Few possible bugs

[dsifford]

Hey Frank,

I've been tracking uncaught exceptions across all installations of my application and these three seem to be coming up regularly:

(Edit 10/15/16 - Fixed.) dsifford/academic-bloggers-toolkit#175
(Edit 10/16/16 - Wontfix.) dsifford/academic-bloggers-toolkit#176
dsifford/academic-bloggers-toolkit#177

I've played around with this and I just can't seem to figure out what could be happening? Any ideas?

[fbennett]

The first seems like an error on bad input (a citation object should always have a "properties" property, and the processor won't strip it).

On the other two, the only thing I can think of is that the "sys" object might be getting clobbered during instantiation there, if the user has rapidly switched between styles and the processor is running async.

[dsifford]

@fbennett Thanks for the feedback. Totally throwing a hail mary at this point. I've been getting these three in particular the most (generally about ~3 times a week from a range of different sites). No idea what could be going on...

I've tried reproducing a few times but can't.

What's interesting is that, to my knowledge, there is no way citeproc could be getting citation object without "properties"...

This line makes a call to this function just before the call to processCitationCluster is made and that function always returns an object containing a properties parameter.

Suuuuper super strange.

I'll keep you posted if I'm able to turn something up.

[dsifford]

Update: Figured out dsifford/academic-bloggers-toolkit#175

This one turned up when people altered the html of citations. In some cases, the processor would be getting a citationId that didn't exist (so it would throw an error) and in other cases, the processor would be getting a citation object without an ID altogether (which would obviously also trigger an error).

I've added two checks for this on my end: one to guard against extra citations that don't belong, and the other to guard against citations that were stripped of their ID.

We can call that one good! Nothing you can do except maybe skip it on your end to prevent the exception. Your call there.

I'll keep you posted on the other two when I figure them out.

[dsifford]

Update: Figured out dsifford/academic-bloggers-toolkit#176

The gist: safari browser is terrible and makes me miserable.

Longer story: Safari blocks client-side http requests when the user has an invalid client certificate installed in their apple toolchain (even when the client cert is optional). There's nothing I can do to prevent this besides passionately urge all users to abandon safari completely (which would be a dream come true if that happened).