Clean up treatment of column in text_epigraphy

[edstratford]

Now that the database is increasingly being managed from the front-end, there are some tweaks that will better make the data more resilient and open.

From now on, all 'section's will have a 'column' rows as a child, and all current children of the sections will fall under a column. In the vast majority of cases, this means that there will just be another row for each section. But the tablet renderer on the frontend already needs to deal with this, and so it is best if it is consistent throughout.

In conjunction, the column column will be retained, but all 0 values will be updated to 1, (though all rows that are type section or epigraphicUnit should be NULL -- and epigrpahicUnit rows should have side as null as well.)

Finally, some columns that are higher than 2 need to be reviewed. These are probably from different types of editions. Will need to review whether within the project, the columns are always numbered within a side, or if different numbering systems are possible.

[edstratford]

Jon will write the script to add column rows as child to all sections, and update parent_uuid of all current children to sections (lines and regions ... and anything (hopefully nothing) else, to the new uuid.

[edstratford]

Prof S will consult with harrison on doing a migration to delete column column and reinstate with value of 1 in all cases, then edit to NULL out ephigraphicUnits and sections, and restore columns > 1 from already saved Google sheet. This must be executed ahead of script by Jon.

[oaregithub]

Prof S totally messed up the database when adding the columns manually himself. Through confusion, he added multiple columns in some places and caused ruptures/reduplications in the object_on_tablet sequences, and -- when adding too many columns also disrupted good parent-child relationships in this position.