Data BUG (LHB-CA) Meayla/Tipcha indistinction

Page 1 of 1 (10 items)
This post has 9 Replies | 0 Followers

Posts 1896
Forum MVP
Reuben Helmuth | Forum Activity | Posted: Wed, Jan 16 2019 4:18 AM

All Meayla's are additionally tagged as Tipcha and all Tipcha's are additionally tagged as Meayla. This results in 35,747 errors. According to Eli, Helmut Richter's website was used as the reference, but apparently the Resolving Ambiguities section wasn't read (this section also covers the first issue I raised in the previously mentioned thread).

Unless I did something wrong, here's a search that finds the Meayla's that are erroneously tagged as Tipcha, and here's one that finds the converse.

Posts 1259
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Tue, Feb 19 2019 1:03 PM

HI Reuben.

Haven't forgotten these, it's just taken awhile to get to them. Thanks for reporting them. Not sure exactly when updates will be released, but I'm digging.

Also: I only have the ability to change the Cantillation resources (graphs, database, supplemental data) and not the Cantillation interactive, so issues will persist there until someone else can prioritize them. 

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 2465
Lee | Forum Activity | Replied: Tue, Feb 19 2019 5:41 PM

Hoping the bugs will be ironed out soon.

Posts 1896
Forum MVP
Reuben Helmuth | Forum Activity | Replied: Wed, Feb 20 2019 12:09 AM

Rick Brannan (Faithlife):
Haven't forgotten these, it's just taken awhile to get to them.

Thanks for letting me know, Rick! I was considering bumping one of these yesterday but chose to bump Workflow development requests instead. :-) 

Posts 1259
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Wed, Feb 20 2019 9:22 AM

Reuben & Lee — 

I want to follow up since I'm a bit out of my depth with Hebrew Cantillations. Here's what I'm planning to address the Meayla/Tipcha ambiguity:

My understanding is that presently, every member of the set Meayla+Tipcha is tagged as both Meayla and Tipcha. To correct this, my understanding is that if a visual word has Meayla, Tipcha, Athnah, and Sof Pasuq tagged, the Tipcha should be removed (so, Meayla and Sof Pasuq + Athnah remain). Otherwise, if a visual word has both Tipcha and Meayla, the Meayla should be removed and the Tipcha remain.

Is that your understanding as well?

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 1896
Forum MVP
Reuben Helmuth | Forum Activity | Replied: Wed, Feb 20 2019 12:47 PM

Rick Brannan (Faithlife):
every member of the set Meayla+Tipcha

I'm not sure exactly what you mean by "set"... The same symbol is either one or the other (never both), and yes, every occurrence of this mark is tagged with both names.

To correct this, if a visual word (with this cantillation mark) contains an emperor (Atnach or Sof Pasuq), then the mark is a Meayla. If the visual word does NOT have an emperor the mark is a Tipcha. It seems like your logic should work, but I wanted to emphasize that the key difference is the presence of an emperor in the same visual word.

Here are screenshots of the searches that should find the erroneous entries.

  

Posts 2465
Lee | Forum Activity | Replied: Wed, Feb 20 2019 12:50 PM

Rick and Reuben: If you're both in compatible time zones, it's best to set up a call.

Posts 1259
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Wed, Feb 20 2019 3:07 PM

Reuben Helmuth:
I'm not sure exactly what you mean by "set"... The same symbol is either one or the other (never both), and yes, every occurrence of this mark is tagged with both names.

I mean if you collect all of the 'true' meaylas and 'true' tipchas, that's the set, and everything in that set is double-tagged.

I think we're on the same page.

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 1259
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Mon, Apr 15 2019 11:49 AM

Hi Reuben.

I'm back working on this as we've been able to update the tooling (there were authentication issues when connecting with necessary servers).

I have a version of the data that returns 44 hits for your first error query, and no hits for your second error query. I believe the 44 hits are in multi-text-segment instances that I'll need to think a little more about, but think decent progress has been made on this one.

I'll respond to others as I'm able. If all goes well, I hope this data will make it into the April 30 resource update (cross your fingers)

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 1896
Forum MVP
Reuben Helmuth | Forum Activity | Replied: Tue, Apr 16 2019 5:20 AM

Rick Brannan (Faithlife):
If all goes well, I hope this data will make it into the April 30 resource update (cross your fingers)

Awesome. 🤞🏽

Page 1 of 1 (10 items) | RSS