The OCR is flaky. The first email has "Release in Part" and the 'Part' is coming up as 'Fart' after OCR. And in the second email the word "fact" is coming up as "fart".
If you download the PDFs you can search by text and it matches on your search term.
I have to say that I am sad no one actually said "fart".
Search works fine for me. Do a select all/copy on the document text and paste into a text editor - you'll see that in both cases the OCR has misinterpreted other words as "fart" due to overlapping with redaction boxes.
Can anyone else confirm this? And does anyone have any idea why this would happen? Some kind of partial text match?
One of the 2 results definitely doesn't contain the substring "fart" anywhere: "will be available in 10 minutes, per ops. They will connect you as soon as she calls in. "