Experimental parquet decoder with first-class selection pushdown support

parquet/Cargo.toml

@@ -69,6 +69,7 @@ paste = { version = "1.0" }
 half = { version = "2.1", default-features = false, features = ["num-traits"] }
 sysinfo = { version = "0.33.0", optional = true, default-features = false, features = ["system"] }
 crc32fast = { version = "1.4.2", optional = true, default-features = false }
+simdutf8 = "0.1.5"
 
 [dev-dependencies]
 base64 = { version = "0.22", default-features = false, features = ["std"] }

parquet/src/arrow/array_reader/byte_view_array.rs

@@ -463,6 +463,8 @@ impl ByteViewArrayDecoderDictionary {
             }
         }
 
+        output.views.reserve(len);
+
         // Calculate the offset of the dictionary buffers in the output buffers
         // For example if the 2nd buffer in the dictionary is the 5th buffer in the output buffers,
         // then the base_buffer_idx is 5 - 2 = 3
@@ -683,7 +685,7 @@ impl ByteViewArrayDecoderDelta {
 
 /// Check that `val` is a valid UTF-8 sequence
 pub fn check_valid_utf8(val: &[u8]) -> Result<()> {
-    match std::str::from_utf8(val) {
+    match simdutf8::basic::from_utf8(val) {
         Ok(_) => Ok(()),
         Err(e) => Err(general_err!("encountered non UTF-8 data: {}", e)),
     }