Updated DateTime to hold timestamp in microseconds, while making date…

… field precision configurable

CHANGELOG.md

@@ -1,3 +1,10 @@
+Tantivy 0.19
+================================
+- Updated [Date Field Type](https://github.com/quickwit-oss/tantivy/pull/1396)
+  The `DateTime` type has been updated to hold timestamps with microseconds precision.
+  `DateOptions` and `DatePrecision` have been added to configure Date fields. The precision is used to hint on fast values compression. Otherwise, seconds precision is used everywhere else (i.e terms, indexing).
+
+
 Tantivy 0.18
 ================================
 - For date values `chrono` has been replaced with `time` (@uklotzde) #1304 :

Cargo.toml

@@ -49,7 +49,7 @@ thiserror = "1.0.30"
 htmlescape = "0.3.1"
 fail = "0.5.0"
 murmurhash32 = "0.2.0"
-time = { version = "0.3.9", features = ["serde-well-known"] }
+time = { version = "0.3.10", features = ["serde-well-known"] }
 smallvec = "1.8.0"
 rayon = "1.5.2"
 lru = "0.7.5"

examples/date_time_field.rs

@@ -0,0 +1,69 @@
+// # DateTime field example
+//
+// This example shows how the DateTime field can be used
+
+use tantivy::collector::TopDocs;
+use tantivy::query::QueryParser;
+use tantivy::schema::{Cardinality, DateOptions, Schema, Value, INDEXED, STORED, STRING};
+use tantivy::Index;
+
+fn main() -> tantivy::Result<()> {
+    // # Defining the schema
+    let mut schema_builder = Schema::builder();
+    let opts = DateOptions::from(INDEXED)
+        .set_stored()
+        .set_fast(Cardinality::SingleValue)
+        .set_precision(tantivy::DatePrecision::Seconds);
+    let occurred_at = schema_builder.add_date_field("occurred_at", opts);
+    let event_type = schema_builder.add_text_field("event", STRING | STORED);
+    let schema = schema_builder.build();
+
+    // # Indexing documents
+    let index = Index::create_in_ram(schema.clone());
+
+    let mut index_writer = index.writer(50_000_000)?;
+    let doc = schema.parse_document(
+        r#"{
+        "occurred_at": "2022-06-22T12:53:50.53Z",
+        "event": "pull-request"
+    }"#,
+    )?;
+    index_writer.add_document(doc)?;
+    let doc = schema.parse_document(
+        r#"{
+        "occurred_at": "2022-06-22T13:00:00.22Z",
+        "event": "comment"
+    }"#,
+    )?;
+    index_writer.add_document(doc)?;
+    index_writer.commit()?;
+
+    let reader = index.reader()?;
+    let searcher = reader.searcher();
+
+    // # Default fields: event_type
+    let query_parser = QueryParser::for_index(&index, vec![event_type]);
+    {
+        let query = query_parser.parse_query("event:comment")?;
+        let count_docs = searcher.search(&*query, &TopDocs::with_limit(5))?;
+        assert_eq!(count_docs.len(), 1);
+    }
+    {
+        let query = query_parser
+            .parse_query(r#"occurred_at:[2022-06-22T12:58:00Z TO 2022-06-23T00:00:00Z}"#)?;
+        let count_docs = searcher.search(&*query, &TopDocs::with_limit(4))?;
+        assert_eq!(count_docs.len(), 1);
+        for (_score, doc_address) in count_docs {
+            let retrieved_doc = searcher.doc(doc_address)?;
+            assert!(matches!(
+                retrieved_doc.get_first(occurred_at),
+                Some(Value::Date(_))
+            ));
+            assert_eq!(
+                schema.to_json(&retrieved_doc),
+                r#"{"event":["comment"],"occurred_at":["2022-06-22T13:00:00.22Z"]}"#
+            );
+        }
+    }
+    Ok(())
+}

fastfield_codecs/src/bitpacked.rs

@@ -14,7 +14,7 @@ pub struct BitpackedFastFieldReader {
     pub max_value_u64: u64,
 }
 
-impl<'data> FastFieldCodecReader for BitpackedFastFieldReader {
+impl FastFieldCodecReader for BitpackedFastFieldReader {
     /// Opens a fast field given a file.
     fn open_from_bytes(bytes: &[u8]) -> io::Result<Self> {
         let (_data, mut footer) = bytes.split_at(bytes.len() - 16);

query-grammar/src/query_grammar.rs

@@ -575,7 +575,7 @@ mod test {
         for special_char in SPECIAL_CHARS.iter() {
             let query = &format!("\\{special_char}my\\{special_char}field:a");
             assert_eq!(
-                super::field_name().parse(&query),
+                super::field_name().parse(query),
                 Ok((format!("{special_char}my{special_char}field"), "a"))
             );
         }

src/aggregation/intermediate_agg_result.rs

@@ -36,7 +36,10 @@ pub struct IntermediateAggregationResults {
 
 impl IntermediateAggregationResults {
     /// Convert intermediate result and its aggregation request to the final result.
-    pub fn into_final_bucket_result(self, req: Aggregations) -> crate::Result<AggregationResults> {
+    pub(crate) fn into_final_bucket_result(
+        self,
+        req: Aggregations,
+    ) -> crate::Result<AggregationResults> {
         self.into_final_bucket_result_internal(&(req.into()))
     }
 

