Duplicating numbers and problem with whitespaces

[ScreamDev1]

Hi, i have some issues with parsing pdf in Russia. First of all it's some problem with witespaces and also for some values of numeric i can see dupicates in parsed text. Attaching pdf and some screenshots, Thanks

obj39453_pd32-000296.pdf

[GreyWyvern]

I believe the "duplicate" numbers here are because the document is overprinting them to make the bold. Perhaps because it doesn't have a bold font? Regardless, the duplicated numbers and text all show up bold in the PDF. Copy-pasting from Adobe only copies one of them though so it's smart enough to do that?

Here's one of the "duplicated" sections:

BT
/F7 12 Tf
1 0 0 -1 0 0 Tm
456 -584 Td
<002D> Tj
8 0 Td
<0048> Tj
8 0 Td
<0046> Tj
8 0 Td
<0003> Tj
1 0 0 -1 0 0 Tm
/Span << /ActualText <> >> BDC
455.500000 -584 Td
<002D> Tj
8 0 Td
<0048> Tj
8 0 Td
<0046> Tj
8 0 Td
<0003> Tj
EMC
ET

First it sets the text matrix to 1 0 0 -1 0 0 then positions it at 456 -584 and prints four characters. Then it resets the text matrix to 1 0 0 -1 0 0 and positions it almost on top of the previous start position, just a fraction to the right (or left?) at 455.500000 -584 to make the characters fatter and thus bolder. Then it prints the same four characters. Overprinting.

Edit: Looking at the code I posted again, it seems like Adobe is honouring the BDC ... EMC block and using the contents of /ActualText <> as the text to be copied. We should probably look to have PdfParser honour these blocks too in the future.