fix numeric order, refactor Document (#2209)

fix numeric order to prefer i64
rename and move Document stuff

src/schema/document/de.rs

@@ -902,7 +902,7 @@ mod tests {
             crate::schema::OwnedValue::Array(vec![
                 crate::schema::OwnedValue::Null,
                 crate::schema::OwnedValue::Str(String::from("Hello, world")),
-                crate::schema::OwnedValue::U64(12345),
+                crate::schema::OwnedValue::I64(12345),
             ]),
         );
     }

src/schema/document/default_document.rs

@@ -12,8 +12,8 @@ use crate::schema::field_value::FieldValueIter;
 use crate::schema::{Facet, Field, FieldValue, NamedFieldDocument, OwnedValue, Schema};
 use crate::tokenizer::PreTokenizedString;
 
-/// Tantivy's Document is the object that can be indexed and then searched for.
-/// It provides a default implementation of the `Document` trait.
+/// TantivyDocument provides a default implementation of the `Document` trait.
+/// It is the object that can be indexed and then searched for.
 ///
 /// Documents are fundamentally a collection of unordered couples `(field, value)`.
 /// In this list, one field may appear more than once.
@@ -256,7 +256,7 @@ impl DocParsingError {
 
 #[cfg(test)]
 mod tests {
-    use crate::schema::document::default_doc_type::TantivyDocument;
+    use crate::schema::document::default_document::TantivyDocument;
     use crate::schema::*;
 
     #[test]

src/schema/document/existing_type_impls.rs

@@ -25,10 +25,10 @@ impl<'a> Value<'a> for &'a serde_json::Value {
             serde_json::Value::Null => ReferenceValue::Null,
             serde_json::Value::Bool(value) => ReferenceValue::Bool(*value),
             serde_json::Value::Number(number) => {
-                if let Some(val) = number.as_u64() {
-                    ReferenceValue::U64(val)
-                } else if let Some(val) = number.as_i64() {
+                if let Some(val) = number.as_i64() {
                     ReferenceValue::I64(val)
+                } else if let Some(val) = number.as_u64() {
+                    ReferenceValue::U64(val)
                 } else if let Some(val) = number.as_f64() {
                     ReferenceValue::F64(val)
                 } else {

src/schema/document/mod.rs

@@ -154,25 +154,25 @@
 //! TODO: Complete this section...
 
 mod de;
-mod default_doc_type;
+mod default_document;
 mod existing_type_impls;
+mod owned_value;
 mod se;
+mod value;
 
 use std::collections::BTreeMap;
-use std::fmt::Debug;
 use std::mem;
-use std::net::Ipv6Addr;
 
 pub(crate) use self::de::BinaryDocumentDeserializer;
 pub use self::de::{
     ArrayAccess, DeserializeError, DocumentDeserialize, DocumentDeserializer, ObjectAccess,
     ValueDeserialize, ValueDeserializer, ValueType, ValueVisitor,
 };
-pub use self::default_doc_type::{DocParsingError, TantivyDocument};
+pub use self::default_document::{DocParsingError, TantivyDocument};
+pub use self::owned_value::OwnedValue;
 pub(crate) use self::se::BinaryDocumentSerializer;
+pub use self::value::{ReferenceValue, Value};
 use super::*;
-use crate::tokenizer::PreTokenizedString;
-use crate::DateTime;
 
 /// The core trait representing a document within the index.
 pub trait Document: DocumentDeserialize + Send + Sync + 'static {
@@ -245,314 +245,6 @@ pub trait Document: DocumentDeserialize + Send + Sync + 'static {
     }
 }
 
