1285 cust auth dev

[RomaMokych]

I’ve implemented eq, not eq, any, none, contains* restrictions.
Also I’ve added some base logic for authorities validation.

That logic is not integrated into transaction validating. I’m going to implement it in the beginning of the next week.

Also I’m going to improve various type support as arguments for restrictions. Now only several types are supported.

[RomaMokych]

@sschiessl-bcp

Can you please give us a general update in terms of what is completed, what needs review / what is outstanding and what your estimate for the remaining tasks is?

Sure. It is left to:

1. Finish with attribute assert - will push it today, it is under internal review now.
2. Fix authority validation logic in custom authorities validations - will push it today, it is under internal review now.
3. ordered_non_unqinue replacement with oredered_unique noticed by Abit - we are in the middle of discussion now. I will experiment with Abit proposal, but I'm still not sure if it will work or not. I think we will finish with today/tomorrow.
4. Several minor fixes, some additional tests. Going to finish them today/tomorrow.

Will require review:

1. attribute assert - will push it today, it is under internal review now.
2. authority validation logic in custom authorities validations - will push it today, it is under internal review now.
3. Fee calculation logic - as I understood current solution is fine. But we still need to decide how units be calculated for each type.

Everything else is done.

[xeroc]

3\. ordered_non_unqinue replacement with oredered_unique noticed by Abit - we are in the middle of discussion now. I will experiment with Abit proposal, but I'm still not sure if it will work or not. I think we will finish with today/tomorrow.

I have recently done something similar for another project:
peerplays-network/peerplays@9946eab
Works perfectly by extending the index into a composite_key with the object_id

3\. 

Everything else is done.

Excellent news. I'll look for someone to do the review.
please ping us when you have pushed the rest of the code. Thanks

[RomaMokych]

3\. ordered_non_unqinue replacement with oredered_unique noticed by Abit - we are in the middle of discussion now. I will experiment with Abit proposal, but I'm still not sure if it will work or not. I think we will finish with today/tomorrow.

I have recently done something similar for another project:
peerplays-network/peerplays@9946eab
Works perfectly by extending the index into a composite_key with the object_id

3\. 

Everything else is done.

Excellent news. I'll look for someone to do the review.
please ping us when you have pushed the rest of the code. Thanks

I've pushed changes for attribute_assert. Please review. Changes are in the restrictions* files.
New restriction was called attribute_assert_restriction.

Also I've pushed fix for authority validation (authority which is contained in the custom_authority).
And I've updated algorithm of authority validation. So now, either custom authority is validated or normal authority if custom one is absent.
These changes are in transaction.cpp file. Please ask any questions.

Also I've removed restriction.hpp and restriction.cpp files with old implementation of restrictions.

Today I plan to do couple of small fixes: I've found issue in API and I'll add couple of tests.
This should be ready till the end of the day.

@abitmore @xeroc @sschiessl-bcp

[sschiessl-bcp]

Today I plan to do couple of small fixes: I've found issue in API and I'll add couple of tests.

Awesome! Please ping when a new holistic review can be done

[RomaMokych]

@sschiessl-bcp @xeroc @abitmore
Guys, you can do holistic review now. I've pushed all latest changes with fixes of authority validation and attribute_assert

[abitmore]

either custom authority is validated or normal authority if custom one is absent.

I haven't review the code yet, just comment here to confirm one thing: custom authority should be checked first, if nothing matches (but not only when "absent"), need to check normal authority.

[sschiessl-bcp]

either custom authority is validated or normal authority if custom one is absent.

I haven't review the code yet, just comment here to confirm one thing: custom authority should be checked first, if nothing matches (but not only when "absent"), need to check normal authority.

@RomaMokych ?

[RomaMokych]

I haven't review the code yet, just comment here to confirm one thing: custom authority should be checked first, if nothing matches (but not only when "absent"), need to check normal authority.

I'll check. I think I'll need to fix this case.
For now I'm failing validation if custom_authorities present and none is matched.

Thank you @abitmore

Will update it in couple of hours. But this is small change you still can review feature.

UPD: updated it. Added tests for such case. Updated validation logic a little (transaction.cpp).

[nathanielhourt]

CMake fails to parse because libraries/chain/operation_type_to_id.cpp is missing

[RomaMokych]

CMake fails to parse because libraries/chain/operation_type_to_id.cpp is missing

Fixed it. @nathanhourt

[xeroc]

@nathanhourt Are you taking the lead in project managing this feature? There's some mileage already done and we'd rather sprint to the finish line!