I have a problem with the substring search in the HTML5 output. I am using RH2019.0.12 new UI, but the problem also occured with update 11. The search settings for my output are as follows:
Use topic description = true
Enable auto complete = true
Auto correct search query = false
Show definitions from glossary = true
Enable substring search = true
Add special characters for search = false
Generate XML sitemap = false
In my output, I have several topics containing the word "Primäreinzelkosten". If I search for "primär", I only get results containing exactly the word "primär", but not the "primäreinzelkosten". If I search for "primä", I get no results at all. If I search for "prim", I get all topics containing "primär", "primäreinzelkosten" and "primrose" or whatever. So the substring search seems to have a problem with the "ä". The substring search works as expected for example when searching for topics containing the word "Kostenblock". They are found whether I search for "koste", "kosten", or "kostenb".
My project language is set to German, my output language is also set to German. I tried outputs with the following Encoding settings: Default Encoding, Westeurope(ISO), UTF-16, UTF-8 without BOM. I also tried the outputs in different browsers (Chrome, Edge, Firefox). No matter what, I do not get any results when searching for "primä".
Does anyone know if there are further settings that I can check in order to make this work? Or is this a bug? In this case, I will create a bug regort.
Thanks and regards
I'm thinking of the old CHM format that has issues if not used on a machine with the same language operating system.
Are you testing on a machine with German as the OS. I'm doubtful with this format but worth a try.
yes, my machine runs a German OS.
I just tried putting both the words below in separate topics and generated with both a frameless and a responsive skin. Worked both times.
Try in a test project and see what you get.
I just ran the test, same issue with a brand new project and default skin Azure Blue:
I have sent you a project where it seems to be working. Test that as a first step.
The email bounced so sending it via WeTransfer. Meantime this is what I get with your settings.
Thanks Peter, file came through now. Could you please run a test with your project searching for "primä"? "Prim" has been working for me as well.
Peter, it is working with your sample. However, there is a difference in the settings. You are using Project Lanuage EN (UK), I am using German as project language and output language. With these settings, it is not working. I just ran another test with project language German and output language Default. This seems to be working. The problem with this setting is, that the RH-texts like "inlcude all words in search" etc. are not displayed in German but in English.
I will wait for feedback from the techcomm guys and then possibly raise a bug.
Thank you very much for your support!
So a character that is used in the German language breaks the search when on a German language system. 🙂
Let us know what Adobe come back with.
With primä I get two results, which is correct.
Try generating my project set to German. I just attempted that and it is hanging at Starting Generation. Before changing the language it was fine.
Yes, this is another problem. You have to create a synonyms file for German to make the output work. Just write "a - a" or something in it.
Searching on primär with language to German, I get only one result like you. Unless Adobe come back with something, I suggest you add my project to any bug report as it is a simple example of the issue.