Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat: support serialize/deserialize DataFile into avro bytes #797

Merged
merged 3 commits into from
Jan 2, 2025
+179 −49
Merged

feat: support serialize/deserialize DataFile into avro bytes #797

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
b94c978
add interface to help serialize/deserialize DataFile
Dec 18, 2024
9cdd762
refine interface and add test
Dec 31, 2024
92871af
rename interface
Jan 2, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
149 changes: 106 additions & 43 deletions crates/iceberg/src/spec/manifest.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -656,6 +656,38 @@ mod _const_schema {
})
};

fn data_file_fields_v2(partition_type: StructType) -> Vec<NestedFieldRef> {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
fn data_file_fields_v2(partition_type: StructType) -> Vec<NestedFieldRef> {
fn data_file_fields_v2(partition_type: &StructType) -> Vec<NestedFieldRef> {

vec![
CONTENT.clone(),
FILE_PATH.clone(),
FILE_FORMAT.clone(),
Arc::new(NestedField::required(
102,
"partition",
Type::Struct(partition_type),
)),
RECORD_COUNT.clone(),
FILE_SIZE_IN_BYTES.clone(),
COLUMN_SIZES.clone(),
VALUE_COUNTS.clone(),
NULL_VALUE_COUNTS.clone(),
NAN_VALUE_COUNTS.clone(),
LOWER_BOUNDS.clone(),
UPPER_BOUNDS.clone(),
KEY_METADATA.clone(),
SPLIT_OFFSETS.clone(),
EQUALITY_IDS.clone(),
SORT_ORDER_ID.clone(),
]
}

pub(super) fn data_file_schema_v2(partition_type: StructType) -> Result<AvroSchema, Error> {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
pub(super) fn data_file_schema_v2(partition_type: StructType) -> Result<AvroSchema, Error> {
pub(super) fn data_file_schema_v2(partition_type: &StructType) -> Result<AvroSchema, Error> {

let schema = Schema::builder()
.with_fields(data_file_fields_v2(partition_type))
.build()?;
schema_to_avro_schema("data_file", &schema)
}

pub(super) fn manifest_schema_v2(partition_type: StructType) -> Result<AvroSchema, Error> {
let fields = vec![
STATUS.clone(),
Expand All @@ -665,62 +697,52 @@ mod _const_schema {
Arc::new(NestedField::required(
2,
"data_file",
Type::Struct(StructType::new(vec![
CONTENT.clone(),
FILE_PATH.clone(),
FILE_FORMAT.clone(),
Arc::new(NestedField::required(
102,
"partition",
Type::Struct(partition_type),
)),
RECORD_COUNT.clone(),
FILE_SIZE_IN_BYTES.clone(),
COLUMN_SIZES.clone(),
VALUE_COUNTS.clone(),
NULL_VALUE_COUNTS.clone(),
NAN_VALUE_COUNTS.clone(),
LOWER_BOUNDS.clone(),
UPPER_BOUNDS.clone(),
KEY_METADATA.clone(),
SPLIT_OFFSETS.clone(),
EQUALITY_IDS.clone(),
SORT_ORDER_ID.clone(),
])),
Type::Struct(StructType::new(data_file_fields_v2(partition_type))),
)),
];
let schema = Schema::builder().with_fields(fields).build()?;
schema_to_avro_schema("manifest_entry", &schema)
}

fn data_file_fields_v1(partition_type: StructType) -> Vec<NestedFieldRef> {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
fn data_file_fields_v1(partition_type: StructType) -> Vec<NestedFieldRef> {
fn data_file_fields_v1(partition_type: &StructType) -> Vec<NestedFieldRef> {

