pub fn read_columns_indexes<R: Read + Seek>(
reader: &mut R,
chunks: &[ColumnChunkMetaData],
fields: &[Field]
) -> Result<Vec<FieldPageStatistics>, Error>
Expand description
Reads the column indexes from the reader assuming a valid set of derived Arrow fields for all parquet the columns in the file.
It returns one FieldPageStatistics
per field in fields
This function is expected to be used to filter out parquet pages.
§Implementation
This function is IO-bounded and calls reader.read_exact
exactly once.
§Error
Errors iff the indexes can’t be read or their deserialization to arrow is incorrect (e.g. invalid utf-8)