跳到主要内容

Saving & Reusing Requests

An operation has its own request, and as many named ones as you save. Selecting an operation in the service tree loads its own request; editing it and moving away, or sending it, is what saves it. Save as in the toolbar keeps what is on screen under a name you choose, and the named requests appear as children of the operation — the quick set you flick between while checking a service.

What is saved, and when​

Aethrix writes the request down at two moments: when you move away from the operation in the tree, and after every send. Coming back restores it, so a half-composed request survives a restart — including the checks you have configured on it (see the request editor for the filling, assertion and extraction tabs).

Nothing is ever created behind your back: there is no timer writing drafts, and no second copy of your work appearing under a generated name.

What a saved request holds​

  • HTTP method and URL
  • Request headers
  • Request body
  • Authentication (credential profile and inline credentials)
  • Protocol-specific options (SOAP version, content type, Kafka SASL/security, gRPC deadline, and so on)
  • Filling rules, checks and extractions configured on the request

Your view-mode preference is not part of it — reopening an operation rebuilds the object tree from the request text and the operation's schema.

Several requests on one operation​

Save as names what is on screen and puts it under the operation as a child node. From there:

  • Select the child node to open that request; select the operation itself to open its own.
  • Right-click a child node to Rename it or Delete it. The operation's own request has no name to change and cannot be deleted — it is what the node opens with.
  • A name is used once per operation. Offering a name that is already taken, or the reserved word Default, is refused and the prompt reopens rather than writing over a request you can see in the tree.

Nothing is named for you: no draft, no timestamp, no copy appears unless you ask for it, and leaving a node or sending writes only to the row you were looking at.

When a second version stops being a scratch pad and starts being something you want to run again — with a suite, a data set and a report — put it in Test Automation: Copy as automation step (the request editor's Copy menu) places the request on screen into a test case, where duplicating the step and varying it is what the step model is for.

See Also​