src/collector/histogram_collector.rs

@@ -72,8 +72,7 @@ impl HistogramComputer {
             return;
         }
         let delta = value - self.min_value;
-        let delta_u64 = delta.to_u64();
-        let bucket_id: usize = self.divider.divide(delta_u64) as usize;
+        let bucket_id: usize = self.divider.divide(delta) as usize;
         if bucket_id < self.counts.len() {
             self.counts[bucket_id] += 1;
         }
@@ -287,7 +286,7 @@ mod tests {
             DateTime::from_primitive(
                 Date::from_calendar_date(1980, Month::January, 1)?.with_hms(0, 0, 0)?,
             ),
-            3600 * 24 * 365, // it is just for a unit test... sorry leap years.
+            3_600_000_000 * 24 * 365, // it is just for a unit test... sorry leap years.
             10,
         );
         let week_histogram = searcher.search(&all_query, &week_histogram_collector)?;

src/fastfield/mod.rs

@@ -52,11 +52,13 @@ pub trait MultiValueLength {
     fn get_total_len(&self) -> u64;
 }
 
-/// Trait for types that are allowed for fast fields: (u64, i64 and f64).
+/// Trait for types that are allowed for fast fields:
+/// (u64, i64 and f64, bool, DateTime).
 pub trait FastValue: Clone + Copy + Send + Sync + PartialOrd + 'static {
     /// Converts a value from u64
     ///
     /// Internally all fast field values are encoded as u64.
+    /// **Note: To be used for converting encoded Term, Posting values.**
     fn from_u64(val: u64) -> Self;
 
     /// Converts a value to u64.
@@ -189,24 +191,27 @@ impl FastValue for bool {
 }
 
 impl FastValue for DateTime {
-    fn from_u64(timestamp_u64: u64) -> Self {
-        let unix_timestamp = i64::from_u64(timestamp_u64);
-        Self::from_unix_timestamp(unix_timestamp)
+    /// Converts a timestamp microseconds into DateTime.
+    ///
+    /// **Note the timestamps is expected to be in microseconds.**
+    fn from_u64(timestamp_micros_u64: u64) -> Self {
+        let timestamp_micros = i64::from_u64(timestamp_micros_u64);
+        Self::from_timestamp_micros(timestamp_micros)
     }
 
     fn to_u64(&self) -> u64 {
-        self.into_unix_timestamp().to_u64()
+        common::i64_to_u64(self.into_timestamp_micros())
     }
 
     fn fast_field_cardinality(field_type: &FieldType) -> Option<Cardinality> {
         match *field_type {
-            FieldType::Date(ref integer_options) => integer_options.get_fastfield_cardinality(),
+            FieldType::Date(ref options) => options.get_fastfield_cardinality(),
             _ => None,
         }
     }
 
     fn as_u64(&self) -> u64 {
-        self.into_unix_timestamp().as_u64()
+        self.into_timestamp_micros().as_u64()
     }
 
     fn to_type() -> Type {
@@ -261,9 +266,9 @@ mod tests {
     use super::*;
     use crate::directory::{CompositeFile, Directory, RamDirectory, WritePtr};
     use crate::merge_policy::NoMergePolicy;
-    use crate::schema::{Document, Field, NumericOptions, Schema, FAST, STRING, TEXT};
+    use crate::schema::{Document, Field, Schema, FAST, STRING, TEXT};
     use crate::time::OffsetDateTime;
-    use crate::{Index, SegmentId, SegmentReader};
+    use crate::{DateOptions, DatePrecision, Index, SegmentId, SegmentReader};
 
     pub static SCHEMA: Lazy<Schema> = Lazy::new(|| {
         let mut schema_builder = Schema::builder();
@@ -559,8 +564,8 @@ mod tests {
     }
 
     #[test]
-    fn test_default_datetime() {
-        assert_eq!(0, DateTime::make_zero().into_unix_timestamp());
+    fn test_default_date() {
+        assert_eq!(0, DateTime::make_zero().into_timestamp_secs());
     }
 
