Current State of Mass Importation

[scott]

So--should I hold off importing the stations for the lines I'm working on (mostly just the lines listed on my Tohoku page)?

[Guest ___]

By all means if you want to start importing stations go for it. I finished all the lines, and frankly my mouse hand and left elbow are sore as can be from days of importing.

 

It's everybody's wiki, feel free to have a go at it.

[scott]

OK--thanks, I'll keep going. I just didn't want to foul anything up that you were working on.

 

I may try a couple of massive exports, like all the JR East stations at once, if I can.

[Darren Jeffries]

Scott,

 

I started at R

 

I can tell that there is a gap between Kofu station and Rifu Station

[scott]

I've been having trouble with the import process timing out--apparently the limit is 60 seconds.

[Darren Jeffries]

I've been having trouble with the import process timing out--apparently the limit is 60 seconds.

Yep, it allows for about 300k of a xml file to be imported... dont know how that relates to articles but it is a fraction of the whole EJRC

[CaptOblivious]

There is a maximum runtime for PHP scripts of 60 seconds (I bumped it up from 30), to keep them from "running away". I will temporarily bump this to 120 seconds while y'all are doing these large imports, but keep in mind making it too big means that actually broken scripts could bring the site to its knees :D

[scott]

That's right, we're the first crack in the wall. :)

 

OK, I think all the JR East and JR Hokkaido stations are in.

[Guest ___]

Scott, you know if the page of the station comes from a redirect then that page will not import right? (I learned this the hard way)

[CaptOblivious]

daaaaaang y'all. Donbot can't keep up with your imports!!

[scott]

Scott, you know if the page of the station comes from a redirect then that page will not import right? (I learned this the hard way)

 

I'm going to pretend that I know what you mean just long enough to try to figure it out. <sheepish grin>

[Guest ___]

http://wiki.jnsforum.com/wiki/JNSwiki:Importing

 

 

Note the circle, that's a redirect. It's a separate page that uses the code #REDIRECT

 

In this case, the West Japan Railway Company is the main page for the aerticle, but if someone enters i nthe search box JR West, Wiki does not know what you are looking for so you need a redirect. In roder to tewll wikipedia that what you are really looking for is West Japan Railway Company, you need to create a new page called JR West, then enter the following code only in to the page

 

#REDIRECT[[West Japan Railway Company]]

 

Now when you enter JR West in to the search box, wiki will automatically forward you to West Japan Railway Company, the correct page.

[scott]

I guess I understood what a redirect was, but not how it affects exports. I was using the "all files in category X" export option, so maybe it's not a problem.

[scott]

Oh, and I imported most of the JR Kyushu stations.

 

One thing I've noticed is that JNSwiki seems to be pickier about accented (wc?) vowels than Wikipedia. I had to change the links to a bunch of the stations that said something like [[Mojiko Station|Mojikō]] to [[Mojikō Station|Mojikō]], even though they apparently worked in Wikipedia with the unaccented vowels.

 

Does this mean that we need to create a redirect to all pages that have accented vowels in their names (Tohoku>Tōhoku, Kyushu>Kyūshū, etc.)? That could be a lot of work!

[Guest ___]

Which is exactly where the redirects are coming from.

[scott]

...meaning that we have to either create pages to redirect from, or let everybody know they need to type the correct accents/diacritics to find what they're looking for?

[scott]

OK, I've read the help page on redirects, and it sounds like you have to manually create a redirect page to each page with spelling issues in the name. There must be hundreds of those already in JNSwiki; I can't imagine going back and making a new page for every one.

 

Is there some sort of setting in the search function that can lump multiple forms of a vowel into one character for searching purposes?

[Guest ___]

Manual redirects are a pain in the tail to say the least.

 

Just a quick note, there's still huge holes in the stations. I'm working slowly to manually bring them in, but man, there's a lot of them.

[scott]

Manual redirects are a pain in the tail to say the least.

 

...which is why I'm wondering if there isn't some sort wild-card feature to the "search" portion of the wiki software that could lump "ō" with "o", etc.

 

Just a quick note, there's still huge holes in the stations. I'm working slowly to manually bring them in, but man, there's a lot of them.

 

What's the easiest way to see what's missing? Are there entire lines/companies left out, or is it more random?

[Guest ___]

Look at the top of the page you are going to export and look to see if there's a line under the title that reads redirected from xxxxxx

 

It's a pain, but redirects aren't tagged in any manner that connects them to a category, so if you export a category the redirects are seldom if ever included. It seems that if you export a category or group of articles and the redirect is not manually added to the export, then the original destination article gets omitted on the import to JNS.

 

It took me forever to figure that out.

[scott]

So--are there any big holes that need filling now?

 

I'm not sure if it's better to keep uploading stuff, or to go back to adding to my pet page(s).

[Guest ___]

Looks like it. Damned if I could tell you which lines off the top of my head though. Should have asked there Kirin's or ten minuets ago, (I'm celebrating tonight)