Magic encoding broken on second call of the same script

Bug description

Automatic encoding does not work on second call of the same script

Steps to reproduce

Start the provided script twice. Character encoding is broken on the second call. Here is the script: The Euro and German umlaut symbols are working fine on first call after device reset but are broken on subsequent calls of the script.

Unfortunately it will sometimes need a few more tries before it will get broken.

from escpos.printer import Usb
p = Usb(0x0416, 0x5011,0, profile="POS-5890")
p.text("\n")
p.set(align=u'left', custom_size=True, width=2, height=2, font=1)
p.text("SUMME          €10,47\n\n")
p.text("Bar:           €15,00\n")
p.text("Rückgeld:       €4,53\n")
p.cut()

Device info

Printer: Excelcan ZJ-5890K python-escpos version: 3.0a3 python version: 3.5.3 (distro version) operating system: Debian 9.x

[patkan]

You can pass on arguments that control magic_encode when you are creating the instance of your printer. Issue something like Usb(0x0416, 0x5011,0, profile="POS-5890", magic_encode_args=PUT THEM HERE)

You can put there a dictionary with the arguments for magic_encode. The class is described here.

If you set encoding to a suitable codepage and disabled to True, then it will print just in this codepage. The rest (unicode-input, automatic encoding into the set codepage) remains unchanged.

You can also call p.magic.force_encoding(encoding) for the same result. encoding is the name of your codepage. In this case, please try “CP858”.

[giggls]

Hm looks like some USB bus timing problem and or broken printer firmware. Files generated this way by python escpos are identical. Printing a file written using sys.stdout.buffer.write(p.output) using “cat file >/dev/usb/lp0” is producing the same Euro sign errors. Does anybody have a ZJ-5890K device where this code works? I wonder if there is a firmware update for the device.