Oxygen XML Forum

I'm not sure how to explain this, so I will do my best. When I create, Webhelp, I have it saved to a local file C:\Users\my.name\Documents\Documentation_Projects\Output. Within Documentation_Projects are several "webhelp output guides" that I am working on. For the sake of explanation, let's say the first one is Hardware Help and another one is Software Help. I use the transform to create the webhelp for Hardware Help. I then zip the output file that has files as illustrated in the attachment. Because the output goes to the same location each time I use the transform, it contains other folders, such as Software Help, but I removed that folder before zipping up the output folder and sending to our Dev team to deploy on a web server.

I noticed that if I put in a search term that is not in my Hardware Help but may be in the Software Help, the results still display it. However, because the Software Help is not yet deployed on the Web Server, the clicking a results link obviously results in an error indicating the browser could not find the file.

My question is why and how does the output that we deployed, that I thought only contain files related to Hardware Help, recognize a term from files that were not included in the zipped output folder.

Hello,

The search engine indexes all the files that are present in the WebHelp Responsive output folder.
Therefore, there could be two solutions:

• if you need all the files (corresponding to both Hardware Help and Software Help) to be generated, but you need to filter the Software related topics only from search, you should try using the "webhelp.search.custom.exclude.file" parameter, which allows you to specify specific file patterns that match the files you don't want to be indexed.
  For more details on how to exclude certain files from search indexing, see "Transformation Parameter Method" from the "How to Exclude Certain DITA Topics from Search Results" section in the WebHelp Responsive User-Guide.

• if you do not need specific topics to appear in the Hardware output at all, you could set a profiling attribute on them (say product="hardware") and then edit the transformation scenario and filter them using a DITVAL file that you could pass to the scenario in the "Filters" tab. This way, the HTML files corresponding to the topics marked with the product="software" profiling attribute wont be generated and implicitly won't be indexed by the search indexer.
  Your DITAVAL filter could contain something like:

  Code: Select all

  <val>
      <prop action="exclude" att="product" val="software"/>
  </val>

  More details on using profiling attributes to filter content available in the "DITA Profiling / Conditional Text" section from the Editor User-Guide.

See especially the "Creating and Editing Profiling Attributes in DITA" and "Filtering Profiling Values with a DITAVAL File" subsections.

I hope this helps.

Regards,
Costin

Would it be better to have a unique transform for each webhelp?

Hi,

Yes, it is better to have each WebHelp output stored in its own directory (different directories for each output).

Regards,
Ionela