improve test 1864 by using latin1 encoded strings#2715
Conversation
…d on all the platforms
Codecov Report
@@ Coverage Diff @@
## master #2715 +/- ##
==========================================
+ Coverage 93.4% 93.59% +0.18%
==========================================
Files 61 61
Lines 12236 12498 +262
==========================================
+ Hits 11429 11697 +268
+ Misses 807 801 -6
Continue to review full report at Codecov.
|
mattdowle
left a comment
mattdowle
left a comment
There was a problem hiding this comment.
test(1864.3, DT1[J(utf8_strings)], DT1[J(latin1_strings)])
| setkey(DT2, y) | ||
| test(1864.1, DT1$x, sort(c(utf8_strings, utf8_strings), method = "radix")) | ||
| test(1864.2, DT2$y, sort(c(utf8_strings, utf8_strings), method = "radix")) | ||
| test(1864.3, DT1[J(utf8_strings)], DT1[J(latin1_strings)]) |
There was a problem hiding this comment.
Imagine that, for reason of a strange bug, both DT1[J(utf8_strings)] and DT1[J(latin1_strings)] returned NULL. The test would still pass. I like this pattern for that reason :
test(1864.3, DT1[J(utf8_strings)], ans<- data.table( .. known answer .. ) )
test(1864.4, DT1[J(latin1_strings)], ans)
There was a problem hiding this comment.
Perhaps use the integers in z to construct ans.
There was a problem hiding this comment.
You're right. Using an explicit answer is indeed better.
3 participants
So it can be tested on all the platforms. It improves the test for PR #2566.