Missing DEPENDS causes strange error

[timcanham]

F´ Version	4.x
Affected Component

---

Problem Description

If a module CMakeLists.txt has an empty DEPENDS section in the register_fprime_library function call, the fprime-util generate step will emit a bizarre error:

CMake Error at lib/fprime/cmake/utilities.cmake:788 (message):
   ----------------------------------------
   F Prime/CMake target 'TRUE' not available to deployment 'SomeDeployment'. 'TRUE' must:
       1. Must be defined somewhere in the F Prime project
       2. Must be defined before 'SomeDeployment' deployment (register_fprime_deployment)
   'TRUE' is undefined, or included via `add_fprime_subdirectory` after `register_fprime_deployment`. in:
       /home/tcanham/source/SomeDirectory/Something/CMakeLists.txt
   ----------------------------------------

Call Stack (most recent call first):
  lib/fprime/cmake/utilities.cmake:737 (fprime_cmake_clear_message)
  lib/fprime/cmake/target/target.cmake:75 (fprime_cmake_fatal_error)
  lib/fprime/cmake/target/target.cmake:136 (check_unknown_links)
  lib/fprime/cmake/target/target.cmake:181 (setup_single_target)
  lib/fprime/cmake/API.cmake:161 (setup_module_targets)
  lib/fprime/cmake/API.cmake:346 (fprime_attach_custom_targets)
  SomeDeploymentDirectory/CMakeLists.txt:19 (register_fprime_deployment)

(Fake directories and names were inserted into the error text)

Context / Environment

Execute fprime-util version-check and share the output.

Operating System: Linux
CPU Architecture: x86_64
Platform: Linux-5.15.133.1-microsoft-standard-WSL2-x86_64-with-glibc2.39
Python version: 3.12.3
CMake version: 3.26.0
Pip version: 26.1.1
Pip packages:
    fprime-tools==4.2.1
    fprime-gds==4.2.1
    fprime-fpp==3.2.0
Project submodules:
    https://github.com/nasa/fprime.git @ v4.2.2

How to Reproduce

1. Take any module directory with no DEPENDS in the register_fprime_library
2. Add DEPENDS with no string
3. Run fprime-util generate

Expected Behavior

It would be nice to have an error that is more informative like "Empty DEPENDS, dummy!"

[thomas-bc]

@LeStarch I wonder if this is something the native CMake cmake_parse_arguments() would handle more gracefully. I'll assign Devin to try and replacing our adhoc implementation with that, and see if it helps.