Why does pdfa.exe pause for a long time before working?

Q:

I’m running the following command:

pdfa.exe -c -z -o output input*.pdf

The command pauses for a long time before processing documents. For ~100,000 documents, I’m seeing a delay of about a half hour. What’s going on in this delay, and how can I avoid it?

A:

The delay is coming from the processing of the wildcard parameter, which happens in the VC runtime. The runtime expands the wildcard and selection-sorts the resulting list of filenames.

The delay can be avoided by omitting the wildcard, so that PDFNet enumerates the list of files itself:

pdfa.exe -c -z -o output input