PDA

View Full Version : WTM numeral search issue



David Kummerow
05-20-2008, 07:46 PM
I've encountered the following issue when attempting to search on numerals in the WTM:

1. When accent-sensitive searching is enabled, a search for *@u* results in nothing being retreived. Further, WTM lists numerals in the browse window as @a, eg רביעי in Gen 1:19 is listed as y[iybir>@amsa+SxxxExHxNxZ9Rx.

2. When accent-sensitive searching is disabled, a search for *@u* results in a successful retrieval of numerals. Further, WTM now lists numerals in the browse windows as @u, eg רביעי in Gen 1:19 is listed as y[iybir>@uomsa+SxxxExHxNxRx.

I'm wondering if others have the same issue? Or if I'm doing something wrong?

Thanks,
David Kummerow.

Philip Brown
05-20-2008, 10:09 PM
Hi, David,

With vowel-sensitive + accent-sensitive searching, I get all the correct hits.
With vowel-sensitive only, all the correct hits.
With accent-sensitive only (vowel OFF), I get an error message that the search item .*@u* doesn't exist. Not sure why.

David Kummerow
05-20-2008, 10:35 PM
Hi Philip,

Thanks for testing. I hadn't checked the correlation with vowel-sensitive searching, and my results are the same as yours. To correct what I wrote above:

1. When accent-sensitive searching is enabled along with vowel-sensitive searching disabled, a search for *@u* results in nothing being retreived. Further, WTM lists numerals in the browse window as @a, eg רביעי in Gen 1:19 is listed as y[iybir>@amsa+SxxxExHxNxZ9Rx.

2. When accent-sensitive searching is disabled when vowel-sensitive searching is also disabled, a search for *@u* results in a successful retrieval of numerals. Further, WTM now lists numerals in the browse windows as @u, eg רביעי in Gen 1:19 is listed as y[iybir>@uomsa+SxxxExHxNxRx.

Is this what you get too?

Thanks,
David.

David Kummerow
05-21-2008, 09:20 PM
I've encountered further issues relating to combinations of vowel-sensitive and accent-sensitive searching:

1. If accent-sensitive searching is enabled and vowel-sensitive searching is enabled, when hovering over a word in WTT or WTM accent information is displayed, eg the first two words of Gen 1:1:

בְּ particle preposition רֵאשִׁית noun common feminine singular absolute accent:Tipha, Majela, or Tarha

ברא verb qal perfect 3rd person masculine singular homonym 1 accent:Munah

2. However, if accent-sensitive searching is enabled and vowel-sensitive searching is disabled, when hovering over verbs accent information is not displayed but for non-verbs it still is displayed (HALOT data on verbs is also lost), eg the first two words of Gen 1:1:

בְּ particle preposition רֵאשִׁית noun common feminine singular absolute accent:Tipha, Majela, or Tarha

ברא verb qal perfect 3rd person masculine singular


What might I be doing wrong to cause inconsistent retrieval of data? Is anyone else able to reproduce the above?

Thanks,
David.

Philip Brown
05-22-2008, 08:23 AM
Hi, David,

I suspect that this is a programming issue. Since no masoretic texts exist, that I know of, without vowel-pointing but with accents, the programmer's assumption was that if a person was concerned about the accents, they would also be concerned about the vowel points.

Perhaps linking the two toggles so that anytime accent sensitive searching is turned on, vowel-point sensitive searching would also be turned on might make that connection explicit and more transparent.

What do you think?

David Kummerow
05-22-2008, 09:09 AM
Yes, I think that would be a good idea. I few things are quite stange though, especially the fact that non-verbs still display the correct data and that depending on what is checked numerals appear tagged as either @u or @a.

Regards,
David.

MBushell
05-28-2008, 05:10 PM
Hi David and Philip,
This is now fixed and I posted a new Hebrew database update to take care of it. There are actually 4 separate Hebrew databases corresponding to the four combinations of accents on and off and vowels on and off. The accents only on combination was being built from an earlier edition fo the WTM database, before they added the numeral tags apparently. I guess few people search on accents with volwel point sensitivity turned off. Otherwise this would have been caught.
God bless,
Mike

David Kummerow
05-28-2008, 07:29 PM
Hi Mike,

Many thanks for the fix! I can confirm that accent and HALOT display on verbs as mentioned above is repaired. A weird glitch is still happening with adjectives. I'll explain the steps with some screen shots:

Step 1. Start with Gen 1:19 with both accent- and vowel-sensitive search disabled. Numerals are tagged with u.

http://img146.imageshack.us/img146/2489/72990338ji8.png


Step 2. Enable vowel-sensitive search and accent-searching -- in that order. Numerals are tagged with u.

http://img146.imageshack.us/img146/4761/35275428dp2.png


Step 3. Disable vowel-sensitive searching, leaving accent-sensitive search enabled. Numerals are tagged with u.

http://img232.imageshack.us/img232/7128/69842681nw3.png


Step 4. Back to start, with vowel-sensitive and accent-sensitive search disabled. See image from step 1 (I couldn't repost it as it seems posts are limited to four images only).


Step 5. Now enable accent-sensitive searching only. Adjectives are not tagged with u but with a.

http://img98.imageshack.us/img98/4224/78352738tb0.png


Thanks for looking into this.

Thanks,
David.

David Kummerow
05-29-2008, 09:17 PM
Issue fixed as of the latest update posted today.

Thanks,
david.