#!/usr/bin/env ruby

require 'optparse'
require 'methadone'
require 'fasta_read'

class App
  include Methadone::Main
  include Methadone::CLILogging

  main do |assembly, chromosome, cstart, cend|
    path = "#{assembly}/#{snp_or_unmasked(options)[0]}/#{chromosome + snp_or_unmasked(options)[1]}.fa"
    exit_now_with!(assembly_or_chromosome(assembly, chromosome)) unless File.exist?(path)
    fasta = IO.read(path)
    debug("About to read #{assembly} #{snp_or_unmasked(options)[0]} chr#{chromosome}:#{cstart}-#{cend}")
    sequence = FastaRead::Sequence.new(fasta, chromosome, cstart, cend).process
    if options[:output]
      File.open(options[:output], "w") do |file|
        file.puts sequence
      end
    else
      puts sequence
    end
    info "#{sequence.length} base pairs"
  end

  def self.snp_or_unmasked(options)
    options[:snp] ? ["snp", ".subst"] : ["unmasked", ""]
  end

  def self.assembly_or_chromosome(assembly, chromosome)
    Dir.exist?("#{assembly}") ? [chromosome, "chromosome"] : [assembly, "assembly"]
  end

  def self.exit_now_with!(value_and_argument)
    exit_now!(1, "the '#{value_and_argument[0]}' #{value_and_argument[1]} doesn't exist in directory structure")
  end
  # supplemental methods here

  # Declare command-line interface here

  description <<-EOF
    Extract DNA Fasta sequence from assembly files.

    The program depends on being run at the top of a directory tree containing .fa files. The .fa files should be of the form where each file maps to a single chomosome.
    The directory tree will have separate branches for SNPs and unmasked files.

    For example, provided the following tree the command expects to be run inside the 'fasta' directory:

        └── fasta
            ├── hg19
            │   ├── snp
            │   │   ├── chr1.subst.fa
            |   │   ├── chr2.subst.fa
            |   │   ├── chr3.subst.fa
            |   │   └── ....
            │   └── unmasked
            │       ├── chr1.fa
            |       ├── chr2.fa
            |       ├── chr3.fa
            |       └── ....
            └── mm10
                ├── snp
                │   ├── chr1.subst.fa
                │   ├── chr2.subst.fa
                │   ├── chr3.subst.fa
                │   └── ....
                └── unmasked
                    ├── chr1.fa
                    ├── chr2.fa
                    ├── chr3.fa
                    └── ....
    EOF

  #
  # Accept flags via:
  # on("--flag VAL","Some flag")
  # options[flag] will contain VAL
  #
  # Specify switches via:
  on "-o OUTPUTFILE", "--output", "outputs sequence to a file"
  on "--snp", "return the sequence from the SNP-masked assembly"

  # on("--[no-]switch","Some switch")
  #
  # Or, just call OptionParser methods on opts
  #
  # Require an argument
  arg :assembly,   "assembly name (hg19, mm10, etc.)"
  arg :chromosome, "id of chromosome (chr1-chr22 or chrx/chry)"
  arg :cstart,     "start coordinate (inclusive) within the chromosome"
  arg :cend,       "end coordinate within the chromosome"
  #
  # # Make an argument optional
  # arg :optional_arg, :optional

  version FastaRead::VERSION

  use_log_level_option

  go!
end