Struct mentat_transaction::query::Known

pub struct Known<'s, 'c> {
    pub schema: &'s Schema,
    pub cache: Option<&'c (CachedAttributes + 'c)>,
}

A convenience wrapper around things known in memory: the schema and caches. We use a trait object here to avoid making dozens of functions generic over the type of the cache. If performance becomes a concern, we should hard-code specific kinds of cache right here, and/or eliminate the Option.

Fields

schema: &'s Schema

cache: Option<&'c (CachedAttributes + 'c)>

Methods

impl<'s, 'c> Known<'s, 'c>

pub fn for_schema(s: &'s Schema) -> Known<'s, 'static>

pub fn new(
    s: &'s Schema,
    c: Option<&'c (CachedAttributes + 'c)>
) -> Known<'s, 'c>

impl<'s, 'c> Known<'s, 'c>

This is CachedAttributes, but with handy generic parameters. Why not make the trait generic? Because then we can't use it as a trait object in Known.

pub fn is_attribute_cached_reverse<U>(&self, entid: U) -> bool where
    U: Into<i64>, 

pub fn is_attribute_cached_forward<U>(&self, entid: U) -> bool where
    U: Into<i64>, 

pub fn get_values_for_entid<U, V>(
    &self,
    schema: &Schema,
    attribute: U,
    entid: V
) -> Option<&Vec<TypedValue>> where
    U: Into<i64>,
    V: Into<i64>, 

pub fn get_value_for_entid<U, V>(
    &self,
    schema: &Schema,
    attribute: U,
    entid: V
) -> Option<&TypedValue> where
    U: Into<i64>,
    V: Into<i64>, 

pub fn get_entid_for_value<U>(
    &self,
    attribute: U,
    value: &TypedValue
) -> Option<i64> where
    U: Into<i64>, 

pub fn get_entids_for_value<U>(
    &self,
    attribute: U,
    value: &TypedValue
) -> Option<&BTreeSet<i64>> where
    U: Into<i64>, 

Trait Implementations

impl<'s, 'c> Copy for Known<'s, 'c>

impl<'s, 'c> Clone for Known<'s, 'c>

fn clone(&self) -> Known<'s, 'c>

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.