Gimp-Forum.net - Scripting questions

Help converting a basic functional plugin

Thu, 12 Mar 2026 00:52:00 +0000

First off, I am not a coder, but I have been trying for the last year to convert a GIMP 2 plugin to GIMP 3, even trying AI and it simply just won't work for me. At my age (70+) I am really struggling to succeed with this task of understanding how to write a GIMP 3 plugin.  Years ago I somehow managed over several months to write a plugin in GIMP 2.10 to do some basic repetitive tasks, like adding solid color layers with masks, in groups and set other various parameters. The first thing that happens in the plugin is the original image layer is copied and desaturated.

I posted here a year ago and Ofnuts kindly wrote an example showing how to desaturate the original layer. That code worked perfectly.  I have tried for the last year to add to that plugin like trying to add solid color layers with masks, into groups and set other various parameters and have completely failed because of my lack of understanding and ineptness. Even AI has not been able to come up with something that will show up in GIMP3.  I'm at the breaking point where I'm about to give up on using GIMP 3 and just go back to GIMP 2 so my basic plugin will work. I am back here now asking if someone could point me to someone or a website or a forum, anything!, where they take requests or would rewrite a gimp2 plugin into gimp 3 format. Any help would be appreciated.

I am really struggling with rewriting this simple task plug-in in GIMP 3 format...I can't even get anything to work or show up in the GIMP menus (nor can AI even write working code for GIMP3) even after copying the basic code from another GIMP 3 plugin example. I'll upload my GIMP2 plugin so you can see what I'm trying to do, but I am really hoping there is a GIMP forum group somewhere that would specifically rewrite code for us non-coding users.  I just cannot simply understand how to write GIMP3 python code structure when I cannot understand the examples to do basic tasks like adding a new layer. Thanks for any feedback you can provide or direction you can point me in.

  ColorPrep.py (Size: 16.1 KB / Downloads: 59)

migrating 2.8 bash scripts to gimp 3

Mon, 09 Mar 2026 11:53:32 +0000

Hello,
new to this forum I hope someone here could help me a little bit.

As I upgraded from debian 12 to debian 13, gimp was upgraded to version 3.
I have a set of short scripts that allowed me to invoke gimp (at least until version 2.8) from a linux bash console. Now they fail, so I have to adapt them to recent changes.

The one question I still stumble on is the ability to set a drawable in my image. The former syntax does not work anymore :

Code:

(drawable (car(gimp-image-get-active-drawable image)))

so that I can't do anything on a specific layer e.g. desaturate :

Code:

(gimp-drawable-desaturate drawable 0)

How can I instruct gimp that, let's say, the drawable is layer [0] ?

Thanks in advance,
Vince.

Gimp 3.0 - How to draw stuff with python scripts (pixel by pixel)

Wed, 04 Mar 2026 17:15:02 +0000

I need help writing some basic script that will draw, for example, a circle, or a rectangle of randomly colored pixels, at a specified X and Y on the selected layer. For now I got this :

Code:

#!/usr/bin/env python3

# -*- coding: utf-8 -*-



import gi

gi.require_version("Gimp", "3.0")

gi.require_version("GimpUi", "3.0")

gi.require_version("Gegl", "0.4")

from gi.repository import Gimp, GimpUi, Gegl, GObject, GLib

import sys





class DrawCirclePlugin(Gimp.PlugIn):



    def do_set_i18n(self, name):

        return False



    def do_query_procedures(self):

        return ["python-fu-draw-circle"]



    def do_create_procedure(self, name):

        Gegl.init(None)



        procedure = Gimp.ImageProcedure.new(

            self, name,

            Gimp.PDBProcType.PLUGIN,

            self.run,

            None,

        )



        procedure.set_image_types("*")

        procedure.set_sensitivity_mask(Gimp.ProcedureSensitivityMask.DRAWABLE)

        procedure.set_menu_label("Draw Circle")

        procedure.add_menu_path("/Filters/Draw")

        procedure.set_documentation(

            "Draw a filled/outlined circle of a chosen color and radius",

            "Draws a circle centred at (center_x, center_y) with the given "

            "radius, fill color, and optional stroke outline.",

            name,

        )

        procedure.set_attribution("Your Name", "Your Name", "2024")



        procedure.add_int_argument(

            "center-x", "Center X", "X coordinate of the circle center (px)",

            0, GLib.MAXINT, 100, GObject.ParamFlags.READWRITE,

        )

        procedure.add_int_argument(

            "center-y", "Center Y", "Y coordinate of the circle center (px)",

            0, GLib.MAXINT, 100, GObject.ParamFlags.READWRITE,

        )

        procedure.add_int_argument(

            "radius", "Radius", "Radius of the circle in pixels",

            1, GLib.MAXINT, 50, GObject.ParamFlags.READWRITE,

        )

        default_color = Gegl.Color.new("red")

        procedure.add_color_argument(

            "color", "Fill Color", "Color used to fill (and stroke) the circle",

            True, default_color, GObject.ParamFlags.READWRITE,

        )

        procedure.add_boolean_argument(

            "fill", "Fill circle", "Fill the interior of the circle",

            True, GObject.ParamFlags.READWRITE,

        )

        procedure.add_boolean_argument(

            "stroke", "Draw outline", "Stroke an outline around the circle",

            False, GObject.ParamFlags.READWRITE,

        )

        procedure.add_int_argument(

            "stroke-width", "Outline width", "Outline thickness in pixels",

            1, 500, 2, GObject.ParamFlags.READWRITE,

        )



        return procedure



    def run(self, procedure, run_mode, image, drawables, config, run_data):

