Polish a few corner cases

[eroux]

This is the final PR for me. @hmottestad I've integrated and completed your PR (#207) in it. It doesn't do anything with the @requireAll flag, but at least it should have a less buggy behavior. The flag is passed as an argument of filterNode in JsonLdApi.java, but this function requires some heavy lifting to properly align with the Frame Matching algorithm of JSON-LD 1.1. If you want to really handle the flag, that's where you need to change things.

[coveralls]

Coverage increased (+0.008%) to 88.657% when pulling 54e2a28 on BuddhistDigitalResourceCenter:fix-new into 5589124 on jsonld-java:master.

[eroux]

@ansell I think the Changelog could be clearer if it was organized by version instead of date, and was in a separate file. I tend to follow these guidelines which really help me... (an example), but that not crucial.

[ansell]

Just need to preserve the custom test added in #168, otherwise looks good.

[ansell]

My impression was that this test file was intended to be a custom test. However, it seems to have mistakenly landed in core/src/test/resources/json-ld.org/ instead of core/src/test/resources/custom. The PR which included it was #168 by @hmottestad for background reference.

Could you move it to core/src/test/resources/custom as part of this PR?

[eroux]

Thanks for your comments, the test is back as a custom test

[coveralls]

Coverage increased (+0.07%) to 88.719% when pulling 227461c on BuddhistDigitalResourceCenter:fix-new into 5589124 on jsonld-java:master.

[eroux]

The error in travis-ci is a network error and, I think, can be ignored (nothing to do with this PR)

[ansell]

Thanks, the changes look good and the Travis-CI error retrieving from purl.org is intermittent (it worked for JDK8), and not likely related to JDK9 in that case given it worked previously. I will merge this in and release 0.11.0 today.

[eroux]

Thanks, great to see a new release!