SourceVITA49 Component Bugfixes

[clink-HTL]

Discovered a few bugs/issues while working with and inspecting the SourceVITA49 component. Created fixes for the issues that were found.

Issues that were discovered:

• The functionality for processing a context packet's INS geolocation field would attempt to retrieve the GPS geolocation rather than the INS one.
• The output SRI mode was not being updated based on the contents of the data payload format field of context packets. This resulted in an incorrect mode for the SRI in certain cases.
• When sockets opened to listen for incoming data, the component was not closing existing sockets before replacing them. This resulted in sockets being left open when its attachment changed.
• Within the 'attach' method, the component claims (in a log message) to ignore new incoming attachments if it already has one. This was not the case, however, and the attachment settings were being overwritten with those of the new attachment.
• The 'sriLock' mutex that protects the component's current SRI information was not being locked in a couple places where that information was being modified.

See commit messages and diffs for further detail.

[jkb-axios]

Thank you for providing the pull request. I reviewed the commits and they mostly look good. One change we'll have to make is to the attach method, which should throw an AttachError exception rather than returning an attach ID. Also, the additions to the process_context method will be combined with code about 15 lines later that uses PayloadFormat as well.

If you would like credit in the form of us merging in any of your commits, please review the guidelines provided on the Community page for accepting contributions. Otherwise, we can implement the fixes and commit them separately.

[clink-HTL]

Thank you for providing the pull request. I reviewed the commits and they mostly look good. One change we'll have to make is to the attach method, which should throw an AttachError exception rather than returning an attach ID. Also, the additions to the process_context method will be combined with code about 15 lines later that uses PayloadFormat as well.

Those changes sound good to me.

I don't have any need to be credited for the commits, so I think the best way forward is to let you implement and commit them separately.

[jkb-axios]

That works. Thanks again for providing the pull request.

[htl-software]

Is there any status update on these changes?