SCI Companion documentation

[Kawa]

If it's well-formed, it should be easy enough to process into a nice bare intermediate. And if it was originally generated from source data, it should at least be regular (and thus parseable), if not well-formed.

[Collector]

I have them. The are what I put online. http://sierrahelp.com/SCI/SCIStudio3Help/ If it looks like what you need I can put together a zip for you.

[troflip]

That just looks like html to me.

[troflip]

I realized I had some of the SCI0 kernel functions documented by myself a while back, so I'm using that. The missing once I'm copying manually from Brian's docs. Oooph... the grammar and spelling mistakes are killing me!

[Collector]

What I have on line is from what Brian had on his site. I only played around with the formatting and made it a little more CSS compliant. I always assumed that it is what Brian used for compiling his CHM, but modified fro online use.

[troflip]

Yeah, ideally there would be something more "core" without all the html formatting. It's probably not *too* hard to just copy paste the raw text for the missing pages (mainly compiler and language features) and give them a little structure. Certainly less work than writing something to scrape that html and come up with something that works in restructuredText (which is the document format I'm using for the documentation).

[lskovlun]

Certainly less work than writing something to scrape that html and come up with something that works in restructuredText (which is the document format I'm using for the documentation).

Have you given pandoc a go? It supports numerous formats...
http://pandoc.org/installing.html

[Collector]

My modifications and formatting to the help file are done mostly via CSS. It wouldn't take that much to do a 'site wide' search and replace. The vast majority of that style is for syntax highlighting. You would have to scrap the index and replace the HTML TOC, but most of the pages themselves should not be that involved to make usable.

[troflip]

Great idea with pandoc Lars... unfortunately, it did a terrible job :-(. It would be more work to fix up that output up than just grabbing the raw text from a browser window. All Brian's paragraphs are html tables, so it tried to make an actual table and completely messed that up. And it left in some <div> tags, lol. Quite a mess.

Collector, I'm not sure you understand what I'm trying to do. I just want to extract the text and hopefully some document structure. That's really not as trivial as search and replace. I mean, I'm sure it's possible with some regular expressions magic, but it would end up being more work than just doing everything manually.

My hope (I guess) was that Brian had used some "mark down" document format to generate the html files, and those files were still around somewhere. But maybe he didn't, he might have just written them in an html editor like FrontPage or something.

[Kawa]

I just want to extract the text and hopefully some document structure. That's really not as trivial as search and replace.

Code: [Select]

<name>CelHigh</name>
<call>number CelHigh(number view, number loop, number cel)</call>
<desc>By specifying a the view, loop and cel as a parameter, it looks up the height of the cel and returns it.</desc>
<samp>var celWidth, celHeight)
  = celWidth CelWide(100 1 3) // Looks up the width of loop 1, cel 3 in VIEW.100
  = celHeight CelHigh(100 1 3) // Looks up the width of loop
  1, cel 3 in VIEW.100</samp>That's what I produced just now with a small C# app that I fed a single page. Making it emit Markdown is trivial.

I mean, I'm sure it's possible with some regular expressions magic, but it would end up being more work than just doing everything manually.

The reason you shouldn't use RegEx on HTML is because HTML isn't a regular language. The reason you could RegEx your way through these docs is that they're reasonably regular.

[Collector]

When making CHM files I always just did simple HTML files to compose the content. It has been a while since I did anything with the Studio help file, but now that you mention it I remember that he did use a bunch of tables just for lay out. I did remove a lot of them, but I only had patience to remove the ones that were completely unnecessary. I suspect that you are right about FrontPage. Looking at his old sites his HTML leaves a lot to be desired.

[troflip]

At any rate, it wasn't too much trouble to just copy the text from the browser (then all those tables go away) and add the little bit of needed structure. I've completed "porting" over the stuff I need. The only thing I'm not going to do is the SCI0 template game class system.

In doing so, I've found a number of broken or incorrect links in his docs. It looks like he did a lot of that stuff manually, when it could have been automated. Then again, it was 2003, or whatever.

Hopefully my docs will be a little better in that respect, but I'm sure I'll still end up with typos and such .

[Collector]

Yeah, when I was redoing the online version I gave up on correcting typos and misspellings since all of the SCI and programming stuff made too many things to trip up a spell checker. It would have to be throughly proofread and manually done.

[troflip]

Updated docs, if anyone's interested. Includes docs for all the SCI1.1 class system and kernels, and the compiler. I used a new theme, and they look really slick now I think.

[Kawa]

Looks neat. I like the new logo.