Partly flesh out query algebrizer. (#243) r=nalexander

This commit is contained in:
Richard Newman 2017-02-03 16:52:03 -08:00
parent f36a78e61e
commit 5af7082165
5 changed files with 836 additions and 7 deletions

View file

@ -3,5 +3,8 @@ name = "mentat_query_algebrizer"
version = "0.0.1"
[dependencies]
[dependencies.mentat_core]
path = "../core"
[dependencies.mentat_query]
path = "../query"
path = "../query"

View file

@ -1,5 +1,5 @@
This crate turns a parsed query, as defined by the `query` crate and produced by the `query-parser` crate, into an *algebrized tree*, also called a *query processor tree*.
This is something of a wooly definition: a query algebrizer in a traditional relational database is the component that combines the schema — including column type constraints — with the query, resolving names, and that sort of thing. Much of that work is unnecessary in our model; for example, we don't need to resolve column aliases, deal with table names, or that sort of thing. But the similarity is strong enough to give us the name of this crate.
This is something of a wooly definition: a query algebrizer in a traditional relational database is the component that combines the schema — including column type constraints — with the query, resolving names and that sort of thing. Much of that work is unnecessary in our model; for example, we don't need to resolve column aliases, deal with table names, or that sort of thing. But the similarity is strong enough to give us the name of this crate.
The result of this process is traditionally handed to the *query optimizer* to yield an *execution plan*. In our case the execution plan is deterministically derived from the algebrized tree, and the real optimization (such as it is) takes place within the underlying SQLite database.

805
query-algebrizer/src/cc.rs Normal file
View file

@ -0,0 +1,805 @@
// Copyright 2016 Mozilla
//
// Licensed under the Apache License, Version 2.0 (the "License"); you may not use
// this file except in compliance with the License. You may obtain a copy of the
// License at http://www.apache.org/licenses/LICENSE-2.0
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#![allow(dead_code, unused_imports)]
extern crate mentat_core;
extern crate mentat_query;
use ::std::collections::BTreeMap;
use ::std::collections::btree_map::Entry;
use self::mentat_core::{
Attribute,
Entid,
Schema,
TypedValue,
ValueType,
};
use self::mentat_query::{
IdentOrEntid,
NamespacedKeyword,
NonIntegerConstant,
Pattern,
PatternNonValuePlace,
PatternValuePlace,
PlainSymbol,
SrcVar,
Variable,
};
/// This enum models the fixed set of default tables we have -- two
/// tables and two views.
#[derive(PartialEq, Eq, Clone, Copy, Debug)]
pub enum DatomsTable {
Datoms, // The non-fulltext datoms table.
FulltextValues, // The virtual table mapping IDs to strings.
FulltextDatoms, // The fulltext-datoms view.
AllDatoms, // Fulltext and non-fulltext datoms.
}
impl DatomsTable {
fn name(&self) -> &'static str {
match *self {
DatomsTable::Datoms => "datoms",
DatomsTable::FulltextValues => "fulltext_values",
DatomsTable::FulltextDatoms => "fulltext_datoms",
DatomsTable::AllDatoms => "all_datoms",
}
}
}
/// One of the named columns of our tables.
#[derive(PartialEq, Eq, Clone, Debug)]
pub enum DatomsColumn {
Entity,
Attribute,
Value,
Tx,
ValueTypeTag,
}
/// A specific instance of a table within a query. E.g., "datoms123".
pub type TableAlias = String;
/// The association between a table and its alias. E.g., AllDatoms, "all_datoms123".
#[derive(PartialEq, Eq, Debug)]
pub struct SourceAlias(DatomsTable, TableAlias);
/// A particular column of a particular aliased table. E.g., "datoms123", Attribute.
#[derive(PartialEq, Eq, Clone, Debug)]
pub struct QualifiedAlias(TableAlias, DatomsColumn);
impl QualifiedAlias {
fn for_type_tag(&self) -> QualifiedAlias {
QualifiedAlias(self.0.clone(), DatomsColumn::ValueTypeTag)
}
#[inline]
fn is_value(&self) -> bool {
self.1 == DatomsColumn::Value
}
}
/// A thing that's capable of aliasing a table name for us.
