Paxmex

This gem parses your Amex data files into human readable data.

Installation

From RubyGems.org

% gem install paxmex

Available Methods

• parse_eptrn(file_path)
• parse_epraw(file_path)
• parse_epa(file_path)

The first two methods return a readable hash in the following format:

{
  "DATA_FILE_TRAILER_RECORD" => {
    "DF_TRL_RECORD_TYPE" => "DFTRL",
    "DF_TRL_DATE" => #<Date: 2013-04-05>,
    "DF_TRL_TIME" => "0435",
    "DF_TRL_FILE_ID" => 0,
    "DF_TRL_FILE_NAME" => "LUMOS LABS INC",
    "DF_TRL_RECIPIENT_KEY" => "00000000002754170029          0000000000",
    "DF_TRL_RECORD_COUNT" => 4
  },
  "DATA_FILE_HEADER_RECORD" => {
    "DF_HDR_RECORD_TYPE" => "DFHDR",
    "DF_HDR_DATE" => #<Date: 2013-04-05>,
    "DF_HDR_TIME" => "0435",
    "DF_HDR_FILE_ID" => 0,
    "DF_HDR_FILE_NAME" => "LUMOS LABS INC"
  },
  ...
}

The last method (parse_epa) returns nearly the same, but it contains nested records (according to the schema definition):

{
  "TRAILER_RECORD" => {
    ...
  },
  "HEADER_RECORD" => {
    ...
  },
  "PAYMENT_SUMMARY" => [
    {
      ...
      "SETTLEMENT_SE_ACCOUNT_NUMBER" => "1234567891",
      "SETTLEMENT_ACCOUNT_NAME_CODE" => "002",
      :children => {
        "SUMMARY_OF_CHARGE" => [
          {
            ...
            "SOC_DATE" => #<Date: 2014-18-07>,
            "DISCOUNT_AMOUNT" => #<BigDecimal: '-0.385E1'>,
            :children => {
              "RECORD_OF_CHARGE" => [
                {
                  ...
                  "CHARGE_AMOUNT" => #<BigDecimal: '0.135E3'>,
                  "CHARGE_DATE" => #<Date: 2014-17-07>,
                  "6-DIGIT_CHARGE_AUTHORISATION_CODE" => "123456"
                },
                ...
              ]
              ...

Values are parsed from their representation into a corresponding native Ruby type:

• Alphanumeric values become strings (with whitespace stripped)
• Numeric values become Fixnums
• Julian strings become Date objects
• Date strings become Date objects
• Time strings become Time objects
• Alphanumerically represented decimals become BigDecimal objects (e.g: '0000000011A' -> BigDecimal.new(1.11, 7))

If you'd like the raw values to be returned instead, you can set the raw_values option to true, e.g.:

Paxmex.parse_eptrn(path_to_file, raw_values: true)
Paxmex.parse_epraw(path_to_file, raw_values: true)
Paxmex.parse_epa(path_to_file, raw_values: true)

User-defined schema

If you need to parse a different format (i.e. not EPRAW, EPTRN or EPA), write your own schema definition and use it like this:

parser = Parser.new(path_to_raw_file, path_to_schema_file)
result = parser.parse

Example

require 'paxmex'

# Use default schema definitions
Paxmex.parse_eptrn('/path/to/amex/eptrn/raw/file')
Paxmex.parse_epraw('/path/to/amex/epraw/raw/file')
Paxmex.parse_epa('/path/to/amex/epa/raw/file')

# Use your own schema definition
parser = Parser.new('/path/to/raw/file', '/path/to/your/schema.yml')
result = parser.parse

The raw input files for either methods are data report files provided by American Express. These files are in either EPRAW, EPTRN or EPA format so use the relevant method to parse them. We have provided dummy EPRAW, EPTRN and EPA files in spec/support. Output and key-value pairs vary depending on whether you choose to parse an EPTRN, EPRAW or EPA file. If you need to parse a file in another format, you can write your own YAML schema for this purpose. We would be happy if you help us improving this project by sharing your schemas.

Contributing

Fork and submit a pull request and make sure you add a test for any feature you add.

License

LICENSE: (The MIT License)