Parsing Behaviour in Source Citations

[David Crouch]

Hi,

I have a source template, as follows:

<[AUTHOR], >"[RECORD TITLE]," [PAGES], [ENTRY]; imaged, "Medway Council Heritage Services catalogues," [RECORD TYPE], [ITAL:]Medway Council Heritage Services[:ITAL] ([URL] : accessed [ACCESS DATE]), [FILE REFERENCE]<; citing  <[REPOSITORY REFERENCE] >< held by the [REPOSITORY]>>.

And here is an example of how I used it:

{AUTHOR}

St. Margaret of Antioch Parish (Darenth)

[RECORD TITLE]

Register of Burials [1941-1992]

N.B. I interpret the guidance from Elizabeth Shown Mills that I put the dates in rounded brackets if that is actually part of the title and I put the dates in square brackets when I am annotating to title to help the reader to select the correct register from a series of registers.

The TMG Preview of the citation is:

 St. Margaret of Antioch Parish (Darenth), "Register of Burials [1941-1992]," p. 52, entry no. 416, George Odel Levings, 03 Jan 1959; imaged, "Medway Council Heritage Services catalogues," database and images, Medway Council Heritage Services (https://cityark.medway.gov.uk/Details/archive/110003192 : accessed 22 Aug 2022), PDF file P109-1E-01.pdf, p. 28 of 54; citing  P109/1E/1  held by the Medway Archives Centre.

When I produce a website using Second Site, the source entry is:

[S2249] St. Margaret of Antioch Parish (Darenth), "Register of Burials [1941-1992]," p. 52, entry no. 416, George Odel Levings, 03 Jan 1959; imaged, "Medway Council Heritage Services catalogues," database and images, Medway Council Heritage Services (https://cityark.medway.gov.uk/Details/archive/110003192 : accessed 22 Aug 2022), PDF file P109-1E-01.pdf, p. 28 of 54; citing P109/1E/1 held by the Medway Archives Centre.

which looks perfectly reasonable.

When I convert the same database to GEDCOM, I get the following source entry:

0 @S2249@ SOUR
1 REFN 2249
1 TITL St. Margaret of Antioch Parish (Darenth), "Register of Burials unknown 1941-1992," p. 52, entry no. 416, George Odel Levings, 03 Jan 1959; imaged, "Medway Council Heritage Services catalogues," database and images, <i>Medway Council Heritag
2 CONC e Services</i> (https://cityark.medway.gov.uk/Details/archive/110003192 : accessed 22 Aug 2022), PDF file P109-1E-01.pdf, p. 28 of 54; citing  P109/1E/1 held by the Medway Archives Centre.
1 REPO @R42@
2 CALN P109/1E/1
1 OBJE @O9195@
1 OBJE @O9196@

I am having two problems:

1. The square brackets around the dates in the register title have been replaced with "unknown 1941-1992", and

2. Despite having the Printer Code for Italics set to "Remove printer code", I am getting <i> </i> around the website name.

I have tried to figure out both of these behaviours without success so I am hoping someone can provide some guidance.

Thanks!

David

[John Cardinal]

David,

 

In TMG, [ and ] have a special meaning. TMG is ignoring it in this case, but it won’t always do so, and other programs may or may respond the same way. I suggest you escape the special characters, i.e., replace “Register of Burials [1941-1992]” with “Register of Burials \[1941-1992\]”. The backslashes tell the parser to ignore the special meaning of [ and ].

 

Regarding the removal of the italic codes, I will look into it.

 

John Cardinal

 

Products: ORA, GedSite, Gedcom Publisher, Second Site, TMG Utility, TMG to GEDCOM

Web Hosting for Genealogists: Family History Hosting

[David Crouch]

Hi John,

Thank you for the quick reply.  Is there an easy way to bulk search and replace "[" with "\[" in the Record Title field of the Source Definition?  Otherwise, I will have to manually trawl through several thousand source definitions.  Is this something that the TMG Utility could perform?

Thanks!

David

[John Cardinal]

David,

 

You can probably use the Find and Replace feature in TMG Utility. I strongly suggest that you do a dry run first where you have the “Log Only” option checked. If you use “[:CR:]” or some other printer code in a source element value, that will get broken. So, review the list of possible changes to see if there are any issues. If there are, you can either:

 

1. Use the Find and Replace feature and then manually cleanup the sources that it breaks,

 

or

 

2. Do all the edits manually, but use the Find and Replace log as a guide to which sources need changing.

 

This should get you started:

 

TMG Utility v7.5.0

Copyright (C) 1999-2023 by John Cardinal. All Rights Reserved. Do not redistribute.

 

Function=Find and Replace

Project=test_ (C:\TMG Projects\test\test_)

Dataset=test / My Data Set

Log Only=True

Started at 16:46:18

 

Options:

  Field: Source Subfields

  Find what: (^|[^\\])(\[|\])

  Replace with: $1\$2

  Contents must match Find What exactly: Unchecked

  Use Pattern Matching: Checked

  Match case: Unchecked

  Find Whole Word Only: <na>

 

-------------

Source#, Info

14, 'Mackenzie, P. [Paul] W.', 'Mackenzie, P.\[Paul\] W.'

 

Finished at 16:46:18

Elapsed Time=0 seconds

Records=22

Changed fields=1

 

Copy/paste the “Find what” and “Replace with” values. They are not for the feint of heart.

[David Crouch]

Hi John,

Thank you for your e-mail and helpful suggestions.  I am glad that you provided the template for the find and replace.  It took me a long time to understand what you were doing.  I have run the changes in "Log Only" mode and generated 15 pages of output.  Everything looks fine but I will look it over again tomorrow morning with fresh eyes before removing the "Log Only" mode.

Thank you for all your assistance.

Be Well,

David

[John Cardinal]

David,

 

A handy trick for understanding a regular expression (regex) written by someone else is to copy and paste it into https://regex101.com. It provides an explanation pane:

 

 

For a regex used by TMG Utility, you should set the “Flavor” to “.NET C#)”. TMGU predates .NET, so that’s a bit of a stretch, but the regex used by TMGU follows the same rules as in .NET (C#).

 

For a regex used in JavaScript, set the flavor to “ECMAScript (JavaScript)”. That’s the right choice for working with a regex used in ORA.

 

https://regex101.com does more than just explain, of course.

 

John Cardinal

 

Products: ORA, GedSite, Gedcom Publisher, Second Site, TMG Utility, TMG to GEDCOM

Web Hosting for Genealogists: Family History Hosting

 

From: t2...@googlegroups.com <t2...@googlegroups.com> On Behalf Of David Crouch
Sent: Monday, October 16, 2023 7:56 PM
To: TMG to GEDCOM <t2...@googlegroups.com>
Subject: Re: [T2G] Parsing Behaviour in Source Citations

 

Hi John,

 

Thank you for your e-mail and helpful suggestions.  I am glad that you provided the template for the find and replace.  It took me a long time to understand what you were doing.  I have run the changes in "Log Only" mode and generated 15 pages of output.  Everything looks fine but I will look it over again tomorrow morning with fresh eyes before removing the "Log Only" mode.