lastRowID() doesn't work in -header mode

[bartve]

lastRowID() doesn't handle -header mode yet.

[WebReflection]

lastRowID is a special case that returns anything but the last inserted row id. Can I ask you which problem are you trying to solve? Is it about parsing? in such case maybe the parseInt could be probably used as default … or something else? Thanks

[WebReflection]

uh … never mind, it's about the mode. I'll check that, sorry for the confusion (however, having it as integer is probably better too)

[WebReflection]

latest has tests in place and should be fixed. Feel free to re-open if you have problems, thanks.

[bartve]

Thanks for fixing! Wouldn't this simpler code work just the same? Just curious why you've made your code so elaborate with type checks and hasOwnProperty etc.

var row = result[0], id;
// You can iterate both an array and object the same way, 
// with objects there's no way of telling the order, but it's always 1 row, so it doesn't matter.
for (var key in row) {
    id = row[key]; // Could do a pareseInt(row[key],10)
    break; // Not even needed as there's a LIMIT 1 in the query.
}
(callback || log).call(self, id);

Not sure about the parseInt as the ROWID is a signed 64 bit integer (http://www.sqlite.org/autoinc.html) and Javascript doesn't seem to handle the full range of that http://stackoverflow.com/q/9643626 Or do we only need a 32 bit Javascript int as in SQLite it's signed and usually only positive ROWIDs are used? I'm not too familiar in this territory.

[WebReflection]

that works for me and for you, then somebody might have the brilliant idea to enrich the Array.prototype that might show up in there instead of 0 or somebody else might have change the Object.prototype without forcing non enumerability. dblite would like to ensure consistent behavior across environments, node.js versions, etc etc, and since this is usually a one shot operation I think my change is safe enough and not so bad in terms of RAM or CPU cycles. About the parseInt, yes, I've realized I should not do that in here.

[bartve]

Thanks for the comments. It wasn't criticism, just interested in why the fluffy code.

[WebReflection]

I could have gone smarter with a row[Object.keys(row).shift()] more explicit, safe, and concise than yours but that would have created a pointless Array for the GC and I was happy to avoid it ^_^

might be premature or too extreme measure though, I might update later on. Right now it works which is what matters ;-)