generics

generics

Generic code helpers for Go 1.18+ onwards

Overview

The generics package is a series of helpers for writing generic Go code that provides a series of functions that don't exist in the mainline package set.

About ZeroFlucs

ZeroFlucs is a B2B provider of pricing technology for Sportsbooks/wagering service providers globally. We use Open-Source software through our platform stack. This, along with other projects is made available through our zeroflucs-given Github profile on MIT licensing. To learn more you can visit:

• The ZeroFlucs Website - For information about our products and services.
• The ZeroFlucs Team Blog - For more content and posts from the ZeroFlucs team.
• ZeroFlucs-Given - For more OSS contributions.

Why Does this Exist?

When writing Go code for Go 1.17 or below, we've all written more than our fair share of methods to check "does this slice contain a thing", or "give me the first item matching a predicate". This package contains a roll-up of helpers and methods, as well as generic collection types that enable a lot of this boilerplate code to be removed.

Key attributes:

• Context support for filters/mappers (Optional)
• Does not mutate input during operation.

All code is covered 100% by tests for expected behaviour. Where filters or mappers are used methods are provided with and without context support.

Slice Queries

package generics_test

import (
	"fmt"

	"github.com/zeroflucs-given/generics/query"
)

func main() {
	inputs := []int{
		1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
	}
	query := query.Slice(inputs)

	result := query.
		Skip(5).
		TakeUntil(func(index int, v int) bool {
			return v >= 15
		}).
		Filter(func(index int, v int) bool {
			return v%2 == 0 // Evens only
		}).
		Reverse().
        ToSlice() // Back to a clean slice type

	fmt.Println(result) // [14, 12, 10, 8, 6]
}

But what about Contexts?

Don't worry, we've got you. Use .WithContext(ctx) on the query chain and the entire subtree is now context aware. Any materialization functions that emit an actual value will now return an extra error argument. Operations are lazy, and will skip over the remainder of any query chain once the first error has occured.

package generics_test

import (
	"context"
	"fmt"

	"github.com/zeroflucs-given/generics/query"
)

func main() {
	inputs := []int{
		1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
	}
	query := query.Slice(inputs)
	ctx := context.Todo()

	result, err := query.
		WithContext(ctx).
		Skip(5).
		TakeUntil(func(ctx context.Context, index int, v int) (bool, error) {
			return v >= 15, nil
		}).
		Filter(func(ctx context.Context, index int, v int) (bool, error)  {
			return v%2 == 0, nil // Evens only
		}).
		Reverse().
        ToSlice() // Back to a clean slice type

	fmt.Println(result) // [14, 12, 10, 8, 6]
	fmt.Println(err) // nil
}

When dealing with slice operations, the following design rules apply:

• Operations that return a single value will return the type default if no item matches (i.e. 0 for numeric types, empty string or nil for objects)
• Operations that accept multiple filters combine these filters into a logical AND by default.
• If no filters are applied, every item is assumed to pass.
• Context aware operations only exist where it makes sense (i.e. Take first 5 doesn't need a context, whereas take with filtering does)

Operation Detail

All / AllContext

Returns true if all items in the slice match the filter. Will return true for an empty slice as no items fail the predicate.

Any / AnyContext

Returns true if any item in the slice passes.

Combinations

Returns all combinations (note: not permutations) of items of length N over the slice.

Chunk

Uniformly distributes (or "chunks") a slice over n buckets.

ChunkMap

Uniformly distributes (or "chunks") a map over n buckets.

CombinationsFiltered

Returns all combinations of items of length N over the slice, where the members of the slice can be filtered. The return type contains references back to the original input list indicies.

Concatenate

Joins N slices of items together in the given order. Allocates a new slice.

Contains

Returns true if the slice contains the specified value T, false otherwise. Uses standard equality operator to compare types.

Count / CountContext

Returns the count of items matching the input filters.

