In this case, I would probably get each line of data to memory or a cursor as quickly as possible to speed up the parsing (ie: FileToStr if you can, or FGETS/FREADS to a cursor if you can't). However, I believe the low-level file functions would be easier to use here.
FGETS( ) returns a maximum of 8,192 bytes. (If you need more characters per line you can FREAD up to 65,535 bytes but using FREAD will require more programming to set the read length correctly for each line and to manually position your file pointer.)
I haven't tested this code but I think it would look something like this:
LOCAL lnFileHandle, lnFileSize, lcLineData, lnFldLoop, lcDestFld, laDataArray[1]
USE DestTable IN 0 &&I'm assuming that the number and order of fields match text file you are importing
lnFileHandle = FOPEN("test.txt") lnFileSize = FSEEK(lnFileHandle, 0, 2)
IF lnFileSize <= 0 WAIT WINDOW "Error! Empty file!" NOWAIT ELSE
DO WHILE !FEOF(lnFileHandle)
lcLineData = FGETS(lnFileHandle, 8192)
FOR lnFldLoop = 1 TO ALINES(laDataArray, lcLineData, "|")
lcDestFld = FIELD(lnFldLoop,'DestTable') REPLACE lcDestFld.&lcDestFld WITH laDataArray(lnFldLoop) IN DestTable
ENDFOR
ENDDO
ENDIF
= FCLOSE(lnFileHandle) && Close the file
You can get field level specific if you want to convert imported character data into numeric, date, binary or memo fields by doing direct replacements instead the field loop process shown above.
The bigger challenge that I see is that when a field is bigger than 254 characters, it is typcally reserved for user input/notes/memos/etc and invariably a user will enter delimiter values inside of the field which can affect your parse routine.
Paul H. Tarver
-----Original Message----- From: Matt Wiedeman [mailto:Matt.Wiedeman@nahealth.com] Sent: Thursday, April 20, 2017 9:58 AM To: profoxtech@leafe.com Subject: Text file import
Hello everyone,
I need to setup a job to import a pipe delimited text file. This is easy enough but one of the fields is larger than 254 characters. If I use a memo field, it does not import that field. I started to setup a routine to step through each character and store the fields manually but I would rather not do it that way.
Does anyone have a function or tip they can share to resolve this situation?
[excessive quoting removed by server]