3 ms·
Great article from a classic series that every programmer should read. Here's an odd thing -- when I copy'n'pasted the awk implementation the pasted copy had e
by emmelaich 10y ago
Great article from a classic series that every programmer should read.
Here's an odd thing -- when I copy'n'pasted the awk
implementation the pasted copy had errors reminiscent
of OCR. For example the (j became Cj, the ARGV[2] become
ARGV121.
What's going on here? Is the website detecting the copy
and make the browser get something else? Is it Chrome -
or something else doing some guessing?
- vilhelm_s 10y agoI think the document is scanned from a paper copy. The pixels shown on the screen is from the scanned image, but the text for copy-and-paste was created by OCR when the document was scanned and prepared. You often see this with pdf files, I guess the software used to create them has OCR functionality built in. If you use "inspect element" in your web browser you can see how it is displayed. Each page is a <div> which contains first a big image (with the scanned document page), and then a <span> for each word, which are positioned using absolute positioning to be placed over the corresponding word in the image. The <span>s each have "color: transparent", so you can't see them but you can still select them.