ZENOTME commented on code in PR #383:
URL: https://github.com/apache/iceberg-rust/pull/383#discussion_r1650086826
##########
crates/iceberg/src/writer/file_writer/parquet_writer.rs:
##########
@@ -78,106 +89,403 @@ impl<T: LocationGenerator, F: FileNameGenerator>
FileWriterBuilder for ParquetWr
type R = ParquetWriter;
async fn build(self) -> crate::Result<Self::R> {
- // Fetch field id from schema
- let field_ids = self
- .schema
- .fields()
- .iter()
- .map(|field| {
- field
- .metadata()
- .get(PARQUET_FIELD_ID_META_KEY)
- .ok_or_else(|| {
- Error::new(
- crate::ErrorKind::Unexpected,
- "Field id not found in arrow schema metadata.",
- )
- })?
- .parse::<i32>()
- .map_err(|err| {
- Error::new(crate::ErrorKind::Unexpected, "Failed to
parse field id.")
- .with_source(err)
- })
- })
- .collect::<crate::Result<Vec<_>>>()?;
-
+ let arrow_schema: ArrowSchemaRef =
Arc::new(self.schema.as_ref().try_into()?);
let written_size = Arc::new(AtomicI64::new(0));
let out_file = self.file_io.new_output(
self.location_generator
.generate_location(&self.file_name_generator.generate_file_name()),
)?;
let inner_writer = TrackWriter::new(out_file.writer().await?,
written_size.clone());
let async_writer = AsyncFileWriter::new(inner_writer);
- let writer = AsyncArrowWriter::try_new(async_writer,
self.schema.clone(), Some(self.props))
- .map_err(|err| {
- Error::new(
- crate::ErrorKind::Unexpected,
- "Failed to build parquet writer.",
- )
- .with_source(err)
- })?;
+ let writer =
+ AsyncArrowWriter::try_new(async_writer, arrow_schema.clone(),
Some(self.props))
+ .map_err(|err| {
+ Error::new(ErrorKind::Unexpected, "Failed to build parquet
writer.")
+ .with_source(err)
+ })?;
Ok(ParquetWriter {
+ schema: self.schema.clone(),
writer,
written_size,
current_row_num: 0,
out_file,
- field_ids,
})
}
}
+#[derive(Default)]
+struct IndexByParquetPathName {
Review Comment:
Oh I see.
> I think we can use following path: parquet schema -> arrow schema ->
iceberg schema, and lookup by path by field name?
But in here I think we still need to search field id using path_in_schema
for a column.🤔
```
for column_chunk in row_group.columns.iter() {
let field_id = get_field_id(column_chunk_metadata.path_in_schema)
}
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For additional commands, e-mail: issues-h...@iceberg.apache.org
