_batch.into();
             let arrow_array: ArrowArrayRef = Arc::new(struct_arrow);
             let vortex_array = Array::from_arrow(arrow_array.clone(), false);
-            let vortex_as_arrow = vortex_array.into_canonical().unwrap().into_arrow();
+            let vortex_as_arrow = vortex_array.into_canonical().unwrap().into_arrow().unwrap();
             assert_eq!(vortex_as_arrow.deref(), arrow_array.deref());
         }
     }
@@ -309,7 +310,7 @@ mod test {
             let vortex_array = Array::from_arrow(arrow_array.clone(), false);
 
             let compressed = compressor.compress(&vortex_array).unwrap();
-            let compressed_as_arrow = compressed.into_canonical().unwrap().into_arrow();
+            let compressed_as_arrow = compressed.into_canonical().unwrap().into_arrow().unwrap();
             assert_eq!(compressed_as_arrow.deref(), arrow_array.deref());
         }
     }

bench-vortex/src/reader.rs

@@ -20,6 +20,7 @@ use object_store::ObjectStore;
 use parquet::arrow::arrow_reader::{ArrowReaderOptions, ParquetRecordBatchReaderBuilder};
 use parquet::arrow::async_reader::{AsyncFileReader, ParquetObjectReader};
 use parquet::arrow::ParquetRecordBatchStreamBuilder;
+use parquet::file::metadata::RowGroupMetaData;
 use serde::{Deserialize, Serialize};
 use stream::StreamExt;
 use vortex::array::{ChunkedArray, PrimitiveArray};
@@ -58,7 +59,7 @@ pub async fn open_vortex(path: &Path) -> VortexResult<Array> {
         .into_array_stream()
         .collect_chunked()
         .await
-        .map(|a| a.into_array())
+        .map(IntoArray::into_array)
 }
 
 pub async fn rewrite_parquet_as_vortex<W: VortexWrite>(
@@ -101,7 +102,7 @@ pub fn compress_parquet_to_vortex(parquet_path: &Path) -> VortexResult<ChunkedAr
     let chunks = reader
         .map(|batch_result| batch_result.unwrap())
         .map(|record_batch| {
-            let vortex_array = Array::from(record_batch);
+            let vortex_array = Array::try_from(record_batch).unwrap();
             compressor.compress(&vortex_array).unwrap()
         })
         .collect_vec();
@@ -219,7 +220,7 @@ async fn parquet_take_from_stream<T: AsyncFileReader + Unpin + Send + 'static>(
             .metadata()
             .row_groups()
             .iter()
-            .map(|rg| rg.num_rows())
+            .map(RowGroupMetaData::num_rows)
             .scan(0i64, |acc, x| {
                 *acc += x;
                 Some(*acc)
@@ -238,7 +239,7 @@ async fn parquet_take_from_stream<T: AsyncFileReader + Unpin + Send + 'static>(
     let row_group_indices = row_groups
         .keys()
         .sorted()
-        .map(|i| row_groups.get(i).unwrap().clone())
+        .map(|i| row_groups[i].clone())
         .collect_vec();
 
     let reader = builder

bench-vortex/src/tpch/dbgen.rs

@@ -175,7 +175,7 @@ fn get_or_cache_toolchain(
         zip_file
             .url()
             .path_segments()
-            .and_then(|segments| segments.last())
+            .and_then(Iterator::last)
             .unwrap(),
     );
 

bench-vortex/src/tpch/mod.rs

@@ -247,8 +247,8 @@ async fn register_vortex_file(
         let sts = record_batches
             .iter()
             .cloned()
-            .map(Array::from)
-            .map(|a| a.into_struct().unwrap())
+            .map(Array::try_from)
+            .map(|a| a.unwrap().into_struct().unwrap())
             .collect::<Vec<_>>();
 
         let mut arrays_map: HashMap<Arc<str>, Vec<Array>> = HashMap::default();
@@ -272,7 +272,7 @@ async fn register_vortex_file(
             .iter()
             .map(|field| {
                 let name: Arc<str> = field.name().as_str().into();
-                let dtype = types_map.get(&name).unwrap().clone();
+                let dtype = types_map[&name].clone();
                 let chunks = arrays_map.remove(&name).unwrap();
 
                 (

clippy.toml

@@ -1 +1,2 @@
+allow-expect-in-tests = true
 allow-unwrap-in-tests = true

encodings/alp/benches/alp_compress.rs

@@ -1,3 +1,5 @@
+#![allow(clippy::unwrap_used)]
+
 use arrow::array::{as_primitive_array, ArrowNativeTypeOp, ArrowPrimitiveType};
 use arrow::datatypes::{Float32Type, Float64Type};
 use divan::{black_box, Bencher};
@@ -49,7 +51,7 @@ where
 
 fn alp_canonicalize_sum<T: ArrowPrimitiveType>(array: ALPArray) -> T::Native {
     let array = array.into_canonical().unwrap().into_arrow();
-    let arrow_primitive = as_primitive_array::<T>(array.as_ref());
+    let arrow_primitive = as_primitive_array::<T>(array.as_ref().unwrap());
     arrow_primitive
         .iter()
         .fold(T::default_value(), |acc, value| {

encodings/alp/src/alp.rs

@@ -4,6 +4,7 @@ use std::mem::size_of;
 use itertools::Itertools;
 use num_traits::{Float, NumCast, PrimInt, Zero};
 use serde::{Deserialize, Serialize};
+use vortex_error::vortex_panic;
 
 const SAMPLE_SIZE: usize = 32;
 
@@ -20,7 +21,7 @@ impl Display for Exponents {
 }
 
 pub trait ALPFloat: Float + Display + 'static {
-    type ALPInt: PrimInt;
+    type ALPInt: PrimInt + Display;
 
     const FRACTIONAL_BITS: u8;
     const MAX_EXPONENT: u8;
@@ -118,7 +119,14 @@ pub trait ALPFloat: Float + Display + 'static {
 
     #[inline]
     fn decode_single(encoded: Self::ALPInt, exponents: Exponents) -> Self {
-        let encoded_float: Self = Self::from(encoded).unwrap();
+        let encoded_float: Self = Self::from(encoded).unwrap_or_else(|| {
+            vortex_panic!(
+                "Failed to convert encoded value {} from {} to {} in ALPFloat::decode_single",
+                encoded,
+                std::any::type_name::<Self::ALPInt>(),
+                std::any::type_name::<Self>()
+            )
+        });
         encoded_float * Self::F10[exponents.f as usize] * Self::IF10[exponents.e as usize]
     }
 }