shorter

type Names ¶

type Names struct {
	// contains filtered or unexported fields
}

`Names` maintains a mapping between long names and short ids in a Mongo collection.

`Shorten()` returns a known id or creates a new `<IdPrefix>xX...` id based on a hash of `<UriPrefix>:<namespace>:<name>`. Each shortener can be assigned a different uppercase `IdPrefix` to ensure that ids are globally unique. `x` is always a lowercase alphanumeric character to separate the prefix from the hash. The rest `X...` of the hash can contain uppercase or lowercase alphanumeric characters. The `UriPrefix` is usually left at its default value `ngn`, for NoG Name.

`Find()` works like `Shorten()` but returns an error if the name does not yet exist.

The shortener uses approximately `MinAbbrvLen*6 - 1 = 65` bits of a secure hash for `xX`, so that ids are likely to be unique even without the prefix. A name shortener creates longer ids if it detects name collisions.

func NewNames(conn *mgo.Session, cfg *NamesConfig) (*Names, error)

func (ns *Names) Find(namespace, name string) (string, error)

func (ns *Names) ListShortAll(namespace string) ([]string, error)

func (ns *Names) Shorten(namespace, name string) (string, error)

type NamesConfig ¶

type NamesConfig struct {
	Collection string
	IdPrefix   string
	UriPrefix  string
}