If you need to transfer data from CSV or Fixed Width files, it is likely that they are being created with a varying filename to identify the date and time of creation. In the CSV and Fixed Width Datastores there is a function to store wildcards so that the filename used in the process is constant but each of the actual files can be processed.
Once a process is run on a file (determined by the LastModifiedDate) it cannot process again, preventing any data loads duplicated to your source.
If you need to process a file again, simply edit it so that the LastModifiedDate is changed — then the process will identify the file as new and will transfer the data again.
[ ] Within the brackets the type of characteristics expected are specified, for example, [0–9] means you can see any numbers within that range, [A–Z] allows any alphabetical characters within that range.
{ } Within the braces, the first number denotes the maximum amount of characters, the second number denotes the maximum number of characters expected.
^ Literal to start or end an expected string.
3. In the datastore go to the last tab — Filename Wildcards.
4. Click + New Wildcard
5. Give your filename an alias — this is how the file will appear in the Datastore.
6. In the value field save an expression that allocates variables [0-9] for numbers and [A-Z] for letters. Any other characters are expressed as they appear. End the string with the file extension as it appears in windows explorer. Select Save.
7. Click Scan > select Save
Your datastore now appears with a constant filename.
If a process executes and all the files have already transferred the error message will say "your datastore is empty".
If a process is created using a filename and then subsequently a wildcard is applied - your datastore will scan with an error saying that the file no longer exists but is being used in a process. To fix - just recreate a process using the wildcard filename and delete the previous one.
