Function arrow2::io::parquet::read::get_column_iterator

pub fn get_column_iterator<R>(
    reader: R,
    metadata: &FileMetaData,
    row_group: usize,
    field: usize,
    page_filter: Option<Arc<dyn Fn(&Descriptor, &DataPageHeader) + Sync + Send + 'static>>,
    scratch: Vec<u8, Global>,
    max_header_size: usize
) -> ColumnIterator<R> where
    R: Read + Seek, 

Available on crate feature io_parquet only.

Returns a [ColumnIterator] of column chunks corresponding to field.

Contrarily to [get_page_iterator] that returns a single iterator of pages, this iterator returns multiple iterators, one per physical column of the field. For primitive fields (e.g. i64), [ColumnIterator] yields exactly one column. For complex fields, it yields multiple columns. max_header_size is the maximum number of bytes thrift is allowed to allocate to read a page header.