Adjustments needed to spines

FrankensteinVariorum / fv-collation

first-stage collation processing in the Frankenstein Variorum Project. For post processing and Variorum development, see our GitHub organization: https://github.com/FrankensteinVariorum

https://frankensteinvariorum.github.io/fv-collation/

GNU Affero General Public License v3.0

9 stars 2 forks source link

Adjustments needed to spines #58

Closed raffazizzi closed 5 years ago

raffazizzi commented 5 years ago

Assessment based on spine_C10

do not use <ref>s
currently there aren't any multiple <ptr>s within an <rdg>: those are needed when there are multiple <seg>s that need to be referred together. Is this still missing from the processing?
Related to the above, we had specific IDs for s split over hierarchies and they looked like this: C10_app2-f1818__Pt1. This may be obsolete now (given @part) and not needed for processing

PS this is how spine_C10.xml looked in CDMX

ebeshero commented 5 years ago

@raffazizzi Here's a spinal readjustment for C-10: https://github.com/PghFrankenstein/Pittsburgh_Frankenstein/blob/Text_Processing/collateXPrep/standoff_Spine/spine_C10-recon.xml

(Note: I renamed the file in case we need to debug it. The old one is still in place at spine_C10.xml) A number of <app> elements disappeared this round because in the up-conversion process to make <seg> hotspots, I eliminated all that came from invariant <app>s. So this spine should be more streamlined.

ebeshero commented 5 years ago

@raffazizzi All of my C10 edition files are up-converted and ready for pointing. They are in the same directory as before for Mexico City: https://github.com/PghFrankenstein/Pittsburgh_Frankenstein/tree/Text_Processing/collateXPrep/bridge-P5-C10

ebeshero commented 5 years ago

@raffazizzi However, at the moment we still only have C10. We'll figure the rest out for the full novel, but not in time for this talk.

ebeshero commented 5 years ago

@raffazizzi Well! I figured out what I was doing wrong, reprocessed everything, set the proper <seg> markers and upconverted the entire novel! w00t!!!!

SO, that changes things a bit. I reconstructed the spine files for all 33 "chunks" here: https://github.com/PghFrankenstein/Pittsburgh_Frankenstein/tree/Text_Processing/collateXPrep/standoff_Spine

The old spine from July 23 is in a different directory now in case we want to look at it (orig_standoff_Spine). As I mentioned earlier, the new spine files produced in this later stage will not contain the invariant <app> elements that the first spine did.

The complete novel edition files are now ready (yaaayyyy!) here: https://github.com/PghFrankenstein/Pittsburgh_Frankenstein/tree/Text_Processing/collateXPrep/bridge-P5

Take a look and let's see if this spine will work! Currently the spine I generated is pointing at my local versions of the S-GA files, so I think you now run your script to build the S-GA pointers...I hope it works!!!

raffazizzi commented 5 years ago

@ebeshero Thanks for all this work and hurrah for figuring the <seg>s out!

The links in the spine are incorrect, but it should be simple to fix in the processing. Example from C21:

<rdg wit="#f1818"><ptr target="https://raw.githubusercontent.com/PghFrankenstein/Pittsburgh_Frankenstein/Text_Processing/collateXPrep/bridge-P5-C10/#C21_app1-f1818__Pt1"/>

should be <rdg wit="#f1818"><ptr target="https://raw.githubusercontent.com/PghFrankenstein/Pittsburgh_Frankenstein/Text_Processing/collateXPrep/bridge-P5/P5-f1818_C22.xml#C21_app1-f1818__Pt1"/>

ebeshero commented 5 years ago

Thanks for catching that, @raffazizzi !

ebeshero commented 5 years ago

Closing here because we're now generating spines in the fv-postProcessing repo, and will check the status of spines again there.