Upgrade rdf-toolkit.jar to version 1.14.2 to suppress NamedIndividual assertions. Fixes #774.

[rjyounes]

@uscholdm Please test this in two ways and let me know if you have any questions or problems:

1. Run manually. Note that the serialization files have moved from tools to tools/serializer. Then revert the ttl files to test the pre-commit hook (step 2).
2. Copy the modified pre-commit hook into .git/hooks, make a small change so that a commit is executed, and make sure the pre-commit hook removes the NamedIndividual assertions in the ttl files.

[uscholdm]

I’ll do test 1. I don’t like to use the precommit hook because it often obscures differences. E.g. if a few things change names, they get tossed to another place in the file. If I addd a new feature with a few things bunched together, they get split up all over the place making it hard to review. In these cases I prefer to do a commit before serializing and then an immediate commit after that says serialize only. 90% of the time I commit after serializing. M.

[rjyounes]

Why do you want to do a commit before serializing? That means if someone diffs that commit against the preceding, they will get noise. That defeats the purpose.

[uscholdm]

Why do you want to do a commit before serializing?

I don't know how to explain it differently in a way that will be easier to understand.

That means if someone diffs that commit against the preceding, they will get noise.

No, it makes it easier to see what the actual changes were in cases where serializing jumbles things all up, which is quite common. The idea is to do two commits one right after the other. The first commit makes it easy to see what the actual differences are, the second commit is for all future development. The second commit (serialize only) makes it obvious how much the serializer jumbles things up.