/// This exists so that we can obtain predictable names in tests.
pub type TableAliaser = Box<FnMut(DatomsTable) -> TableAlias>;
pub fn default_table_aliaser() -> TableAliaser {
let mut i = -1;
Box::new(move |table| {
i += 1;
format!("{}{:02}", table.name(), i)
})
}
#[derive(PartialEq, Eq, Debug)]
pub enum ColumnConstraint {
EqualsEntity(QualifiedAlias, Entid),
EqualsValue(QualifiedAlias, TypedValue),
EqualsColumn(QualifiedAlias, QualifiedAlias),
}
/// A `ConjoiningClauses` (CC) is a collection of clauses that are combined with `JOIN`.
/// The topmost form in a query is a `ConjoiningClauses`.
///
/// - Ordinary pattern clauses turn into `FROM` parts and `WHERE` parts using `=`.
/// - Predicate clauses turn into the same, but with other functions.
/// - `not` turns into `NOT EXISTS` with `WHERE` clauses inside the subquery to
/// bind it to the outer variables, or adds simple `WHERE` clauses to the outer
/// clause.
/// - `not-join` is similar, but with explicit binding.
/// - `or` turns into a collection of `UNION`s inside a subquery, or a simple
/// alternation.
/// `or`'s documentation states that all clauses must include the same vars,
/// but that's an over-simplification: all clauses must refer to the external
/// unification vars.
/// The entire `UNION`-set is `JOIN`ed to any surrounding expressions per the `rule-vars`
/// clause, or the intersection of the vars in the two sides of the `JOIN`.
///
/// Not yet done:
/// - Function clauses with bindings turn into:
/// * Subqueries. Perhaps less efficient? Certainly clearer.
/// * Projection expressions, if only used for output.
/// * Inline expressions?
///---------------------------------------------------------------------------------------
pub struct ConjoiningClauses {
/// `true` if this set of clauses cannot yield results in the context of the current schema.
is_known_empty: bool,
/// A function used to generate an alias for a table -- e.g., from "datoms" to "datoms123".
aliaser: TableAliaser,
/// A vector of source/alias pairs used to construct a SQL `FROM` list.
from: Vec<SourceAlias>,
/// A list of fragments that can be joined by `AND`.
wheres: Vec<ColumnConstraint>,
/// A map from var to qualified columns. Used to project.
bindings: BTreeMap<Variable, Vec<QualifiedAlias>>,
/// A map from var to type. Whenever a var maps unambiguously to two different types, it cannot
/// yield results, so we don't represent that case here. If a var isn't present in the map, it
/// means that its type is not known in advance.
known_types: BTreeMap<Variable, ValueType>,
/// A mapping, similar to `bindings`, but used to pull type tags out of the store at runtime.
/// If a var isn't present in `known_types`, it should be present here.
extracted_types: BTreeMap<Variable, QualifiedAlias>,
}
impl ::std::fmt::Debug for ConjoiningClauses {
fn fmt(&self, fmt: &mut ::std::fmt::Formatter) -> ::std::fmt::Result {
// TODO: improve this representation.
fmt.debug_struct("ConjoiningClauses")
.field("is_known_empty", &self.is_known_empty)
.field("from", &self.from)
.field("wheres", &self.wheres)
.field("bindings", &self.bindings)
.field("known_types", &self.known_types)
.field("extracted_types", &self.extracted_types)
.finish()
}
}
/// Basics.
impl Default for ConjoiningClauses {
fn default() -> ConjoiningClauses {
ConjoiningClauses {
is_known_empty: false,
aliaser: default_table_aliaser(),
from: vec![],
wheres: vec![],
bindings: BTreeMap::new(),
known_types: BTreeMap::new(),
extracted_types: BTreeMap::new(),
}
}
}
impl ConjoiningClauses {
pub fn bind_column_to_var(&mut self, table: TableAlias, column: DatomsColumn, var: Variable) {
let alias = QualifiedAlias(table, column);
// If this is a value, and we don't already know its type or where
// to get its type, record that we can get it from this table.
