prohibit duplicate key columns

[ben-schwen]

Closes #4888
Closes #4891

[github-actions]

• HEAD=duplicated_key_columns slower P<0.001 for memrecycle regression fixed in #5463
  

Generated via commit 7253f69

Download link for the artifact containing the test results: ↓ atime-results.zip

Task	Duration
R setup and installing dependencies	2 minutes and 41 seconds
Installing different package versions	47 seconds
Running and plotting the test cases	5 minutes and 23 seconds

[MichaelChirico]

Gemini identified some remaining ways for ambiguity to creep in:

dt = data.table(a=1:2, a.1=3:4, val=10:11)
dt[, .(a.1, sum(val)), keyby=.(a, a)]
# Key: <a, a.1>
#        a   a.1   a.1    V2
#    <int> <int> <int> <int>
# 1:     1     1     1    10
# 2:     2     2     2    11

dt = data.table(a=1:2, b=3:4, key="a")
dt[, .(a, a)]
# Key: <a>
#        a     a
#    <int> <int>
# 1:     1     1
# 2:     2     2
subset(dt, select=c(a, a))
# Key: <a>
#        a     a
#    <int> <int>
# 1:     1     1
# 2:     2     2

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.04%. Comparing base (d4974e9) to head (7253f69).
⚠️ Report is 2 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #7760   +/-   ##
=======================================
  Coverage   99.04%   99.04%           
=======================================
  Files          87       87           
  Lines       17064    17087   +23     
=======================================
+ Hits        16901    16924   +23     
  Misses        163      163           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[ben-schwen]

Gemini identified some remaining ways for ambiguity to creep in:

dt = data.table(a=1:2, a.1=3:4, val=10:11)
dt[, .(a.1, sum(val)), keyby=.(a, a)]
# Key: <a, a.1>
#        a   a.1   a.1    V2
#    <int> <int> <int> <int>
# 1:     1     1     1    10
# 2:     2     2     2    11

dt = data.table(a=1:2, b=3:4, key="a")
dt[, .(a, a)]
# Key: <a>
#        a     a
#    <int> <int>
# 1:     1     1
# 2:     2     2
subset(dt, select=c(a, a))
# Key: <a>
#        a     a
#    <int> <int>
# 1:     1     1
# 2:     2     2

I've added these cases, but I'm sure we will encounter special versions of these again

[MichaelChirico]

Yep, sharing because those look pretty easy to encounter in practice. Remaining ones will be more baroque. A fresh session finds nothing, so I think this is good now.

[MichaelChirico]

test is somewhat confusing, is CJ(..., sorted=FALSE) not enough?

[MichaelChirico]

save names(ans)[seq_along(grpcols)] to a variable