[Bug]: Retractions are not counted in total filament used

[vehystrix]

The problem

The filament used counter is incorrect, because it only counts positive values for the extruder. So depending on the model and the retraction settings, the total count will be off by a lot.

To reproduce, I sliced a simple cube with a retraction length set to an absurd 10mm.
Total extrusion length (sum of all positive G1 Exxx) of 7358.76547mm
Total retraction length (sum of all negative G1 E-xxx) of -3230mm
For a net total of 4128.76547mm of filament used

Screenshot of the dashboard when the print is almost done:

Or, as reported by the slicer:

The GCODE I used:
Cube_0.2mm_PLA_Elegoo Neptune 3 Plus_54m30s.txt

I calculated the totals using the following powershell commands:
Extrusion: Get-Content -Path 'Cube_0.2mm_PLA_Elegoo Neptune 3 Plus_54m30s.txt' | Select-String -Pattern 'G1.*E([0-9.]+)' | ForEach-Object {$_.Matches.Groups[1].Value} | Measure-Object -sum
Retraction: Get-Content -Path 'Cube_0.2mm_PLA_Elegoo Neptune 3 Plus_54m30s.txt' | Select-String -Pattern 'G1.*E(-[0-9.]+)' | ForEach-Object {$_.Matches.Groups[1].Value} | Measure-Object -sum
Net total: Get-Content -Path 'Cube_0.2mm_PLA_Elegoo Neptune 3 Plus_54m30s.txt' | Select-String -Pattern 'G1.*E([-0-9.]+)' | ForEach-Object {$_.Matches.Groups[1].Value} | Measure-Object -sum

I took a quick look at the dashboard code, and from what I can see here only positive values for e are processed for some reason

Version of OctoPrint

1.10.2

Platform

Intel x64

Operating system running OctoPrint

Debian linux

Dashboard version

1.19.12

Python version

3.11.2

Printer model & used firmware incl. version

Neptune 3 Plus on Marlin 2.1.1

Browser and version of browser, operating system running browser

Firefox 131.0.3 on Windows 10

Logs and other files needed for analysis

• Systeminfo Bundle (always include!) See here if you don't know where to find that.
• GCODE file with which to reproduce (always include if the issue happens while printing!)
• Screenshots and/or videos showing the problem (always include in case of issues with the user interface)

File uploads & additional information

No response

[georgandersen]

Thanks for explaining the issue.
I am using a retraction of 5mm, so that's why extruded filament always is more than calculated.