-/// A single field value.
-pub trait Value<'a>: Send + Sync + Debug {
-    /// The child value type returned by this doc value.
-    type ChildValue: Value<'a>;
-    /// The iterator for walking through the elements within the array.
-    type ArrayIter: Iterator<Item = ReferenceValue<'a, Self::ChildValue>>;
-    /// The visitor walking through the key-value pairs within
-    /// the object.
-    type ObjectIter: Iterator<Item = (&'a str, ReferenceValue<'a, Self::ChildValue>)>;
-
-    /// Returns the field value represented by an enum which borrows it's data.
-    fn as_value(&self) -> ReferenceValue<'a, Self>;
-
-    #[inline]
-    /// Returns if the value is `null` or not.
-    fn is_null(&self) -> bool {
-        matches!(self.as_value(), ReferenceValue::Null)
-    }
-
-    #[inline]
-    /// If the Value is a String, returns the associated str. Returns None otherwise.
-    fn as_str(&self) -> Option<&'a str> {
-        if let ReferenceValue::Str(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a u64, returns the associated u64. Returns None otherwise.
-    fn as_u64(&self) -> Option<u64> {
-        if let ReferenceValue::U64(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a i64, returns the associated i64. Returns None otherwise.
-    fn as_i64(&self) -> Option<i64> {
-        if let ReferenceValue::I64(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a f64, returns the associated f64. Returns None otherwise.
-    fn as_f64(&self) -> Option<f64> {
-        if let ReferenceValue::F64(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a datetime, returns the associated datetime. Returns None otherwise.
-    fn as_datetime(&self) -> Option<DateTime> {
-        if let ReferenceValue::Date(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a IP address, returns the associated IP. Returns None otherwise.
-    fn as_ip_addr(&self) -> Option<Ipv6Addr> {
-        if let ReferenceValue::IpAddr(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a bool, returns the associated bool. Returns None otherwise.
-    fn as_bool(&self) -> Option<bool> {
-        if let ReferenceValue::Bool(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a pre-tokenized string, returns the associated string. Returns None
-    /// otherwise.
-    fn as_pre_tokenized_text(&self) -> Option<&'a PreTokenizedString> {
-        if let ReferenceValue::PreTokStr(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a bytes value, returns the associated set of bytes. Returns None otherwise.
-    fn as_bytes(&self) -> Option<&'a [u8]> {
-        if let ReferenceValue::Bytes(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a facet, returns the associated facet. Returns None otherwise.
-    fn as_facet(&self) -> Option<&'a Facet> {
-        if let ReferenceValue::Facet(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// Returns the iterator over the array if the Value is an array.
-    fn as_array(&self) -> Option<Self::ArrayIter> {
-        if let ReferenceValue::Array(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// Returns the iterator over the object if the Value is an object.
-    fn as_object(&self) -> Option<Self::ObjectIter> {
-        if let ReferenceValue::Object(val) = self.as_value() {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// Returns true if the Value is an array.
-    fn is_array(&self) -> bool {
-        matches!(self.as_value(), ReferenceValue::Object(_))
-    }
-
-    #[inline]
-    /// Returns true if the Value is an object.
-    fn is_object(&self) -> bool {
-        matches!(self.as_value(), ReferenceValue::Object(_))
-    }
-}
-
-/// A enum representing a value for tantivy to index.
-pub enum ReferenceValue<'a, V>
-where V: Value<'a> + ?Sized
-{
-    /// A null value.
-    Null,
-    /// The str type is used for any text information.
-    Str(&'a str),
-    /// Unsigned 64-bits Integer `u64`
-    U64(u64),
-    /// Signed 64-bits Integer `i64`
-    I64(i64),
-    /// 64-bits Float `f64`
-    F64(f64),
-    /// Date/time with nanoseconds precision
-    Date(DateTime),
-    /// Facet
-    Facet(&'a Facet),
-    /// Arbitrarily sized byte array
-    Bytes(&'a [u8]),
-    /// IpV6 Address. Internally there is no IpV4, it needs to be converted to `Ipv6Addr`.
-    IpAddr(Ipv6Addr),
-    /// Bool value
-    Bool(bool),
-    /// Pre-tokenized str type,
-    PreTokStr(&'a PreTokenizedString),
-    /// A an array containing multiple values.
-    Array(V::ArrayIter),
-    /// A nested / dynamic object.
-    Object(V::ObjectIter),
-}
-
-impl<'a, V> ReferenceValue<'a, V>
-where V: Value<'a>
-{
-    #[inline]
-    /// Returns if the value is `null` or not.
-    pub fn is_null(&self) -> bool {
-        matches!(self, Self::Null)
-    }
-
-    #[inline]
-    /// If the Value is a String, returns the associated str. Returns None otherwise.
-    pub fn as_str(&self) -> Option<&'a str> {
-        if let Self::Str(val) = self {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a u64, returns the associated u64. Returns None otherwise.
-    pub fn as_u64(&self) -> Option<u64> {
-        if let Self::U64(val) = self {
-            Some(*val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a i64, returns the associated i64. Returns None otherwise.
-    pub fn as_i64(&self) -> Option<i64> {
-        if let Self::I64(val) = self {
-            Some(*val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a f64, returns the associated f64. Returns None otherwise.
-    pub fn as_f64(&self) -> Option<f64> {
-        if let Self::F64(val) = self {
-            Some(*val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a datetime, returns the associated datetime. Returns None otherwise.
-    pub fn as_datetime(&self) -> Option<DateTime> {
-        if let Self::Date(val) = self {
-            Some(*val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a IP address, returns the associated IP. Returns None otherwise.
-    pub fn as_ip_addr(&self) -> Option<Ipv6Addr> {
-        if let Self::IpAddr(val) = self {
-            Some(*val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a bool, returns the associated bool. Returns None otherwise.
-    pub fn as_bool(&self) -> Option<bool> {
-        if let Self::Bool(val) = self {
-            Some(*val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a pre-tokenized string, returns the associated string. Returns None
-    /// otherwise.
-    pub fn as_pre_tokenized_text(&self) -> Option<&'a PreTokenizedString> {
-        if let Self::PreTokStr(val) = self {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a bytes value, returns the associated set of bytes. Returns None otherwise.
-    pub fn as_bytes(&self) -> Option<&'a [u8]> {
-        if let Self::Bytes(val) = self {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// If the Value is a facet, returns the associated facet. Returns None otherwise.
-    pub fn as_facet(&self) -> Option<&'a Facet> {
-        if let Self::Facet(val) = self {
-            Some(val)
-        } else {
-            None
-        }
-    }
-
-    #[inline]
-    /// Returns true if the Value is an array.
-    pub fn is_array(&self) -> bool {
-        matches!(self, Self::Object(_))
-    }
-
-    #[inline]
-    /// Returns true if the Value is an object.
-    pub fn is_object(&self) -> bool {
-        matches!(self, Self::Object(_))
-    }
-}
-
 pub(crate) mod type_codes {
     pub const TEXT_CODE: u8 = 0;
     pub const U64_CODE: u8 = 1;

