Detect printer model and media type #53

vulpes2 · 2024-07-23T19:32:50Z

I've implemented some rudimentary printer model and media detection stuff. This should be a good base for adding support for media verification and printer model detection before printing later on. Tested working on Linux with a QL-700 and PT-P700.

$ ./cli.py -b linux_kernel discover
INFO:brother_ql:Probing device at file:///dev/usb/lp0
INFO:brother_ql.backends.helpers:Sending status information request to the printer.
INFO:brother_ql.backends.helpers:Printer Series Code: 0x34
INFO:brother_ql.backends.helpers:Printer Model Code: 0x35
INFO:brother_ql.backends.helpers:Printer Status Type: Reply to status request 
INFO:brother_ql.backends.helpers:Printer Phase Type: Waiting to receive)
INFO:brother_ql.backends.helpers:Printer Errors: []
INFO:brother_ql.backends.helpers:Media Type: Die-cut labels
INFO:brother_ql.backends.helpers:Media Size: 29 x 90 mm
INFO:brother_ql:Found a label printer at: file:///dev/usb/lp0 (model: QL-700)
$ ./cli.py discover
INFO:brother_ql:Defaulting to pyusb as backend for discovery
INFO:brother_ql:Probing device at usb://0x04f9:0x2061/000XXXXXXXXX
INFO:brother_ql.backends.helpers:Sending status information request to the printer.
INFO:brother_ql.backends.helpers:Printer Series Code: 0x30
INFO:brother_ql.backends.helpers:Printer Model Code: 0x67
INFO:brother_ql.backends.helpers:Printer Status Type: Reply to status request 
INFO:brother_ql.backends.helpers:Printer Phase Type: Waiting to receive)
INFO:brother_ql.backends.helpers:Printer Errors: []
INFO:brother_ql.backends.helpers:Media Type: Laminated tape
INFO:brother_ql.backends.helpers:Media Size: 12 x 0 mm
INFO:brother_ql:Found a label printer at: usb://0x04f9:0x2061/000XXXXXXXXX (model: PT-P700)
INFO:brother_ql:Probing device at usb://0x04f9:0x2042/000XXXXXXXXX
INFO:brother_ql.backends.helpers:Sending status information request to the printer.
INFO:brother_ql.backends.helpers:Printer Series Code: 0x34
INFO:brother_ql.backends.helpers:Printer Model Code: 0x35
INFO:brother_ql.backends.helpers:Printer Status Type: Reply to status request 
INFO:brother_ql.backends.helpers:Printer Phase Type: Waiting to receive)
INFO:brother_ql.backends.helpers:Printer Errors: []
INFO:brother_ql.backends.helpers:Media Type: Die-cut labels
INFO:brother_ql.backends.helpers:Media Size: 29 x 90 mm
INFO:brother_ql:Found a label printer at: usb://0x04f9:0x2042/000XXXXXXXXX (model: QL-700)

Changes made:

Added a status command to query printer info and media info
Added some hardware info in model data
Detect printer model during discovery
Added hardware info for PT-P700 and PT-P750W
Fixed auto-cut for PT printers

Fixes pklaus#123
A good staring point for addressing #39
Fixes #54

matmair · 2024-07-23T21:04:12Z

Hi there @vulpes2. As just laid out in #51 (comment) I really appreciate the effort here. This change will be gladly welcomed into the fork.

I would ask to only format new code with tools like ruff/black to keep backports of the PR as simple as possible.
When working in similar situations in the past I found the (VS Code) settings python.linting.ignorePatterns or ignore(see https://code.visualstudio.com/docs/python/settings-reference#_python-language-server-settings) helpful. From my experience, most IDEs and linters provide similar tools to improve the signal-to-noise ratio.

matmair · 2024-07-23T21:05:52Z

I am sorry that my intention regarding back portability was not stated explicitly, it only covers existing code. New additions can follow modern Python behaviour and styling habits.

vulpes2 · 2024-07-23T21:14:26Z

Would it be fine if I rebased the feature changes without the formatting changes?

matmair · 2024-07-23T21:24:01Z

Yes! Please (at) me once this is ready for review.

pyusb is cross-platform and works out of the box after the correct udev rules have been set up, while /dev/usb/lp0 is still typically inaccessible by regular users unless the user is in a specific user group.

vulpes2 · 2024-07-23T21:42:50Z

@matmair All done, I've only reformatted models.py since it's just a list of hardware info, everything else was left as-is.

matmair · 2024-07-24T06:12:30Z

I will review/test this with a few printers and report back in the next few days!

matmair

I did a pure code only review without functional tests, I will ping you once I had the possibility to test the code with real hardware.

