config.yml

development:
  adapter: mysql2
  encoding: utf8
  database: my_database
  username: root
  roles:
    - admin
    - developer
    - guest
  password:
  users:
    - 
      name: pineapple
      role: admin
    -
      name: umbrella
      role: developer

output

development_adapter=("mysql2")
development_encoding=("utf8")
development_database=("my_database")
development_username=("root")
development_roles+=("admin")
development_roles+=("developer")
development_roles+=("guest")
development_users__name+=("pineapple")
development_users__role+=("admin")
development_users__name+=("umbrella")
development_users__role+=("developer")

parse_yaml2.sh

## derived from https://gist.github.com/epiloque/8cf512c6d64641bde388
## works for arrays of hashes, as long as the hashes do not have arrays
parse_yaml2() {
    local prefix=$2
    local s
    local w
    local fs
    s='[[:space:]]*'
    w='[a-zA-Z0-9_]*'
    fs="$(echo @|tr @ '\034')"
    sed -ne "s|^\($s\)\($w\)$s:$s\"\(.*\)\"$s\$|\1$fs\2$fs\3|p" \
        -e "s|^\($s\)\($w\)$s[:-]$s\(.*\)$s\$|\1$fs\2$fs\3|p" "$1" |
    awk -F"$fs" '{
      indent = length($1)/2;
      if (length($2) == 0) { conj[indent]="+";} else {conj[indent]="";}
      vname[indent] = $2;
      for (i in vname) {if (i > indent) {delete vname[i]}}
      if (length($3) > 0) {
              vn=""; for (i=0; i<indent; i++) {vn=(vn)(vname[i])("_")}
              printf("%s%s%s%s=(\"%s\")\n", "'"$prefix"'",vn, $2, conj[indent-1],$3);
      }
    }' | sed 's/_=/+=/g'
}

This sort of works for arrays of hashes, except the hashes cannot themselves contain arrays.
Also you need to code around it. Basically you get N bash arrays , one for each property in the hash.

this one does not handle string with doublequotes inside

This is a really great function. I needed mine to work with dashes "-" in the keynames so I added \- to the regex on line 9:
w='[a-zA-Z0-9_\-]*'
This results in illegal Bash variable names. Back to the drawing board. Still a great function though.

@SomeGuy54321 I have the same requirement, if you make that change to line 9, you can add these to the final sed:

sed 's/_/__/g' | sed 's/-/_/g'

So the convention becomes space = double underscore and dash = underscore.