Hi Danielo, > I'm not sure what a tiddler "pointing" to a file means to you. Could you > elaborate this? Is a tiddler that has a field that points to the file > location? Is a tiddler that embeds that file? >
a reference in a field — yes embedded file — certainly not The meta files are needed for allow the user to add tags and fields to the > tiddler without writing that information to the original file. I think I > will not use meta files but something similar to tiddlywiki.files is. > That's what I mean, you simply use tiddlers, not those meta-files we currently have. You never touch the actual file. You merely reference it via some path reference stored at the associated tiddler. Yes, I'm pulling those files into the wiki. That's the idea. That is the > only way to provide text search, which is my main goal. > I would not do that. I'd rather hope and go look for some nodejs search module that you can talk to so as to "search in files". Not sure if it exists, but I'd be damned if it didn't. > I used to include lots of code snippets in my wiki, and sometimes I would > love to be able to search within my "full" source code folders in the same > way. Maybe a too narrow scenario? > The way I see it, you want to leverage nodejs. So, the focus really is on finding a module that TiddlyWiki can hook up with so as to do some searching in files, rather than merely the meta-data you wish to save in an associated tiddler. > I think i'm starting to get your point, Your idea is to create the index, > maybe on startup or whatever, and then, on the tiddlers, just embed those > files. > I don't see a single file being embedded. There would (have to) be a way for TiddlyWiki to talk some module about querying the local fs and be given search results, e.g. "search for this text / in this folder / these files / of that type"... Isn't it? So the only thing available for searching is the title, and maybe > the extra information you add to the tiddler that links to the original > file. > Yes, if you'd only have a json index being included in your TiddlyWiki on startup, then searching directly from within TiddlyWiki would be constrained to just that. I'd even go so far as to see that nodejs module not only return search results, but also dynamically updating that folder index, so it's not just happening on startup, but whenever new files are added as well. So, some daemon would perhaps return al files changed after timestamp x... and add those to the lot. As for all files deleted after timestamp x, that may be more difficult and only a job for the next startup... to clen "meta"-tiddlers for files that are no longer present in the local-fs, or even moved... there could be some code to check if stuff has perhaps been moved. It is definitively a different approach, and in some way I like it very > much. > So, yes... not sure if that's even possible or how... to have TiddlyWiki talk to nodejs as it would for syncing files... only this time, actually ask for search results... and then some node-module doing the local filesystem-access and returning the proper response. Best wishes, Tobias. -- You received this message because you are subscribed to the Google Groups "TiddlyWikiDev" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/tiddlywikidev. For more options, visit https://groups.google.com/d/optout.
