Obviously the issue is a specifics of “Create links from URLs” feature of Acroread.

If it is enabled, Acroread scans a loaded document for strings, look like URLs, and turns them into hyperlinks.

For example, a code

is turned into a link, even it is a plain text for a web browsers.

Unfortunately the feature is not clever enough, to recognize line-broken URL strings.

A workaround would be to define c regular hyperlink, using HTML means:

PD4ML should convert such code correctly even with very long URLs.