Copy link to clipboard
Copied
Hi, Is there any way to enable substring search in CHM output? If someone has (finally) found a different solution for a compressed, offline HTML output, I’d love to hear about it.
Copy link to clipboard
Copied
I assume you are using Robohelp RH2022. The process is similar for Robohelp Classic versions. If you're not using Robohelp, then you'll need to check on the forums for whichever tool you're using to generate the chm. 🙂
In your chm Window Settings, select the Adv. Search checkbox. If you don't have a window definition, create one and make sure that setting is selected. Make sure you select that window as the Default Window in the CHM output preset.
You'll then get some checkboxes at the bottom of the Search tab. I think "Match similar words" is the one you want.
This site provides more details about chm search. https://help-info.de/kb/HTMLHelp/100003-hh-faq-search-for-word-or-phrase-in-chm-file/#to-refine-a-fu...
Copy link to clipboard
Copied
Thanks so much. I am indeed using RH 2022. The advanced search option is an improvement, but it still requires users to use wildcards in order to find a substring. Is there any way to automate that so that substrings will appear in the results even without adding an asterisk * before and after the search term?
Copy link to clipboard
Copied
You're talking about CHMs, right? These are an extinct form of output developmentally - MS dropped them years decades ago.
Copy link to clipboard
Copied
I know. But what good alternative is there for an offline compressed output? My company insists that the output must be compressed (i.e., a project with tens-hundreds of HTML files is not an option).
Copy link to clipboard
Copied
There's nothing really out there - I use a self-extracting exe program to package my HTML5 up for local installs. That squishes it down to a reasonable size in one file.
Copy link to clipboard
Copied
Thanks, that sounds like an excellent solution. If you wouldn't mind sharing what program you use/recommend and any other details about the process, I'd be very appreciative.
Copy link to clipboard
Copied
I use a couple - one's a really old program that isn't made anymore called Wise, but the other is WinRAR. They both work much the same way - you tell it to pick up the folder containing all the HTML5 files and compress it into a self-extracting executable. The Wise program has a way of creating a front-end UI so the person running it can point it at a location on their LAN server and run it (which unzips all the files/folders into that specified location). Probably any installer program like InstallShield would work too.
Copy link to clipboard
Copied
@Jeff_Coatsworth CHM on a server?
________________________________________________________
My site www.grainge.org includes many free Authoring and RoboHelp resources that may be of help.
Copy link to clipboard
Copied
Nope - HTML5 files
Copy link to clipboard
Copied
Doh! Of course.
________________________________________________________
My site www.grainge.org includes many free Authoring and RoboHelp resources that may be of help.
Copy link to clipboard
Copied
Thanks for your help!
Copy link to clipboard
Copied
I understand your predicament. It somewhat amuses me that your company insist on you using a format that relies on the Microsoft Help Compiler that was last updated some twenty years ago and is no longer available to download from Microsoft's site. They pulled it a few years back.
It had security issues as described in CHMs on a Network (grainge.org). Are your IT people aware of this and would the developers use tools like this for their work? See if that helps you argue the case, if you want to.
Because CHMs rely on the Microsoft compiler, there are no tricks other than the one @Amebr has described.
________________________________________________________
My site www.grainge.org includes many free Authoring and RoboHelp resources that may be of help.
Copy link to clipboard
Copied
Is the issue the developers don't want to ship multiple files or they don't want multiple files on the users' machines?
A exe will ship one file but the developers will have to then extract the files.
________________________________________________________
My site www.grainge.org includes many free Authoring and RoboHelp resources that may be of help.
Copy link to clipboard
Copied
Hmm, I just tried it in an old chm i found on my PC (no idea what it was created using) and the find simialr words setting works in that one, but when I copy the text from that chm and generate one in RH the same search doesn't not work the same. So it almost seems like RH is not creating the search to spec somehow. However, getting any changes to how RH generates chms may be difficult.
Copy link to clipboard
Copied
Thanks for trying that out. Where I'm specifically encountering a problem is in our programmers guide. People search for the name of an interface, but without the "I" that precedes it and the version number that follows it (e.g., the topic and interface name is IMeshLayer80 and they search for MeshLayer. Even when Find similar words is selected - the search comes up blank. One solution I guess is to create index entries for each of these abbreviated interface names so that they'll show up in the Index. But I'd like to hide the Search tab then (so that people don't perform failed searches). Is there a way to hide the Search tab in the CHM?
Copy link to clipboard
Copied
I believe it can be deselected in the Window settings.
Will you be swapping one problem for another if an index term doesn't find what the user wants?
________________________________________________________
My site www.grainge.org includes many free Authoring and RoboHelp resources that may be of help.
Copy link to clipboard
Copied
You're right. Thanks for that advice. Sorry I hadn't checked that myself. If I create a very-comprehensive index, I may be better off. Otherwise, you're right, I might be swapping one problem for another. Thanks for all your help.
Copy link to clipboard
Copied
Ah yeah, when chm was developed I don't think any searches worked like that without wildcards. 20 years ago the similar words setting was cutting edge and i'm pretty sure it was only similar based on grammar rules like plural and tense (user vs users vs using) and even then was not hugely accurate. 🙂