src/schema/document/se.rs

@@ -501,7 +501,7 @@ mod tests {
             length elements.len(),
             type_codes::NULL_CODE => (),
             type_codes::TEXT_CODE => String::from("Hello, world"),
-            type_codes::U64_CODE => 12345u64,
+            type_codes::I64_CODE => 12345i64,
         );
         assert_eq!(
             result, expected,

src/schema/document/value.rs

@@ -0,0 +1,316 @@
+use std::fmt::Debug;
+use std::net::Ipv6Addr;
+
+use common::DateTime;
+
+use crate::schema::Facet;
+use crate::tokenizer::PreTokenizedString;
+
+/// A single field value.
+pub trait Value<'a>: Send + Sync + Debug {
+    /// The child value type returned by this doc value.
+    type ChildValue: Value<'a>;
+    /// The iterator for walking through the elements within the array.
+    type ArrayIter: Iterator<Item = ReferenceValue<'a, Self::ChildValue>>;
+    /// The visitor walking through the key-value pairs within
+    /// the object.
+    type ObjectIter: Iterator<Item = (&'a str, ReferenceValue<'a, Self::ChildValue>)>;
+
+    /// Returns the field value represented by an enum which borrows it's data.
+    fn as_value(&self) -> ReferenceValue<'a, Self>;
+
+    #[inline]
+    /// Returns if the value is `null` or not.
+    fn is_null(&self) -> bool {
+        matches!(self.as_value(), ReferenceValue::Null)
+    }
+
+    #[inline]
+    /// If the Value is a String, returns the associated str. Returns None otherwise.
+    fn as_str(&self) -> Option<&'a str> {
+        if let ReferenceValue::Str(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a u64, returns the associated u64. Returns None otherwise.
+    fn as_u64(&self) -> Option<u64> {
+        if let ReferenceValue::U64(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a i64, returns the associated i64. Returns None otherwise.
+    fn as_i64(&self) -> Option<i64> {
+        if let ReferenceValue::I64(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a f64, returns the associated f64. Returns None otherwise.
+    fn as_f64(&self) -> Option<f64> {
+        if let ReferenceValue::F64(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a datetime, returns the associated datetime. Returns None otherwise.
+    fn as_datetime(&self) -> Option<DateTime> {
+        if let ReferenceValue::Date(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a IP address, returns the associated IP. Returns None otherwise.
+    fn as_ip_addr(&self) -> Option<Ipv6Addr> {
+        if let ReferenceValue::IpAddr(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a bool, returns the associated bool. Returns None otherwise.
+    fn as_bool(&self) -> Option<bool> {
+        if let ReferenceValue::Bool(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a pre-tokenized string, returns the associated string. Returns None
+    /// otherwise.
+    fn as_pre_tokenized_text(&self) -> Option<&'a PreTokenizedString> {
+        if let ReferenceValue::PreTokStr(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a bytes value, returns the associated set of bytes. Returns None otherwise.
+    fn as_bytes(&self) -> Option<&'a [u8]> {
+        if let ReferenceValue::Bytes(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a facet, returns the associated facet. Returns None otherwise.
+    fn as_facet(&self) -> Option<&'a Facet> {
+        if let ReferenceValue::Facet(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// Returns the iterator over the array if the Value is an array.
+    fn as_array(&self) -> Option<Self::ArrayIter> {
+        if let ReferenceValue::Array(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// Returns the iterator over the object if the Value is an object.
+    fn as_object(&self) -> Option<Self::ObjectIter> {
+        if let ReferenceValue::Object(val) = self.as_value() {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// Returns true if the Value is an array.
+    fn is_array(&self) -> bool {
+        matches!(self.as_value(), ReferenceValue::Object(_))
