Struct fst::raw::OpBuilder

pub struct OpBuilder<'f> { /* fields omitted */ }

A builder for collecting fst streams on which to perform set operations on the keys of fsts.

Set operations include intersection, union, difference and symmetric difference. The result of each set operation is itself a stream that emits pairs of keys and a sequence of each occurrence of that key in the participating streams. This information allows one to perform set operations on fsts and customize how conflicting output values are handled.

All set operations work efficiently on an arbitrary number of streams with memory proportional to the number of streams.

The algorithmic complexity of all set operations is O(n1 + n2 + n3 + ...) where n1, n2, n3, ... correspond to the number of elements in each stream.

The 'f lifetime parameter refers to the lifetime of the underlying set.

Methods

impl<'f> OpBuilder<'f>

fn new() -> Self

Create a new set operation builder.

fn add<I, S>(self, stream: I) -> Self where
    I: for<'a> IntoStreamer<'a, Into = S, Item = (&'a [u8], Output)>,
    S: 'f + for<'a> Streamer<'a, Item = (&'a [u8], Output)>, 

Add a stream to this set operation.

This is useful for a chaining style pattern, e.g., builder.add(stream1).add(stream2).union().

The stream must emit a lexicographically ordered sequence of key-value pairs.

fn push<I, S>(&mut self, stream: I) where
    I: for<'a> IntoStreamer<'a, Into = S, Item = (&'a [u8], Output)>,
    S: 'f + for<'a> Streamer<'a, Item = (&'a [u8], Output)>, 

Add a stream to this set operation.

The stream must emit a lexicographically ordered sequence of key-value pairs.

fn union(self) -> Union<'f>

Performs a union operation on all streams that have been added.

Note that this returns a stream of (&[u8], &[IndexedValue]). The first element of the tuple is the byte string key. The second element of the tuple is a list of all occurrences of that key in participating streams. The IndexedValue contains an index and the value associated with that key in that stream. The index uniquely identifies each stream, which is an integer that is auto-incremented when a stream is added to this operation (starting at 0).

fn intersection(self) -> Intersection<'f>

Performs an intersection operation on all streams that have been added.

Note that this returns a stream of (&[u8], &[IndexedValue]). The first element of the tuple is the byte string key. The second element of the tuple is a list of all occurrences of that key in participating streams. The IndexedValue contains an index and the value associated with that key in that stream. The index uniquely identifies each stream, which is an integer that is auto-incremented when a stream is added to this operation (starting at 0).

fn difference(self) -> Difference<'f>

Performs a difference operation with respect to the first stream added. That is, this returns a stream of all elements in the first stream that don't exist in any other stream that has been added.

Note that this returns a stream of (&[u8], &[IndexedValue]). The first element of the tuple is the byte string key. The second element of the tuple is a list of all occurrences of that key in participating streams. The IndexedValue contains an index and the value associated with that key in that stream. The index uniquely identifies each stream, which is an integer that is auto-incremented when a stream is added to this operation (starting at 0).

fn symmetric_difference(self) -> SymmetricDifference<'f>

Performs a symmetric difference operation on all of the streams that have been added.

When there are only two streams, then the keys returned correspond to keys that are in either stream but not in both streams.

More generally, for any number of streams, keys that occur in an odd number of streams are returned.

Note that this returns a stream of (&[u8], &[IndexedValue]). The first element of the tuple is the byte string key. The second element of the tuple is a list of all occurrences of that key in participating streams. The IndexedValue contains an index and the value associated with that key in that stream. The index uniquely identifies each stream, which is an integer that is auto-incremented when a stream is added to this operation (starting at 0).

Trait Implementations

impl<'f, I, S> Extend<I> for OpBuilder<'f> where
    I: for<'a> IntoStreamer<'a, Into = S, Item = (&'a [u8], Output)>,
    S: 'f + for<'a> Streamer<'a, Item = (&'a [u8], Output)>, 

fn extend<T>(&mut self, it: T) where
    T: IntoIterator<Item = I>, 

Extends a collection with the contents of an iterator.

impl<'f, I, S> FromIterator<I> for OpBuilder<'f> where
    I: for<'a> IntoStreamer<'a, Into = S, Item = (&'a [u8], Output)>,
    S: 'f + for<'a> Streamer<'a, Item = (&'a [u8], Output)>, 

fn from_iter<T>(it: T) -> Self where
    T: IntoIterator<Item = I>, 

Creates a value from an iterator.