Hi Benoit,

Did you already check the topic TXT File Handling in the Uniface Library? It should sufficiently answer your question. Here's an extract:

***

[...]

Interface Definitions for Comma-Separated Files

Use interface definitions when handling comma-separated text files. For example, a text file contains the following:

’string1’,’string2’,1200,’string3'
'string4’,’string5’,123,’string6’

The file contains two records. Each record is made of two strings, a number, and another string. The strings are enclosed in quotes, and the fields are separated by commas. The field interface definitions in the entity are shown in the table:

Field6 is the primary key because the other fields are variable-length character strings, and a variable-length field cannot be a primary key. The VI in the interface definition indicates that a string identifier is used with the record part after the backslash (\\) separator.

The first field starts after a quote character, and the second field starts after the comma. To prevent possible misinterpretation of a single comma, field5 is added to prevent the last comma being added to field4.

[...]

***

Hope this helps.

Best regards,
Daniel

*** Usual disclaimer ***

Hi Benoit,

uniface requires a key field for all non-database entities.

BUT with TXT, this pro-forma field does NOT have to meet the requirements of a real Primary Key.

Duplicates are allowed etc. so any non-variable-length field will do the job.

Success, Uli