Hey there. I'm working on a service that works with Octoprint, and currently rely on it to analyze the gcode file for print dimensions and filament usage. This process is often very slow though, so I was wondering if it could be possible to use the Octoprint metadata analysis code on my server instead, and whether or not that'd improve the while thing time-wise?
I'm not an GCODE expert, so I don't wanna throw myself into writing a custom analyser, but I also don't wanna go through too much trouble with trying to "extract" the Octoprint one, if it relies too heavily on a lot of other Octoprint components and classes from other parts of Octoprint...
Do you guys think it's something I should try to do? Or are there some open sourced gcode analysis libraries out there that I'm just missing, that would be better to use?
Just an idea I had that I'm now looking for a little bit of feedback on, from people who know more about Octoprint than me
IMHO before I start developing new stuff I looked around for such a ready to use solution, like you do.
BUT in this case I would like to ask some questions:
When do you need such information? After upload, file selection, after print?
What is "slow" in your case? 2 seconds, 5 minutes, ... Maybe the performance could be increased by upgrading your raspi (if you use one)
How is your "system-landscape", look a like? E.g. gcode.ws is executed on the clients browser, octoprint analyse the gcode on the python server side.
So, to give you a good answer we need more informations
Another plugin example would be print time genius. It uses an external processor that is more efficient and I believe it calculates print size and such.
@jneilliii I should've specified; I was looking for a GCODE-analyzer for the data like filament usage, estimated print time and max- min coordinates for the code. I needed this data to see if the model would fit the 3D-printer's bed, and to check if the printer has enough filament to finish the print. I used OctoPrint to analyze the file before, but for the analysis to finish, it could take up to 6 minutes...
I ended up tearing gcode.ws apart, and stripped it of all the visuals I didn't need. Ended up with a result that was fine; a client-side GCODE analysis. A server-side analysis would've been just fine as well, but by the looks of it I would've had to code it from the ground up, which I was not really up to.
It took about 6 hours to modify the code to fit my needs, but I now have a much faster analysis of GCODE! The file that took our poor Raspberries (3B) ~6 minutes to analyze now takes about 10 seconds (it's the second file I upload in the gif, the one that isn't basically instant).
The PrettyGcode viewer looks super cool as well, and I can definitely get some use out of that - but I ended up with what I wanted to!
The gif above shows the live analysis of the GCODE-files I upload. One file that is all good - printer has enough filament and it fits the bed, the second GCODE is much larger, takes some moer time to analyze, and doesn't fit the bed.
The file that took our poor Raspberries (3B) ~6 minutes to analyze now takes about 10 seconds (it's the second file I upload in the gif, the one that isn't basically instant).