let needs_type_extraction =
alias.is_value() &&
!self.known_types.contains_key(&var) &&
!self.extracted_types.contains_key(&var);
if needs_type_extraction {
self.extracted_types.insert(var.clone(), alias.for_type_tag());
}
self.bindings.entry(var).or_insert(vec![]).push(alias);
}
pub fn constrain_column_to_constant(&mut self, table: TableAlias, column: DatomsColumn, constant: TypedValue) {
self.wheres.push(ColumnConstraint::EqualsValue(QualifiedAlias(table, column), constant))
}
pub fn constrain_column_to_entity(&mut self, table: TableAlias, column: DatomsColumn, entity: Entid) {
self.wheres.push(ColumnConstraint::EqualsEntity(QualifiedAlias(table, column), entity))
}
pub fn constrain_attribute(&mut self, table: TableAlias, attribute: Entid) {
self.constrain_column_to_entity(table, DatomsColumn::Attribute, attribute)
}
/// Constrains the var if there's no existing type.
/// Returns `false` if it's impossible for this type to apply (because there's a conflicting
/// type already known).
fn constrain_var_to_type(&mut self, variable: Variable, this_type: ValueType) -> bool {
// Is there an existing binding for this variable?
let types_entry = self.known_types.entry(variable);
match types_entry {
// If so, the types must match.
Entry::Occupied(entry) =>
*entry.get() == this_type,
// If not, record the one we just determined.
Entry::Vacant(entry) => {
entry.insert(this_type);
true
}
}
}
/// Ensure that the given place has the correct types to be a tx-id.
/// Right now this is mostly unimplemented: we fail hard if anything but a placeholder is
/// present.
fn constrain_to_tx(&mut self, tx: &PatternNonValuePlace) {
match *tx {
PatternNonValuePlace::Placeholder => (),
_ => unimplemented!(), // TODO
}
}
/// Ensure that the given place can be an entity, and is congruent with existing types.
/// This is used for `entity` and `attribute` places in a pattern.
fn constrain_to_ref(&mut self, value: &PatternNonValuePlace) {
// If it's a variable, record that it has the right type.
// Ident or attribute resolution errors (the only other check we need to do) will be done
// by the caller.
if let &PatternNonValuePlace::Variable(ref v) = value {
if !self.constrain_var_to_type(v.clone(), ValueType::Ref) {
self.mark_known_empty("Couldn't constrain var to Ref.");
}
}
}
fn mark_known_empty(&mut self, why: &str) {
self.is_known_empty = true;
println!("{}", why); // TODO: proper logging.
}
fn entid_for_ident<'s, 'a>(&self, schema: &'s Schema, ident: &'a NamespacedKeyword) -> Option<Entid> {
schema.get_entid(&ident.to_string()) // TODO: #291.
}
fn table_for_attribute_and_value<'s, 'a>(&self, attribute: &'s Attribute, value: &'a PatternValuePlace) -> Option<DatomsTable> {
if attribute.fulltext {
match value {
&PatternValuePlace::Placeholder =>
Some(DatomsTable::Datoms), // We don't need the value.
// TODO: an existing non-string binding can cause this pattern to fail.
&PatternValuePlace::Variable(_) =>
Some(DatomsTable::AllDatoms),
&PatternValuePlace::Constant(NonIntegerConstant::Text(_)) =>
Some(DatomsTable::AllDatoms),
_ => {
// We can't succeed if there's a non-string constant value for a fulltext
// field.
None
}
}
} else {
Some(DatomsTable::Datoms)
}
}
fn table_for_places<'s, 'a>(&self, schema: &'s Schema, attribute: &'a PatternNonValuePlace, value: &'a PatternValuePlace) -> Option<DatomsTable> {
match attribute {
// TODO: In a non-prepared context, check if a var is known by external binding, and
// choose the table accordingly, as if it weren't a variable. #279.
// TODO: In a prepared context, defer this decision until a second algebrizing phase.
// #278.
