autotelik/datashift

# Copyright:: (c) Autotelik Media Ltd 2011
# Author ::   Tom Statter
# Date ::     Aug 2010
# License::   MIT
#
#  Details::  Base class for loaders, providing a process hook which populates a model,
#             based on a method map and supplied value from a file - i.e a single column/row's string value.
#             Note that although a single column, the string can be formatted to contain multiple values.
#
#             Tightly coupled with Binder classes (in lib/engine) which contains full details of
#             a file's column and it's correlated AR associations.
#
module DataShift

  require 'datashift/binder'

  module  Querying

    include DataShift::Logging
    extend DataShift::Logging

    include Delimiters
    extend Delimiters

    # Return arrays 'where' clause field(s), and the value(s) to search for
    #
    # If not contained within inbound_data, check the Inbound Column headings
    #
    # Where field embedded in row ,takes precedence over field in column heading
    #
    # Treat rest of the node as the value to use in the where clause e.g
    #   price:0.99 =>  Product.where( price: 0.99)
    #
    # Column headings will be used, if the row only contains data e.g
    #   0.99
    #
    def self.where_field_and_values(method_binding, inbound_data)

      part1, part2 = inbound_data.split(name_value_delim)

      heading_lookups = method_binding.inbound_column.lookup_list

      if part1.blank? && part2.blank?

        # Column completely empty - check for lookup supplied via the
        # inbound column headers/config

        # TODO: - how best to support multipel lookup fields?
        # part1 = heading_lookups.collect {|lookup| lookup.field }
        part1 = heading_lookups.find_by_operator
        part2 = heading_lookups.collect(&:value)

      elsif part2.blank?

        # Only **value(s)** in column, so use field from header/config field

        logger.debug("No part2e [#{part1}][#{part1.class}] - [#{part2}][#{part2.class}]")

        part2 = part1.split(multi_value_delim)
        part1 = method_binding.inbound_column.find_by_operator

      else
        part2 = part2.split(multi_value_delim)

      end

      logger.debug("Where clause [#{part1}][#{part1.class}] - [#{part2}][#{part2.class}]")

      [part1, [*part2]]
    end

    # Options:
    #
    #   :case_sensitive   : Default is a case insensitive lookup.
    #   :use_like         : Attempts a lookup using ike and x% rather than equality
    #
    def search_for_record(klazz, field, search_term, options = {})

      begin
        return klazz.send("find_by_#{field}", search_term) if options[:case_sensitive]

        return klazz.where("#{field} like ?", "#{search_term}%").first if options[:use_like]

        return klazz.where("lower(#{field}) = ?", search_term.downcase).first
      rescue StandardError => e
        logger.error("Querying - Failed to find a record for [#{search_term}] on #{klazz}.#{field}")
        logger.error e.inspect
        logger.error e.backtrace.last
      end

      nil
    end

    # Find a record for model klazz, looking up on field containing search_terms
    # Responds to global Options :
    #
    #   :add_prefix     : Add a prefix to each search term
    #   :case_sensitive : Default is a case insensitive lookup.
    #   :use_like       : Attempts a lookup using like and x% rather than equality
    #
    # Returns nil if no record found
    def get_record_by(klazz, field, search_term, split_on = ' ', options = {})

      split_on_prefix = options[:add_prefix]

      find_search_term = split_on_prefix ? "#{split_on_prefix}#{search_term}" : search_term

      logger.info("Scanning for record where #{klazz}.#{field} ~=  #{find_search_term}")

      begin
        record = search_for_record(klazz, field, find_search_term)

        unless record
          logger.info("Nothing found - trying split file_name to terms on [#{split_on}]")

          # try individual portions of search_term, front -> back i.e "A_B_C_D" => A, B, C etc
          search_term.split(split_on).each do |str|
            find_search_term = split_on_prefix ? "#{split_on_prefix}#{str}" : str
            logger.info("Scanning by term for record where #{field} ~=  #{find_search_term}")
            record = search_for_record(klazz, field, find_search_term, options)
            break if record
          end
        end

        # this time try incrementally scanning i.e "A_B_C_D" => A, A_B, A_B_C etc
        unless record
          search_term.split(split_on).inject('') do |str, term|
            z = split_on_prefix ? "#{split_on_prefix}#{str}#{split_on}#{term}" : "#{str}#{split_on}#{term}"
            record = search_for_record(klazz, field, z, options)
            break if record
            term
          end
        end

        return record
      rescue StandardError => e
        logger.error("Exception attempting to find a record for [#{search_term}] on #{klazz}.#{field}")
        logger.error e.backtrace
        logger.error e.inspect
        return nil
      end
    end

    def get_record_by!(klazz, field, search_terms, split_on = ' ', options = {} )
      x = get_record_by(klazz, field, search_terms, split_on, options)

      raise RecordNotFound, "No #{klazz} record found for [#{search_terms}] on #{field}" unless x

      x
    end

    def find_or_new( klass, condition_hash = {} )
      records = klass.where(condition_hash).all

      return records.first if records.any?

      klass.new
    end

  end

end