[mymonin]

Closes #7025

This should fix ARMBUILD-Ubuntu-24.04-arm compilation failures which can be observed in other PRs like #7024 (cdash build error trace) and #7021 (cdash build error trace).

PR Checklist

• No API changes were made (or the changes have been approved)
• No major design changes were made (or the changes have been approved)
• Added test (or behavior not changed)
• Updated API documentation (or API not changed)
• Added license to new files (if any)
• Added Python wrapping to new files (if any) as described in ITK Software Guide Section 9.5
• Added ITK examples for all new major features (if any)

Refer to the ITK Software Guide for
further development details if necessary.

[greptile-apps]

Confidence Score: 5/5

[Medium impact] The PR appears safe to merge.

Summary

Adds the direct itkMacro.h include needed by itkIndent.h.

• ITK's print helpers can use Indent without a header loop.

_{Reviews (2) · Last reviewed commit: "COMP: Fix missing direct include for mac..." · Reviewed by Greptile}

[dzenanz]

Looks good on a glance. What commit introduced this build error?

[mymonin]

Looks good on a glance. What commit introduced this build error?

Based on the git history of itkIndent.h, I assume it is 70ddeb8, which I referred to in the description of #7025 (closed by this PR).

[hjmjohnson]

Thanks @mymonin for tracking this down and for the clear repro in #7025. I'd like to take a different route, so I opened #7028, which keeps your diagnosis and your itkIndent.h-first regression include (credited as co-author).

Two concerns with the forward declarations here:

1. The forward-declared operator<< lacks ITKCommon_EXPORT. The real declaration is friend ITKCommon_EXPORT std::ostream & operator<<(std::ostream &, const Indent &); (itkIndent.h). In a file that includes itkIndent.h first, the plain declaration is seen before the exported one, which MSVC shared builds can reject as inconsistent DLL linkage (C2375/C4273).
2. Two declarations of the same signature to keep in sync. Item 1 is that drift already happening. I'd rather remove the cycle than paper over it.

#7028 restores @blowekamp's earlier fix (79f01ca, 4f524e8), where itkIndent.h does not include itkMacro.h, by including only ITKCommonExport.h and itkConfigure.h and marking the legacy New()/Delete() with [[deprecated]] instead of itkLegacyMacro. It also adds an #error in itkPrintHelper.h that names the fix if itkIndent.h ever pulls in itkMacro.h again.

[mymonin]

I agree #7028 is a better fix if we can afford moving the macro definition to a more suitable header. Avoiding any circular include dependency makes the header logic much easier to reason about.

[hjmjohnson]

@mymonin 7028 was harder to complete than I expected. Can you please verify (on that PR) that the proposed solution works for you?

[hjmjohnson]

Superceeded.