Yeah, I need to revisit the docs! I’ll update it. Thanks for noting that!
zoedsoupe
keathley
zoedsoupe
announcing new features and performance improvements in Peri, our schema validation library! Here’s a quick overview:
New Features:
1. Transform Type
- Apply a transformation function to your data post-validation.
- Example:
defschema :user, %{ name: :string, age: {:integer, {:transform, &(&1 * 2)}} }
2. Default Values
- Set default values for fields when they are missing.
- Example:
defschema :user, %{ name: {:string, {:default, "Anonymous"}} }
3. Dependent Types
- Validate fields based on the value of other fields.
- Example:
defschema :user, %{ age: :integer, license: {:dependent, :age, fn age -> age >= 18 end, :string} }
Performance Improvements:
- Enhanced
Peri.Parserfor faster schema traversal and error handling.
Check out the latest version on GitHub and let us know your thoughts!
Happy coding!
2
bgoosman
Hey, nice library. Works well so far. I’m finding myself writing both 1) Peri schema 2) defstruct. Feels like I’m duplicating work. Any suggestions?
defmodule SightxrApi.Job.WorkflowSchemas do
import Peri
defschema(:parse_documents, %{
"files" => {:required, {:list, get_schema(:file)}},
"out_dirs" => {:list, :string},
"parser" => {:enum, ["tika", "azure_document_intelligence"]}
})
# ... omitted for brevity
end
And the defstruct:
defmodule SightxrApi.Job.ParseDocumentsWorkflow do
defmodule Options do
defstruct [:files, :parser, :out_dirs]
def new(args) do
%__MODULE__{
files: args["files"],
parser: args["parser"],
out_dirs: args["out_dirs"]
}
end
end
# ... omitted for brevity
end
1
shunji
not sure if this helps, but in my structs, to reduce duplication, I am using the non-macro API:
@schema %{
value: {:required, {:float, {:gt, 0}}},
unit: {:required, {:enum, [:meters]}}
}
@keys Map.keys(@schema)
defstruct @keys
def new(params) do
with {:ok, data} <- Peri.validate(@schema, params) do
{:ok, struct(__MODULE__, data)}
end
end
```
2