Numbers don't display correctly if I copy and paste

[Ted0987]

Hi all, first time poster here. I just downloaded the free portable version 2.5 (Build 201.0). It's working pretty good, but for some reason when I copy some text and paste it into a Word document, the numbers don't display properly. I'm using Candara for my Word document, but the numbers will only display properly if I use Arial Narrow, Times New Roman Special G1, Symbol, or a few others. Word defaults to MingLiu-HKSCS for the numbers.

It works okay with Foxit Reader so I don't think it's an issue with Word. I thought it might be the portable version, but I installed the regular version and it does it too. It does it on XP and Vista. The PDF document permissions allow copying.

Anyone have any idea what's up with this? Thanks.

[John - Tracker Supp]

HI,

to give you a definitive answer we would need to see the (a) PDF file concerned - could you please zip and post an example here please ?

Many thanks

[Ted0987]

Ahh, okay. I did get some good news. It's only doing it with this one PDF so it's not as critical as I thought. But here's the PDF. If anyone can figure out why it's doing it, that would be great. If it happens again then I'll know what's up. Appreciate it.

[Paul - PDF-XChange]

Thanks Ted,

is this the only PDF you have that uses that font? We'll look into this but I would be interested to know.

Sincerely

[Ted0987]

Actually, yeah. It's the only one. All my other PDFs use other fonts.

[Paul - PDF-XChange]

good to know - thanks. Once we have more information for you we'll update here.

hth

[Lzcat - Tracker Supp]

Hi Ted.
Regarding your file - we copy numbers correctly, when Adobe does not
To display numbers which we "don't copy" uses embedded fonts with built-in non-standard encoding, where codes corresponding to standard digits are mapped to other characters (for example the standard letter 0 (zero) is by default mapped to a glyph named zero and the corresponding unicode value is 0x0030, while your font maps it into a glyph named zerooldstyle with an unicode value of 0xF730). Since the font uses a built-in encoding we must determine the unicode regarding to resulting glyph name, which in our case is 0xF73* (regarding to digit), so for the text we copy the correct unicode values. But unfotunately most fonts do not contain corresponding glyphs and display default characters. Adobe fixes the unicode values to general one (Replacing 0xF730 with 0x0030 and so on), so numbers are displayed in most fonts. You may easily check that - just open notepad, change the font to Adobe Garamond Pro and paste number 1883 from page 5 copied from both Viewer and from Adobe. You will see that the text copied using the Viewer looks the same as in the PDF, while the one copied with Adobe is a bit different. But when you change the font to previous you will see another difference - text copied from the Viewer became unreadable.
For now we have no plan to change such behaviour in 2.5.* release, but will see what can be done in next version.