Alita Glossary -> Gone WIKI :)

[crazyankan]

Looking forward for the next version.
I have some problems from Viromania to William S. Burroughts. cant click on them

[Sergio Nova]

Last edition available.

http://www.easy-share.com/1907317630/Alita_uptoLO13.zip

Comments, critics and suggestions welcome.

[^Ripper]

Awesome work as usual
The latest version is on the site.

[Sergio Nova]

Awesome work as usual
The latest version is on the site.

Thanks a lor, ^Ripper.

[Cailon]

@^ripper: did you delete even the thread for your wiki-idea?

I just found this JoJo's-wikipedia and its open for all (so they say): http://jjba.wikia.com/wiki/Main_Page
It seems, they have no problems with vandalism (otherwise it wouldn't be open for everyone). Also, despite of the huuuuge JoJo-universe, they have 'only' about 200 articles. On the other hand though, its a start.

So maybe a BAA wiki could work, too. It sure would grow very slow but if we e.g. put a link to it in the release and open it for everyone, maybe people will come to fill it!?

[Sergio Nova]

Looking forward for the next version.
I have some problems from Viromania to William S. Burroughts. cant click on them

I hope it is solved now.

[Sergio Nova]

Here you have the glossary up to GLO14.
Please comment any point you consider necessary.

http://www.mediafire.com/?gci32leo7rpk2ob

Thanks.

[^Ripper]

Updated file on site.
Great work as always.

As for the old post about the wiki part.
Well i haven't abandoned the whole idea of a wiki.
Its just if people aren't going to use it, why make one.
A few weeks ago someone mentioned adding a few new forums.
Well its still empty

If people WANT a wiki and have time working on them I'll add one (of your choice).
Hey Sergio isn't there a program to convert that huge database of yours into a wiki?

[Sergio Nova]

Hey Sergio isn't there a program to convert that huge database of yours into a wiki?

Actually, the same file made to generate the ebook format can generate others like HTML, for example. Would it be the case?
I will send you the HTML.

[^Ripper]

There seems to be a html -> wiki text converter.
Still it does one page at a time, and every page has do be done manually.

So lots of work turning yours into a wiki itself.
Let me dig some more into this.

[Sergio Nova]

The possible output formats are: HTML Help, Winhelp, VS Help (I have no idea about such a format), browser based (that I prefer to name HTML, simply), Adobe PDF, MS Word (for God's sake!), eBook (the format I distribute here) and XML (Help & Manual schema).

I will search the web to try to find a solution.

Meantime, does this make sense?: http://www.uni-bonn.de/~manfear/html2wiki-tables.php

http://xpt.sourceforge.net/tools/wiki_import/

[^Ripper]

Well i installed the wiki (easy)
And the translation crap (fucking annoying as hell due to dependencies that have dependencies that have dependencies....)

Anyway the html version has its downs since it used JavaScript for html links.
Can you upload an XML version?
Also the html output, do you have options to disable JavaScript linking of sorts?

[edit]
Actually the resulting translation pulp is so cluterred with html tags/javascript/meta tags n other crap that it might be easyer to just copy/paste all articles manually
Anyway i just want to see if the xml version does a better job

Also how unlikely as it may seem, can you provide a msword version too
[/edit]

[Sergio Nova]

*removed by ^Ripper

This is the first time I generate such files (I mean, I had certainly generated RTF, but not for this glossary) and I have no idea how to work with XML.

[^Ripper]

Downloading

I've removed the link, dunno if you want your exports to remain private or not.
Let me see what i can do.

It would be nice if the clean page linking stays intact so the wiki can re-use them.

Oh btw, see the page rape @ http://stuff.rippersanime.com/Wiki/inde ... entChanges

[^Ripper]

Well....

The 3rd time import did make it "readable" but still useless.
Importing goes fairly well but makes articles with wrong topics and converts certain characters wrong.

I had to strip all the html tags from the html version to make it import like it did now.
However only flat text, some in color/bold remains with absolutely no hyperlinking whatsoever.
The hyperlinks can be added manually easily, but still manual labor.

