feat: Allow datafusion-ffi to opt out of proto parquet

.github/workflows/rust.yml

@@ -161,6 +161,26 @@ jobs:
       - name: Check datafusion-proto (avro)
         run: cargo check --profile ci --no-default-features -p datafusion-proto --features=avro
 
+  # Check datafusion-ffi features
+  #
+  # Ensure via `cargo check` that the crate can be built with a
+  # subset of the features packages enabled.
+  linux-datafusion-ffi-features:
+    name: cargo check datafusion-ffi features
+    needs: linux-build-lib
+    runs-on: ${{ vars.USE_RUNS_ON == 'true' && format('runs-on={0},family=m8a+m7a+c8a,cpu=16,image=ubuntu24-full-x64,extras=s3-cache,disk=large,tag=datafusion', github.run_id) || 'ubuntu-latest' }}
+    container:
+      image: amd64/rust
+    steps:
+      - uses: runs-on/action@d141ef83eb66d096ce8afc767e09115a65c63b60  # v2.1.2
+      - uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10  # v6.0.3
+      - name: Setup Rust toolchain
+        uses: ./.github/actions/setup-builder
+        with:
+          rust-version: stable
+      - name: Check datafusion-ffi (no-default-features)
+        run: cargo check --profile ci --no-default-features -p datafusion-ffi
+
 
   # Check datafusion crate features
   #

Cargo.toml

@@ -151,7 +151,7 @@ datafusion-physical-expr-adapter = { path = "datafusion/physical-expr-adapter",
 datafusion-physical-expr-common = { path = "datafusion/physical-expr-common", version = "54.0.0", default-features = false }
 datafusion-physical-optimizer = { path = "datafusion/physical-optimizer", version = "54.0.0" }
 datafusion-physical-plan = { path = "datafusion/physical-plan", version = "54.0.0" }
-datafusion-proto = { path = "datafusion/proto", version = "54.0.0" }
+datafusion-proto = { path = "datafusion/proto", version = "54.0.0", default-features = false }
 datafusion-proto-common = { path = "datafusion/proto-common", version = "54.0.0" }
 datafusion-proto-models = { path = "datafusion/proto-models", version = "54.0.0" }
 datafusion-pruning = { path = "datafusion/pruning", version = "54.0.0" }

benchmarks/Cargo.toml

@@ -63,7 +63,7 @@ tokio = { workspace = true, features = ["rt-multi-thread", "parking_lot"] }
 tokio-util = { version = "0.7.17" }
 
 [dev-dependencies]
-datafusion-proto = { workspace = true }
+datafusion-proto = { workspace = true, features = ["parquet"] }
 tempfile = { workspace = true }
 
 [[bench]]

datafusion-examples/Cargo.toml

@@ -53,7 +53,7 @@ dashmap = { workspace = true }
 base64 = "0.22.1"
 datafusion-expr = { workspace = true }
 datafusion-physical-expr-adapter = { workspace = true }
-datafusion-proto = { workspace = true }
+datafusion-proto = { workspace = true, features = ["parquet"] }
 datafusion-sql = { workspace = true }
 env_logger = { workspace = true }
 futures = { workspace = true }

datafusion/ffi/Cargo.toml

@@ -63,7 +63,7 @@ datafusion-physical-expr = { workspace = true }
 datafusion-physical-expr-common = { workspace = true }
 datafusion-physical-optimizer = { workspace = true }
 datafusion-physical-plan = { workspace = true }
-datafusion-proto = { workspace = true }
+datafusion-proto = { workspace = true, default-features = false }
 datafusion-proto-common = { workspace = true }
 datafusion-session = { workspace = true }
 futures = { workspace = true }
@@ -83,10 +83,12 @@ datafusion-functions-window = { workspace = true }
 doc-comment = { workspace = true }
 
 [features]
+default = ["parquet"]
 integration-tests = [
     "datafusion-functions",
     "datafusion-functions-aggregate",
     "datafusion-functions-table",
     "datafusion-functions-window",
 ]
+parquet = ["datafusion-proto/parquet"]
 tarpaulin_include = [] # Exists only to prevent warnings on stable and still have accurate coverage

datafusion/ffi/src/session/mod.rs

@@ -258,6 +258,7 @@ fn table_options_to_rhash(mut options: TableOptions) -> SVec<(SString, SString)>
             "datafusion_ffi.table_current_format".into(),
             match current_format {
                 ConfigFileType::JSON => "json",
+                #[cfg(feature = "parquet")]
                 ConfigFileType::PARQUET => "parquet",
                 ConfigFileType::CSV => "csv",
             }
@@ -476,13 +477,15 @@ fn table_options_from_rhashmap(options: SVec<(SString, SString)>) -> TableOption
     let formats = [
         ConfigFileType::CSV,
         ConfigFileType::JSON,
+        #[cfg(feature = "parquet")]
         ConfigFileType::PARQUET,
     ];
     for format in formats {
         // It is imperative that if new enum variants are added below that they be
         // included in the formats list above and in the extension check below.
         let format_name = match &format {
             ConfigFileType::CSV => "csv",
+            #[cfg(feature = "parquet")]
             ConfigFileType::PARQUET => "parquet",
             ConfigFileType::JSON => "json",
         };
@@ -504,7 +507,6 @@ fn table_options_from_rhashmap(options: SVec<(SString, SString)>) -> TableOption
                 .unwrap_or_else(|err| log::warn!("Error parsing table options: {err}"));
         }
     }
-
     let extension_options: HashMap<String, String> = options
         .iter()
         .filter_map(|(k, v)| {
@@ -525,6 +527,7 @@ fn table_options_from_rhashmap(options: SVec<(SString, SString)>) -> TableOption
     table_options.current_format =
         current_format.and_then(|format| match format.as_str() {
             "csv" => Some(ConfigFileType::CSV),
+            #[cfg(feature = "parquet")]
             "parquet" => Some(ConfigFileType::PARQUET),
             "json" => Some(ConfigFileType::JSON),
             _ => None,
@@ -661,7 +664,10 @@ mod tests {
         let mut table_options = TableOptions::default();
         table_options.csv.has_header = Some(true);
         table_options.json.schema_infer_max_rec = Some(10);
-        table_options.parquet.global.coerce_int96 = Some("123456789".into());
+        #[cfg(feature = "parquet")]
+        {
+            table_options.parquet.global.coerce_int96 = Some("123456789".into());
+        }
         table_options.current_format = Some(ConfigFileType::JSON);
 
         let state = SessionStateBuilder::new_from_existing(ctx.state())

datafusion/proto/src/logical_plan/file_formats.rs

@@ -18,7 +18,9 @@
 use std::sync::Arc;
 
 use super::LogicalExtensionCodec;
-use crate::convert::{FromProto, TryFromProto};
+use crate::convert::FromProto;
+#[cfg(feature = "parquet")]
+use crate::convert::TryFromProto;
 use crate::protobuf::{
     CsvOptions as CsvOptionsProto, CsvQuoteStyle as CsvQuoteStyleProto,
     JsonOptions as JsonOptionsProto,