Regular Expressions 101

Community Patterns

Parse Docker image reference components (URI / tag / name / registry / host)

0

Regular Expression
ECMAScript (JavaScript)

/
^(?<Name>(?<=^)(?:(?<Domain>(?:(?:localhost|[\w-]+(?:\.[\w-]+)+)(?::\d+)?)|[\w]+:\d+)\/)?\/?(?<Namespace>(?:(?:[a-z0-9]+(?:(?:[._]|__|[-]*)[a-z0-9]+)*)\/)*)(?<Repo>[a-z0-9-]+))[:@]?(?<Reference>(?<=:)(?<Tag>[\w][\w.-]{0,127})|(?<=@)(?<Digest>[A-Za-z][A-Za-z0-9]*(?:[-_+.][A-Za-z][A-Za-z0-9]*)*[:][0-9A-Fa-f]{32,}))?
/
gm

Description

This ES2018+ Javascript RegExp expression should parse the components of any valid Docker image reference URI.

It captures:

  • Name
    • ?Host
      • ?Domain
      • ?Port
    • ?Namespace
    • Repo
  • ?Reference
    • ?Tag
    • ?Digest

This expression uses named capture groups and lookarounds, which are available from ES2018. Be sure to check compatibility! See here for a good reference: https://www.smashingmagazine.com/2019/02/regexp-features-regular-expressions/#compatibility-table

The expression itself is based on a literal transcription of the RegEx from the official Go implementation / spec: https://pkg.go.dev/github.com/docker/distribution/reference

There is one notable deviation here from the spec -- I have separated the 'Path' component into two parts -- 'Namespace' and 'Repo'. This is helpful for registries which have container paths more than two levels deep (e.g. GitLab's.) So, in the case of registry.xyz/myorg/mygroup/myproject:latest, it would extract myorg/mygroup as "namespace" and myproject as "repo".

Hope this is helpful for someone!

Submitted by https://github.com/Propolisa - 2 years ago (Last modified 2 years ago)