In the built documentation folder of an application with Wolfram Workbench, there are two folders /Documentation/English/Index and /Documentation/English/SpellIndex which both contain three files: _g.cfs, segments_z and segments.gen.

For what purpose are these files included? Can they be modified an any meaningful way? What if they are removed?

Applications seem to continue to work like normal when the files are edited.

  • $\begingroup$ This seems to contain info on how to generate these. If you can separate that information out from that long post and summarize it in a self-answer, that would be quite useful. You can also try to spelunk the functions that create them. $\endgroup$ – Szabolcs Nov 30 '16 at 16:46
  • $\begingroup$ Unrelated: You create packages, so you may be interested in the IntelliJ plugin and giving feedback on how to develop it. $\endgroup$ – Szabolcs Nov 30 '16 at 16:48
  • $\begingroup$ @Szabolcs I checked it out at one point, but thought that it wouldn't work for me since I write packages that extensively use greek characters α, β, γ. I need the Mathematica front-end .m editor to render these correctly since I don't think I would like to use \[Alpha], \[Beta], \[Gamma] ... . I figured that it would be far too much effort to add this as a new feature, so I didn't bother requesting it. $\endgroup$ – QuantumDot Nov 30 '16 at 17:24
  • $\begingroup$ I thought this created search indexes, to be used when searching the docs. Greek letters in formulas should make no difference. Do you write in ελληνικά? $\endgroup$ – Szabolcs Nov 30 '16 at 17:31
  • 2
    $\begingroup$ No, what I mean is that some of my package function names are Kramersβ and RegularizedRiemannζ, and I have algebraic symbols like α, β, γ. I assumed they would be rendered in the plugin like Kramers\[Beta], etc, (just like in Workbench), so it wouldn't work for me. $\endgroup$ – QuantumDot Nov 30 '16 at 17:51

I believe these files are created by the Workbench (using the DocumentationSearch package) and contain search indices for a package's documentation.

I think Index is for the search indices, and SpellIndex for the "did you mean" feature:

enter image description here

There is information on creating these indices manually in the following QAs:

Since M11.2, a different system is used which creates a directory called TextSearchIndex.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.