Extend DocumentDeserialize with a stateful variant DocumentDeserializ…

…eSeed

This is modelled on Serde's `DeserializeSeed` include how the relevant API entry
points gain a `_seed` variant. It can be used for example to obtain runtime
field ID values when deserializing a struct field by field without relying on
the order of the fields as written to/read from the document store.

src/core/searcher.rs

@@ -1,12 +1,13 @@
 use std::collections::BTreeMap;
+use std::marker::PhantomData;
 use std::sync::Arc;
 use std::{fmt, io};
 
 use crate::collector::Collector;
 use crate::core::Executor;
 use crate::index::{SegmentId, SegmentReader};
 use crate::query::{Bm25StatisticsProvider, EnableScoring, Query};
-use crate::schema::document::DocumentDeserialize;
+use crate::schema::document::{DocumentDeserialize, DocumentDeserializeSeed};
 use crate::schema::{Schema, Term};
 use crate::space_usage::SearcherSpaceUsage;
 use crate::store::{CacheStats, StoreReader};
@@ -86,8 +87,17 @@ impl Searcher {
     /// The searcher uses the segment ordinal to route the
     /// request to the right `Segment`.
     pub fn doc<D: DocumentDeserialize>(&self, doc_address: DocAddress) -> crate::Result<D> {
+        self.doc_seed(doc_address, PhantomData)
+    }
+
+    /// A stateful variant of [`doc`][Self::doc].`
+    pub fn doc_seed<T: DocumentDeserializeSeed>(
+        &self,
+        doc_address: DocAddress,
+        seed: T,
+    ) -> crate::Result<T::Value> {
         let store_reader = &self.inner.store_readers[doc_address.segment_ord as usize];
-        store_reader.get(doc_address.doc_id)
+        store_reader.get_seed(doc_address.doc_id, seed)
     }
 
     /// The cache stats for the underlying store reader.
@@ -109,9 +119,21 @@ impl Searcher {
         &self,
         doc_address: DocAddress,
     ) -> crate::Result<D> {
+        self.doc_async_seed(doc_address, PhantomData).await
+    }
+
+    #[cfg(feature = "quickwit")]
+    /// A stateful variant of [`doc_async`][Self::doc_async].
+    pub async fn doc_async_seed<T: DocumentDeserializeSeed>(
+        &self,
+        doc_address: DocAddress,
+        seed: T,
+    ) -> crate::Result<T::Value> {
         let executor = self.inner.index.search_executor();
         let store_reader = &self.inner.store_readers[doc_address.segment_ord as usize];
-        store_reader.get_async(doc_address.doc_id, executor).await
+        store_reader
+            .get_async_seed(doc_address.doc_id, executor, seed)
+            .await
     }
 
     /// Access the schema associated with the index of this searcher.

src/schema/document/de.rs

@@ -69,6 +69,28 @@ pub trait DocumentDeserialize: Sized {
     where D: DocumentDeserializer<'de>;
 }
 
+/// A stateful extension of [`DocumentDeserialize`].
+pub trait DocumentDeserializeSeed: Sized {
+    /// The type produced by using this seed.
+    type Value;
+
+    /// Attempts to deserialize `Self::Value` from the given `seed` and `deserializer`.
+    fn deserialize<'de, D>(self, deserializer: D) -> Result<Self::Value, DeserializeError>
+    where D: DocumentDeserializer<'de>;
+}
+
+impl<T> DocumentDeserializeSeed for PhantomData<T>
+where T: DocumentDeserialize
+{
+    /// The type produced by using this seed.
+    type Value = T;
+
+    fn deserialize<'de, D>(self, deserializer: D) -> Result<Self::Value, DeserializeError>
+    where D: DocumentDeserializer<'de> {
+        <T as DocumentDeserialize>::deserialize(deserializer)
+    }
+}
+
 /// A deserializer that can walk through each entry in the document.
 pub trait DocumentDeserializer<'de> {
     /// A indicator as to how many values are in the document.

src/schema/document/default_document.rs

@@ -603,7 +603,7 @@ impl<'a> Iterator for CompactDocObjectIter<'a> {
             container: self.container,
             value,
         };
-        return Some((key, value));
+        Some((key, value))
     }
 }
 
