I wondered myself about OCR (probably print as PDF) as well the proofreader's spell check afterwards. A spell check might not find proper names with misspellings. For instance spell check might see a misspelling of a common name like James but maybe not Comey and a lazy proofer might not notice.
Strange to me though is when comparing the possible "rn" to "m" or "m" to "rn" issues i ran into this. Notice the "rn" in attorney looks very dissimilar to the "rn" found by the search tool in "Corney".
Also notice when I zoomed in on the "m" in Corney and compared it to the "m" in James, they do look a bit different and they are different in width. Notice the first leg or the Comey "m" looks like the first leg of the Attorney "n" and the middle leg of the James "m".
I don't know what this tells us other than IMO, the proof reader failed to correct the OCR errors of Comey converted to Corney with some bizarre character combination that the search tools recognizes as "rn" instead of "m".
A final spitball may be a combination of fonts used by the OCR tool which confuses the search tool.