brother_ql/brother_ql_print.py

brother_ql/cli.py

brother_ql/models.py

brother_ql/backends/helpers.py

matmair · 2024-07-29T21:04:38Z

I finally got time to test this with a few printers and labels; it looks great. So far I have only tested with usb-based printers. Am I understanding this right that it will only work with USB printers?

vulpes2 · 2024-07-29T22:18:30Z

This might also be possible on network printers, but I do not own such a device to test it, nor is brother_ql designed for two-way communication with network printers. In the send function you can see it doesn't even try to get a response if the network backend is selected, so some additional refactoring would be required before we can even attempt to implement it.

mkildemaa · 2024-08-02T11:51:31Z

Note: PT-P700 seems to use slightly different command parameters, so the cutter isn't being activated right now. Will fix this when I have more time.

Just for information, maybe it helps: I am currently figuring out the same thing. It seems to me PT-P700 and PT-P710BT use same command parameters for cutting as QL-800 because the cutter works for PT-P700 and PT-P710BT when running the command as model QL-800 instead of PT-P750W.

vulpes2 · 2024-08-03T07:15:01Z

@mkildemaa I've pushed a fix for the PT-P700, it seems to work for me. Please test on other PT printers and report back if you have access to them. I don't mind addressing other bits in the same PR if I'm able to.

There are a few issues I've fixed:

conversion.py is calling add_print() with last_page set to true for all pages. If a block of raster data is not the final page, the print command should be 0x0C instead of 0x1A. The correct sequence for a multi-page printing job is [raster data] 0x0C [raster data] 0x0C [raster data (final page)] 0x1A. The incorrect implementation works fine on QL printers, but will waste a huge amount of tape on PT printers, because PT printers will feed a blank piece of tape at the beginning of every print job. A print job is considered as terminated once 0x0A has been received.
The auto-cut bit is at the same offset for QL and PT printers, so it was never correctly set in the first place for PT printers. The documentation is misleading because they used different starting offsets.
Chain printing (exclusive to PT printers) is now disabled by default. When chain printing is enabled, the printer does not feed and cut the last printed page. This is a useful behavior for printing multiple labels as a long continuous strip (label can optionally be half-cut on supported printers), but should not be the default. P-Touch disables it by default, and there is no code to support it in brother_ql, so it should remain disabled until support is implemented.

Auto-cut definitions from the command reference

PT-H500/P700/E500 Raster Command Reference

QL-600/QL-710W/QL-720NW Raster Command Reference

vulpes2 · 2024-08-03T08:51:14Z

Uncompressed data was still not printing properly on my PT-P700, I was getting garbled graphics or blank labels. Turns out the compression mode must be set with every single print job on a supported printer, otherwise the compression mode is unknown to a printer that supports both. There is no default mode specified in any of the raster command reference manuals, so this should be specified for both QL and PT printers for better reliability.

kornpow · 2024-08-03T20:55:58Z

Ran this today and get this error:

./cli.py -b linux_kernel discover
  File "/app/brother_ql-inventree/brother_ql/./cli.py", line 61
    logger.info(f"Probing device at {device["identifier"]}")
                                             ^^^^^^^^^^
SyntaxError: f-string: unmatched '['

kornpow

After a couple of changes to the f-strings, I was able to run the code.

Tested with 3 tapes for the QL-800 and 1 tape for the QL-1110NWB!

brother_ql/cli.py

brother_ql/backends/helpers.py

kornpow · 2024-08-03T23:25:30Z

Output from my test runs:

INFO:brother_ql.backends.helpers:Media Type: Continuous length tape
INFO:brother_ql.backends.helpers:Media Size: 12 x 0 mm
INFO:brother_ql:Found a label printer at: usb://0x04f9:0x209b/000L8Z690042 (model: QL-800)
...
...
INFO:brother_ql.backends.helpers:Media Type: Continuous length tape
INFO:brother_ql.backends.helpers:Media Size: 62 x 0 mm
INFO:brother_ql:Found a label printer at: usb://0x04f9:0x209b/000L8Z690042 (model: QL-800)
...
...
INFO:brother_ql.backends.helpers:Media Type: Die-cut labels
INFO:brother_ql.backends.helpers:Media Size: 62 x 29 mm
INFO:brother_ql:Found a label printer at: usb://0x04f9:0x209b/000L8Z690042 (model: QL-800)
...
...
INFO:brother_ql.backends.helpers:Media Type: Continuous length tape
INFO:brother_ql.backends.helpers:Media Size: 102 x 0 mm
INFO:brother_ql:Found a label printer at: usb://0x04f9:0x20a8/000F1Z265904 (model: QL-1110NWB)

kornpow · 2024-08-03T23:27:51Z

I really think you should make this code a little bit more specific to the task at hand "detect printer and media type".
Then create another PR on top of this one which has the PT cutter included, if you would like.

