I was pretty excited to try this out in the office when I got home from Tom Rizzo's session at Vegas Dev Connections.
I setup a prototype web to crawl a small folder on our file server and immediately ran into an issue. It seems that when Word or Excel filenames contain a pound sign (#), the search result icons show the generic iExplore icon and the files do not launch properly. An example filename would be \\fileserver\documents\Form #123.xls.
When a Word link with a pound sign is clicked, the open or save dialog is presented, but the file simply will not load. When an Excel link with a pound sign is clicked, the open or save dialog is presented, an error message is thrown ("\\fileserver\documents\Form cannot be found. Check your spelling, or try a different path”), but the file loads anyway. I thought maybe the HREF parameter was interpreting the pound sign as a bookmark, instead of part of the filename.
I copied the anchor tags for some offending files to a plain html file and verified that these files will not load from there either. From there, I tried to see if there was some escape character (%23, ##, etc) I could use to interpret the pound sign and get the files to load so I could tweak the Search Result's XSLT file, but I found none.
Does anyone have a workaround for this behavior?