vec![
FILE_PATH.clone(),
FILE_FORMAT.clone(),
Arc::new(NestedField::required(
102,
"partition",
Type::Struct(partition_type),
)),
RECORD_COUNT.clone(),
FILE_SIZE_IN_BYTES.clone(),
BLOCK_SIZE_IN_BYTES.clone(),
COLUMN_SIZES.clone(),
VALUE_COUNTS.clone(),
NULL_VALUE_COUNTS.clone(),
NAN_VALUE_COUNTS.clone(),
LOWER_BOUNDS.clone(),
UPPER_BOUNDS.clone(),
KEY_METADATA.clone(),
SPLIT_OFFSETS.clone(),
SORT_ORDER_ID.clone(),
]
}

pub(super) fn data_file_schema_v1(partition_type: StructType) -> Result<AvroSchema, Error> {
let schema = Schema::builder()
.with_fields(data_file_fields_v1(partition_type))
.build()?;
schema_to_avro_schema("data_file", &schema)
}

pub(super) fn manifest_schema_v1(partition_type: StructType) -> Result<AvroSchema, Error> {
let fields = vec![
STATUS.clone(),
SNAPSHOT_ID_V1.clone(),
Arc::new(NestedField::required(
2,
"data_file",
Type::Struct(StructType::new(vec![
FILE_PATH.clone(),
FILE_FORMAT.clone(),
Arc::new(NestedField::required(
102,
"partition",
Type::Struct(partition_type),
)),
RECORD_COUNT.clone(),
FILE_SIZE_IN_BYTES.clone(),
BLOCK_SIZE_IN_BYTES.clone(),
COLUMN_SIZES.clone(),
VALUE_COUNTS.clone(),
NULL_VALUE_COUNTS.clone(),
NAN_VALUE_COUNTS.clone(),
LOWER_BOUNDS.clone(),
UPPER_BOUNDS.clone(),
KEY_METADATA.clone(),
SPLIT_OFFSETS.clone(),
SORT_ORDER_ID.clone(),
])),
Type::Struct(StructType::new(data_file_fields_v1(partition_type))),
)),
];
let schema = Schema::builder().with_fields(fields).build()?;
Expand Down Expand Up @@ -1189,6 +1211,47 @@ impl DataFile {
self.sort_order_id
}
}

/// Convert data files to avro bytes.
Xuanwo marked this conversation as resolved.
Show resolved Hide resolved
pub fn data_files_to_avro(
Xuanwo marked this conversation as resolved.
Show resolved Hide resolved
data_files: impl IntoIterator<Item = DataFile>,
partition_type: &StructType,
version: FormatVersion,
) -> Result<Vec<u8>> {
let avro_schema = match version {
FormatVersion::V1 => _const_schema::data_file_schema_v1(partition_type.clone()).unwrap(),
FormatVersion::V2 => _const_schema::data_file_schema_v2(partition_type.clone()).unwrap(),
};
let mut writer = AvroWriter::new(&avro_schema, Vec::new());

for data_file in data_files {
let value = to_value(_serde::DataFile::try_from(data_file, partition_type, true)?)?
.resolve(&avro_schema)?;
writer.append(value)?;
}

Ok(writer.into_inner()?)
}

/// Parse data files from avro bytes.
pub fn parse_from_avro(
Xuanwo marked this conversation as resolved.
Show resolved Hide resolved
bytes: &[u8],
schema: &Schema,
partition_type: &StructType,
version: FormatVersion,
) -> Result<Vec<DataFile>> {
let avro_schema = match version {
FormatVersion::V1 => _const_schema::data_file_schema_v1(partition_type.clone()).unwrap(),
FormatVersion::V2 => _const_schema::data_file_schema_v2(partition_type.clone()).unwrap(),
};

let reader = AvroReader::with_schema(&avro_schema, bytes)?;
reader
.into_iter()
.map(|value| from_value::<_serde::DataFile>(&value?)?.try_into(partition_type, schema))
.collect::<Result<Vec<_>>>()
}

/// Type of content stored by the data file: data, equality deletes, or
/// position deletes (all v1 files are data files)
#[derive(Debug, PartialEq, Eq, Clone, Copy, Serialize, Deserialize)]
Expand Down
Loading