Home/ Questions/Q 7248757
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-28T22:10:57+00:00

I’m trying to parse some data in a fixed format text file where each

  • 0

I’m trying to parse some data in a fixed format text file where each “record” is spread over a number of lines, as so …

 MAILBOX: 10013      Created: 01/20/09  4:39 pm
    MSGS: 0         UNPLAYED: 0           URGENT: 0          RECEIPT: 0
  LCOS: RBC Standard    : 20            FCOS: RBC Standard      : 20 
  GCOS: Default GCOS 1  : 1             NCOS: Default           : 1 
  TCOS: Default TCOS 1  : 1             RCOS:                   : 1 
BAD LOGS: 0         LAST LOG: NEVER                             MINS:      0.0
  PASSWD: Y            TUTOR: N              DAY: M            NIGHT: M       
    NAME:                                   CODE: 
   EXTEN: 10013                            INDEX: 0
ATTEN DN:                                  INDEX: 0         
DISTRIBUTION LISTS WITH CHANGE RIGHTS:
    all
DISTRIBUTION LISTS WITH REVIEW RIGHTS:
    all

I’ve used File Helpers before for single line records, and it’s been very useful. Checking it’s documentation, it does have a MultiRecordEngine feature, but this is going to mean …

  • a class for each line … not a problem
  • calculating the exact size of each fixed format field … painful and open to error
  • logic to check each line

And a further wrinkle I found was the fixed format is actually not fixed, i.e. there are different format lines depending on the target record, so some have 21 lines, some 22, 23, 24, etc.

I have found a Java flat file parsing library, FFP, but I’m a .NET, C#, PowerShell coder

Are there better ways of handling this sort of parsing ?

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    What you need is a lexer. Your record is too big to use a single Regex to parse, so you have to write one regex for each line, and a state machine to validate that the lines follows in the right order.

    Or you can use a general purpose lexer/parser to generate the code for you. Wikipedia has long list. The Gold parser looks like a good candidate.

    I would not try to do the lexing/parsing in PowerShell. I would rather write the code as C# or F# and use the assembly from PowerShell.

    Edit: I’ve just looked at FileHelpers library. You could create a Multirecord Engine with a .NET Type that matches each line in you source record. All you have to do then is parse the result array for valid order and create objects.

    • 0