As support this method within my reviews in aniDB I'd like to have the ability to link to other aniDB reviews (of the same reviewer) within the content of my reviews. Just that - no complete HTML code support.
Let's make an example. I want to link to my review for Aa! Megami-sama The Movie (id: 206) within my review from Aa! Megami-sama OVA (id: 27).
The meta code might look something like
Code: Select all
[review:206]<a href="http://anidb.info/perl-bin/animedb.pl?s ... att&aid=[b]206[/b]#2452">AA! Megami-sama - The Movie</a>
Deriving the "AA! Megami-sama - The Movie" string from the id "206" should be no problem.
The tricky thing would be to derive the "2452" from two other data - the "206" (which is the ID of the anime being reviewed) and the "73514" (which happens to be the userid of the user who wrote this review - this information is available in the context of the content where the meta tag were to be interpreted).
Apparently the link targets for reviews are created by some auto-increment column used as primary key for the reviews; it would have been much more flexible to use the pair of reviewer userid and series id as the primary key, and then generate the links from this pair of data (there can be no more than one review per reviewer in a review page, right? So <a name="userid"></a> would work fine as unique link target instead of <a name="review_primary_key"></a> ).
As things are now it might be difficult to change the (assumed) SQL table format to provide the additional information required for the link generation I'd like to have, but there's hope even for this (if my feature request were considered useful, that is). But if the required database changes were considered too complex then I'd be fine with an alternative syntax like
Code: Select all
[review:206#2452]AA! Megami-sama - The Movie[/review]It would be cool if the generator would check whether the review that serves as link target actually exists (and generate some string like "invalid reference to review for anime 206 by user 73514" instead if it doesn't) but I'm fine with not doing so as to save CPU time for the additional database access (a link to an unknown target would be handled appropriately by all known web browsers).