They want to add additional Administrative Division Places that have the same complex hierarchy as one that was already define once for an adjacent Place. So they are wondering if there is a way to clone a Place hierarchy? Or do they have to do it manually all over again?
They didn’t give an example. So here’s one based on my own home county in western Pennsylvania wuth 16 present-day townships. Those 16 have very similar enclosing county place trees. (5 will be identical, 11 more will have a slightly different variant)
The Details
My county is the result of several reorganizations over time. And the townships would have been “enclosed” by 4 different counties and 2 nations during the period that I’d be recording Events.
Until 1784, it was in Six Nations of the Iroquois Confederacy (under the Shawnee). Then its territory was "purchased by Pennsylvania to expand the fledgling USA. Then from 1784 to 1849, the originally organized county was subdivided 3 times and amalgamated once.
So my home township (in the Northern half of the county) of Hickory is now in Lawrence county, formerly Mercer county, formerly, Allegheny county, formerly Westmoreland county, formerly the Shawnee nation.
Or:
Before 1784, Shawnee nation
From 1784 to 1788, Westmoreland county (Pennsylvania, USA)
From 1788 to 1800, Allegheny county
From 1800 to 1849, Mercer county
From 1849, Lawrence county.
When I add the other 4 townships in the northern side, they would have identical enclosing hierarchies. (The other southern 11 townships would be similar except with Beaver county instead of Mercer county.
So after the 1st, I’d want to duplicate townships 5 times, change “Mercer” in the last one and duplicate that 10 more times.
They can semi‑do it via CSV.
Do you remember my old project with CSV import way‑back‑then?
when we talked about OpenRefine etc. as a “data washer”
I cloned the hierarchy for many entries by using Excel and auto‑generating place IDs [Pxx‑xxx‑xxxx], based on the Norwegian “Land” – “Fylke” – “Kommune” digits.
But this had no information in text format for each entity, just the hierarchal structure…
I managed to import 86000 and some place entries before the import stopped responding in gramps 5.x.x, 300K did not work.
This was after the “csv fix” was made.
Just a tip if they have a lot of places.
PS: it only worked way‑back‑then via the import gramplet — I don’t know how it works today, just sharing the tip.
Thanks for the response! Using CSV with the Import Text gramplet was my first impulse too. (Although the other information you mention about the weirdness is also interesting/concerning. And I did/do not recall the OpenRefine data-washer example. Time for some research and experiments.)
The Places CSV support even includes the Date header. And it looks like you should be able to stack the Enclosed_by entries by using the Square brackets for the ID too.
Place
place - a reference to this place
title - title of place
name - name of place
type - type of place (eg, City, County, State, etc.)
latitude - latitude of place
longitude - longitude of place
code - postal code, etc.
enclosed_by - the reference to another place > that encloses this one
date - date that the enclosed_by place was in effect
But the tough part seems to be the “avoiding rework” side of the “cloning” idea. It is and obvious timesaver to compose the Place creation CSV data to do all 16 townships at one time. But the efficency in singleton cloning requires being able copy the existing hierarchy.
The Place view mode export does not include a Enclosing Place column. Even if it did, it wouldn’t handle more than 1 entry nor their couplings with date ranges. And exporting a selected Place to XML is not a feature
(Even when using the Isotammi addon XML exporter that supports a full Place tree export. That option is not available in the built-in XML exporter.)
I suspect that it will require writing a SuperTool script to do the cloning. sigh.
The script could be similar to the examples by @PLegoux to list the Enclosing places to a table. So you could use the Copy button in its GUI to clipboard the data to paste into the Import Text gramplet.
Or script could do the whole cloning. Hmmm. Might even be able to spawn a New Place editor dialog with the Place Type and Enclosing Place data pre-filled.
It is just as easy (and faster) to just manually add the additional places using the clipboard to hold the enclosing place records with the date information. These can be easily dropped into the new records.
The user (and others) may be interested in the Atlas of Historical County Boundaries. It has recently experienced a major change which I discovered today.
The site displays the changes in the state and county boundaries through time. starting in 1629 Massachusetts. On each state specific page there is a PDF available that tracks each county’s changes over time with date information and name changes. It overlays these changes over the county as it exists today. The maps do not include any community information but can often be gleaned from the pages by tracking sections of the existing county with changes in the overlays.
As I said the site recently changed. What no longer exists on each state’s page are the changes over time in a text format. The text format was often enough for me to understand the changes. The PDF maps became useful when a county changed from more than one other county. I added the text information to each county record in Gramps as an easy reference.
NOTE: I had placed a URL link on each state’s Gramps record. Checking it and expecting a need to update the links I found that they still work. So until someone cleans the servers both versions of each state’s page exists.
Yes, that works if the fix from Gramps 5.x is implemented in newer versions, that was how I got my hierarchy of Norwegian places correctly into Gramps back then…
But I do not know if it works when importing with the text import from the views, I have never tried that, only the full csv import gramplet.
And I got multiple lowest level places in without problems as long as the square brackets was used…