&PatternNonValuePlace::Placeholder | &PatternNonValuePlace::Variable(_) =>
// If the value is known to be non-textual, we can simply use the regular datoms
// table (TODO: and exclude on `index_fulltext`!).
//
// If the value is a placeholder too, then we can walk the non-value-joined view,
// because we don't care about retrieving the fulltext value.
//
// If the value is a variable or string, we must use `all_datoms`, or do the join
// ourselves, because we'll need to either extract or compare on the string.
Some(
match value {
// TODO: see if the variable is projected, aggregated, or compared elsewhere in
// the query. If it's not, we don't need to use all_datoms here.
&PatternValuePlace::Variable(_) =>
DatomsTable::AllDatoms, // TODO: check types.
&PatternValuePlace::Constant(NonIntegerConstant::Text(_)) =>
DatomsTable::AllDatoms,
_ =>
DatomsTable::Datoms,
}),
&PatternNonValuePlace::Entid(id) =>
schema.attribute_for_entid(id)
.and_then(|attribute| self.table_for_attribute_and_value(attribute, value)),
&PatternNonValuePlace::Ident(ref kw) =>
schema.attribute_for_ident(kw)
.and_then(|attribute| self.table_for_attribute_and_value(attribute, value)),
}
}
/// Produce a (table, alias) pair to handle the provided pattern.
/// This is a mutating method because it mutates the aliaser function!
/// Note that if this function decides that a pattern cannot match, it will flip
/// `is_known_empty`.
fn alias_table<'s, 'a>(&mut self, schema: &'s Schema, pattern: &'a Pattern) -> Option<SourceAlias> {
self.table_for_places(schema, &pattern.attribute, &pattern.value)
.map(|table| SourceAlias(table, (self.aliaser)(table)))
}
fn get_attribute<'s, 'a>(&self, schema: &'s Schema, pattern: &'a Pattern) -> Option<&'s Attribute> {
match pattern.attribute {
PatternNonValuePlace::Entid(id) =>
schema.attribute_for_entid(id),
PatternNonValuePlace::Ident(ref kw) =>
schema.attribute_for_ident(kw),
_ =>
None,
}
}
fn get_value_type<'s, 'a>(&self, schema: &'s Schema, pattern: &'a Pattern) -> Option<ValueType> {
self.get_attribute(schema, pattern).map(|x| x.value_type)
}
}
/// Expansions.
impl ConjoiningClauses {
/// Take the contents of `bindings` and generate inter-constraints for the appropriate
/// columns into `wheres`.
///
/// For example, a bindings map associating a var to three places in the query, like
///
/// ```edn
/// {?foo [datoms12.e datoms13.v datoms14.e]}
/// ```
///
/// produces two additional constraints:
///
/// ```example
/// datoms12.e = datoms13.v
/// datoms12.e = datoms14.e
/// ```
pub fn expand_bindings(&mut self) {
for cols in self.bindings.values() {
if cols.len() > 1 {
let ref primary = cols[0];
let secondaries = cols.iter().skip(1);
for secondary in secondaries {
// TODO: if both primary and secondary are .v, should we make sure
// the type tag columns also match?
// We don't do so in the ClojureScript version.
self.wheres.push(ColumnConstraint::EqualsColumn(primary.clone(), secondary.clone()));
}
}
}
}
/// When a CC has accumulated all patterns, generate value_type_tag entries in `wheres`
/// to refine value types for which two things are true:
///
/// - There are two or more different types with the same SQLite representation. E.g.,
/// ValueType::Boolean shares a representation with Integer and Ref.
/// - There is no attribute constraint present in the CC.
///
/// It's possible at this point for the space of acceptable type tags to not intersect: e.g.,
/// for the query
///
/// ```edn
/// [:find ?x :where
/// [?x ?y true]
/// [?z ?y ?x]]
/// ```
///
/// where `?y` must simultaneously be a ref-typed attribute and a boolean-typed attribute. This
/// function deduces that and calls `self.mark_known_empty`. #293.
pub fn expand_type_tags(&mut self) {
// TODO.
}
}
/// Application.
impl ConjoiningClauses {
/// Apply the constraints in the provided pattern to this CC.
