snowflake: fix timestamps at extreme ranges #3220
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
rockwotj
commented
Feb 28, 2025
| @@ -145,7 +145,7 @@ func fls128(n Num) int { | |||
| if n.hi != 0 { | |||
| return 127 - bits.LeadingZeros64(uint64(n.hi)) | |||
| } | |||
| return 64 - bits.LeadingZeros64(n.lo) | |||
| return 63 - bits.LeadingZeros64(n.lo) | |||
There was a problem hiding this comment.
This was the bug fix, the rest is mostly tests or refactoring/clean up :)
There was a problem hiding this comment.
Heh, that reminded me of https://pkg.go.dev/math/rand#Int63 :)
rockwotj
force-pushed
the
snowflake-tz
branch
from
1634668 to
fedf8ff
Compare
rockwotj
changed the title
mihaitodor
approved these changes
Feb 28, 2025
Comment on lines
61
to
79
| // Compare returns -1 if a < b, 0 if a == b, and 1 if a > b. | ||
| func Compare(a, b Num) int { | ||
| if a.hi < b.hi || (a.hi == b.hi && a.lo < b.lo) { | ||
| return -1 | ||
| } | ||
| if a.hi == b.hi && a.lo == b.lo { | ||
| return 0 | ||
| } | ||
| return 1 | ||
| } | ||
|
|
||
| // CompareUnsigned returns -1 if |a| < |b|, 0 if a == b, and 1 if |a| > |b|. | ||
| func CompareUnsigned(a, b Num) int { | ||
| if uint64(a.hi) < uint64(b.hi) || (a.hi == b.hi && a.lo < b.lo) { | ||
| return -1 | ||
| } | ||
| if a.hi == b.hi && a.lo == b.lo { | ||
| return 0 | ||
| } | ||
| return 1 | ||
| } |
There was a problem hiding this comment.
I wonder if we can leverage https://pkg.go.dev/cmp#Compare for these. Not sure if that's possible
There was a problem hiding this comment.
Yes we can make this much more readable :) Thanks for this!
| @@ -145,7 +145,7 @@ func fls128(n Num) int { | |||
| if n.hi != 0 { | |||
| return 127 - bits.LeadingZeros64(uint64(n.hi)) | |||
| } | |||
| return 64 - bits.LeadingZeros64(n.lo) | |||
| return 63 - bits.LeadingZeros64(n.lo) | |||
There was a problem hiding this comment.
Heh, that reminded me of https://pkg.go.dev/math/rand#Int63 :)
rockwotj
force-pushed
the
snowflake-tz
branch
from
6262bc7 to
ac75908
Compare
This all boiled down to an amazing off by one error in calculating the MSB when there are less than 64 bits in the int128. Added a bunch of tests around these values at every layer of the stack, each was really an artifact of my debugging :)
When preserving logical types, preserve time values as timestamps instead of duration strings, timestamps are much more natural to work with in bloblang.
rockwotj
force-pushed
the
snowflake-tz
branch
from
ac75908 to
ca5139b
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fix a bug where timestamps at the extreme ends of the spectrum could be encoded incorrectly due to a bug in
int128.Div. Also fix Time types in avro to be something more usable.