I’m still using his index data (mostly), so some of the links may no longer be accurate. It’s definitely missing all the features that have been added since 2007. But I will try to incrementally improve it over the course of this week. If I don’t, well, at least the links work.
Whenever a new release of I7 comes out won’t all the links break? It looks like the page numbers are just increasing numbers, so if an extra page is added early they’ll all be put out of sync. Ick.
The first script scans the I7 doc pages and generates a mapping of page titles to filenames. The second script uses that mapping and the index file to generate the final output. So when a new I7 comes out, I just have to re-run both of them.
(And then adjust for any pages whose names have changed. But this is a small fraction. Even between 3L95 and 6E72, most of the index data and page titles remained usable.)
I’m sure it can be overcome fairly easily, but the biggest problem I think is that any links to the documentation will be broken with the next release.
Okay, more serious note: Is it worth having index entries that refer to the recipe book? What about to specific examples in manual pages (as opposed to just linking to the page)? Obviously it would be nice to have these capabilities, but I’m asking whether they’d really be used – is there an index topic for which the best answer is to the recipe book.
I would vote yes, at least for some cases. My impression is that a lot of newbie questions are based around trying to create certain kinds of common things (cars, cell phones, elevators, etc) which there are all Recipe Book examples for. I certainly don’t think every example needs to be indexed this way, and of course someone would have to hand-curate the most useful ones. But time constraints aside it seems worthwhile.
There does seem to be a few examples that keep being mentioned in forums on a fairly regular basis. The one with noses and the possessive case is one. The one with rooms-inside-rooms was another. Maybe about 3 or 4 others. I’m blanking.
I’m still updating the index file. I’ll finish doing that – it may be another week or two – and link to all the source code and index data that I’m using.
After that, of course, I’ll go on to something else and forget about it. Somebody remind me in April, or whenever the next I7 release is.
There are actually two things this could be referring to, I guess – one is the Walls and Noses example in the recipe book, about how to refine the wording of a parser question, and the other is the example in Punctuation Removal that shows how to parse things like “Joe’s tie” on the basis of the fact that it’s a tie Joe happens at the moment to be wearing. Examples in the built-in extensions aren’t really visible unless the user already knows to go look at that extension.
Ah, the Punctuation Removal one is the one that I’ve unknowingly posted in about five threads (after Jesse showed it to me) – except using an “After reading the command” to turn apostrophe-s into space-s instead of using Punctuation Removal. Good to know.
eblong.com/zarf/i7index/ is now up to date. I’ve added a link to the data and scripts that I used to generate it.
It doesn’t have any Recipe Book entries, however. I’ve just covered Writing With Inform, as best I can. Feel free to use this thread to shower me with suggestions for additions.
@zarf , this is so useful, but of course, would be more useful if all the links weren’t broken again.
I don’t know how automated your process for generating th index is, but would it be possible to either a) update it again, or b) pass it to whomever is generating the IF7 docs to provide on the docs page (preferable)?
I’ll update this thread as I go.