write categorical to parquet without casting to utf8

polars/polars-arrow/Cargo.toml

@@ -9,7 +9,7 @@ description = "Arrow interfaces for Polars DataFrame library"
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 
 [dependencies]
-arrow = { package = "arrow2", git = "https://github.com/jorgecarleitao/arrow2", rev = "25ddb0d5fc2dfea80f3ac52eba5c26c294743da8", default-features = false }
+arrow = { package = "arrow2", git = "https://github.com/jorgecarleitao/arrow2", rev = "eead22fcff1f7ecad9bf48e83616a6a08f487bd4", default-features = false }
 #arrow = { package = "arrow2", git = "https://github.com/ritchie46/arrow2", branch="dev", default-features = false }
 #arrow = { package = "arrow2", version = "0.5.3", default-features=false}
 thiserror = "^1.0"

polars/polars-core/Cargo.toml

@@ -123,7 +123,7 @@ docs-selection = [
 ]
 
 [dependencies]
-arrow = { package = "arrow2", git = "https://github.com/jorgecarleitao/arrow2", rev = "25ddb0d5fc2dfea80f3ac52eba5c26c294743da8", default-features = false, features=["compute"] }
+arrow = { package = "arrow2", git = "https://github.com/jorgecarleitao/arrow2", rev = "eead22fcff1f7ecad9bf48e83616a6a08f487bd4", default-features = false, features=["compute"] }
 #arrow = { package = "arrow2", git = "https://github.com/ritchie46/arrow2", default-features = false, features=["compute"], branch="dev" }
 #arrow = { package = "arrow2", version="0.5.3", default-features = false, features=["compute"]}
 polars-arrow = {version = "0.16.0", path = "../polars-arrow"}

polars/polars-io/Cargo.toml

@@ -28,7 +28,7 @@ temporal = ["polars-core/dtype-date", "polars-core/dtype-datetime"]
 private = []
 
 [dependencies]
-arrow = { package = "arrow2", git = "https://github.com/jorgecarleitao/arrow2", rev = "25ddb0d5fc2dfea80f3ac52eba5c26c294743da8", default-features = false }
+arrow = { package = "arrow2", git = "https://github.com/jorgecarleitao/arrow2", rev = "eead22fcff1f7ecad9bf48e83616a6a08f487bd4", default-features = false }
 #arrow = { package = "arrow2", git = "https://github.com/ritchie46/arrow2", default-features = false, branch="dev"}
 #arrow = { package = "arrow2", version="0.5.3", --default-features=false }
 polars-core = {version = "0.16.0", path = "../polars-core", features = ["private"], default-features=false}

polars/polars-io/src/parquet.rs

@@ -155,21 +155,6 @@ where
 
     /// Write the given DataFrame in the the writer `W`.
     pub fn finish(mut self, df: &DataFrame) -> Result<()> {
-        // temp coerce cat to utf8 until supported in https://github.com/jorgecarleitao/parquet2/issues/57 is fixed
-
-        let columns = df
-            .get_columns()
-            .iter()
-            .map(|s| {
-                if let DataType::Categorical = s.dtype() {
-                    s.cast(&DataType::Utf8).unwrap()
-                } else {
-                    s.clone()
-                }
-            })
-            .collect();
-        let df = DataFrame::new_no_checks(columns);
-
         let mut fields = df.schema().to_arrow().fields().clone();
 
         // date64 is not supported by parquet and will be be truncated to date32