Cutting, auto-cutting, etc definitely seems out of scope?

kornpow · 2024-08-04T03:01:51Z

I am also able to use this code within the python scripts instead of the CLI. It only worked for pyusb backend in my testing.

from brother_ql.backends.helpers import status

>>> status("usb://04f9:209b")
{'series_code': 52, 'model_code': 56, 'status_type': 'Reply to status request', 'phase_type': 'Waiting to receive', 'media_type': 'Die-cut labels', 'media_width': 62, 'media_length': 29, 'errors': []}

>>> status("/dev/usb/lp1")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/app/brother_ql-inventree/brother_ql/backends/helpers.py", line 139, in status
    result = interpret_response(data)
             ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/app/brother_ql-inventree/brother_ql/reader.py", line 165, in interpret_response
    raise NameError('Insufficient amount of data received', hex_format(data))
NameError: ('Insufficient amount of data received', '')

vulpes2 · 2024-08-04T04:57:29Z

It looks like python 3.12 has extended the f-string syntax so it was valid syntax only in python 3.12. Tested the fix with python 3.11 and everything seems to work. https://docs.python.org/3/whatsnew/3.12.html

About mixing the cutter related changes, I have very limited time to work on this and this PR is moving very slowly because none of us are doing this for a living, so I'd rather put it here instead of constantly rebasing a second PR just for those changes.

vulpes2 · 2024-08-04T05:00:54Z

Unfortunately I also can't reproduce the other error you have been seeing:

>>> status("usb://04f9:2042")
{'series_code': 52, 'model_code': 53, 'status_type': 'Reply to status request', 'phase_type': 'Waiting to receive', 'media_type': 'Continuous length tape', 'media_width': 62, 'media_length': 0, 'errors': []}
>>> status("/dev/usb/lp0")
{'series_code': 52, 'model_code': 53, 'status_type': 'Reply to status request', 'phase_type': 'Waiting to receive', 'media_type': 'Continuous length tape', 'media_width': 62, 'media_length': 0, 'errors': []}

vulpes2 · 2024-08-04T15:17:30Z

@matmair I believe all the concerns have been addressed, can you review this again? Would be great if we can get this merged soon.

matmair · 2024-08-04T20:33:15Z

@vulpes2 I am very happy with this, thank you for your contribution.
Do you want to tackle #55 before I cut a new release?

mkildemaa · 2024-08-04T20:45:22Z

@mkildemaa I've pushed a fix for the PT-P700, it seems to work for me. Please test on other PT printers and report back if you have access to them. I don't mind addressing other bits in the same PR if I'm able to.

There are a few issues I've fixed:

conversion.py is calling add_print() with last_page set to true for all pages. If a block of raster data is not the final page, the print command should be 0x0C instead of 0x1A. The correct sequence for a multi-page printing job is [raster data] 0x0C [raster data] 0x0C [raster data (final page)] 0x1A. The incorrect implementation works fine on QL printers, but will waste a huge amount of tape on PT printers, because PT printers will feed a blank piece of tape at the beginning of every print job. A print job is considered as terminated once 0x0A has been received.

The auto-cut bit is at the same offset for QL and PT printers, so it was never correctly set in the first place for PT printers. The documentation is misleading because they used different starting offsets.

Chain printing (exclusive to PT printers) is now disabled by default. When chain printing is enabled, the printer does not feed and cut the last printed page. This is a useful behavior for printing multiple labels as a long continuous strip (label can optionally be half-cut on supported printers), but should not be the default. P-Touch disables it by default, and there is no code to support it in brother_ql, so it should remain disabled until support is implemented.

Auto-cut definitions from the command reference

Sorry for the delay in answering. Much appreciated! All my printers are in the office in which the next time I will be there is tomorrow. I see the PR is now merged so I will report back only if there are any issues and if there is silence from my part then everything worked.

vulpes2 · 2024-08-05T05:32:35Z

Do you want to tackle #55 before I cut a new release?