        Gimp.message("A")

        if run_mode == Gimp.RunMode.INTERACTIVE:

            GimpUi.init("draw_circle")

            dialog = GimpUi.ProcedureDialog.new(procedure, config, "Draw Circle")

            dialog.fill(None)

            if not dialog.run():

                dialog.destroy()

                return procedure.new_return_values(

                    Gimp.PDBStatusType.CANCEL, GLib.Error()

                )

            dialog.destroy()

        drawable = drawables[0]

        image.undo_group_start()





        # draw stuff here





        image.undo_group_end()



        return procedure.new_return_values(Gimp.PDBStatusType.SUCCESS, GLib.Error())





Gimp.main(DrawCirclePlugin.__gtype__, sys.argv)

I see that Drawable class have a method SetPixel, but in the docs it says that undo doesn't work on that. So, how to modify drawable pixels so that undo works ?

Writing a simple python plugin

Thu, 22 Jan 2026 02:54:45 +0000

Why is it so hard to create a plugin. I am trying to write a simple plugin to export the current image to a PNG file and then close the view. I am stuck at the save the image. Plugin attached.

exportandclose.py (Size: 4.07 KB / Downloads: 200)

GIMP 3.x Python Plugin Issue: Reliable Program Exit and Image Closing

Fri, 12 Dec 2025 11:26:12 +0000

Hello GIMP Community,

I am facing an issue in a GIMP 3.0 Python plugin: reliably closing the program and the last active image when the sequential workflow is finished.

My script is running in the GUI Mode, and I need to perform two actions in order:

Close the last open image container.
Force GIMP to quit the application entirely.

The old methods from v2 do not work anymore and I am unsure if what I am trying to achieve is actually possible.

Here is my script so far (German annotations):

Code:

#!/usr/bin/env python3

# -*- coding: utf-8 -*-



import sys

import gi

gi.require_version('Gimp', '3.0')

gi.require_version('GimpUi', '3.0')

from gi.repository import Gimp, GimpUi, GObject, GLib, Gio

import os

import json

import tempfile

import traceback



# Pfad für die temporäre Konfigurationsdatei

CONFIG_PATH = os.path.join(tempfile.gettempdir(), 'gimp_spore_workflow_config.json')



class ExportAndNextWorkflow(Gimp.PlugIn):



    def do_query_procedures(self):

        return ['python-fu-export-and-next']



    def do_set_i18n(self, procname):

        return False, 'gimp30-python', None



    def do_create_procedure(self, name):

        procedure = Gimp.ImageProcedure.new(

            self, name,

            Gimp.PDBProcType.PLUGIN,

            self.run, None

        )



        procedure.set_image_types("*")

        procedure.set_sensitivity_mask(Gimp.ProcedureSensitivityMask.DRAWABLE)

        procedure.set_menu_label("1. Exportieren & Nächstes Bild")

        procedure.add_menu_path('/Auswahl/')



        procedure.set_documentation(

            "Exportiert die Auswahl, speichert W/H und lädt das nächste Bild.",

            "Exportiert die aktuelle Auswahl und bereitet den nächsten Schritt vor.",

            name

        )

        procedure.set_attribution("Author", "Author", "2024")



        return procedure



    # --- HILFSFUNKTIONEN ---



    def load_config(self):

        if os.path.exists(CONFIG_PATH):

            try:

                with open(CONFIG_PATH, 'r') as f:

                    return json.load(f)

            except:

                pass

        return None



    def save_config(self, width, height, current_path=None):

        data = {'width': int(width), 'height': int(height)}

        try:

            existing_data = self.load_config()

            if existing_data:

                data.update(existing_data)



            if current_path is not None:

                data['last_path'] = current_path



            with open(CONFIG_PATH, 'w') as f:

                json.dump(data, f, indent=4)

        except Exception as e:

            pass



    def get_next_file(self, current_path):

        directory = os.path.dirname(current_path)

        basename = os.path.basename(current_path)



        extensions = ('.jpg', '.jpeg', '.png', '.tif', '.tiff', '.xcf')

        files = [f for f in os.listdir(directory) if f.lower().endswith(extensions)]

        files.sort()



        try:

            index = files.index(basename)

            if index + 1 < len(files):

                return os.path.join(directory, files[index + 1])

        except ValueError:

            pass



        return None



    # --- HAUPT-RUN-FUNKTION ---



    def run(self, procedure, run_mode, image, n_drawables, drawables, config, run_data=None):



        # 0. Initialisierung & Auswahlprüfung

        selection = image.get_selection()

        success, non_empty, x1, y1, x2, y2 = selection.bounds(image)



        if not non_empty:

            error = GLib.Error.new_literal(Gimp.PlugIn.error_quark(), "Keine Auswahl vorhanden.", 0)

            return procedure.new_return_values(Gimp.PDBStatusType.CALLING_ERROR, error)



        target_w = x2 - x1

        target_h = y2 - y1



        # 1. Pfad ermitteln

        current_path = None

        current_file = image.get_file()



        if current_file is not None:

            current_path = current_file.get_path()

        else:

            config_data = self.load_config()

            if config_data and 'last_path' in config_data:

                current_path = config_data['last_path']



        if current_path is None:

            error = GLib.Error.new_literal(Gimp.PlugIn.error_quark(),

                                            "Bild hat keinen Dateipfad (Konfiguration leer).", 0)

            return procedure.new_return_values(Gimp.PDBStatusType.EXECUTION_ERROR, error)



        directory = os.path.dirname(current_path)

        basename = os.path.splitext(os.path.basename(current_path))[0]



