require "strscan"

module FBO
  class Parser
    TAG_PATTERN = /<\w+>/m
    PATTERN = /<([A-Z]+)>.*<\/\1>/m

    def initialize(file, tag_pattern: TAG_PATTERN)
      @file = file
      @tag_pattern = tag_pattern
    end

    def notices
      if @notices.nil?
        parse_notices(split_file_contents)
      end
      @notices
    end

    private

    def split_file_contents
      contents = @file.read
      contents = cleanup_contents(contents)
      scanner = StringScanner.new(contents)
      text_notices = []

      while !scanner.eos?
        initial_tag = scanner.check_until(@tag_pattern)
        if initial_tag
          initial_tag.strip!
        else
          break
        end

        text_notice = scanner.scan_until(/#{ closing_tag_for(initial_tag) }\s*/)
        if text_notice
          text_notices << text_notice.strip
        else
          break
        end
      end
      text_notices
    end

    def parse_notices(text_notices)
      @notices = []
      text_notices.each do |text|
        if FBO::Parser::PresolicitationHandler.is_presolicitation?(text)
          @notices << FBO::Parser::PresolicitationHandler.parse(text)
        elsif FBO::Parser::CombinedSolicitationHandler.is_combined_solicitation?(text)
          @notices << FBO::Parser::CombinedSolicitationHandler.parse(text)
        elsif FBO::Parser::SourcesSoughtHandler.is_sources_sought?(text)
          @notices << FBO::Parser::SourcesSoughtHandler.parse(text)
        elsif FBO::Parser::AmendmentHandler.is_amendment?(text)
          @notices << FBO::Parser::AmendmentHandler.parse(text)
        elsif FBO::Parser::ModificationHandler.is_modification?(text)
          @notices << FBO::Parser::ModificationHandler.parse(text)
        elsif FBO::Parser::AwardHandler.is_award?(text)
          @notices << FBO::Parser::AwardHandler.parse(text)
        else
          @notices << FBO::Parser::UnknownHandler.parse(text)
        end
      end
    end

    def cleanup_contents(contents)
      contents
        .encode('UTF-16le', :invalid => :replace, :replace => '')
        .encode('UTF-8')
        .gsub(/\r\n/, "\n")
        .gsub(/^M/, "")
    end

    def closing_tag_for(tag)
      tag.sub(/</, "</")
    end
  end
end