Compared to the original Lucene format, this skip list is only storing the document identifiers and the file pointer of the block within the .doc file.
The {@link MultiLevelSkipListWriter} implementation is based on documentcount, but it is used here with block count instead of document count. In order to make it compatible with block, this class is converting document count into block count.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|