I generated a Responsive HTML5 help, and after I put it online, when I tried to look for a word, I got only 2 results, whereas when I searched the same word in the Word document generated from the same project there were many more.
I activated the search options to look also for substrings, and I got 3 results.
Is there anything else I should do in order to have this functionality work properly?
I am using RH11 in an Win7 environment.
Can you provide some specifics? For example, what specific words? Where did they appear? Are some of them actual links that lead to other topics or URLs or suchlike?
Does the same behavior occur if you generate and test a WebHelp output?
Thank you so much for the quick response.
Actually, right now, after I checked that option with the substrings, the search is considerably better, however there are still occurrences of the word that are not discovered.
I was looking for a simple word "filter". In the Word documentation there are some 27 matches, however in the HTML5 help there are only 3. None of the occurrences of the word throughout the documentation is a link to another topic.
I tested the output before I committed it to the web, and it was pretty much the same result.
Thank you again
And how does it behave if you generate WebHelp?
Thank you for your response.
I did not generate WebHelp.
I will do a test to see how it behaves.
Also, what point version is your RH11 install? Check to see if you're up to date with your patches.
It is updated with all patches.
I'm seeing the same thing, only to a greater degree. There are queries that produce NO results in the Responsive HTML5 layout, however, the same queries produce dozens in the WebHelp Pro layout. Further, searching in Responsive HTML often becomes unresponsive, producing no results for queries that shortly before produced results or, worse, returning nothing at all, not even a "no results" message.
Searching in Responsive HTML5 seems very inconsistent and lacking.
Rocky, what version of RH are you using?
RoboHelp 2015 with RoboHelp Server 10
I think there's been reports of issues with the last RH2015 patch that just came out - what point version are you running?
Ok, bogaem, I just tested a project in RH18.104.22.1681 and created both Responsive HTML5 and WebHelp. Searching for terms like "filter" or "setup" returned exactly the same numbers in either output. How did you make out if you created WebHelp and searched for your terms (you don't have to upload it to a webserver to test; just looking in your \!SSL!\WebHelp\ folder will do).
First of all the results say for instance "3 result(s) found", however there are actually 3 topics in which occurrences of that search word and not the total number of occurrences found in the whole help. When I search for the same search word in the Word document, it would tell me for instance "67 matches" which is clearer. Could be also that because it gives me the number of html files in which the search word was found instead of the total number of occurrences that it is perhaps confusing. I tried to add-up "manually" the number of occurrences found in each html file (topic), and somehow it matched the number that should have been returned.
Also, when the results of the search come, they show the names of the html files where the occurrences of the search word were found, and not the topic. Is there a way to change that, i.e. to get the names of the topics, that are "nicer", instead of the names of the html files?
Pretty sure that RH has never reported the number of times it found a word - just the number of topics that have the search term. Trying to compare Word's "word hits" with RH's "topic hits" is a bit of comparing apples & oranges.
Not sure how you tweak the search results to give you topic titles instead of topic filenames - I think it can be done, but I could be wrong. I haven't started working with HTML5 in my projects yet.
Yeah, I think I understand what you mean.
As I said, the number of occurrences is very close.
The only thing that bothers me a little bit now is those names of html files instead of topic names.
If you could find something about this issue, it would be great.
Should I open another thread about that?
bogaem - I would search around in the forums first to see if there's any way to control the filename vs topic title - it may depend on where the source content is coming from - I know that if you bring in FM material, you can use marker text to control aspects of the resulting RH filenames.
What do you mean by "the source the content is coming from"?
I just have a RH project and I generated a responsive HTML5, that is all.
@bogaem - ok, your project is authored in RH, not brought in from Word or FM. Not sure if there's any controls exposed in RH that would affect the use of the filename vs. the topic name or heading title in the HTML5 output. You may have to contact RH tech support.
Thank you, however I will stick with this forum for answers.
I will see if I can start another thread, as this one is, from my point of view, exhausted.
Thank you again
Since the OP seems satisfied with the explanation, should I start a new thread for my complete lack of search results or is the thread title close enough?
@rocky_xyzzy - I would first check to see if you're getting the same results with one of the sample projects or not. If not, then it's something specific to your project. In that case I would check to see if the choice of browser has any impact and if where the help is located plays any role (web server or local workstation).
The same project works fine, which is not terribly surprising. I would have been stunned if basic search functionality was broken in a sample project, as even the most rudimentary product QA would have caught that. So, yes, it has something to do with our project.
All of our projects' search functionality work fine when the projects are published as WebHelp Pro. They're all significantly broken (from any browser) when published as Responsive HTML5.
I would have been stunned if basic search functionality was broken in a sample project, as even the most rudimentary product QA would have caught that.
Not necessarily. RoboHelp is a different animal in this regard. Sure, there are sample projects included. But updates to the software as well as your environment can affect things. Sure, when first released, the samples *SHOULD* provide a nice basic guideline as to whether things work or not. But even the samples can break down the road. This is because as RoboHelp is updated with Service Releases, the service release can introduce changes to the code that break things. Even in a sample project. This is because the coding used is only applied during the generation process.
Quick question here. You mention WebHelp Pro. Are you also publishing to the RoboHelp Server product?
Yes, we're publishing from RoboHelp 2015 to RoboHelp Server 10