///
/// This is a single-pass process, which means it is naturally incomplete, failing to take into
/// account all information spread across two patterns.
///
/// If the constraints cannot be satisfied -- for example, if this pattern includes a numeric
/// attribute and a string value -- then the `is_known_empty` field on the CC is flipped and
/// the function returns.
///
/// A pattern being impossible to satisfy isn't necessarily a bad thing -- this query might
/// have branched clauses that apply to different knowledge bases, and might refer to
/// vocabulary that isn't (yet) used in this one.
///
/// Most of the work done by this function depends on the schema and ident maps in the DB. If
/// these change, then any work done is invalid.
///
/// There's a lot more we can do here and later by examining the
/// attribute:
///
/// - If it's unique, and we have patterns like
///
/// [?x :foo/unique 5] [?x :foo/unique ?y]
///
/// then we can either prove impossibility (the two objects must
/// be equal) or deduce identity and simplify the query.
///
/// - The same, if it's cardinality-one and the entity is known.
///
/// - If there's a value index on this attribute, we might want to
/// run this pattern early in the query.
///
/// - A unique-valued attribute can sometimes be rewritten into an
/// existence subquery instead of a join.
fn apply_pattern_clause_for_alias<'s>(&mut self, schema: &'s Schema, pattern: &Pattern, alias: &SourceAlias) {
if self.is_known_empty {
return;
}
// Process each place in turn, applying constraints.
// Both `e` and `a` must be entities, which is equivalent here
// to being typed as Ref.
// Sorry for the duplication; Rust makes it a pain to abstract this.
// The transaction part of a pattern must be an entid, variable, or placeholder.
self.constrain_to_tx(&pattern.tx);
self.constrain_to_ref(&pattern.entity);
self.constrain_to_ref(&pattern.attribute);
let ref col = alias.1;
match pattern.entity {
PatternNonValuePlace::Placeholder =>
// Placeholders don't contribute any bindings, nor do
// they constrain the query -- there's no need to produce
// IS NOT NULL, because we don't store nulls in our schema.
(),
PatternNonValuePlace::Variable(ref v) =>
self.bind_column_to_var(col.clone(), DatomsColumn::Entity, v.clone()),
PatternNonValuePlace::Entid(entid) =>
self.constrain_column_to_entity(col.clone(), DatomsColumn::Entity, entid),
PatternNonValuePlace::Ident(ref ident) => {
if let Some(entid) = self.entid_for_ident(schema, ident) {
self.constrain_column_to_entity(col.clone(), DatomsColumn::Entity, entid)
} else {
// A resolution failure means we're done here.
self.mark_known_empty("Entity ident didn't resolve.");
return;
}
}
}
match pattern.attribute {
PatternNonValuePlace::Placeholder =>
(),
PatternNonValuePlace::Variable(ref v) =>
self.bind_column_to_var(col.clone(), DatomsColumn::Attribute, v.clone()),
PatternNonValuePlace::Entid(entid) => {
if !schema.is_attribute(entid) {
// Furthermore, that entid must resolve to an attribute. If it doesn't, this
// query is meaningless.
self.mark_known_empty("Attribute entid isn't an attribute.");
return;
}
self.constrain_column_to_entity(col.clone(), DatomsColumn::Attribute, entid)
},
PatternNonValuePlace::Ident(ref ident) => {
if let Some(entid) = self.entid_for_ident(schema, ident) {
self.constrain_column_to_entity(col.clone(), DatomsColumn::Attribute, entid);
if !schema.is_attribute(entid) {
self.mark_known_empty("Attribute ident isn't an attribute.");
return;
}
} else {
// A resolution failure means we're done here.
self.mark_known_empty("Attribute ident didn't resolve.");
return;
}
}
}
// Determine if the pattern's value type is known.
// We do so by examining the value place and the attribute.
// At this point it's possible that the type of the value is
// inconsistent with the attribute; in that case this pattern
// cannot return results, and we short-circuit.
let value_type = self.get_value_type(schema, pattern);
match pattern.value {
PatternValuePlace::Placeholder =>
(),
PatternValuePlace::Variable(ref v) => {
if let Some(this_type) = value_type {
// Wouldn't it be nice if we didn't need to clone in the found case?
