Okay after a bit of testing I've found that the problem with the loop is that it will successfully complete one pass through the entire keytable but will remain stuck on the last index. Basically, map_next() keeps returning the very last index even after that's already been done once, and thus generating an infinite loop.
Just in case anyone else is interested (doubt it), I've managed to solve the problem by modifying my loop to this:
Code:
new szKey[512];
new szLastKey[512];
map_first(table, "", szKey, strlen(szKey));
while(map_isfilled(table, szKey) && !equal(szLastKey, szKey))
{
// do whatever
copy(szLastKey, 511, szKey);
map_next(table, szKey, szKey, strlen(szKey));
}
Now unfortunately, this also causes a slower runtime. The original (failed) loop was O(n), which happens to be great. While the equal method here I assume is a bit-by-bit comparison - which shouldn't affect the overall performance too much - the copy method I'm not as certain about but I won't waste my time looking it up in the source code.
Still, if anyone has an alternative to this that changes it to pure O(n) efficiency, I'd be quite interested. Not that I'd be using the keytable to store more than a few thousand values, but until you hit the limit where you can't make it any faster...