        # Zielordner für Exporte definieren

        target_dir = os.path.join(directory, "Auswahlen")



        # --- TEIL A: EXPORTIEREN ---



        try:

            # 2. Zielordner erstellen (falls nicht vorhanden)

            if not os.path.exists(target_dir):

                os.makedirs(target_dir)



            # 3. Nächste freie Export-Nummer finden (suche im Zielordner)

            number = 1

            while True:

                new_filename = f"[{number}] {basename}.jpg"

                new_filepath = os.path.join(target_dir, new_filename)

                if not os.path.exists(new_filepath):

                    break

                number += 1



            # 4. Auswahl kopieren, neues Bild erstellen, speichern

            drawable = image.get_layers()[-1]

            Gimp.edit_copy([drawable])

            new_image = Gimp.edit_paste_as_new_image()



            # Speichere direkt in den Auswahlen-Ordner

            new_file = Gio.file_new_for_path(new_filepath)

            Gimp.file_save(Gimp.RunMode.NONINTERACTIVE, new_image, new_file)

            new_image.delete()



            # 5. Auswahlgröße speichern (für Script 2)

            self.save_config(target_w, target_h)



        except Exception as e:

            if 'new_image' in locals():

                try: new_image.delete()

                except: pass



            error_msg = f"Fehler beim Export: {str(e)}\n{traceback.format_exc()}"

            error = GLib.Error.new_literal(Gimp.PlugIn.error_quark(), error_msg, 0)

            return procedure.new_return_values(Gimp.PDBStatusType.EXECUTION_ERROR, error)



        # --- TEIL B: NÄCHSTES BILD LADEN (MIT ENDPRÜFUNG) ---



        # 6. Nächstes Bild ermitteln

        next_path = self.get_next_file(current_path)



        if not next_path:

            # ENDE DER SEQUENZ ERREICHT!



            # Letzten Pfad zurücksetzen.

            self.save_config(target_w, target_h, current_path=None)



            # Aktuelles Bild aus der Anzeige entfernen

            image.delete()



            # GIMP 3.0 Methode zum Aufruf der PDB-Prozedur "gimp-quit"

            try:

                # Korrektur: Verwendung von Gimp.TYPE_INT anstelle von GObject.TYPE_INT

                Gimp.plug_in_manager_run(

                    "gimp-quit",

                    Gimp.RunMode.NONINTERACTIVE,

                    GObject.Value.new(Gimp.TYPE_INT, 0)

                )

            except Exception as e:

                Gimp.message(f"WARNUNG: GIMP 3.0 Quit-Aufruf fehlgeschlagen: {str(e)}")



            # Korrigierter Rückgabestatus:

            return procedure.new_return_values(Gimp.PDBStatusType.SUCCESS, None)



        try:

            # 7. Normaler Ablauf: Bild laden und Pfad speichern

            new_gfile = Gio.file_new_for_path(next_path)

            new_image = Gimp.file_load(Gimp.RunMode.NONINTERACTIVE, new_gfile)



            self.save_config(target_w, target_h, current_path=next_path)



            # 8. Anzeige erstellen und altes Bild schließen

            Gimp.Display.new(new_image)

            image.delete()

            Gimp.displays_flush()



        except Exception as e:

            error_msg = f"Fehler beim Wechseln: {str(e)}\n{traceback.format_exc()}"

            error = GLib.Error.new_literal(Gimp.PlugIn.error_quark(), error_msg, 0)

            return procedure.new_return_values(Gimp.PDBStatusType.EXECUTION_ERROR, error)



        # Korrigierter Rückgabestatus:

        return procedure.new_return_values(Gimp.PDBStatusType.SUCCESS, None)



Gimp.main(ExportAndNextWorkflow.__gtype__, sys.argv)

Does anyone have an idea whether what I am trying to do is possible?

help with running an scm script as batch from bash

Wed, 10 Dec 2025 15:15:21 +0000

Hello,

with Gimp 2.8 I have an scm script (I found somewhere) that runs perfect on one image when called from the menu within gimp. Now I'd like to run it for a whole directory and wrote a bash script for this.

It works in general, but when running the script on a single image within the Gimp, it creates a group with two layers and places this group above the original image layer. But when running it as a batch with my bash script, the original image layer is inserted within the new group.

Neither grok nor chat gpt could help me successfully :-)

Can you hint me what to change, so that the original image layer stays as a separate layer outside (below) the new layer-group?

(B.t.w., I have literally no idea of Gimp scripting, am just capable to use and adapt a little scripts I find somewhere :-) )

This is my bash script:

Code:

#!/bin/bash

INPUT_DIR="..."

