On Sat, Aug 20, 2011 at 07:57:20PM -0400, Dave Reisner wrote: > On Sun, Aug 21, 2011 at 01:51:44AM +0200, Lukas Fleischer wrote: > > Coerce "ID", "FirstSubmitted" and "LastModified" into integers to ensure > > json_encode() serializes them as numeric values. This means that there > > will be a minor API break. There doesn't seem to be a better way to do > > this properly, though. Fixes FS#25693. > > > > Signed-off-by: Lukas Fleischer <[email protected]> > > --- > > web/lib/aurjson.class.php | 3 +++ > > 1 files changed, 3 insertions(+), 0 deletions(-) > > > > diff --git a/web/lib/aurjson.class.php b/web/lib/aurjson.class.php > > index 277c824..ad99529 100644 > > --- a/web/lib/aurjson.class.php > > +++ b/web/lib/aurjson.class.php > > @@ -126,6 +126,9 @@ class AurJSON { > > while ( $row = mysql_fetch_assoc($result) ) { > > $name = $row['Name']; > > $row['URLPath'] = URL_DIR . substr($name, 0, 2) . "/" . > > $name . "/" . $name . ".tar.gz"; > > + $row['ID'] = intval($row['ID']); > > + $row['FirstSubmitted'] = intval($row['FirstSubmitted']); > > + $row['LastModified'] = intval($row['LastModified']); > > > > if ($type == 'info') { > > $search_data = $row; > > -- > > 1.7.6 > > > > I suppose I was a bit hasty in my bug report. If we're going to break > things, can we break even more? There's no reason we couldn't also > coerce: CategoryID, NumVotes, and OutOfDate (which might be better as a > boolean).
Yeah, if we break stuff, we should do it properly :) I'll have a look at the other fields. And I'm already looking forward to getting death threats from all the AUR helper developers around...
