assume glue is glue::glue

[MichaelChirico]

Without this, #2031 throws false positive lints.

I'm not sure why we didn't do this to begin with.

• If someone's using a function called glue() that doesn't come from {glue}, they should set interpret_glue = FALSE.
• If they're using two glue() functions, one of which is unrelated to glue::glue()
  • That's weird, why would you do that?
  • Even so, the only (?) implication is that we'll try and interpret the string input, if there is one, which
    • I guess has a small efficiency impact
    • Shouldn't error
    • Could conceivably cause a false negative if that glue() mask also uses {-interpretation-like formatting that happens to look exactly like a glue::glue() interpolation? Which again, why

[codecov-commenter]

Codecov Report

Merging #2032 (6fb9fb9) into main (2b62a3a) will decrease coverage by 0.01%.
The diff coverage is n/a.

❗ Current head 6fb9fb9 differs from pull request most recent head 105d814. Consider uploading reports for the commit 105d814 to get more accurate results

@@            Coverage Diff             @@
##             main    #2032      +/-   ##
==========================================
- Coverage   98.55%   98.55%   -0.01%     
==========================================
  Files         113      113              
  Lines        4995     4994       -1     
==========================================
- Hits         4923     4922       -1     
  Misses         72       72              

Files Changed	Coverage Δ
R/object_usage_linter.R	99.41% <ø> (-0.01%)	⬇️