OUTPUT_SUFFIX=".xcf"

BLUR=15

VISIBLE="FALSE"

for file in "$INPUT_DIR"/*.{jpg,png,tif,bmp}; do

  if [ -f "$file" ]; then

    output_file="${file%.*}${OUTPUT_SUFFIX}"

gimp -i -b "

  (let*

  (

    (filename \"$file\")

    (image (car (gimp-file-load RUN-NONINTERACTIVE filename filename)))

    ;; Bottom layer = original image

    (layer-info (gimp-image-get-layers image))

    (drawable (vector-ref (cadr layer-info) 0))

    ;; Extract name

    (fname (car (reverse (strbreakup filename \"/\"))))

    (basename (car (strbreakup fname \".\")))

  )

  ;; Rename to match GUI

  (gimp-item-set-name drawable basename)

  ;; select this layer before FS script

  (gimp-image-set-active-layer image drawable)

  ;; Run FS

  (script-fu-hk-frequency-separation image drawable $BLUR $VISIBLE)

  ;; Save

  (gimp-xcf-save RUN-NONINTERACTIVE image drawable \"$output_file\" \"$output_file\")

  (gimp-image-delete image)

)

" -b '(gimp-quit 0)'

    echo "Processed: $file -> $output_file"

  fi

done

This is the scm script:

Code:

; This plugin was tested with Gimp 2.8.3

; This plugin will not work for Gimp 2.6.x because of the unknown layer group

;

; frequency-separartion.scm

; seperates a layer or the whole visible into a

; layer group with high- and low frequency parts

;

; This program is free software; you can redistribute it and/or modify

; it under the terms of the GNU General Public License as published by

; the Free Software Foundation; either version 2 of the License, or

; (at your option) any later version.

;

; This program is distributed in the hope that it will be useful,

; but WITHOUT ANY WARRANTY; without even the implied warranty of

; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

; GNU General Public License for more details.

;

; You should have received a copy of the GNU General Public License

; along with this program; if not, write to the Free Software

; Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.

;



(define LList '())



(define (makelist E) (list E))

(define (isempty L) (null? L))

(define (add E L) (cons E L))



(define (Ith i L)

  (if (= i 1)

    (car L)

    (Ith (- i 1) (cdr L))))





(define (fs-set-layers-visible inImage)

        (let*

          (

            (num-layers (length LList))

            (theLayer 0)

          )

          (while (> num-layers 0)

            (set! theLayer (Ith num-layers LList))

            (gimp-drawable-set-visible theLayer TRUE)

            (set! num-layers (- num-layers 1))

          )



;          (gimp-displays-flush)

          (set! LList (list))

        )

)



(define (fs-set-layers-invisible inImage inLayer)

        (let*

          (

            (layers (gimp-image-get-layers inImage))

            (num-layers (car layers))

            (layer-array (cadr layers))

            (theLayer 0)

          )



        (while (> num-layers 0)

           (set! num-layers (- num-layers 1))

           (set! theLayer (vector-ref layer-array num-layers))

           (if (= (car (gimp-drawable-get-visible theLayer) ) TRUE)

              (if (<> theLayer inLayer)

                 (begin

                   (gimp-drawable-set-visible theLayer FALSE)

                   (set! LList (cons theLayer LList))

                 )

              )

           )

        )

;        (gimp-displays-flush)

        )

)



(define (script-fu-hk-frequency-separation image layer blur visible)

    (let*

      (

        ; test, whether seperating a layer group

        (isgroup (car (gimp-item-is-group layer)))

        ; get parent, maybe layer group

        (parent (car (gimp-item-get-parent layer)))

        ; get layername

        (layername (car (gimp-item-get-name layer)))

        ; gets position of layer

        (position (car (gimp-image-get-item-position image layer)))

        ; creates layergroup

        (fs-layergroup (car (gimp-layer-group-new image)))

        ; variables

        (helplayer 0)

        (lowlayer 0)

        (highlayer 0)

        (forceVisible 0)

        (viewmode 0)

        (opacity 0)

      )



      ; FALSE=freqsep just the layer

      ; TRUE=freqsep from new visible

      (if (= visible FALSE)

        (begin

          (if (= isgroup 1)

             (begin

               (set! forceVisible 1)

             )

             (begin

               (set! forceVisible 0)

             )

          )

        )

        (begin

           (set! forceVisible 1)

        )

      )



      (gimp-image-undo-group-start image)

      (if (= forceVisible 0)

        (begin

          (fs-set-layers-invisible image layer)

        )

      )

      ;rename layergroup

      (gimp-item-set-name fs-layergroup "frequency separation")

      ;add layergroup as first to the image layer-stack

      (gimp-image-insert-layer image fs-layergroup 0 0)



      ; 0=freqsep just the layer

      ; 1=freqsep from new visible

      (if (= forceVisible 0)

        (begin

          (set! viewmode (car (gimp-layer-get-mode layer)))

          (set! opacity (car (gimp-layer-get-opacity layer)))

          (gimp-image-reorder-item image layer fs-layergroup -1)

          (gimp-layer-set-mode layer NORMAL-MODE)

          (gimp-layer-set-opacity layer 100)

          (set! helplayer (car (gimp-layer-copy layer 1)))

          (set! lowlayer (car (gimp-layer-copy layer 1)))

          (gimp-drawable-set-name helplayer (string-append "help-" layername))

          (gimp-drawable-set-name lowlayer (string-append "low-" layername))

        )

        (begin

          (set! helplayer (car (gimp-layer-new-from-visible image image "helplayer")))

          (set! lowlayer (car (gimp-layer-new-from-visible image image "low")))

        )

      )

      ; adding layers to layergroup

      (gimp-image-insert-layer image helplayer fs-layergroup -1)

      (gimp-image-insert-layer image lowlayer fs-layergroup -1)



      ; despeckle the low frequency layer

      (plug-in-gauss 1 image lowlayer blur blur 1)

      ; setting layer mode to get the high frequency layer

      (gimp-layer-set-mode lowlayer GRAIN-EXTRACT-MODE)