That would be ideal, thanks.

vulpes2 force-pushed the status-query branch from dc05d83 to 1ff3c8b Compare July 23, 2024 20:49

vulpes2 added 5 commits July 23, 2024 21:30

implement status check to retrieve printer and media info

d56136a

add hardware info for known printers

0ecd6e8

printer model detection during discovery

6d0b108

use pyusb as the default backend instead of linux_kernel

b30f71e

pyusb is cross-platform and works out of the box after the correct udev rules have been set up, while /dev/usb/lp0 is still typically inaccessible by regular users unless the user is in a specific user group.

add hardware info for PT-P700 and PT-P750W

a0ee20d

vulpes2 force-pushed the status-query branch from 1ff3c8b to a0ee20d Compare July 23, 2024 21:41

matmair reviewed Jul 24, 2024

View reviewed changes

vulpes2 requested a review from matmair July 29, 2024 22:28

restore removed functions during refactor

e44b154

vulpes2 force-pushed the status-query branch from aef867e to e44b154 Compare July 29, 2024 22:36

fill in hardware info for all known models

de284d1

vulpes2 force-pushed the status-query branch from 1e35404 to de284d1 Compare July 29, 2024 23:00

vulpes2 force-pushed the status-query branch 2 times, most recently from 60c869c to 789dda6 Compare August 3, 2024 03:46

revert default backend to linux_kernel

1e681b3

vulpes2 force-pushed the status-query branch from 789dda6 to 1e681b3 Compare August 3, 2024 03:52

fix cutter setting on PT printers

a22180d

always send compression mode on supported printers

9bf9fbb

kornpow reviewed Aug 3, 2024

View reviewed changes

brother_ql/cli.py Outdated Show resolved Hide resolved

brother_ql/cli.py Outdated Show resolved Hide resolved

brother_ql/backends/helpers.py Outdated Show resolved Hide resolved

fix f-strings

e929e77

vulpes2 mentioned this pull request Aug 4, 2024

Enable Auto-Power On #55

Closed

vulpes2 requested a review from kornpow August 4, 2024 15:14

matmair merged commit 273629c into matmair:master Aug 4, 2024

matmair mentioned this pull request Aug 4, 2024

Get currently loaded tape #39

Closed

vulpes2 deleted the status-query branch August 5, 2024 05:24

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Detect printer model and media type #53

Detect printer model and media type #53

vulpes2 commented Jul 23, 2024 •

edited

Loading

matmair commented Jul 23, 2024

matmair commented Jul 23, 2024

vulpes2 commented Jul 23, 2024

matmair commented Jul 23, 2024

vulpes2 commented Jul 23, 2024

matmair commented Jul 24, 2024

matmair left a comment

matmair commented Jul 29, 2024 •

edited

Loading

vulpes2 commented Jul 29, 2024

mkildemaa commented Aug 2, 2024 •

edited

Loading

vulpes2 commented Aug 3, 2024

PT-H500/P700/E500 Raster Command Reference

QL-600/QL-710W/QL-720NW Raster Command Reference

vulpes2 commented Aug 3, 2024

kornpow commented Aug 3, 2024

kornpow left a comment

kornpow commented Aug 3, 2024

kornpow commented Aug 3, 2024 •

edited

Loading

kornpow commented Aug 4, 2024

vulpes2 commented Aug 4, 2024 •

edited

Loading

vulpes2 commented Aug 4, 2024

vulpes2 commented Aug 4, 2024

matmair commented Aug 4, 2024

mkildemaa commented Aug 4, 2024

vulpes2 commented Aug 5, 2024

Detect printer model and media type #53

Detect printer model and media type #53

Conversation

vulpes2 commented Jul 23, 2024 • edited Loading

matmair commented Jul 23, 2024

matmair commented Jul 23, 2024

vulpes2 commented Jul 23, 2024

matmair commented Jul 23, 2024

vulpes2 commented Jul 23, 2024

matmair commented Jul 24, 2024

matmair left a comment

Choose a reason for hiding this comment

matmair commented Jul 29, 2024 • edited Loading

vulpes2 commented Jul 29, 2024

mkildemaa commented Aug 2, 2024 • edited Loading

vulpes2 commented Aug 3, 2024

PT-H500/P700/E500 Raster Command Reference

QL-600/QL-710W/QL-720NW Raster Command Reference

vulpes2 commented Aug 3, 2024

kornpow commented Aug 3, 2024

kornpow left a comment

Choose a reason for hiding this comment

kornpow commented Aug 3, 2024

kornpow commented Aug 3, 2024 • edited Loading

kornpow commented Aug 4, 2024

vulpes2 commented Aug 4, 2024 • edited Loading

vulpes2 commented Aug 4, 2024

vulpes2 commented Aug 4, 2024

matmair commented Aug 4, 2024

mkildemaa commented Aug 4, 2024

vulpes2 commented Aug 5, 2024

vulpes2 commented Jul 23, 2024 •

edited

Loading

matmair commented Jul 29, 2024 •

edited

Loading

mkildemaa commented Aug 2, 2024 •

edited

Loading

kornpow commented Aug 3, 2024 •

edited

Loading

vulpes2 commented Aug 4, 2024 •

edited

Loading