How to workaround slow down in loading chm file into C# .NET program from background security software

I recommend you in this case use a different technique. At the form load time load the contents of your chm file into memory. Then, use WebBrowser.DocumentText or WebBrowser.DocumentStream property to set the text of the WebBrowser to the required section of the chm document.

Thanks chaau for your help.
Hope you don't mind, I am afraid I need some more direction from you.

>> load the contents of your chm file into memory
Did you mean calling the 'WebBrowser.Navigate(.)' method and keep the content in the 'WebBrowser.Document'.

>>  use WebBrowser.DocumentText or WebBrowser.DocumentStream property to set the
>>  text of the WebBrowser to the required section of the chm document
Could you tell me how to do this or point me to some reference on how I might do it.

Thank you very much.

for the code wise, may want to consider async/wait pattern - some sort of "DoNavigationAsync" (can try google for example if any similar scheme) that executes asynchronously (and checking for WebBrowserReadyState.Complete") or some "WaitingFor" function to check for delay on Navigate completion. in short, you want WebBrowser.navigate to be asynchrounous, so you need to setup an event handler to handle the navigation completed event

actually if the chm is not going to change or is the same chm most of the time then maybe we can disable the security scanning - maybe there is some exception rule for the security scanning to avoid (just like some system or page file that it is not scan as CPU resource is too huge - the scan is done probably in off peak user hr).

will also be good if the security scan can be called via API if that is available for the security s/w. at least that can give you the API to decide if scan can be enabled. I was thinking of checking if chm is changed via some hash check - meaning generate a hashed string which is based on MD5 or SHA1 or some simple checksum to ascertain its original state - then decide the scan. However, this hashing will need to be maintained in some config and done before hand also it may not be any faster esp if the chm is huge (hashing req loading whole file as well), unless there is some signature we can use as "hash" to determine chm is in good state.

I think there are a couple of ways to work around this.  Both involve mining the chm file for all the help text.
Classic approach:
1. get the contents ahead of time, during development, for all controls.
2. use OnHover event to display the text.

or
CSS approach:
.HoverText {
  visibility: hidden;
  opacity: 0;
}
.input:hover .HoverText {
  visibility: visible;
  opacity: 1;
}

<p class="HoverText">This is text about a control from the chm file</p>

You could also transfer all the mined hover text into your database, replacing the real-time mining of the chm file with database calls.

You could also include the hover text in the content of the web page as a JSON object and use javascript to render the hover text from the local content.  Be aware this approach may cause page bloat.

Probably this can be useful thought using HelpNavigator and HelpKeyword
http://www.codeproject.com/Articles/6661/Integrate-Help-Into-Your-NET-Apps

Choose an appropriate HelpNavigator Enum value so that the desired section of the help file will be opened and queried using your HelpKeyword value. The latter should hold a context-specific keyword that you had included in the Help file before compilation. It could be an index entry, a search value or the name of a file that is part of the compilation.

It would be interesting to know that when you invoke a web page (whether local or remote) using the Help.ShowHelp method, you are opening your browser window in an external process. This means that when your client application process dies, the browser process holds its ground unless you had earlier closed it yourself. It can be embarrassing if your Help file hangs around after the application that called it has shut down. However, it does not happen with .chm files that are opened with the HelpProvider component.

As a work around, I suppose we could use an instance of the Process class found in the System.Diagnostics namespace.

Thanks btan and akimark for your help and directions.

In reply to btan:
>> "... DoNavigationAsync..."
The help paragraph that shows up in the help panel (located at the bottom of the program) is meant to give immediate information to user about the intention associated with the UI control when they hover it. Therefore, navigating async will frustrate a user.

>>  " .. disable the security scanning..."
This is not possible in my case as the software is meant for external customers, and we have no prior knowledge what kind of security s/w they use.

>> "...using HelpNavigator and HelpKeyword.."
What happened is my program has been working until recently the security s/w is replaced, then I realize that some security s/w can be a problem. Therefore, I am hoping not to rewrite the existing code.

In reply to akimark:
>> "...get the contents ahead of time, during development, for all controls.."
There're thousands of Controls and each has its own help text. Things continue to change. So getting the contents during development, which I think you meant hard coding, is not possible.

Ideally, as per the suggestion from the first expert chaau, preloading the chm content into memory, and access it during run-time is the ideal solution in my case as I can keep the existing architecture and don't need to rewrite much. Unfortunately, I don't know how to do it and I am seeking more directions on this approach. I would appreciate if either of you could fill me in on how to do that. Thanks in advance.

A CHM file contains its own file system. You can handle read/write streaming using the IStorage interface, which supports the creation and management of structured storage objects. Loading into the memory will also be an one time slow process and subsequently it will be via the memory access - security scan will still be enforced too. Nonetheless, below is a HTMLhelper lib which you may be interested ... (esp on the "The data dumping" section). Note that this is kind of old archive and istorage seems to be deprecated.

http://www.codeproject.com/Articles/6765/HtmlHelp-library-and-example-viewer

I guess we are looking at memory mapped file with more MSDN info below - main thing is to map view for specific access but note scanning may still be existence and whether it is fast based on various security s/w can differs as well...

http://msdn.microsoft.com/en-us/library/dd997372(v=vs.110).aspx

maybe eventually it is offline check of file or exception if that is alright....

Thanks aikimark and btan.