      ; getting the high frequency layer. that works, because the lauergroup is topmost

      (set! highlayer (car (gimp-layer-new-from-visible image image (string-append "high-" layername))))

      ; add it it to the layergroup

      (gimp-image-insert-layer image highlayer fs-layergroup -1)

      ; remove the helping layer

      (gimp-image-remove-layer image helplayer)

      ; make the low frequency layer normal

      (gimp-layer-set-mode lowlayer NORMAL-MODE)

      ; make the high frequency layer grain-merge, so the result is the image

      (gimp-layer-set-mode highlayer GRAIN-MERGE-MODE)

      ; put the layergroup just above the layer, that was selected

      (gimp-image-reorder-item image fs-layergroup parent position)





      (if (= forceVisible 0)

         (begin

           ; restore viewmode and opacity, now for the whole layergroup.

           (gimp-layer-set-mode fs-layergroup viewmode)

           (gimp-layer-set-opacity fs-layergroup opacity)

           (fs-set-layers-visible image)

         )

      )



      (gimp-image-set-active-layer image lowlayer)



      (gimp-displays-flush)

        (gimp-image-undo-group-end image)

  )

)



(script-fu-register "script-fu-hk-frequency-separation"

"Frequency separation..."

"frequency-separation.scm

divides layer in high and low frequencies"

"H.Kuhse"

"H.Kuhse"

"2013-01-04"

"*"

SF-IMAGE "Image" 0

SF-DRAWABLE "Layer" 0

SF-ADJUSTMENT _"Blur:" '(15 1 50 1 1 0 0)

SF-TOGGLE _"from visible (instead selected layer)" 1

)



(script-fu-menu-register

  "script-fu-hk-frequency-separation"

  "/Filters"

)

GIMP 3 script-fu not loading and run with error when Loaded through Console

Thu, 04 Dec 2025 21:06:09 +0000

new to v3 and having scripting issues.

My test case is a published "hello World"script. (https://testing.developer.gimp.org/resou...script-fu/)

On MacOS I have given GIMP full access in system settings. I have placed this .scm in a subdirectory with the same name as the executable .scm file. In Folders, this script directory is the first listed.

It does not show up when I restart GIMP.

I've read that I can load it in the Console, and when I do this, I get the error:

"Error: eval: unbound variable: (/Users/mark/GIMP_scriptfu/mh_hello_world3/mh_hello_world.scm : 35) [] "

On line 35 in the script, it holds the copyright year. How is this causing this script to error out?

I have tried different example scripts and am getting similar errors.

I am new to this. Thanks for any help!

Code:

#!/usr/bin/env gimp-script-fu-interpreter-3.0



(define (script-fu-zemarmot-hello-world

         image

         drawables

         font

         compute-size

         size

         text)

  (script-fu-use-v3)

  (let* ((layer (gimp-text-layer-new image text font size UNIT-PIXEL)))



    (gimp-image-undo-group-start image)



    (gimp-image-insert-layer image layer -1 0)

    (if (= compute-size TRUE)

      (let* ((image-width (gimp-image-get-width image))

             (layer-width (gimp-drawable-get-width layer)))

        (begin

          (set! size (* size (/ image-width layer-width)))

          (gimp-text-layer-set-font-size layer size UNIT-PIXEL)

        )

      )

    )



    (gimp-image-undo-group-end image)

  )

)



(script-fu-register-filter "script-fu-zemarmot-hello-world"

  "Script-Fu v3 Hello World"

  "Official Hello World Tutorial in Script-Fu v3"

  "Jehan"

  "Jehan, Zemarmot project"

  "2025"

  "*"

  SF-ONE-OR-MORE-DRAWABLE

  SF-FONT       "Font"               "Sans-serif"

  SF-TOGGLE     "Compute Ideal Size" #f

  SF-ADJUSTMENT "Font size (pixels)" '(20 1 1000 1 10 0 1)

  SF-STRING     "Text"               "Hello World!"

)



(script-fu-menu-register "script-fu-zemarmot-hello-world" "/Hello W_orlds")

Gimp 3.0.6 Python plug-in bug with text layer.set_color(color)?

Tue, 18 Nov 2025 14:43:20 +0000

My environment: Apple Mac mini M2, macOS Tahoe 26.1.

I want to set the color of a text layer from a Python plug-in. In Gimp 2 I used the plug-in "gimp-text-layer-set-color".

In Gimp 3 I find the layer method "set_color(color)", but the color remains to be the foreground color. The same with the PDB plug-in.
First try with layer method:

Code:

textColor = Gegl.Color.new("red") # e.g.

textLayer = Gimp.TextLayer.new(image, "Text", someFont, sometextSize, Gimp.Unit.pixel())

textLayer.set_name("Text")

textLayer.set_color(textColor)

image.insert_layer(textLayer, parent, position)

The text is black.
Next try with pdb proc:

Code:

textColor = Gegl.Color.new("red") # e.g.

textLayer = Gimp.TextLayer.new(image, "Text", someFont, sometextSize, Gimp.Unit.pixel())

textLayer.set_name("Text")

pdb = Gimp.get_pdb()

textLayerSetColor = pdb.lookup_procedure("gimp-text-layer-set-color")

pdbConfig = textLayerSetColor.create_config()

pdbConfig.set_property("layer", textLayer)

pdbConfig.set_property("color", textColor)

textLayerSetColor.run(pdbConfig)

image.insert_layer(textLayer, parent, position)

The text is black.
Next try with foreground:

Code:

textColor = Gegl.Color.new("red") # e.g.

foreground = Gimp.context_get_foreground()

Gimp.context_set_foreground(textColor)

textLayer = Gimp.TextLayer.new(image, "Text", someFont, sometextSize, Gimp.Unit.pixel())

textLayer.set_name("Text")

image.insert_layer(textLayer, parent, position)

Gimp.context_set_foreground(foreground)

Success! The text is red.

