Export confluence docs as static web-based help

[Robert Messer]

We are considering using Confluence for technical documentation. We like using a wiki-type product for documentation, because various internal users and programmers can all easily contribute to the docs. But at the end of the day, when we deploy our product, we want to have the ability to easily provide a link to a static html help system. I do see that the Confluence product allows export to html, for example Confluence's own docs have a link to a html versioon here: http://confluence.atlassian.com/display/ALLDOC/Confluence+Documentation+Directory#ConfluenceDocumentationDirectory-downloads

But I downloaded the html and opened it up, and it seems to be a bunch of losely formatted html documents, with no navigation or search options. Am I missing something, or is there some way to deploy confluence docs without having our clients access our confluence site? Of course, I understand that allowing clients to access the wiki and contribute can be useful, but there are times we need to deploy the docs as a stand-alone entity, and so I need to understand how I would best do that. I also need the end result to be searchable and have navigation more like a typical web-based help system, and not just a bunch of html pages. Thanks.

[Colin Goudie]

I would checkout K15t's plugins. Specifically the upcoming Scroll Wiki Exporter

http://confluence.atlassian.com/display/APW/K15t+Software+GmbH

[Cassandra Targett]

I have the same interest as the original poster. One thing I don't understand about your response is how you add the hierarchy information (which pages are parents of other pages) via this method. Is that done with the ssr.exe you're using?

[Jim Birch]

I export our Info Services space every month as a hopefully-never-used disaster recovery tool. Our Confluence contains most of the how-to for our system, including eg how to rebuild the Confluence server after some real disaster.

This process creates a working copy of a space's information as a navigable html file set. It doesn't include searching, except by external facilities, but the index page is available and useful if the space is heirarchical. Page and attachments links work.

There are no breadcrumbs in the export which is a problem. This should be quite doable. There's a feature request for breadcrumbs in the html export that you could vote for at https://jira.atlassian.com/browse/CONF-5337. It may be possible to do something with the export velocity template, I haven't looked. If you have good page crosslinking and indexes this may not be such a problem.

1. Do the html export

2. Expand the zip

3. Replace the drab exported site.css file with a better looking one from your live confluence website.

4. Add a file ((Jump-To-Index)).html file that redirects to index.html while sorting to the top of the file display. Or use another mechanism to get your users to the index.

5. If you use any rollups with hidden text blocks run a multifile editor like ssr.exe to display any hidden text.

(in dos) for %%f in (*.html) do ssr.exe 0 "display: none" "display: table" "%%f"

(This isn't required with the newer expand macro but there is a hidden text problem with the Customware cloak macro on exported pages.)

6. Distribute. In our case this means rezip and place on a couple of strategic usb keys.

Our space exports to a ~2Gb zip or 4Gb expanded. The thing is big but it works.

[Cassandra Targett]

I have the same need as the original poster, and I don't understand how this method adds the hierarchy to the html pages for navigation. Is that done with the ssr.exe you're using?

[Deleted user]

Did you figure out how to add the hierachy to the exported pages? I am having almost the same issue. I want to export to html and have at least the child pages as links on the page.

[Joe Clark]

One option that you could consider is the "auto-export" plugin for Confluence (https://plugins.atlassian.com/plugin/details/33010).

This plugin automatically produces static HTML content for your Confluence pages whenever changes are made - this content can then be served up immediately by another web server such as Apache, or zipped up for export or other nefarious purposes.

This plugin was originally developed by members of the Apache foundation (but is now kept up-to-date by Atlassian) for their documentation requirements.

[Aleksandar Lazic]

I had a similar question and got the answer from support that there is a feauture request https://jira.atlassian.com/browse/CONF-18265 .

I hope that the feautre comes in any of the future version and maybe also to jira :)

Aleks

[Barbara Green]

We are also using Confluence for technical documentation. Our product interface is web-based and hosted, so I asked the developer to point the Help link to the home page for a space that contains the help topics. This means that I only have to author the content once, and there is no export or re-edit involved.

Considerations we have had so far:

• The Help link is in an included element in the interface. So we either have to give up context-sensitivity, or get rid of the include and call an explicit URL on each interface page.
• The space with the help has to be accessible by anonymous log-in. Otherwise, the interface user would have to log in to Confluence to get the online help, which would be awful.
• Images used in the help (like a heading banner) have to be accessible as well. You can't have a heading banner in a secured space because the anonymous user won't be able to see it. We created a space on our public web server to store the images we need and we just call them from that server.

This technique wouldn't be useful for standalone apps, but so far, it's working okay for our html app. I was previously using a DITA tool to create web help, and I really wanted a DITA->Confluence tool, but that would still mean rendering for the app and rendering for Confluence. As it stands, I author once, and the information available from the interface is always the most current.