// It doesn't matter too much: collisons won't be too frequent.
if !self.constrain_var_to_type(v.clone(), this_type) {
// The types don't match. This pattern cannot succeed.
self.mark_known_empty("Value types don't match.");
return;
}
}
self.bind_column_to_var(col.clone(), DatomsColumn::Value, v.clone());
},
PatternValuePlace::EntidOrInteger(i) =>
// If we know the valueType, then we can determine whether this is an entid or an
// integer. If we don't, then we must generate a more general query with a
// value_type_tag.
if let Some(ValueType::Ref) = value_type {
self.constrain_column_to_entity(col.clone(), DatomsColumn::Value, i);
} else {
unimplemented!();
},
PatternValuePlace::IdentOrKeyword(ref kw) => {
// If we know the valueType, then we can determine whether this is an ident or a
// keyword. If we don't, then we must generate a more general query with a
// value_type_tag.
// We can also speculatively try to resolve it as an ident; if we fail, then we
// know it can only return results if treated as a keyword, and we can treat it as
// such.
if let Some(ValueType::Ref) = value_type {
if let Some(entid) = self.entid_for_ident(schema, kw) {
self.constrain_column_to_entity(col.clone(), DatomsColumn::Value, entid)
} else {
// A resolution failure means we're done here: this attribute must have an
// entity value.
self.mark_known_empty("Value ident didn't resolve.");
return;
}
} else {
unimplemented!();
};
},
PatternValuePlace::Constant(ref c) => {
// TODO: don't allocate.
let typed_value = c.clone().into_typed_value();
if !typed_value.is_congruent_with(value_type) {
// If the attribute and its value don't match, the pattern must fail.
self.mark_known_empty("Value constant not congruent with attribute type.");
return;
}
// TODO: if we don't know the type of the attribute because we don't know the
// attribute, we can actually work backwards to the set of appropriate attributes
// from the type of the value itself! #292.
self.constrain_column_to_constant(col.clone(), DatomsColumn::Value, typed_value);
},
}
}
pub fn apply_pattern<'s, 'p>(&mut self, schema: &'s Schema, pattern: &'p Pattern) {
// For now we only support the default source.
match pattern.source {
Some(SrcVar::DefaultSrc) | None => (),
_ => unimplemented!(),
};
if let Some(alias) = self.alias_table(schema, &pattern) {
self.apply_pattern_clause_for_alias(schema, pattern, &alias);
self.from.push(alias);
} else {
// We didn't determine a table, likely because there was a mismatch
// between an attribute and a value.
// We know we cannot return a result, so we short-circuit here.
self.mark_known_empty("Table aliaser couldn't determine a table.");
}
}
}
#[cfg(test)]
mod testing {
use super::*;
fn associate_ident(schema: &mut Schema, i: &str, e: Entid) {
schema.entid_map.insert(e, i.to_string());
schema.ident_map.insert(i.to_string(), e);
}
fn add_attribute(schema: &mut Schema, e: Entid, a: Attribute) {
schema.schema_map.insert(e, a);
}
#[test]
fn test_unknown_ident() {
let mut cc = ConjoiningClauses::default();
let schema = Schema::default();
cc.apply_pattern(&schema, &Pattern {
source: None,
entity: PatternNonValuePlace::Variable(Variable(PlainSymbol::new("?x"))),
