HIghlighting - 1 problem and 1 question
First the possible problem/bug - I am trying to figure out a Search in beta 3 that now appears possible. I want to count all occurrence of a custom palette being used in a resource (or collection, etc.). I first tried this in beta 2.
Here is my example: I developed a custom palette, then used it three times to mark text in a specific volume of the Pulpit Commentary (you can see the highlight "Custom Palette" below). . Then I searched (Basic) that resource for * in my custom palette, which was available as an option in the search drop down options. It runs for over 10 minutes, then gives me "22 occurrences". That doesn't match anything! It's not number of letters highlighted, I don't know what 22 is. So why is it so slow, is there a more correct way to find occurrences, and why 22 occurrences?
Second, my question - in image #1 below, where you can see my custom palette "Test Palette", I searched for "The". It returns 2 results, but it only shows one in the results. Now they are close together, is this correct? I added another one much further down, and it reports 3 results and then shows 2 results. So does it decide only to return 1 result if they are close together? They are a few sentences apart.
Comments
-
Dominick Sela said:
Then I searched (Basic) that resource for * in my custom palette, which was available as an option in the search drop down options. It runs for over 10 minutes, then gives me "22 occurrences".
* searches are notoriously unoptimised and best avoided. In a single resource on my desktop it can take easily 7 minutes - see http://community.logos.com/forums/p/14925/114066.aspx. I'm doing a search of Church Fathers collection - just finished with 18 results for 18 highlighted words after 9 minutes (Logos4 has a word oriented search engine).
Dominick Sela said:in image #1 below, where you can see my custom palette "Test Palette", I searched for "The". It returns 2 results, but it only shows one in the results. Now they are close together, is this correct? I added another one much further down, and it reports 3 results and then shows 2 results. So does it decide only to return 1 result if they are close together? They are a few sentences apart.
I ran a similar search for "This is the" in a custom style and got correct results which were well spaced.
Don't know why the results were split into two sections, but it was correct!
When I highlighted "this we believe" immediately following the first highlight above and searched for "this" I again got correct results in two sections. If i search in "all text" the results are still split the same way. Therefore not unique to style searching!
Dave
===Windows 11 & Android 13
0 -
Dominick Sela said:
Second, my question - in image #1 below, where you can see my custom palette "Test Palette", I searched for "The". It returns 2 results, but it only shows one in the results. Now they are close together, is this correct? I added another one much further down, and it reports 3 results and then shows 2 results. So does it decide only to return 1 result if they are close together? They are a few sentences apart.
I'm getting the same result. Even though the result text does get truncated, usually it displays enough of the text to show all the found search terms, as in Dave's example. We'll look into it.
0 -
Dominick Sela said:
Second, my question - in image #1 below, where you can see my custom palette "Test Palette", I searched for "The". It returns 2 results, but it only shows one in the results. Now they are close together, is this correct? I added another one much further down, and it reports 3 results and then shows 2 results. So does it decide only to return 1 result if they are close together? They are a few sentences apart.
Basic Search groups its results by "article" (which in most resources is a chapter or logical section within the work). The "2x" by the result in your screenshot shows that there are two hits within that one article. The hits are far enough apart that they can't both be shown within the preview text. This is also where the "2 results in 1 article" number in the section header comes from.
Adding more markup "much further down" most likely put the highlight in a new article, so the search result was shown separately from the two in the first article.
The resource in Dave's screenshot seems highly unusual in that there is a new article almost every paragraph. (You can see this by changing the resource navigation type to "Article" then clicking the Up/Down arrows.)
0 -
Dave Hooton said:
Thanks for that screenshot. I found a typo in it and reported it. "tiffs is the true faith"
No, tiffs are a graphic file format, not the true faith. [;)]
0 -
Rosie Perera said:
No, tiffs are a graphic file format, not the true faith.
The Lord knew about graphics formats[;)]
The "Early Church Fathers" series contains 32 tiffs - reported by email.
Dave
===Windows 11 & Android 13
0 -
Dave Hooton said:
The "Early Church Fathers" series contains 32 tiffs - reported by email.
These are apparently in the print versions, as a web search for 'tiffs' and 'tiff-filled' generates hits that link to excerpts from the Early Church Fathers.
0 -
Melissa Snyder said:Dave Hooton said:
The "Early Church Fathers" series contains 32 tiffs - reported by email.
These are apparently in the print versions, as a web search for 'tiffs' and 'tiff-filled' generates hits that link to excerpts from the Early Church Fathers.
They're not in the print versions, just in incorrectly OCR'ed online versions. Here's one example where it's incorrect in one place and correct in another. After searching for "tiff-filled" and finding the first hit, I then searched for the rest of the sentence up to (but not including) "tiff-filled" and found what it's supposed to be.
Similar searches could probably find the correct word wherever "tiff" was used. I'm guessing sacredtexts.com is a more reliable source for these texts than wherever Logos got them.
0
