Skip to content

Latest commit

 

History

History
44 lines (23 loc) · 1.44 KB

Attribute-Rules.md

File metadata and controls

44 lines (23 loc) · 1.44 KB
Raw

General attribute validation rules.

Most of our customer created entities have a name (exceptions are currently the KafkaUser and KafkaExporter).

After a lot of heated discussions, we are going with the following (copied from Pims comment)

Unicode, but strict rules for validation/sanitization,

NOTE: this is unclear; we can validate a unicode string to only have code points in the ascii range.

entity names

My suggestion is check all STRM entity names to have a default limit and no content validation:

any Unicode string of between 4 and 100 characters.

billing id

The billing-id is still used all over, and if it goes away it will be replaced by project-id. billing-id is often used in k8s label values, but should be moved to annotations.

Do we need something different than the general entity name validation?

doc, description, label and tags attributes

My suggestion would be to be generous with these.

  • doc, description: 0 to 5000 characters
  • label, tags: 0 to 500 characters, no more than 20 labels

names in schemas and event contracts.

Names in schemas are defined by Avro:

  • start with [A-Za-z_]
  • subsequently contain only [A-Za-z0-9_]

I would like to add "have a maximum length of 63 characters"

field paths

This is used in event contracts and masked fields. A field path is one ore more schema names, with forward slash characters in between.