attribute: PatternNonValuePlace::Ident(NamespacedKeyword::new("foo", "bar")),
value: PatternValuePlace::Constant(NonIntegerConstant::Boolean(true)),
tx: PatternNonValuePlace::Placeholder,
});
assert!(cc.is_known_empty);
}
#[test]
fn test_unknown_attribute() {
let mut cc = ConjoiningClauses::default();
let mut schema = Schema::default();
associate_ident(&mut schema, ":foo/bar", 99);
cc.apply_pattern(&schema, &Pattern {
source: None,
entity: PatternNonValuePlace::Variable(Variable(PlainSymbol::new("?x"))),
attribute: PatternNonValuePlace::Ident(NamespacedKeyword::new("foo", "bar")),
value: PatternValuePlace::Constant(NonIntegerConstant::Boolean(true)),
tx: PatternNonValuePlace::Placeholder,
});
assert!(cc.is_known_empty);
}
#[test]
fn test_apply_simple_pattern() {
let mut cc = ConjoiningClauses::default();
let mut schema = Schema::default();
associate_ident(&mut schema, ":foo/bar", 99);
add_attribute(&mut schema, 99, Attribute {
value_type: ValueType::Boolean,
..Default::default()
});
let x = Variable(PlainSymbol::new("?x"));
cc.apply_pattern(&schema, &Pattern {
source: None,
entity: PatternNonValuePlace::Variable(x.clone()),
attribute: PatternNonValuePlace::Ident(NamespacedKeyword::new("foo", "bar")),
value: PatternValuePlace::Constant(NonIntegerConstant::Boolean(true)),
tx: PatternNonValuePlace::Placeholder,
});
// println!("{:#?}", cc);
let d0_e = QualifiedAlias("datoms00".to_string(), DatomsColumn::Entity);
let d0_a = QualifiedAlias("datoms00".to_string(), DatomsColumn::Attribute);
let d0_v = QualifiedAlias("datoms00".to_string(), DatomsColumn::Value);
// After this, we know a lot of things:
assert!(!cc.is_known_empty);
assert_eq!(cc.from, vec![SourceAlias(DatomsTable::Datoms, "datoms00".to_string())]);
// ?x must be a ref.
assert_eq!(cc.known_types.get(&x).unwrap(), &ValueType::Ref);
// ?x is bound to datoms0.e.
assert_eq!(cc.bindings.get(&x).unwrap(), &vec![d0_e.clone()]);
// Our 'where' clauses are two:
// - datoms0.a = 99
// - datoms0.v = true
// No need for a type tag constraint, because the attribute is known.
assert_eq!(cc.wheres, vec![
ColumnConstraint::EqualsEntity(d0_a, 99),
ColumnConstraint::EqualsValue(d0_v, TypedValue::Boolean(true)),
]);
}
#[test]
fn test_apply_unattributed_pattern() {
let mut cc = ConjoiningClauses::default();
let schema = Schema::default();
let x = Variable(PlainSymbol::new("?x"));
cc.apply_pattern(&schema, &Pattern {
source: None,
entity: PatternNonValuePlace::Variable(x.clone()),
attribute: PatternNonValuePlace::Placeholder,
value: PatternValuePlace::Constant(NonIntegerConstant::Boolean(true)),
tx: PatternNonValuePlace::Placeholder,
});
// println!("{:#?}", cc);
let d0_e = QualifiedAlias("datoms00".to_string(), DatomsColumn::Entity);
let d0_v = QualifiedAlias("datoms00".to_string(), DatomsColumn::Value);
assert!(!cc.is_known_empty);
assert_eq!(cc.from, vec![SourceAlias(DatomsTable::Datoms, "datoms00".to_string())]);
// ?x must be a ref.
assert_eq!(cc.known_types.get(&x).unwrap(), &ValueType::Ref);
// ?x is bound to datoms0.e.
assert_eq!(cc.bindings.get(&x).unwrap(), &vec![d0_e.clone()]);
// Our 'where' clauses are two:
// - datoms0.v = true
// - datoms0.value_type_tag = boolean
// TODO: implement expand_type_tags.