Where is the error? With my code or with Gimp?

Problem with Gimp 3 Python plug-in color_argument

Sun, 16 Nov 2025 17:51:06 +0000

I try to migrate my Python plug-in to decorate a jpeg photo with margin, border and description from Gimp 2 to Gimp 3.

it works as intended as long as I don't add a color-argument. What is wrong with this code?

Code:

import sys

import os

from glob import glob

import gi

gi.require_version('Gimp', '3.0')

from gi.repository import Gimp

gi.require_version('GimpUi', '3.0')

from gi.repository import GimpUi

from gi.repository import GObject

from gi.repository import GLib

#from gi.repository import Gtk

from gi.repository import Gio

from gi.repository import Gegl



...



class MarginBorder(Gimp.PlugIn):

  def do_query_procedures(self):

    return [plugInMBProc]

  def do_create_procedure(self, name):

    procedure = None

    if name == plugInMBProc:

      procedure = Gimp.ImageProcedure.new(self, name,

                                          Gimp.PDBProcType.PLUGIN,

                                          margin_border_run, None)

      procedure.set_sensitivity_mask(Gimp.ProcedureSensitivityMask.DRAWABLE |

                                     Gimp.ProcedureSensitivityMask.NO_DRAWABLES)

      procedure.set_menu_label("Rand und Rahmen")

      procedure.set_attribution("Volker", "Volker Lenhardt", "2025")

      procedure.add_menu_path("/Volker")

      procedure.set_documentation("Volkers Rand, Rahmen und Bildtext",

                                  "Fügt dem aktiven Bild die Dekorationsebenen "\

                                  + "Rand, Rahmen und Bildtext hinzu.",

                                  None)

      procedure.add_font_argument("font", # name

                                  "Font", # nick

                                  None, # blurb

                                  True, # none_ok

                                  Gimp.Font.get_by_name(

                                    "URW Gothic L Book"), # default_value

                                  False, # default_to_context 

                                  GObject.ParamFlags.READWRITE) # flags

      procedure.add_string_argument("desc", # name

                                    "Bildbeschreibung", # nick

                                    None, # blurb

                                    "Irgendwo", # value

                                    GObject.ParamFlags.READWRITE) # flags

      procedure.add_double_argument("marginRelation", # name

                                    "Randbreite in Promille der größten Bildausdehnung", # nick

                                    None, # blurb

                                    0.0, # min

                                    30.0, # max

                                    9.0, # value

                                    GObject.ParamFlags.READWRITE) # flags

      procedure.add_double_argument("borderRelation", # name

                                    "Rahmenbreite in Promille der größten Bildausdehnung", # nick

                                    None, # blurb

                                    0.0, # min

                                    10.0, # max

                                    1.8, # value

                                    GObject.ParamFlags.READWRITE) # flags

      procedure.add_color_argument("borderColor", # name

                                   "Rahmenfarbe", # nick

                                   None, # blurb

                                   False, # has_alpha

                                   Gegl.Color.new("black"), # value

                                   GObject.ParamFlags.READWRITE) # flags

    return procedure



Gimp.main(MarginBorder.__gtype__, sys.argv)

If I omit the procedure.add_color_argument all is well, but with it the plug-in doesn't even show up in Gimp.

I can't find any code errors, or do you?

Thanks in advance for your help.

Gimp 3.0 get Metadata with Python

Sat, 08 Nov 2025 17:41:14 +0000

How do I retrieve jpg metadata from a Python plug-in in Gimp 3.0?

I open a jpg photo and need some metadata. With Gimp 2 and Python 2 I used the plug-in "gimp_image_get_metadata".
My code was (abridged):

Code:

import xml.etree.ElementTree as ET

metadata = pdb.gimp_image_get_metadata(self.image)

dateTag  = "exif.photo.datetimeoriginal"

date = self.get_metadata_value(metadata, dateTag)

def get_metadata_value(self, metadata,

                             metadataName):

  ciName = metadataName.lower()

  value = ""

  rootNode = ET.XML(metadata)

  for tag in rootNode.findall('tag'):

    if tag.attrib['name'].lower() == ciName:

      value = tag.text

      break

  return value

The plug-in isn't there any more in Gimp 3. Instead I find the class Gimp.Metadata. This class provides some methods for restricted operations, but none for my needs. I try:

Code:

metadata = image.get_metadata()

I expect to get an xml document, but I don't know how to look for it.
Is there a different approach?

My system: Mac mini (Apple M2), macOS Tahoe (26.0.1)
My Gimp: 3.0.6
My Python: 3.10

Get layer by uncompressed size with script

Tue, 02 Sep 2025 16:40:48 +0000

Hi all !

With Gimp, the most part of time I works on photos to simply add extra layers and export the result (maps, explanations, mechanisms, and so).

In this case I do not alter the original and keep it.

Recently I realized the most part of my XCF projects take more than 20MB per file vs the originals which take 2MB max.

After digging I understood the problem : projects keep the original raw uncompressed photo on a layer which is a waste of storage as I can include again the original.

