P� 16. feb. 2005 kl. 15.06 skrev Ross Gardler:
Sjur Moshagen wrote:
I have now added the encoding parameter at the following locations: $FORREST_HOME/plugins/wiki/input.xmap $FORREST_HOME/plugins/wiki/resources/chaperon/sitemap.xmap
The second of these is not used by Forrest, it is there because this directory is a copy of the Cocoon block.
ok.
In the xmap files you need the namespace.
ok, I just checked that they're there (they are).
Did you do "ant local-deploy" from within the plugin directory? This will deploy your changes to the running instance of Forrest.
No, I didn't (I edited the xmap in the build dir manually). Now I have done it, but still no UTF-8.
This requires some more investigation then. I've not looked into it at all, I just jumped in on the "shot-in-the-dark" suggestion from Thorsten. Please file a bug report and if you continue to investigate keep it updated with your findings.
Configuration of plugins is not possible yet, we have discussed various solutions over on the dev list, but none have been implemented yet. One thing is sure is that it should *not* be central. The whole idea of plugins is to decentralise. Furthermore, it doesn't really make sense to configure the plugin across the board since there may be multiple files each with different encodings.
Ok. But how could one specify the encoding of text files, which inherently has no way of specifying it within the file?
This is really a dev list discussion so please move it over there is you want to contribute your thoughts (nothing has been decided yet). is that a set of plugin properties (that may or may not be part of forrest.properties) are used to define what files to process, and how.
For example (names need to be more carefully thought out):
org.apache.forrest.plugin.wiki.defaultEncoding = UTF-8
org.apache.forrest.plugin.wiki.Latin-1Encoding.include = /path/to/latin/file.wiki, path/to/another/latin/file.wiki, path/to/latin/dir
org.apache.forrest.plugin.wiki.Latin-1Encoding.exclude = path/to/latin/dir/excpetionUTFFile.wiki
In this case default encoding is set to UTF-8. The two files and everything in the direcroty listed in the Latin-1Encoding.include property are all Latin encoded, except the one exception file defined in the Lain-1Encoding.exclude property. There is much discussion about this in the dev archives (search for "selective PDF") and a partial solution in the issue tracker too.
If you are familiar with ANT then these kins of include and exclude patterns should be familiar to you.
But this is future - first we need to get the encoding working in its own:-)
So true!
Ross