assert_eq!(cc.wheres, vec![
ColumnConstraint::EqualsValue(d0_v, TypedValue::Boolean(true)),
]);
}
#[test]
fn test_apply_two_patterns() {
let mut cc = ConjoiningClauses::default();
let mut schema = Schema::default();
associate_ident(&mut schema, ":foo/bar", 99);
associate_ident(&mut schema, ":foo/roz", 98);
add_attribute(&mut schema, 99, Attribute {
value_type: ValueType::Boolean,
..Default::default()
});
add_attribute(&mut schema, 98, Attribute {
value_type: ValueType::String,
unique_identity: true,
..Default::default()
});
let x = Variable(PlainSymbol::new("?x"));
let y = Variable(PlainSymbol::new("?y"));
cc.apply_pattern(&schema, &Pattern {
source: None,
entity: PatternNonValuePlace::Variable(x.clone()),
attribute: PatternNonValuePlace::Ident(NamespacedKeyword::new("foo", "roz")),
value: PatternValuePlace::Constant(NonIntegerConstant::Text("idgoeshere".to_string())),
tx: PatternNonValuePlace::Placeholder,
});
cc.apply_pattern(&schema, &Pattern {
source: None,
entity: PatternNonValuePlace::Variable(x.clone()),
attribute: PatternNonValuePlace::Ident(NamespacedKeyword::new("foo", "bar")),
value: PatternValuePlace::Variable(y.clone()),
tx: PatternNonValuePlace::Placeholder,
});
// Finally, expand bindings to get the overlaps for ?x.
cc.expand_bindings();
println!("{:#?}", cc);
let d0_e = QualifiedAlias("datoms00".to_string(), DatomsColumn::Entity);
let d0_a = QualifiedAlias("datoms00".to_string(), DatomsColumn::Attribute);
let d0_v = QualifiedAlias("datoms00".to_string(), DatomsColumn::Value);
let d1_e = QualifiedAlias("datoms01".to_string(), DatomsColumn::Entity);
let d1_a = QualifiedAlias("datoms01".to_string(), DatomsColumn::Attribute);
assert!(!cc.is_known_empty);
assert_eq!(cc.from, vec![
SourceAlias(DatomsTable::Datoms, "datoms00".to_string()),
SourceAlias(DatomsTable::Datoms, "datoms01".to_string()),
]);
// ?x must be a ref.
assert_eq!(cc.known_types.get(&x).unwrap(), &ValueType::Ref);
// ?x is bound to datoms0.e and datoms1.e.
assert_eq!(cc.bindings.get(&x).unwrap(),
&vec![
d0_e.clone(),
d1_e.clone(),
]);
// Our 'where' clauses are four:
// - datoms0.a = 98 (:foo/roz)
// - datoms0.v = "idgoeshere"
// - datoms1.a = 99 (:foo/bar)
// - datoms1.e = datoms0.e
assert_eq!(cc.wheres, vec![
ColumnConstraint::EqualsEntity(d0_a, 98),
ColumnConstraint::EqualsValue(d0_v, TypedValue::String("idgoeshere".to_string())),
ColumnConstraint::EqualsEntity(d1_a, 99),
ColumnConstraint::EqualsColumn(d0_e, d1_e),
]);
}
}

View file

@ -9,16 +9,31 @@
// specific language governing permissions and limitations under the License.
extern crate mentat_query;
mod cc;
use mentat_query::{
FindQuery,
FindSpec,
SrcVar,
Variable,
WhereClause,
};
pub struct AlgebrizedQuery {
#[allow(dead_code)]
pub struct AlgebraicQuery {
default_source: SrcVar,
find_spec: FindSpec,
has_aggregates: bool,
limit: Option<i64>,
cc: cc::ConjoiningClauses,
}
pub fn algebrize(parsed: FindQuery) -> AlgebrizedQuery {
AlgebrizedQuery {}
#[allow(dead_code)]
pub fn algebrize(parsed: FindQuery) -> AlgebraicQuery {
AlgebraicQuery {
default_source: parsed.default_source,
find_spec: parsed.find_spec,
has_aggregates: false, // TODO: we don't parse them yet.
limit: None,
cc: cc::ConjoiningClauses::default(),
}
}

View file

@ -182,6 +182,12 @@ impl FromValue<PatternNonValuePlace> for PatternNonValuePlace {
}
}
#[derive(Clone, Debug, Eq, PartialEq)]
pub enum IdentOrEntid {
Ident(NamespacedKeyword),
Entid(i64),
}
/// The `v` part of a pattern can be much broader: it can represent
/// integers that aren't entity IDs (particularly negative integers),
/// strings, and all the rest. We group those under `Constant`.