Add float_warnings patches for Ruby 2.0.0p64[5 7 8], 2.1.[8 9 10], 2.2.[4 5 6 7 8], 2.3.[0 1 2 3 4 5] and 2.4.[0 1 2]

[knugie]

Latest float_warnings patches.

Passing tests:
# Each of the following `ruby -e` lines prints:
#-e:1: warning: initializing BigDecimal with an instance of Float.
#-e:1: warning: calling .to_d on an instance of Float.

for MAJOR in "2"; do
  for MINOR in "0.0"; do
    for PATCH in "247" "353" "451" "481" "576" "594" "598" "645" "647" "648"; do
      rvm use $MAJOR.$MINOR-p$PATCH-float_warnings --patch patches/ruby/$MAJOR.$MINOR/p$PATCH/float_warnings/01-bigdecimal_float_warning.patch --install
      ruby -e "require 'bigdecimal'; require 'bigdecimal/util'; BigDecimal.new(123.4, 2) && 7.1.to_d"
      rvm remove $MAJOR.$MINOR-p$PATCH-float_warnings
    done
  done
  for MINOR in "1"; do
    for PATCH in "0" "1" "2" "3" "4" "5" "6" "7" "8" "9" "10"; do
      rvm use $MAJOR.$MINOR.$PATCH-float_warnings --patch patches/ruby/$MAJOR.$MINOR.$PATCH/float_warnings/01-bigdecimal_float_warning.patch --install
      ruby -e "require 'bigdecimal'; require 'bigdecimal/util'; BigDecimal.new(123.4, 2) && 7.1.to_d"
      rvm remove $MAJOR.$MINOR.$PATCH-float_warnings
    done
  done
  for MINOR in "2"; do
    for PATCH in "0" "1" "2" "3" "4" "5" "6" "7" "8"; do
      rvm use $MAJOR.$MINOR.$PATCH-float_warnings --patch patches/ruby/$MAJOR.$MINOR.$PATCH/float_warnings/01-bigdecimal_float_warning.patch --install
      ruby -e "require 'bigdecimal'; require 'bigdecimal/util'; BigDecimal.new(123.4, 2) && 7.1.to_d"
      rvm remove $MAJOR.$MINOR.$PATCH-float_warnings
    done
  done
  for MINOR in "3"; do
    for PATCH in "0" "1" "2" "3" "4" "5"; do
      rvm use $MAJOR.$MINOR.$PATCH-float_warnings --patch patches/ruby/$MAJOR.$MINOR.$PATCH/float_warnings/01-bigdecimal_float_warning.patch --install
      ruby -e "require 'bigdecimal'; require 'bigdecimal/util'; BigDecimal.new(123.4, 2) && 7.1.to_d"
      rvm remove $MAJOR.$MINOR.$PATCH-float_warnings
    done
  done
  for MINOR in "4"; do
    for PATCH in "0" "1" "2"; do
      rvm use $MAJOR.$MINOR.$PATCH-float_warnings --patch patches/ruby/$MAJOR.$MINOR.$PATCH/float_warnings/01-bigdecimal_float_warning.patch --install
      ruby -e "require 'bigdecimal'; require 'bigdecimal/util'; BigDecimal.new(123.4, 2) && 7.1.to_d"
      rvm remove $MAJOR.$MINOR.$PATCH-float_warnings
    done
  done
done

[pkuczynski]

Hi @knugie, could you shed some light on the source of this patches and what they supposed to fix?

[knugie]

Hi @pkuczynski, of course. float_warnings is supposed to help Ruby developers identify unexpected behaviour when initializing BigDecimal with a Float. The patches add two warnings. One in bigdecimal C code BigDecimal.new and one in bigdecimal/util Ruby code Float#to_d.

fmi, see my closed rvm PRs: https://github.com/rvm/rvm/pulls?q=is%3Apr+author%3Aknugie+is%3Aclosed
and the original talk that made me create float_warnings: https://slides.com/wolfgangteuber/saving-money-with-ruby/fullscreen#/

Btw, it's totally unrelated to Railsexpressruby. I simply use @skaes' pattern to provide patchsets.
I hope that makes sense, happy to explain in more detail if needed.

[pkuczynski]

Yeah, it makes perfect sense now. I was wondering if it's somehow connected with @skaes. I will merge it in, as it was merged before by @mpapis. This is not exactly what RVM intended to do, but I see a value in this feature...