     fn get_vals_for_docs(ff: &MultiValuedFastFieldReader<u64>, docs: Range<u32>) -> Vec<u64> {
@@ -766,10 +771,15 @@ mod tests {
     fn test_datefastfield() -> crate::Result<()> {
         use crate::fastfield::FastValue;
         let mut schema_builder = Schema::builder();
-        let date_field = schema_builder.add_date_field("date", FAST);
+        let date_field = schema_builder.add_date_field(
+            "date",
+            DateOptions::from(FAST).set_precision(DatePrecision::Microseconds),
+        );
         let multi_date_field = schema_builder.add_date_field(
             "multi_date",
-            NumericOptions::default().set_fast(Cardinality::MultiValues),
+            DateOptions::default()
+                .set_precision(DatePrecision::Microseconds)
+                .set_fast(Cardinality::MultiValues),
         );
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
@@ -797,23 +807,23 @@ mod tests {
         let dates_fast_field = fast_fields.dates(multi_date_field).unwrap();
         let mut dates = vec![];
         {
-            assert_eq!(date_fast_field.get(0u32).into_unix_timestamp(), 1i64);
+            assert_eq!(date_fast_field.get(0u32).into_timestamp_micros(), 1i64);
             dates_fast_field.get_vals(0u32, &mut dates);
             assert_eq!(dates.len(), 2);
-            assert_eq!(dates[0].into_unix_timestamp(), 2i64);
-            assert_eq!(dates[1].into_unix_timestamp(), 3i64);
+            assert_eq!(dates[0].into_timestamp_micros(), 2i64);
+            assert_eq!(dates[1].into_timestamp_micros(), 3i64);
         }
         {
-            assert_eq!(date_fast_field.get(1u32).into_unix_timestamp(), 4i64);
+            assert_eq!(date_fast_field.get(1u32).into_timestamp_micros(), 4i64);
             dates_fast_field.get_vals(1u32, &mut dates);
             assert!(dates.is_empty());
         }
         {
-            assert_eq!(date_fast_field.get(2u32).into_unix_timestamp(), 0i64);
+            assert_eq!(date_fast_field.get(2u32).into_timestamp_micros(), 0i64);
             dates_fast_field.get_vals(2u32, &mut dates);
             assert_eq!(dates.len(), 2);
-            assert_eq!(dates[0].into_unix_timestamp(), 5i64);
-            assert_eq!(dates[1].into_unix_timestamp(), 6i64);
+            assert_eq!(dates[0].into_timestamp_micros(), 5i64);
+            assert_eq!(dates[1].into_timestamp_micros(), 6i64);
         }
         Ok(())
     }

src/fastfield/multivalued/mod.rs

@@ -13,7 +13,7 @@ mod tests {
     use crate::collector::TopDocs;
     use crate::indexer::NoMergePolicy;
     use crate::query::QueryParser;
-    use crate::schema::{Cardinality, Facet, FacetOptions, NumericOptions, Schema};
+    use crate::schema::{Cardinality, DateOptions, Facet, FacetOptions, NumericOptions, Schema};
     use crate::time::format_description::well_known::Rfc3339;
     use crate::time::{Duration, OffsetDateTime};
     use crate::{DateTime, Document, Index, Term};
@@ -58,7 +58,7 @@ mod tests {
         let mut schema_builder = Schema::builder();
         let date_field = schema_builder.add_date_field(
             "multi_date_field",
-            NumericOptions::default()
+            DateOptions::default()
                 .set_fast(Cardinality::MultiValues)
                 .set_indexed()
                 .set_fieldnorm()

src/fastfield/multivalued/writer.rs

@@ -4,12 +4,12 @@ use fnv::FnvHashMap;
 use tantivy_bitpacker::minmax;
 
 use crate::fastfield::serializer::BitpackedFastFieldSerializerLegacy;
-use crate::fastfield::{value_to_u64, CompositeFastFieldSerializer, FastFieldType};
+use crate::fastfield::{value_to_u64, CompositeFastFieldSerializer, FastFieldType, FastValue};
 use crate::indexer::doc_id_mapping::DocIdMapping;
 use crate::postings::UnorderedTermId;
-use crate::schema::{Document, Field};
+use crate::schema::{Document, Field, Value};
 use crate::termdict::TermOrdinal;
-use crate::DocId;
+use crate::{DatePrecision, DocId};
 
 /// Writer for multi-valued (as in, more than one value per document)
 /// int fast field.
@@ -36,16 +36,22 @@ use crate::DocId;
 /// term ids when the segment is getting serialized.
 pub struct MultiValuedFastFieldWriter {
     field: Field,
+    precision_opt: Option<DatePrecision>,
     vals: Vec<UnorderedTermId>,
     doc_index: Vec<u64>,
     fast_field_type: FastFieldType,
 }
 
 impl MultiValuedFastFieldWriter {
     /// Creates a new `MultiValuedFastFieldWriter`
-    pub(crate) fn new(field: Field, fast_field_type: FastFieldType) -> Self {
+    pub(crate) fn new(
+        field: Field,
+        fast_field_type: FastFieldType,
+        precision_opt: Option<DatePrecision>,
+    ) -> Self {
         MultiValuedFastFieldWriter {
             field,
+            precision_opt,
             vals: Vec::new(),
             doc_index: Vec::new(),
             fast_field_type,
@@ -83,7 +89,14 @@ impl MultiValuedFastFieldWriter {
         }
         for field_value in doc.field_values() {
             if field_value.field == self.field {
-                self.add_val(value_to_u64(field_value.value()));
+                let value = field_value.value();
+                let value_u64 = match (self.precision_opt, value) {
+                    (Some(precision), Value::Date(date_val)) => {
+                        date_val.truncate(precision).to_u64()
+                    }
+                    _ => value_to_u64(value),
+                };
+                self.add_val(value_u64);
             }
         }
     }