Cut

Takes a slice and returns the head of the slice, plus a second value for the remainder of the slice.

DefaultIfEmpty

Given a slice, if the slice is empty or nil will create a slice of a single default item.

Distinct

Sorts and removes any duplicate elements from a slice, returning a new copy. The input slice is unchanged.

DistinctFunc

Similar to Distinct but takes a custom comparison function.

DistinctStable

Similar to Distinct but keeps the original order of the elements.

DistinctStableFunc

Similar to DistinctStable but takes a custom hash function. Hash collisions are ignored.

ExecuteOnce

Takes a function to be run at a later time and caches its result for retrieval many times. Subsequent retrievals will block until either their context is cancelled, or the task completes.

Filter / FilterContext

Creates a filtered set of the values in the slice, using a filter function.

First / FirstWithContext

Returns the first item of the slice that matches the filters. If no value matches, returns the type default.

FirstIndexOf

Returns the first index of a value in a typed slice, or -1 if not present.

Group / GroupWithContext

Uses a mapper function to assign input values to buckets.

If

If returns the equivalent value based on the predicate. This is an eager evaluation of both sides and not a true ternary operator.

Last / LastWithContext

Returns the last item of the slice that matches the filter.

LastIndexOf

Returns the index of the last occurrence of a value in the slice, or -1 if not present.

Map / MapWithContext

Runs the specified mapper over each element of the input slice, creating an output slice of a different type.

Mutate

Allows mutation of the slice elements, but the output must be of the same type as the original elements.

PointerTo

Returns a pointer reference to the input. Useful for lots of APIs that use string, integer pointers to differentiate between empty and absent.

PointerOrNil

Returns a pointer to the input, unless the input is the default value for its type (i.e. 0, empty string etc). In that scenario will return nil.

Reverse

Creates a reverse-sorted version of the input slice.

Skip

Skip the first N elements of the slice.

SkipUntil / SkipUntilWithContext

Skip items from the front of the slice until the predicate returns true.

SkipWhile / SkipWhileWithContext

Skips items from the front of the slice until the predicate returns false.

Take

Take the next N elements of the slice.

TakeUntil / TakeUntil

Take items from the slice until the filter function returns true.

TakeWhile / TakeWhileWithContext

Take items from the slice until the filter function returns false.

ToMap / ToMapWithContext

Converts a slice of values to Go map, using mappers for the key and values respectively.

ValueOrDefault

If a pointer is set, will return the dereferenced value. Otherwise returns the default value of the target type.

Slice Aggregations

Min

Returns the minimum value from the input slice. Returns 0 if no values.

Max

Returns the maximum value from the input slice. Returns 0 if no values.

Sum

Returns the total sum of the values. Note that when handling large values, you may overflow your input type.

Map Operations

The following map operation helpers exist in the generics package:

KeyValuesToMap

Assembles a map from a slice of key-value pairs.

Keys

Returns a slice of the key members of the map.

MapValues / MapValuesWithContext

Allows translation of a maps items to other values.

ToKeyValues

Converts a map into a slice of key-value pairs. As per Go handling of maps, the order of output here is not in a guaranteed order.

Values

Return a slice of the value members of the map.

Error Checking

Must

Returns the first value in a pair of (T, error) if there is no error. Otherwise will panic.

Filter Building

The filtering sub-package contains various constructs used to filter values.

True / TrueWithContext

A constant value filter that always returns true. Useful for testing.

False / FalseWithContext

A constant value filter that always returns false. Useful for testing.

And(...filters) / AndWithContext

Creates a composite AND filter for multiple conditions. An empty set of filters is a true.

Or(...filters) / OrWithContext

Creates a composite OR filter for multiple conditions. An empty set of filters is a false.

Not(filter) / NotWithContext

Creates a NOT/inverted version of the specified filter to allow for negative checking.

Wrap

Takes a non-context aware filter, but makes it compatible with code that expects contexts.