+    }
+
+    #[inline]
+    /// Returns true if the Value is an object.
+    fn is_object(&self) -> bool {
+        matches!(self.as_value(), ReferenceValue::Object(_))
+    }
+}
+
+/// A enum representing a value for tantivy to index.
+#[derive(Clone, Debug, PartialEq)]
+pub enum ReferenceValue<'a, V>
+where V: Value<'a> + ?Sized
+{
+    /// A null value.
+    Null,
+    /// The str type is used for any text information.
+    Str(&'a str),
+    /// Unsigned 64-bits Integer `u64`
+    U64(u64),
+    /// Signed 64-bits Integer `i64`
+    I64(i64),
+    /// 64-bits Float `f64`
+    F64(f64),
+    /// Date/time with nanoseconds precision
+    Date(DateTime),
+    /// Facet
+    Facet(&'a Facet),
+    /// Arbitrarily sized byte array
+    Bytes(&'a [u8]),
+    /// IpV6 Address. Internally there is no IpV4, it needs to be converted to `Ipv6Addr`.
+    IpAddr(Ipv6Addr),
+    /// Bool value
+    Bool(bool),
+    /// Pre-tokenized str type,
+    PreTokStr(&'a PreTokenizedString),
+    /// A an array containing multiple values.
+    Array(V::ArrayIter),
+    /// A nested / dynamic object.
+    Object(V::ObjectIter),
+}
+
+impl<'a, V> ReferenceValue<'a, V>
+where V: Value<'a>
+{
+    #[inline]
+    /// Returns if the value is `null` or not.
+    pub fn is_null(&self) -> bool {
+        matches!(self, Self::Null)
+    }
+
+    #[inline]
+    /// If the Value is a String, returns the associated str. Returns None otherwise.
+    pub fn as_str(&self) -> Option<&'a str> {
+        if let Self::Str(val) = self {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a u64, returns the associated u64. Returns None otherwise.
+    pub fn as_u64(&self) -> Option<u64> {
+        if let Self::U64(val) = self {
+            Some(*val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a i64, returns the associated i64. Returns None otherwise.
+    pub fn as_i64(&self) -> Option<i64> {
+        if let Self::I64(val) = self {
+            Some(*val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a f64, returns the associated f64. Returns None otherwise.
+    pub fn as_f64(&self) -> Option<f64> {
+        if let Self::F64(val) = self {
+            Some(*val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a datetime, returns the associated datetime. Returns None otherwise.
+    pub fn as_datetime(&self) -> Option<DateTime> {
+        if let Self::Date(val) = self {
+            Some(*val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a IP address, returns the associated IP. Returns None otherwise.
+    pub fn as_ip_addr(&self) -> Option<Ipv6Addr> {
+        if let Self::IpAddr(val) = self {
+            Some(*val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a bool, returns the associated bool. Returns None otherwise.
+    pub fn as_bool(&self) -> Option<bool> {
+        if let Self::Bool(val) = self {
+            Some(*val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a pre-tokenized string, returns the associated string. Returns None
+    /// otherwise.
+    pub fn as_pre_tokenized_text(&self) -> Option<&'a PreTokenizedString> {
+        if let Self::PreTokStr(val) = self {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a bytes value, returns the associated set of bytes. Returns None otherwise.
+    pub fn as_bytes(&self) -> Option<&'a [u8]> {
+        if let Self::Bytes(val) = self {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// If the Value is a facet, returns the associated facet. Returns None otherwise.
+    pub fn as_facet(&self) -> Option<&'a Facet> {
+        if let Self::Facet(val) = self {
+            Some(val)
+        } else {
+            None
+        }
+    }
+
+    #[inline]
+    /// Returns true if the Value is an array.
+    pub fn is_array(&self) -> bool {
+        matches!(self, Self::Object(_))
+    }
+
+    #[inline]
+    /// Returns true if the Value is an object.
+    pub fn is_object(&self) -> bool {
+        matches!(self, Self::Object(_))
+    }
+}

