Cüneyt wrote: > I know quite a few guys here, they are simply amazing. But the most > useful gems I've ever discovered are mostly hidden in the messages. > Someone posts 3 lines of code, and you suddenly feel the chi-flow > through you, feel enlightened, know the answer to everything ...some > call it simply "the slap-on-forehead effect" or "RTFM-aftershock" :] > Yes!! That's exactly it -- looove your description. I suspect it's some sort of osmosis happening thru being in the company of geniuseses. :)
> That's understandable. BTW, I'm amazed of what you've done there. It's > quite a piece of work and looks really good! > Thank you! All credit really to Bruce, who creates this insanely brilliant tool & inspires us all, methinks. > After all, we need no > browser or fancy JS but only 2 simple things: > > 1. simplified HTML-output or any format which is easy to parse by PP. The > current output is filled with some unneeded stuff, which is very usual > with automatically generated HTML anyway. The important thing is it should > be extremely lightweight in order to minimize the overhead. > > Since the generated XML is 10 MB, it's not feasible to parse it with > XSLT/SomeScriptingLanguage with every request, even with caching. So, the > best solution would be pre-generated files, it means some more work but I > can help you with the XSLTs or programming. > Yep, I came to the same conclusion. Freemind actually does 3 flavours of HTML export: one with an img map as main index, one with JS trees (that's the one I've used as "old format"), and another one.. I've forgotten, think it's just plain vanilla HTML with no folding tree, from memory. But they're all hugely full of guff. (That's my new favourite word, incidentally. :) Hmm... I'm going mad.. thought the source Freemind file was 10mb, it's only 5 mb, actually. (still pretty huge, I guess!) So, what I've actually done is break the main source doc into 5 separate Freemind docs, 4 of which are online as "old format". (The 5th is the intro section, which will come back eventually, but it's full of holes & v. messy at mo.) This made each section much more manageable. So, each exported XML section is parsed with a PowerPro script making use of RegEx plugin to remove all the unnecessary formatting from Freemind XML, leaving a much much much smaller XML file for each section (ranging in size from ~300kb for CL functions, to ~850kb for the Plugins section). I have the parsing script set to a hotkey in PP, so it's all v. smooth. Next, I've got some PHP scripts that are actually doing the XML-XSLT parsing to spit out all the pages. The PHP reads the XML source & decides which pages & sub-pages are needed (based on whatever Actions/Commands/Functions/etc, are there) then creates them. Takes about 30sec (on my 5.y.o. laptop) to output all the Built-in Commands. As for the actual final pages, (the "new format") -- I've used tableless CSS-driven client-side code there, so they're extremely lightweight. So, not sure why I've gone into all that... I guess so you all know more of where I'm at & how I'm doing it. When developing this initially, I didn't think at all about integration with lookup systems, so think I'll revisit the pages & make sure they're standardised in some way to facilitate easy parsing by PowerPro. > 2. a unique URL for each keyword, i.e. something like: > > lookup_url = ?"http://ppsr.info/?lookup="; ++ searched_keyword > > so that there are no pre-request computation. > Yep, great idea, & should be easily done server-side, so I'll try & get this done soon. Are all PowerPro keywords unique? Any plugins/CL functions/etc sharing names? That's the only potential problem I see here. > Possible scenario: I'm writing a script in Ultraedit, type "win" and want > to see a list of possible actions for this keyword. I press > CTRL-SHIFT-SPACE to request the online help for this keyword. The net > plugin fetches ppsr.info/?lookup=win, gets the output and shows it via a > menu or even a cl-generated bar. You navigate with up and down keys > through the functions. With Enter you insert the highlighted function and > with right arrow you send a new request for the highlighted keyword and > navigate deeper. It won't be real AJAX but the feeling would come close. > I don't use UltraEdit so can't help with this directly, but if you develop the plugin I'll work with you to ensure the files can be retrieved simply. > Of course, this requires a reliable Net connection on our side and some > bandwidth on your side. I can help you with the bandwidth though, and a > simple round-robin solution with a PP vector is trivial. > Bandwidth isn't a problem, so don't worry about that. The round-robin solution has me confused but I think that'd be your end, for the UltraEdit plugin, if I understand correctly? > OTOH, if you'd go the browser way, you wouldn't have to do nothing on > your side, and the ouput could be much fancier and you could simply > deliver XML. I can definitely deliver XML, another great idea, I'll do that next. For the moment tho, I want to finish the HTML version first so we've at least got one in entirety asap. > Maybe I'm just talking twaddle and wanting most useless stuff - again - > (I'm saying this so if you don't like the idea I can always call it > brainstorming as a last resort :P) but it definitely would be a good > showcase for PP. > No, that's not twaddle at all! I love the ideas, and am so excited by the interest in this. To be honest, I haven't done anything on PPSR since about May, so I need to revisit & work out where I'm at, then I may come to you for XSLT assistance to get it out the door asap. Thanks heaps Cü :) Karen Attention: PowerPro's Web site has moved: http://www.ppro.org Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/power-pro/ <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