In definitive I want to keep the original photo, the exported one and all that allowed to obtain the exported result.

I imagined an external batch process to open the targeted files, remove the usefulness layer, archive the original with the lighted project.

I discovered script-fu language (LISP scheme) and followed tutorial script-fu and experimented it with success.

But I did not found a way to achieve this : get the uncompressed layer size

I found the following procedures which are not "size related" :
- gimp-image-get-layer-by-name
- gimp-image-get-layer-by-tattoo

Is it possible to do this with script-fu ? Is it possible in another way ?

Extra question, is there a way to attach a file to a XCF project (I suppose no) like we usually do it with the email attachments method ? I hoped this could be achieved with an add-on ...

Thank you in advance for the time you will spend to me .

With adelphity
lnj

Bucket fill whole selection in Python-fu

Thu, 21 Aug 2025 21:45:03 +0000

Hi,

This is a very simple problem, but when I try a bucket fill through a Python-fu call, I always get a "Fill similar colors" result when I need "Fill whole selection". I can't seem to find the relevant setting in the listed context setters for the procedure (I'm looking here). What am I missing?

Thank you!

Python "no return values, shouldn't happen"

Mon, 18 Aug 2025 21:55:34 +0000

I'm creating a python script to load a custom image format (.PRS with the magic string "YB"). But I'm having trouble getting even a skeleton file loading plugin to work. This is what I've been able to reduce my code down to:

Code:

#!/usr/bin/env python3

# -*- coding: utf-8 -*-

import gi

gi.require_version('Gimp', '3.0')

from gi.repository import Gimp, GObject

import sys

load_proc = "file-prs-load"

plugin_binary = "file-prs"

class PrsLoader(Gimp.PlugIn):

    def do_query_procedures(self):

        return [load_proc]

    def do_create_procedure(self, name):

        procedure = None

        if name == load_proc:

            procedure = Gimp.LoadProcedure.new(self, name,

                                                Gimp.PDBProcType.PLUGIN,

                                                self.PrsLoad, None)

            procedure.set_menu_label("PRS image")

            procedure.set_attribution("Nathan", "Nathan", "2025")

            procedure.set_menu_label("PRS image")

            procedure.set_documentation("Loads files of PRS file format",

                                        "Loads files of PRS file format",

                                        None)

            procedure.set_mime_types("image/prs")

            procedure.set_extensions("prs")

            procedure.set_magics("0,string,YB")

        return procedure

    def PrsLoad(self, procedure, run_mode, file, metadata, flags, config, run_data):

        print("PrsLoad called")

        #Return dummy image for now

        image = Gimp.Image.new(100, 100, Gimp.ImageBaseType.RGB)

        return_vals = procedure.new_return_values(Gimp.PDBStatusType.SUCCESS, None)

        val = GObject.Value(Gimp.Image.__gtype__, image)

        return_vals.append(val)

        return return_vals

Gimp.main(PrsLoader.__gtype__, sys.argv)

When I try to open a PRS image file, I get the following errors:

Code:

GIMP Message

Opening '/home/nathan/Pictures/tree2b.prs' failed:

PRS image plug-in could not open image

Code:

[nathan@TARDIS file-prs]$ gimp

set device 'Wayland Pointer' to mode: disabled

[file-prs-load] The catalog directory does not exist: /home/nathan/.config/GIMP/3.0/plug-ins/file-prs/locale

[file-prs-load] Override method set_i18n() for the plug-in to customize or disable localization.

[file-prs-load] Localization disabled

env: ‘gjs’: No such file or directory

gimp: LibGimpBase-WARNING: gimp: gimp_wire_read(): unexpected EOF

[file-prs-load] The catalog directory does not exist: /home/nathan/.config/GIMP/3.0/plug-ins/file-prs/locale

[file-prs-load] Override method set_i18n() for the plug-in to customize or disable localization.

[file-prs-load] Localization disabled

[file-prs-load] The catalog directory does not exist: /home/nathan/.config/GIMP/3.0/plug-ins/file-prs/locale

[file-prs-load] Override method set_i18n() for the plug-in to customize or disable localization.

[file-prs-load] Localization disabled

PrsLoad called

(file-prs.py:2595385): LibGimp-WARNING **: 17:46:44.730: _gimp_procedure_run_array: no return values, shouldn't happen

/home/nathan/.config/GIMP/3.0/plug-ins/file-prs/file-prs.py:46: Warning: g_error_new: assertion 'domain != 0' failed

  Gimp.main(PrsLoader.__gtype__, sys.argv)

INFO: a stray image seems to have been left around by a plug-in: "[Untitled]"[nathan@TARDIS file-prs]$

I'm not too worried about the localization messages at this time.

gegl filters in batch processing

Sun, 10 Aug 2025 20:18:42 +0000

I am trying to call a gegl edge detect filter repeatedly in a while loop as here

Code:

        (while (not (null? filepaths))        

                       ;some stuff    

                       (gimp-drawable-merge-new-filter

            2                         ; The drawable to apply the filter to

            "gegl:edge"                        ; The name of the GEGL operation

            0                                  ; The merge-mode (0 for normal)

            LAYER-MODE-REPLACE                 ; The layer mode (e.g., replace, normal, etc.)

