mentat/query-algebrizer
Richard Newman df90c366af
Partial work from simple aggregates work (#497) r=nalexander
* Pre: make FindQuery, FindSpec, and Element non-Clone.
* Pre: make query translator return a Result.
* Pre: make projection return a Result.
* Pre: refactor query parser in preparation for parsing aggregates.
* Pre: rename PredicateFn -> QueryFunction.
* Pre: expose more about bound variables from CC.
* Pre: move ValueTypeSet to core.
2017-11-30 15:02:07 -08:00
..
src Partial work from simple aggregates work (#497) r=nalexander 2017-11-30 15:02:07 -08:00
tests Partial work from simple aggregates work (#497) r=nalexander 2017-11-30 15:02:07 -08:00
Cargo.toml Partial work from simple aggregates work (#497) r=nalexander 2017-11-30 15:02:07 -08:00
README.md Partly flesh out query algebrizer. (#243) r=nalexander 2017-02-15 16:10:59 -08:00

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.

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.