feat: Implement De/Serialize on Blob

This commit is contained in:
Clément Renault 2018-11-23 18:00:24 +01:00
parent a43a772e9a
commit 0e856db4e6
No known key found for this signature in database
GPG key ID: 0151CDAB43460DAE
7 changed files with 196 additions and 9 deletions

View file

@ -7,31 +7,32 @@ use std::{io, mem};
use byteorder::{NativeEndian, WriteBytesExt};
use fst::raw::MmapReadOnly;
use serde::ser::{Serialize, Serializer};
use crate::DocumentId;
use crate::data::Data;
#[derive(Clone)]
pub struct DocIds {
doc_ids: Data,
data: Data,
}
impl DocIds {
pub unsafe fn from_path<P: AsRef<Path>>(path: P) -> io::Result<Self> {
let mmap = MmapReadOnly::open_path(path)?;
let doc_ids = Data::Mmap(mmap);
Ok(DocIds { doc_ids })
let data = Data::Mmap(mmap);
Ok(DocIds { data })
}
pub fn from_bytes(vec: Vec<u8>) -> Result<Self, Box<Error>> {
// FIXME check if modulo DocumentId
let len = vec.len();
let doc_ids = Data::Shared {
let data = Data::Shared {
vec: Arc::new(vec),
offset: 0,
len: len
};
Ok(DocIds { doc_ids })
Ok(DocIds { data })
}
pub fn contains(&self, doc: DocumentId) -> bool {
@ -40,13 +41,19 @@ impl DocIds {
}
pub fn doc_ids(&self) -> &[DocumentId] {
let slice = &self.doc_ids;
let slice = &self.data;
let ptr = slice.as_ptr() as *const DocumentId;
let len = slice.len() / mem::size_of::<DocumentId>();
unsafe { from_raw_parts(ptr, len) }
}
}
impl Serialize for DocIds {
fn serialize<S: Serializer>(&self, serializer: S) -> Result<S::Ok, S::Error> {
self.data.as_ref().serialize(serializer)
}
}
pub struct DocIdsBuilder<W> {
doc_ids: BTreeSet<DocumentId>, // TODO: prefer a linked-list
wrt: W,

View file

@ -8,6 +8,7 @@ use std::mem;
use fst::raw::MmapReadOnly;
use byteorder::{LittleEndian, ReadBytesExt, WriteBytesExt};
use serde::ser::{Serialize, Serializer, SerializeTuple};
use crate::DocIndex;
use crate::data::Data;
@ -84,6 +85,15 @@ impl DocIndexes {
}
}
impl Serialize for DocIndexes {
fn serialize<S: Serializer>(&self, serializer: S) -> Result<S::Ok, S::Error> {
let mut tuple = serializer.serialize_tuple(2)?;
tuple.serialize_element(self.ranges.as_ref())?;
tuple.serialize_element(self.indexes.as_ref())?;
tuple.end()
}
}
pub struct DocIndexesBuilder<W> {
keys: BTreeMap<String, u64>,
indexes: Vec<Vec<DocIndex>>,

View file

@ -23,6 +23,12 @@ impl Deref for Data {
type Target = [u8];
fn deref(&self) -> &Self::Target {
self.as_ref()
}
}
impl AsRef<[u8]> for Data {
fn as_ref(&self) -> &[u8] {
match self {
Data::Shared { vec, offset, len } => {
&vec[*offset..offset + len]