GetWavePointer_Impl: Fix return memory address

[t-b]

We don't want to prefix it with 08 but instead we want to have 08 as precision which must stand after the percent sign.

While at it we also leverage the power of # and capital X to get a capital 0X prefix.

Bug introduced in b7bb12a4 (AreWavesEqual creates descriptive message in WAVE_DATA mode, 2021-03-04).

[Garados007]

I found an interesting bug but I don't know why the pipeline for b3ca346 succeeded. If I check out this commit (the same goes for main) and run Various.ipf locally I get the following error:

  Entering test case "UTF_Various#TC_WaveName"
  0: is false. Assertion "CHECK(GrepString(str, "^_free_ \\(0x[0-9a-f]+\\)$"))" failed in TC_WaveName (Various.ipf, line 574)
  0: is false. Assertion "CHECK(GrepString(str, "^namedFreeWave \\(0x[0-9a-f]+\\)$"))" failed in TC_WaveName (Various.ipf, line 579)
  Leaving test case "UTF_Various#TC_WaveName"

If I add INFO("name: \"%s\"", s1 = str) right before the CHECK I get more information:

  Entering test case "UTF_Various#TC_WaveName"
  0: is false. Assertion "CHECK(GrepString(str, "^_free_ \\(0x[0-9a-f]+\\)$"))" failed in TC_WaveName (Various.ipf, line 575)
    ⓘ name: "_free_ (0XFD735D0)"
  0: is false. Assertion "CHECK(GrepString(str, "^namedFreeWave \\(0x[0-9a-f]+\\)$"))" failed in TC_WaveName (Various.ipf, line 581)
    ⓘ name: "namedFreeWave (0XFD73E30)"
  Leaving test case "UTF_Various#TC_WaveName"

It seems that the format of the address has been changed. Everything has now capital letters. I will address this issue in PR #355 but I am still puzzled why this was allowed to pass in CI.

A quick side note by myself: I think it is more difficult to read the pointer if the "x" is capital.

[Garados007]

I have looked more in the results of the pipeline run. The test for Various didn't even run till the end. It just ended midway. These are the last lines in the log:

  Entering test case "UTF_Various#TC_MMD_Types:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0"
  Leaving test case "UTF_Various#TC_MMD_Types:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0"
  Entering test case "TC_MD_bck:first"

The test run is considered as succeeded because the failed JU file is never written to disk.

[Garados007]

Found more hints of this bug. I have created #356 to tackle this.

[t-b]

@Garados007 Thanks for noticing! Here Various.pxp finishes locally but does not pass (which is expected).