How to computer the number of Unresolved notes with a ToDo tag across multiple projects with Claude AI

Question

Our team has at least 16 Paratext projects. We need to know the number of Unresolved notes that have the ToDo tag for each book of the NT for each Paratext project. I am working with Claude Cowork to try to get this info and put it into a spreadsheet. I can get the info myself by opening a Notes List window on every Paratext  project and set the filters appropriately. That gets me the numbers today and I don't mind doing it that way one time. But I am going to need these numbers again as we track the teams' progress, so I want Claude to do it for me and be able to do it again. It has been unable to come up with numbers that match what I see in Paratext. It is having a hard time reverse engineering the data structures. So here are its questions as it put them to me: 
The specific questions I'd ask the developers:

1. When a user filters for "Unresolved To Do" notes, how does Paratext determine the current tag of each thread? Is it: (a) global accumulation of all TagAdded/TagRemoved across all users' files, (b) the root comment's tag only, or (c) something else?
2. How does Paratext determine whether a thread is unresolved? Root comment's Status field, any user's Status, or a dedicated thread-level field?
3. Does the filter show threads from all users or only threads the logged-in user has participated in?
4. Does a thread with a deleted root comment still appear in the filter?

Thanks for any help you can give me on this. 

John Nystrom

Aitape West Translation Programme

SIL PNG 

Answer 1

Since it's talking about a "root comment", I'm guessing this is already being done, but just in case, before anything else can be calculated, the comments from all users with the same thread ID need to be gathered and sorted by date. This is Paratext's definition of a project note thread. After that ordered list of comments is created, then you can calculate the other stuff:

1. The current tag is determined by looking at the <TagAdded> and <TagRemoved> elements while looping through the comments. It was originally designed to allow multiple tags on a note, but I don't think that is actually possible through the UI, so you can probably just look at the last comment that has a <TagAdded> or <TagRemoved> element.
2. There is no thread-level data. Threads only exist by having comments with the same thread ID. To determine if a thread is unresolved, the last comment with a non-empty <Status> element needs to be anything but "deleted" (usually "todo").
3. It's all users.
4. I'm not sure what that means. If it means that the first comment has a <Status> of "deleted" and there are no following comments that change the status, then no, that thread would be considered resolved.

Comments

Also, beware that the dates are in a format that includes an offset (i.e. they are in local-time). They are not in UTC. This means that the parsing of the date/time for each comment needs to take that offset into account to make sure sorting is done properly.

Answer 2

Maybe this is simpler than it looks. You can open a notes list for multiple projects.
When Notes List comes up, select all 16 of your projects. Then set your filter to unresolved and todo.

I get 353 Unresolved To Do notes in aArp, aMal, and aSer

Comments

Thank you Steve. Yes, we can do what you show here, and do it all the time. But the goal for my current purpose is to get separate counts for each book in the NT for each of the projects so we can see who is caught up and who is behind, and to be able to get those counts in a spreadsheet that can be updated.

---

As much as I love solving all problems externally using (AI-written) scripts to do things PT was never intended to do...

I think Steve might be onto something. Why not create a Layout with the notes of all 16 projects open in 16 separate windows, then save that as a new Layout. You can load that Layout when you need to see the totals, then load your standard Layout the rest of the time.

---

Mnjames,

I actually do have a Paratext layout that does exactly that. We certainly can look at each project's Notes List and see how many notes are outstanding for the current book. But we need that information for all books as we anticipate producing Mini-Bibles in these languages. We need this info in a form that gives us an overview of the whole team so we can see the big picture and the details. It's not practical for me or another team member to do that exercise for each book and manually type the number into a spreadsheet so we can see the big picture.

Paratext once had a Big Picture view that, if I remember correctly, provided this information, but it was removed several years ago. I loved it and felt our team needed it. The Paratext "Owner" at the time hated it and asked me if we used it often enough to justify having it stay in the UI. I gave in and said "no, go ahead and take it out if it's a problem for the development team," but I now regret saying that. This is the sort of thing administrators of cluster projects need. 

---

Thanks to a suggestion from @mnjames, I now have two Python scripts that accomplish this task and create a spreadsheet. I can run those scripts again at any time to update the spreadsheet. If anybody is interested in these scripts, let me know. I'll copy in the cells of the spreadsheet below. 

By looking at the numbers, we can see who is close to completing their notes on which books and who has a long way to go. This can help us answer questions like, "How long is it going to take to clear the backlog of project notes so we can publish Mini-Bibles for these languages?" and "Who is likely to have which books ready to publish if we do our Mini-Bibles all at the same time on a certain date?" 

Thanks to all who helped me think through this, and to those who encouraged me to learn to use Claude functions I can't do in Paratext itself. 

JN

Book	aArp	aArpBT	aSno	aMal	bBar	bG1BT	bPou	bSum	cGoi	cOGBT	cRbr	cWol	dOpi	dOLBT	Total
MAT	8	10	5	6	89	29	123	617	27	11	39	12	110	320	1406
MRK		14	2	11	59	48	50	131	15	5	24	20	33	33	445
LUK	1	7	39	32	43	37	39	66	112	38	732	126	34	34	1340
JHN	1	9		6	132	72	151	530	89	42	128	107	162	240	1669
ACT	1	10	196	129	253	252	251	1236	280	252	830	651	250	250	4841
ROM	1	92													93
1CO	1														1
2CO	1							1							2
GAL	7	1	5	4	7	8	5	189	6	7	9	8	10	10	276
EPH		2	5	3	6	5	6	98		1	1	6	5	5	143
PHP		1		2		2	1	158				5	4	4	177
COL					6	4	8	67					34	32	151
1TH			3		1	1	1	84	1		1	2	3	3	100
2TH			1	1				55	1				3	3	64
1TI			4	2	4	3	1	15	6	3	4	7	3	3	55
2TI			11	1	1	1	1	19	6	1	4	4	1	1	51
TIT					1	1		7	3	1	1	2			16
PHM			1			1					1	1			4
HEB	15	5													20
JAS		1	1		16	12	5	50			2		5	5	97
1PE			2	2	3	4	4	848	6	6	7	5	4	4	895
2PE			3	2	4	5	7	373	6	5	3	4	4	4	420
1JN			2		1	1	1	21	2				1	1	30
2JN			4		1	1	1	3					1	1	12
3JN			1					1							2
JUD			4	4	5	5	2	199	4	3		2	4	4	236
REV	1	10	3	1	6	3	9	7	14	3	5	4	6	6	78
Total	37	162	292	206	638	495	666	4775	578	378	1791	966	677	963	12624