add setting unicast bit #96
Conversation
Codecov Report
❗ Your organization needs to install the Codecov GitHub app to enable full functionality. @@ Coverage Diff @@
## master #96 +/- ##
==========================================
+ Coverage 97.99% 98.00% +0.01%
==========================================
Files 6 6
Lines 2046 2057 +11
==========================================
+ Hits 2005 2016 +11
Misses 41 41
|
|
See #94 (comment) |
XiXiangFiles
force-pushed
the
feature-qu-tag
branch
from
ab17088 to
eaafaac
Compare
|
@silverwind thx for your reviewing. I modify it to |
| } else { | ||
| q.class = classes.toString(classValue) | ||
| } | ||
|
|
||
| offset += 2 | ||
|
|
||
| const qu = !!(q.class & QU_MASK) |
There was a problem hiding this comment.
I think this section is still problematic (and there doesn't seem to be a test for decoding a question with the QU bit set) - if the bit is set, then we end up attempting a bitwise AND operation on the q.class string value (in line 1648 just below), which wipes out the string value (the result is 0).
Thank you very much for this effort - would you mind also adding a test for the decode function?
There was a problem hiding this comment.
Certainly, this change was quite some time ago. I recall this PR request occurred in 2018 or 2019. This weekend, I will take a look and make some adjustments.
| @@ -1608,8 +1608,12 @@ question.encode = function (q, buf, offset) { | |||
|
|
|||
| buf.writeUInt16BE(types.toType(q.type), offset) | |||
| offset += 2 | |||
| if (q.qu === undefined || q.qu !== true) { | |||
There was a problem hiding this comment.
this check does not make sense
Description:
for the unicast bit issue
I just rebase to the latest. previous pr: #49
related issues
#21