While a picture says more than many words:
I find freemind convenient to generate a nice sitemap of our website. Although in 0.8.1 there is possibilty to export as XHTML with linked mindmap, I wanted not the mindmap but only the imagemap with the image.png , and hence donot need the anchor-links, but Links refering to URLs .
At the beginning I just let some regex run on the XHTML-export of freemind. But thats stupid. Also, I realized that I am not the only one who would find this functionality useful, so I decided to code where it belongs to, that is: inside of freemind itself.
I found freemind a very elaborated kind of software - nice design! Much of it new to me. But, puh, not so easy to start with the "hack". Or "patch". Then, most important to me: the new code should not affect the functionality of the old code, in (nearly) no way.
these are the files:
Only the 'ClickableImageCreator' is an old code, that I customized a bit. I have added a second constructor and a little 'if-else-logic'. The code works just like before.
I tested it with a fresh download of 0.8.1, just copied the files in place, and let freemind run. Seems fine.
Possible enhancements for this patch:
# At least as a sitemap-generator it would be fine to let freemind be runable in textmode, so it could create a sitemap direkt on the webserver, eventually as a cronjob. OK, for that purpose we donot need complete freemind - but it would be convenient as the code is already there ;)
# also possible: import a textfile, wherein a directory-structure exists, just what an 'ls -R > dir_stru.txt' will give.
# I'd dreamed before of the visualisation of filesystem or think of dmoz.org . With freemind it is easy to have some experiments. Also I could imagine to use it for visualisation of rigid taxonmy in e.g. linguistics. For this, a simple imagemap could provide the necessary simplicity.