So I'm still thinking this should be redone manually, page for page.
Just copy/paste FLAT text on a newly made article and then make it look good.

Also i'm not a big fan of mediawiki...as an admin you can't do shit on their 'control panel'.

[^Ripper]

Well the project files helped Sergio.
No more need of hack'n'slashing the code.
It even made it possible to remove the clutter from the templates and providing a true title for each article.
Or at least, for most articles.
However the linking problem still exists as it gets lots in translation.
I'll see if there is a function in mediawiki to fix those

[edit]
Nope...basically we're fucked...
It has to be done manually...
- Import index
- Then click/create every article and paste the plain unformatted text
- Color/fontsizing, upload images, highlight links to create wikilinks (even if they don't exist yet, as it makes it easier to make).

Doing this automatically created a lot of errors to correct :/
Its faster to just make em from scratch.
[/edit]

[Sergio Nova]

Good to know it can help.

[^Ripper]

The mediawiki thing isn't going to happen.
So i brought back the old Tikiwiki from before.
I can write a script that replaces all those tags and possibly fix those links and import them into the database.

Finger crossed

[^Ripper]

Well the conversion has gone better then i imagined.
there are still errors, but much less.
Oh yeah i ditched mediawiki and gone back with tikiwiki
http://www.rippersanime.com/Wiki/
I dug this up from my backups

The main problem is when text is between [ ] tags it messes up as tikiwiki requires it to be [[ ]
Let me see if i can fix this without disturbing the converted links that are [ ] tags.

[edit]
Not too much work and seems to have done the trick!
I'll open up the wiki when i feel its ready for editing.
[/edit]

[Sergio Nova]

It is evolving, but the images disappeared. The same happened to the texts resuming the volumes.

[^Ripper]

Fixed the images, stupid error :/

Anyway THIS is what the import can do.
Nearly all styling tags have been imported.

Things that did not make it:

- Text with variable em size got resized standard 1 em.
- Header size either standard or H1
- Some foreign text, or text with symbols not in the ISO-8859-1 Latin character set. (Check!)
- hack spacings like   or underlines with the same background/forground color.
- All html tags have been converted into tikiwiki tags or just removed.
- Some pages with links need to be redone (Like Zwölf) as the title reads wrong.

So basically, check/compare every page for errors and some minor styling issues.

The index page is flat and will be styled again.
I had to do this to fix all those links in the wiki.

[Sergio Nova]

1. The first topic I opened to check was Alita. Coincidentally, it still lacks the images.
2. How does it not accept Zwölf but accepts Schrödinger’s cat? The special character is the same. Besides, all the characters used in the glossary are listed in ISO-8859-1. Once the lines are repeated, is it not possible simply to erase the first line? That would solve the matter, at least in Zwölf's case, as the second line is correct.

[^Ripper]

The problem is the links.

Code: Select all

<a href="abc.htm">link text</a>

the webhtml publish
Needs to be converted to

Code: Select all

((abc.html|link text))

tikiwiki
This is no problem.
What is, is the pagename vs filename.
They need to be converted otherwise you get wiki/index.php?page=blabla.htm with matching blabla.htm text on top of the page.
This is where zwolf went wrong and others too.
So i need to convert blabla.htm -> Bla Bla (for page title) then convert all blabla.html references in all pages to Bla Bla in the wiki format.
Otherwise no link will work.
Anyway i did this with the index since it had a nice translation of blabla.htm with Blabla as text.
But its incomplete.
So i have to manually extract all the failed ones, clean them up, put them in the array with the rest and run the script again.

Thats for tomorrow.
Coding is fun
Took ~30 perl regex to fix all those tags.

[^Ripper]

Indexed all the pages it seems

The only thing is we have double titles.
As most (if not all) pages have the same wiki page name and topic

Alita

Alita

for example.
Now i can filter those out, so the wikipage name is automatically the topic and only topic.
Again, for tomorrow

Its getting along nicely

[Sergio Nova]

As far as I can see, it is really going well.