src/schema/mod.rs

@@ -126,7 +126,6 @@ mod json_object_options;
 mod named_field_document;
 mod numeric_options;
 mod text_options;
-mod value;
 
 use columnar::ColumnType;
 
@@ -134,7 +133,7 @@ pub use self::bytes_options::BytesOptions;
 #[allow(deprecated)]
 pub use self::date_time_options::DatePrecision;
 pub use self::date_time_options::{DateOptions, DateTimePrecision, DATE_TIME_PRECISION_INDEXED};
-pub use self::document::{DocParsingError, Document, TantivyDocument, Value};
+pub use self::document::{DocParsingError, Document, OwnedValue, TantivyDocument, Value};
 pub(crate) use self::facet::FACET_SEP_BYTE;
 pub use self::facet::{Facet, FacetParseError};
 pub use self::facet_options::FacetOptions;
@@ -153,7 +152,6 @@ pub use self::numeric_options::NumericOptions;
 pub use self::schema::{Schema, SchemaBuilder};
 pub use self::term::{Term, ValueBytes, JSON_END_OF_PATH};
 pub use self::text_options::{TextFieldIndexing, TextOptions, STRING, TEXT};
-pub use self::value::OwnedValue;
 
 /// Validator for a potential `field_name`.
 /// Returns true if the name can be use for a field name.