How to get latest stable Ruby version from command line? - ruby

Question

I'm interested in querying the latest stable version of Ruby from the command line. I'm primarily interested in a string of the form "1.9.3p327".
Update #1
To clarify, the goal is to consistently query the latest stable version, whether that be 1.9.3p327 or 3.0.2p392.

Answer 1

Something like:
curl 'http://ftp.ruby-lang.org/pub/ruby/1.9/' | ruby ./extract-and-print-max-patchlevel.rb
Where the implementation of extract-and-print-max-patchlevel.rb script is an exercise for the reader but here's a start:
#!/usr/bin/env ruby
maxpatch=0
maxstr=nil
STDIN.each_line do |line|
next unless line =~ /1\.9\.3-p(\d+)\b/
patch = $1.to_i
if patch > maxpatch
maxpatch = patch
maxstr = $&
end
end
puts maxstr
Note that it assumes Ruby 1.9.3 is the latest, so you might want to revisit that.

Answer 2

Came back to this after starting to use rbenv. If you have rbenv installed, the following one-liner does the trick:
#!/usr/bin/env ruby
puts `~/.rbenv/bin/rbenv install --list`.split("\n").map{|item| item.strip}.select{|item| item[/^\d*\.\d*\.\d*/]}.reject{|item| (item.include? '-') && !(item =~ /-p\d*$/)}.last

Answer 3

You can play with sed or awk. For example:
curl "http://ftp.ruby-lang.org/pub/ruby/1.9/" | sed -E 's/^.*"(ruby-)(1.9.3-p[0-9]+)(.*)".*$/\2/' | sed -e '/^1.*/!d' | sort | sed '$!d'
I'm not sed expert and here could be better solution but this line works.

Answer 4

Choosing between #maerics' and #YevgeniyAnfilofyev's answers, I found the Ruby one more authentic and turned it into one-liner, assuming that the list from curl is already sorted, so we simply need the last line meeting the condition:
curl 'http://ftp.ruby-lang.org/pub/ruby/1.9/' 2> /dev/null | ruby -e "puts STDIN.lines.map { |x| /1\.9\.3-p\d+\b/.match(x) }.compact.last[0]"
Or, Ruby part expanded:
puts STDIN.lines.map do |x|
/1\.9\.3-p\d+\b/.match(x) # After such map we will get an array of MatchData or nils
end.compact.last[0] # Then we remove nils using Array#compact,
# get last MatchData and retrieve matched string with [0]
Following a good tradition, adding sorting to this script left as an excercise for the reader :)
UPD: Of course, hard-coded 1.9.3 is bad, but if we replace each digit 1, 9 and 3 with \d, the regex seems to become more or less independent. The other problem is that we only look into .../ruby/1.9 folder of that ftp. This may be fixed if we, instead, look into .../ruby folder first, and find all version-numbered folders with regex /\d\.\d\//. Then repeat the above query, joining results from all folders. But, of course, this already cannot made into a one-liner...