Repo naming conventions

[ebeshero]

@djbpitt @cmsmcq As we have begun timing tests on collections of documents, these should be represented and provided for in the repo naming conventions document. Do these collection files need to be all renamed? (I hope not). I'm going to attempt to model a good practice with the Frankenstein files for us to evaluate so that we can decide the best policy for the repo.

[ebeshero]

In particular, should collection files (such as Brown) be stored in the centralized input directory?

[ebeshero]

With this commit d71949f I have created the "input" directory as currently specified in our Repo conventions. I've also moved my XSLT files into new positions when they're either adaptations of David's XSLT3 function, or other kinds of experiments, and I've modified functions where necessary (when processing the small collection of Frankenstein files) to point to their new locations in "input".

What's currently missing: the collection I promised of all the hundreds of Frankenstein files across all collation units. Since this is to be the canonical collection, and I noted some errors I now want to correct in my collation process (prior to feeding to collateX), I want to run that collation again and send fresh output here. That'll happen later today--it takes some hours for the collation of the whole novel to complete!

[ebeshero]

@cmsmcq 's recommendations via e-mail:
"My proposals:

1 raising/input/Frankenstein should contain the files now in raising/frankenData/bridge-P3-C10 (and any others)

2 More generally, each directory in raising/input/* may contain multiple files, which should have informative names.

3 Target files showing what we are trying to (re-) produce from the flattened input should be EITHER

(a) in a parallel raising/targets directory tree (so the target for raising/input/Frankenstein/P3-f1818_C10.xml
is targets/input/Frankenstein/P3-f1818_C10.xml

OR

(b) in the same directory as the input, with the prefix “target.” (so the target for C10 of 1818 would be raising/input/Frankenstein/target.P3-f1818_C10.xml

Having written them both out, I have a mild preference for (b), unless one of you would prefer something else.

4 If the two of you prefer, we can also stipulate that flattened files have the prefix ‘flattened.’ or ‘flat.’.

5 Brown input belongs in raising/input/Brown"

[ebeshero]

@cmsmcq 's recommendations look sensible to me. I like the idea of a special set of targets when processing collections. One modification I’d propose to that is creating a targets directory named to pair with the collection being processed. So if I have a collection named c10-coll (as I do now), I’d want to see a target-c10-coll directory to match it. Does that work? The files inside the new target-c10-coll directory could the preserve their edition-specific names.

[ebeshero]

I'll generate that now, just so we can see how it looks in the repo.
And how about I move the Brown files now, too?

[ebeshero]

@cmsmcq @djbpitt Heads up! I've just moved the brown directory into the new canonical input directory. The files inside it are otherwise unaltered.