Closed nutjob4life closed 3 years ago
@nutjob4life looks good! It is kind of counter-intuitive, but i think the idea behind the design was a trail of validation, e.g. SIP -> AIP label -> AIP products . not sure if that actually happens in validate, but the opportunity is there...
📜 Summary
This resolves #99 by putting the checksum¹ of the AIP label XML file into the
<aip_label_checksum>
² element of the SIP label XML file. Previously this was the checksum³ of the AIP checksum⁴ manifest file and the transfer manifest was ignored. Now, both the checksum⁵ manifest and the transfer manifest are ignored and the AIP label gets the checksum⁶ limelight.🩺 Test Data and/or Report
🧩 Related Issues
99
📝 Footnotes
¹Not a checksum, but a hash. ²Also not a checksum. ³Actually a hash. ⁴It's called a checksum manifest, but it contains the output of message digests, which are hashes. ⁵Again, not a checksum. ⁶"Sum" is right in the name, this is is not at all a sum; a message digest's output is a hash⁷. ⁷Yes this is all rather anal-retentive⁸. ⁸What do you mean, you've never seen footnotes in pull requests before? 🧐