is here someone who is familiar with the Inducks search or exists a help page where the functions of the search mask are explained in detail?
I would like to know which fields of this page are searched by the "keywords" search and what exactly is meant by "match any word", "match all words", "phrase" and "expanded" (see picture).
This does not seem to be as obvious as probably everyone would assume. For this purpose two examples with the story "The Secret of the Ice Sword":
1. in the English description the expression "various tribes there are slaves to a monstrous tyrant" occurs (see picture below). If you enter this into the search mask with "match any word", you get, as expected, a very long list of stories (especially because of "are", "to" and "a"). Under "match all words" this leads to exactly the story "The Secret of the Ice Sword". Under "phrase" it finds nothing and under "expanded" you get a long list, but a different one than under "match any word".
2. in the Italian description the word "Ululand" appears (see picture below). In the keyword search for "Ululand" the selections "match any word", "match all words" and "expanded" lead to no result and the selection "phrase" leads oddly only to the story "The Great Tournament of Argaar"?!?
I think the reason you don't find I TL 1411-BP with "Ululand" is that there are two descriptions in the same language (Italian) and the search query only search in one of the two descriptions. So a bug in the programming (it probably wasn't intended for a story to have several different descriptions in the same language).
I'm not 100% sure about this, but got the same result trying to search for other stories that have two different descriptions in the same language. Ex. I TL 2346-1. Searching for keywords in one of the two Italian descriptions work, but not for the other one.
Edit: and the reason you find I TL 1464-BP using "Ululand" and phrase-seach, is that there it's part of the word "Ululandsbyen" in the Norwegian description. It doesn't find the word even if it is used in one of the two Italian descriptions for that story.
You seem to be right. Under "phrase" every single word is searched for exactly the expression you are looking for. And if there is more than one description for a language, the "match any word", "match all words" and "phrase" search will only search the last description. This search mask should definitely get an update. But what exactly the setting "expanded" does (or should do) is still not clear to me.
Hi A few remarks: - the fields searched for are: appsummary, plotsummary, writsummary, artsummary, inksummary, creatorrefsummary, keywordsummary. The content in these fields is generated from other fields (see e.g. inducks.org/bolderbast/xe23.html) and include titles and descriptions in all languages. I'm not sure how this behaves when we have more than 1 description in a language. I think the software will pick the description in the most trusted source (e.g. a story index > an issue index). - the "expanded" query mode is described here: dev.mysql.com/doc/refman/8.0/en/fulltext-query-expansion.html - keywords of 1 (possibly 2) letters are discarded. We've had to include 3-letters words to handle Inducks abbreviations. - you can download the entire database as text files and see for yourself: inducks.org/inducks/isv/ - the above is assuming no bug. MySQL has many ways to search for keywords and it is not always simple to know what's best.
- I have added "votes" as an option in the sorting menu
- There was indeed a bug in the vote counts, they had not been updated for one year. Corrected. Since we're now counting thousands more votes from 2020, the ranking may change.
Hi A few remarks: - the fields searched for are: appsummary, plotsummary, writsummary, artsummary, inksummary, creatorrefsummary, keywordsummary. The content in these fields is generated from other fields (see e.g. inducks.org/bolderbast/xe23.html) and include titles and descriptions in all languages. I'm not sure how this behaves when we have more than 1 description in a language. I think the software will pick the description in the most trusted source (e.g. a story index > an issue index). - the "expanded" query mode is described here: dev.mysql.com/doc/refman/8.0/en/fulltext-query-expansion.html - keywords of 1 (possibly 2) letters are discarded. We've had to include 3-letters words to handle Inducks abbreviations. - you can download the entire database as text files and see for yourself: inducks.org/inducks/isv/ - the above is assuming no bug. MySQL has many ways to search for keywords and it is not always simple to know what's best.
- I have added "votes" as an option in the sorting menu
- There was indeed a bug in the vote counts, they had not been updated for one year. Corrected. Since we're now counting thousands more votes from 2020, the ranking may change.