Crate asdi

Another Simplistic Datalog Implementation (in Rust).

This package provides a data model to represent Datalog programs in memory, a parser for the textual representation, and some evaluation implementations. For more information see the ASDI book.

The text representation parser is a separate feature, so if you only need to construct and evaluate programs using the API you may opt out of the Pest parser and support.

Status

1. Library API mostly stable, the next effort will be to make the construction API more ergonomic.
2. Library Documentation good top-level documentation but very little else right now.
3. Parser full support for the core language as well as pragmas, require more unit tests.
4. I/O relations are now connected to the file input/output pragmas, the io module includes traits for reading/writing relations as well as basic JSON and CSV support.
5. Evaluation currently have a naïve and a stratified semi-naïve implementation.

Example

```datalog parent(xerces, brooke). parent(brooke, damocles).

ancestor(X, Y) ⟵ parent(X, Y). ancestor(X, Y) ⟵ parent(X, Z) ⋀ parent(Z, Y).

?- ancestor(xerces, X). ```

```rust use asdi::edb::{Attribute, Predicate}; use asdi::idb::{Atom, Query, Term, Variable}; use asdi::Program; use std::str::FromStr;

fn ancestor_example() { // See https://en.wikipedia.org/wiki/Datalog

let mut ancestors = Program::default();

let parent_predicate = Predicate::from_str("parent").unwrap();
{
    let parent = ancestors
        .add_new_relation(
            parent_predicate.clone(),
            vec![Attribute::string(), Attribute::string()],
        )
        .unwrap();
    parent
        .add_as_fact(["xerces".into(), "brooke".into()])
        .unwrap();
    parent
        .add_as_fact(["brooke".into(), "damocles".into()])
        .unwrap();
};

let ancestor_predicate = Predicate::from_str("ancestor").unwrap();

let var_x: Term = Variable::from_str("X").unwrap().into();
let var_y: Term = Variable::from_str("Y").unwrap().into();
let var_z: Term = Variable::from_str("Z").unwrap().into();

ancestors
    .add_new_rule(
        ancestor_predicate.clone(),
        [var_x.clone(), var_y.clone()],
        [Atom::new(parent_predicate.clone(), [var_x.clone(), var_y.clone()]).into()],
    )
    .unwrap();
ancestors
    .add_new_rule(
        ancestor_predicate.clone(),
        [var_x.clone(), var_y.clone()],
        [
            Atom::new(parent_predicate, [var_x.clone(), var_z.clone()]).into(),
            Atom::new(ancestor_predicate.clone(), [var_z, var_y]).into(),
        ],
    )
    .unwrap();

ancestors
    .add_new_query(ancestor_predicate, ["xerces".into(), var_x])
    .unwrap();

println!(">{}<", ancestors);

} ```

Crate features

| Feature | Default | Enables | |------------|--------------|---------------------------------------------------------------| | graphviz | Yes | Graph representation for dependency graphs and stratification | | parser | Yes | Parsing of textual representation | | tabular | Yes | Tabular output for views | | io | Yes | collects all the common I/O formats | | io_csv | Indirectly | Delimited line format support | | io_json | Indirectly | JSON format support | | io_text | Indirectly | Native text format (write only) support |

Changes

Version 0.2.5

• The focus is largely on documentation, both rustdoc and mdbook.
• Book now contains much better concrete documentation and UML for most major model areas.
• Some API changes as the documentation uncovered inconsistencies.
• Significant changes in the visitor and writer areas.

Version 0.2.4

• Completed work to support stratification, including precedence graphs.
• Added new StratifiedEvaluator implementation that performs stratification as well as using the semi-naive evaluation scheme.
• Added Criterion-based benchmarks for evaluation implementations.
• Separated query trait from relational operations.
• Minor changes in API for consistency.

Version 0.2.3

• Added new "*=" operator for regular expressions
• Renamed the variants of LiteralInner/Literal to be Relational and Arithmetic to convey meaning, not just type.
• Renamed Relations to RelationSet and Rules to RuleSet.
• Renamed PredicateSet to NameReferenceSet and used for variables also; this also added AttributeNameRef type for clarity.
• Updated the highlight.js datalog language definition for comment and operator changes.
• Replaced more uses of the term "comparison (operator)?" with arithmetic literal
• Fixed more Clippy warnings.

Version 0.2.2

• Made changes to syntax to align more closely with traditional Datalog, as well as Soufflé:
  • using % instead of # for line comments,
  • added /* and */ for block comments,
  • removed the need for a @ prefix on the boolean constants true and false,
  • using . instead of @ for pragmas.
  • ASDI still uses separate assert and infer pragmas rather than Soufflé's single decl.
  • Also, to keep pragmas regular with other statements they end in a . whereas in Soufflé they do not.
• Fixed parsing errors in CSV and JSON files, mainly parsing strings (include/exclude quotes).

Version 0.2.1

• Added traits for consistency across types.
• Added a predicate cache to reduce duplication.
• Finished the file input and output for relations.
• Cleaned up the error module and added functions for error construction.

Version 0.2.0

• New internal structures and library layout.
• Less cloning, but still too much.
• Finished top-level documentation, more to come.

Version 0.1.0

• Not published, used for own verification only.

Version 0.1.0-dev

• Initial release, mostly to verify CI infrastructure through to crates.io