rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2026-05-18 09:10:41 +00:00
Code Issues Packages Projects Releases Wiki Activity

skip sstable index binary search when ordinal is in same block

Browse Source
This commit is contained in:
trinity-1686a
2026-05-15 14:11:06 +02:00
committed by trinity-1686a
parent c367df37c1
commit ea61a68db4
4 changed files with 40 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
sstable/src/dictionary.rs
View File

@@ -496,10 +496,15 @@ impl<TSSTable: SSTable> Dictionary<TSSTable> {
// Open the block for the first ordinal.
let mut bytes = Vec::new();
let mut current_block_addr = self.sstable_index.get_block_with_ord(ord);
let (mut current_block_addr, block_id) = self.sstable_index.get_and_locate_with_ord(ord);
let mut current_sstable_delta_reader =
self.sstable_delta_reader_block(current_block_addr.clone())?;
let mut current_block_ordinal = current_block_addr.first_ordinal;
let mut current_block_bound = self
.sstable_index
.get_block(block_id + 1)
.map(|block_addr| block_addr.first_ordinal)
.unwrap_or(u64::MAX);
loop {
// move to the ord inside the current block
@@ -528,17 +533,19 @@ impl<TSSTable: SSTable> Dictionary<TSSTable> {
}
};
// TODO optimization: it is silly to do a binary search to get the block every single
// time.
//
// Check if block changed for new term_ord
let new_block_addr = self.sstable_index.get_block_with_ord(next_ord);
if new_block_addr != current_block_addr {
if next_ord >= current_block_bound {
let (new_block_addr, block_id) =
self.sstable_index.get_and_locate_with_ord(next_ord);
current_block_addr = new_block_addr;
current_block_ordinal = current_block_addr.first_ordinal;
current_sstable_delta_reader =
self.sstable_delta_reader_block(current_block_addr.clone())?;
bytes.clear();
current_block_bound = self
.sstable_index
.get_block(block_id + 1)
.map(|block_addr| block_addr.first_ordinal)
.unwrap_or(u64::MAX)
}
ord = next_ord;
}

8
sstable/src/index/mod.rs
View File

@@ -98,6 +98,14 @@ impl SSTableIndex {
}
}
pub(crate) fn get_and_locate_with_ord(&self, ord: TermOrdinal) -> (BlockAddr, u64) {
match self {
SSTableIndex::V2(v2_index) => v2_index.get_and_locate_with_ord(ord),
SSTableIndex::V3(v3_index) => v3_index.get_and_locate_with_ord(ord),
SSTableIndex::V3Empty(v3_empty) => v3_empty.get_and_locate_with_ord(ord),
}
}
pub fn get_block_for_automaton<'a>(
&'a self,
automaton: &'a impl Automaton,

7
sstable/src/index/v2.rs
View File

@@ -77,6 +77,13 @@ impl SSTableIndex {
self.get_block(self.locate_with_ord(ord)).unwrap()
}
pub(crate) fn get_and_locate_with_ord(&self, ord: TermOrdinal) -> (BlockAddr, u64) {
let location = self.locate_with_ord(ord);
// locate_with_ord always returns an index within range
let block_addr = self.get_block(location).unwrap();
(block_addr, location as u64)
}
pub(crate) fn get_block_for_automaton<'a>(
&'a self,
automaton: &'a impl Automaton,

13
sstable/src/index/v3.rs
View File

@@ -68,6 +68,11 @@ impl SSTableIndexV3 {
self.block_addr_store.binary_search_ord(ord).1
}
pub(crate) fn get_and_locate_with_ord(&self, ord: TermOrdinal) -> (BlockAddr, u64) {
let (location, block_addr) = self.block_addr_store.binary_search_ord(ord);
(block_addr, location)
}
pub(crate) fn get_block_for_automaton<'a>(
&'a self,
automaton: &'a impl Automaton,
@@ -138,8 +143,8 @@ impl SSTableIndexV3Empty {
}
/// Get the [`BlockAddr`] of the requested block.
pub(crate) fn get_block(&self, _block_id: u64) -> Option<BlockAddr> {
Some(self.block_addr.clone())
pub(crate) fn get_block(&self, block_id: u64) -> Option<BlockAddr> {
(block_id == 0).then(|| self.block_addr.clone())
}
/// Get the block id of the block that would contain `key`.
@@ -164,6 +169,10 @@ impl SSTableIndexV3Empty {
pub(crate) fn get_block_with_ord(&self, _ord: TermOrdinal) -> BlockAddr {
self.block_addr.clone()
}
pub(crate) fn get_and_locate_with_ord(&self, _ord: TermOrdinal) -> (BlockAddr, u64) {
(self.block_addr.clone(), 0)
}
}
const STORE_BLOCK_LEN: usize = 128;