bug with filename containing unicode/utf8 letters

[SuperUserNameMan]

Hello,
Very helpful project. Thanks a lot for developing and sharing it.

Context :

• Windows 8.1 64bits (french version)
• PdfScribe v1.0.7 downloaded from your github releases.

I found this bug :

• When the filename of the printed pdf contains unicode/utf8 letters, the saved filename might ends-up truncated.

How to replicate :

• Save a printed pdf as : facture_échantillons.pdf

Observed result :

• The pdf will be saved as : facture_฀

[stchan]

looks like including a newer ghostscript dll (v9.27) has fixed this. See release v1.0.8

[SuperUserNameMan]

@stchan : thanks ! it works perfectly with test_âäà_êëéè€_ïî_ôö_ûüù_ÿ_ñ.pdf.

Just for curiosity, I tried with 中国要求重启谈判.pdf and with призывает.pdf, and I got this error in a dialog box : "PDF Scribe : There was a PDF generation error. Ghostscript error code -100."

[stchan]

I'm guessing there's some issue with how parameters need to be encoded for Ghostscript when non-latin alphabets are involved. There was an issue filed for GhostscriptSharp (PdfScribe uses code from a very old version of Matthew Ephraim's wrapper) regarding Japanese characters, though I'm not sure if it's related.

Addressing this will have to take a backseat to #17 though, and that has to wait until I've got more time to work on PdfScribe.

[Zhuangkh]

#40

I submitted a small fix for this problem, call the set_arg_encoding before InitAPI.

[stchan]

Changes merged, included in latest release (1.1.2) - thanks for the fix - closing.