Attention is currently required from: laforge, pespin.

View Change

2 comments:

• File library/s1ap/S1AP_Templates.ttcn:

  • Patch Set #1, Line 734: /*, {

    what's wrong with sending this? You are not explaining why.

    Why was it already commented out here? I think because it's an **optional** IE that requires at least one E-RAB item (i.e. cannot be empty). There needs to be a logic (likely a function) that would be adding this IE conditionally based on the template parameter `rab_failed_items`. For instance, we want to exclude this `protocolIEs` item completely if `istemplatekind(rab_failed_items, "omit")`.

    Furthermore, I'd expect we want to test this case in ttcn3.

    Yes, I would like to add such testcase (if the time permits). But for a normal scenario when eNB reports all E-RABs as set up, this IE shall not be present in the list. This is why I am commenting it out below in `tr_S1AP_InitialCtxSetupResp`.

    For that special scenario with not all E-RABs being set up we can:

    * go for adding the additional `f_gen_protocolIEs()` logic,
     * easy to implement for `?` or `omit`
     * gets really complex for `*`
    * add another pair of templates extending `{ts,tr}_S1AP_InitialCtxSetupResp`
     * not if that would work for a list of `protocolIEs` though
    * add optional IEs conditionally in the testcase itself

  • Patch Set #1, Line 755: protocolIEs := {

    then please use superset() here.

    `superset()` removes the need to put the `*` at the end (which was there before my patches). It does not solve the problem of having optional IEs in a template though. I am fine with migrating to `superset()` in a separate patch.

To view, visit change 38277. To unsubscribe, or for help writing mail filters, visit settings.