@@ -637,7 +637,7 @@ impl<'a> Iterator for CompactDocArrayIter<'a> {
             container: self.container,
             value,
         };
-        return Some(value);
+        Some(value)
     }
 }
 

src/schema/document/mod.rs

@@ -169,8 +169,9 @@ use std::mem;
 
 pub(crate) use self::de::BinaryDocumentDeserializer;
 pub use self::de::{
-    ArrayAccess, DeserializeError, DocumentDeserialize, DocumentDeserializer, ObjectAccess,
-    ValueDeserialize, ValueDeserializer, ValueType, ValueVisitor,
+    ArrayAccess, DeserializeError, DocumentDeserialize, DocumentDeserializeSeed,
+    DocumentDeserializer, ObjectAccess, ValueDeserialize, ValueDeserializer, ValueType,
+    ValueVisitor,
 };
 pub use self::default_document::{
     CompactDocArrayIter, CompactDocObjectIter, CompactDocValue, DocParsingError, TantivyDocument,

src/store/reader.rs

@@ -1,5 +1,6 @@
 use std::io;
 use std::iter::Sum;
+use std::marker::PhantomData;
 use std::num::NonZeroUsize;
 use std::ops::{AddAssign, Range};
 use std::sync::atomic::{AtomicUsize, Ordering};
@@ -14,7 +15,9 @@ use super::Decompressor;
 use crate::directory::FileSlice;
 use crate::error::DataCorruption;
 use crate::fastfield::AliveBitSet;
-use crate::schema::document::{BinaryDocumentDeserializer, DocumentDeserialize};
+use crate::schema::document::{
+    BinaryDocumentDeserializer, DocumentDeserialize, DocumentDeserializeSeed,
+};
 use crate::space_usage::StoreSpaceUsage;
 use crate::store::index::Checkpoint;
 use crate::DocId;
@@ -201,11 +204,21 @@ impl StoreReader {
     /// It should not be called to score documents
     /// for instance.
     pub fn get<D: DocumentDeserialize>(&self, doc_id: DocId) -> crate::Result<D> {
+        self.get_seed(doc_id, PhantomData)
+    }
+
+    /// A stateful version of [`get`][Self::get].
+    pub fn get_seed<T: DocumentDeserializeSeed>(
+        &self,
+        doc_id: DocId,
+        seed: T,
+    ) -> crate::Result<T::Value> {
         let mut doc_bytes = self.get_document_bytes(doc_id)?;
 
         let deserializer = BinaryDocumentDeserializer::from_reader(&mut doc_bytes)
             .map_err(crate::TantivyError::from)?;
-        D::deserialize(deserializer).map_err(crate::TantivyError::from)
+        seed.deserialize(deserializer)
+            .map_err(crate::TantivyError::from)
     }
 
     /// Returns raw bytes of a given document.
@@ -237,16 +250,27 @@ impl StoreReader {
     /// Iterator over all Documents in their order as they are stored in the doc store.
     /// Use this, if you want to extract all Documents from the doc store.
     /// The `alive_bitset` has to be forwarded from the `SegmentReader` or the results may be wrong.
-    pub fn iter<'a: 'b, 'b, D: DocumentDeserialize>(
+    pub fn iter<'a: 'b, 'b, D: DocumentDeserialize + 'b>(
         &'b self,
         alive_bitset: Option<&'a AliveBitSet>,
     ) -> impl Iterator<Item = crate::Result<D>> + 'b {
+        self.iter_seed(alive_bitset, &PhantomData)
+    }
+
+    /// A stateful variant of [`iter`][Self::iter].
+    pub fn iter_seed<'a: 'b, 'b, T: DocumentDeserializeSeed + Clone + 'b>(
+        &'b self,
+        alive_bitset: Option<&'a AliveBitSet>,
+        seed: &'b T,
+    ) -> impl Iterator<Item = crate::Result<T::Value>> + 'b {
         self.iter_raw(alive_bitset).map(|doc_bytes_res| {
             let mut doc_bytes = doc_bytes_res?;
 
             let deserializer = BinaryDocumentDeserializer::from_reader(&mut doc_bytes)
                 .map_err(crate::TantivyError::from)?;
-            D::deserialize(deserializer).map_err(crate::TantivyError::from)
+            seed.clone()
+                .deserialize(deserializer)
+                .map_err(crate::TantivyError::from)
         })
     }
 
@@ -389,11 +413,22 @@ impl StoreReader {
         doc_id: DocId,
         executor: &Executor,
     ) -> crate::Result<D> {
+        self.get_async_seed(doc_id, executor, PhantomData).await
+    }
+
+    /// A stateful variant of [`get_async`][Self::get_async].
+    pub async fn get_async_seed<T: DocumentDeserializeSeed>(
+        &self,
+        doc_id: DocId,
+        executor: &Executor,
+        seed: T,
+    ) -> crate::Result<T::Value> {
         let mut doc_bytes = self.get_document_bytes_async(doc_id, executor).await?;
 
         let deserializer = BinaryDocumentDeserializer::from_reader(&mut doc_bytes)
             .map_err(crate::TantivyError::from)?;
-        D::deserialize(deserializer).map_err(crate::TantivyError::from)
+        seed.deserialize(deserializer)
+            .map_err(crate::TantivyError::from)
     }
 }