Add possibility to set a --skip-modules option without setting a --num-modules

[yassirnajmaoui]

Changes in this pull request

This fixes a problem that occurs when the user specifies a --skip-module option, but no --num-modules option.
This has been broken since we started using module_indices= in commit c0c9518.
In order to fix this, I had to create a custom class that defines a range that doesn't have a "stop" number, but that can still check if an element is in the range.

This is useful if a user wants to display all the scanner, but also wants to skip a set amount of modules.

Checklist before requesting a review

• I have performed a self-review of my code
• I have added docstrings/doxygen in line with the guidance in the developer guide
• The code builds and runs on my machine

Contribution Notes

Please read and adhere to the contribution guidelines.

Please tick the following:

• The content of this Pull Request (the Contribution) is intentionally submitted for inclusion in the ETSI software (the Work) under the terms and conditions of the Apache-2.0 License.

[KrisThielemans]

This seems ok but quite complicated. The following seems to work (but would need a bit more testing):

def openrange(start, step, stop):
     c=start
     while stop == 0 or c < stop:
          yield c
          c += step

module_indices = openrange(0, skip_modules+1, num_modules)

I see though that itertools.islice uses stop=None for "no end condition", so maybe we should do that as well (i.e. while stop is None or ... and default num_modules to None (which is probably the default for argparse already anyway).

[yassirnajmaoui]

This seems ok but quite complicated. The following seems to work (but would need a bit more testing):

def openrange(start, step, stop):
     c=start
     while stop == 0 or c < stop:
          yield c
          c += step

module_indices = openrange(0, skip_modules+1, num_modules)

I see though that itertools.islice uses stop=None for "no end condition", so maybe we should do that as well (i.e. while stop is None or ... and default num_modules to None (which is probably the default for argparse already anyway).

I know it seems complicated for what it's supposed to do, but I haven't found a way to do it with built-in functions that won't do an infinite loop when calling the in operator. I tried itertools.count and it does an infinite loop aswell.

Regarding your code, it doesn't work either, try this:

def openrange(start, step, stop):
     c=start
     while stop == 0 or c < stop:
          yield c
          c += step

r = openrange(0,5,0)

print(0 in r) # Returns True (as it should)
print(5 in r) # Returns True (as it should)
print(11 in r) # Goes through an infinite loop (should be returning False)

[KrisThielemans]

print(11 in r) # Goes through an infinite loop (should be returning False)

ah, yes, in doesn't work with a generator. Sorry!

I guess they should still be possible by reordering the for loops a bit (a pity we have the rep_module loop!), or other horrible work-arounds. However, this might all be over-engineering. Why not simple add

def get_num_modules(scanner_geometry: petsird.ScannerGeometry) -> int:
    """Compute total number of modules in the scanner"""
    num_modules = 0
    for rep_module in scanner_geometry.replicated_modules:
        num_modules += len(rep_volume.transforms)
    return num_modules

such that you know the upper bound. That function should then be added to petsird_helpers.h later.