1395071 - IndexedDB2.0: IDBObjectStore.get() and delete() fails if argument is a binary key

Status: RESOLVED FIXED

(Core :: Storage: IndexedDB, defect)

Description

[David Fahlander]

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.90 Safari/537.36 OPR/47.0.2631.71

Steps to reproduce:

Call IDBObjectStore.get() or IDBObjectStore.delete() using a typed array or an array buffer. It throws directly.

A repro is here: http://dexie.org/test/firefox-bug-binary-keys/



Actual results:

The methods throws DataError even though binary keys should be allowed also for primary keys.


Expected results:

I should have been able to get() or delete() an object by its binary primary key.

Comment 1

[Andrew Overholt [:overholt]]

Bevis, you did work to support this in bug 1271500. Maybe you know what's up here?

Comment 2

[Bevis Tseng[:bevis][:btseng](Exited)]

I'll look into it. keep the NI on me.

Comment 3

[Bevis Tseng[:bevis][:btseng](Exited)]

Thanks for reporting this, David.

After further review, I found that there is another place(IDBKeyRange::FromJSVal) to check the type of the passed JS value that wasn't identified when fixing bug 1271500:
http://searchfox.org/mozilla-central/rev/51b3d67a5ec1758bd2fe7d7b6e75ad6b6b5da223/dom/indexedDB/IDBKeyRange.cpp#101

I'll update a fix to unify the type checking in single place(Key::EncodeJSValInternal):
http://searchfox.org/mozilla-central/rev/51b3d67a5ec1758bd2fe7d7b6e75ad6b6b5da223/dom/indexedDB/Key.cpp#213

Comment 4

[David Fahlander]

I'm happy to help. Thanks for jumping into it so quickly!
/David

Comment 5

[Bevis Tseng[:bevis][:btseng](Exited)]

Attachment: (v1) Patch: Unify the key validation using Key::EncodeJSValInternal.

Comment 6

[Andrea Marchesini [:baku]]

Comment on attachment 8903505 [details] [diff] [review]
(v1) Patch: Unify the key validation using Key::EncodeJSValInternal.

Review of attachment 8903505 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/indexedDB/IDBKeyRange.cpp
@@ +95,3 @@
>  
> +  // Unwrap an IDBKeyRange object if possible.
> +  if (aVal.isObject() &&

if (obj && ...

Comment 7

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/90691dd0694f
Unify the key validation using Key::EncodeJSValInternal. r=baku

Comment 8

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

https://hg.mozilla.org/mozilla-central/rev/90691dd0694f

Comment 9

[Gerry Chang [:gchang]]

Too late for 56. Mass won't fix for 56.