Mica::XML::Fields

Reads an XML document into the same field hash a JSON object body produces, so one schema can validate either format. Web frameworks use it as a request-body parser; it is opt-in (require "mica/xml/fields") because it pulls in stdlib JSON.

The root element stands for the JSON object itself, so its name is not a key. Each child element becomes a field named by its local name:

• an element with only text becomes that text, exactly (<name>Ada</name> gives "name" => "Ada"; an empty element gives "");
• an element with attributes or child elements becomes a nested object; its own text is the "#text" key, kept exactly when it has no child elements, and with surrounding white space removed (and left out when blank) when the text is mixed in among child elements;
• a name that repeats among siblings becomes an array, in document order;
• attributes become "@name" keys of their element's object.

Values stay strings; the schema converts them, as it does for form data.

A request body is untrusted, so read streams it once through Mica::XML::Reader (no DOM is built) and raises Fields::RejectedDocumentError at the first sign of what a field hash cannot carry faithfully or safely: a document type declaration (DTD), refused at the <!DOCTYPE before any declaration is read; sibling elements or attributes that share a local name but not a namespace; or nesting deeper than MAX_DEPTH. Every event and every copied byte is charged to the parse's work meter, so the whole read stays linear in the body's size.

require "mica/xml/fields"

Mica::XML::Fields.read(%(<user id="7"><name>Ada</name><role>admin</role><role>dev</role></user>))
# => {"@id" => "7", "name" => "Ada", "role" => ["admin", "dev"]}