            100.0                              ; The opacity (0.0 to 100.0)

            "amount" 2.0                       ; Parameter: amount of edge detection

            "border-behavior" "none"           ; Parameter: how to handle image borders

            "algorithm" "sobel"                ; Parameter: the edge detection algorithm

            )

                        ;some more stuff

          )

it works but only on the first image in the batch. Is there something I need to do to get it to work on all the images?

file-glob always returns an empty list

Sat, 09 Aug 2025 22:31:17 +0000

Hi, I just set out to write a tinyscheme script that retrieves the list of files from a folder using file-glob and although it looks trivial I can't get it to work.

the line in question is:

(let* (filepaths (cadr (file-glob (string-append input-path "*.png") 0))))

which always returns empty whether I place it in a script or type a version into the console. Gimp just tells me that parameter 1 to cdr needs to be a pair.
Yes, I have verified that the folder pointed to by input-path exists and that there is at least one png file in it.Permissions are ok (I can load the file explicitly)
I have tried unix style and windows style path separator: nothing works.

I can hardly believe something this fundamental is actually broken - is there something about my setup that is wrong? I have a standard 3.04 installation using the windows installer any insights are appreciated.

Gimp 3.0.4 python batch interpreter oddities

Sat, 09 Aug 2025 11:47:12 +0000

Hi,

I use the python-fu-eval batch interpreter on GIMP 3.0.4 AppImage and have noticed some unexpected behaviour. As a trivial example,

User:~$ ~/AppImages/gimp/GIMP-3.0.4-x86_64.AppImage -i --quit --batch-interpreter=python-fu-eval -b 'if True: a="something"; print(a);'
something
batch command executed successfully

But if the "if" statement comes after the declaration of variable "a"

User:~$ ~/AppImages/gimp/GIMP-3.0.4-x86_64.AppImage -i --quit --batch-interpreter=python-fu-eval -b 'a="something";if True: print(a);'
batch command experienced a calling error:
Traceback (most recent call last):
File "/tmp/.mount_GIMP-3bPBAmP/usr/lib/x86_64-linux-gnu/gimp/3.0/plug-ins/python-eval/python-eval.py", line 42, in code_eval
    exec(code, globals())
File "", line 1
    a="something";if True: print(a);
               ^^
SyntaxError: invalid syntax

Similar problems arise with a "with" statement.

I don't understand this. Has anyone else noticed this sort of behaviour?

My OS is linux 6.1.0-37-amd64

Thanks.

"Simple" python script in gimp 3

Mon, 28 Jul 2025 06:08:13 +0000

Hello!
For a game I'd like to create tiles for tileset. The task is simple: get a image with 3+ layers (colors, and heightmap), for each layer create 3 copy, rotate it 60, 180 ,270 degree and offset it 1*,2*,3* width.
After get the heightmap layer duplicate it, add noise and call the normal map filter. (there are some masking, etc). Get the color layers, add some nose. Flattering down each group of layers.
Because I'd like to use it many times, i created (or try to create) a plugin for it.
Looks like easy.

The problems: because its a python, you need to test is, for the typos too. so each iteration need to restart gimp load the image test the script.
There are no usable documentation. I already found the https://lazka.github.io/pgi-docs/#Gimp-3.0, helped a lot, but lack of examples. all examples for the gimp 2.x
For example, how can I do filter->generic->normal map?
There are some undocumented "features": If you rename a layer, in the gimp UI the name stay the old one (its not a big problem) but after you cant export it into .ora and gimp went some unstable state. If remove a layer from image cant add it again to it, because already had.

Now I have 300+ line of code for that "simple" task, about quarter are dead code ("maybe that work") and not working well:

- after I call .transform_rotate_simple(rotate_angle,True,0,0) its stay is "floating selection transform" state and somehow need to anchor it how?
- how can I proper call the filters->generic->normal map filter with preset or how set the values?
- how can I proper call the filters->noise->hsv filter with preset or how set the values?
- how can I invert selection?
- how can I proper generate greyscale noise between color rgb(123,123,123) and rgb (133,133,133), I tried the buffer.set(rect, "RGBA u8", bytes(pixels)) with pixels.extend([gray, gray, gray, 255]) it's worked, but not get the right color (I get lighter about rgb(180,180,180), now I use set_pixel but its slow for 64x64 images
- from where can I get some useful examples for gimp 3 python scripting?

Thanx

Python Snippets for GIMP 3

Sat, 19 Jul 2025 13:21:40 +0000

Hi all,

I’m familiar with Script-Fu, but I know nothing about Python scripting. (GIMP 3)
I’m testing a macro system and would really appreciate some small, self-contained Python snippets that can run with a button press, just something minimal to confirm it's working.

Any suggestions or examples would be cool.

Thanks!

edit_paste pastes at incorrect location after scaling

Thu, 17 Jul 2025 13:41:38 +0000

I am working on an application to scale an image, but the bottom part of the image contains a special pattern that must be copied over and repeated since it is distorted during scaling. I have the following section of code:

Code:

image1 = Gimp.file_load(Gimp.RunMode.NONINTERACTIVE, Gio.file_new_for_path(image1_dir))

image1_width = image1.get_width()

image1.select_rectangle(Gimp.ChannelOps.REPLACE, 0, 1479, image1_width, 57)

image1_dr = image1.get_selected_drawables()

Gimp.edit_copy(image1_dr)

image.scale(image_new_width, image_height)

image1.select_rectangle(Gimp.ChannelOps.REPLACE, 0, 1479, image1_width, 57)

image1_dr = image1.get_selected_drawables()

Gimp.edit_paste(image1_dr[0], True)

image1.select_rectangle(Gimp.ChannelOps.REPLACE, image1_width, 1479, image1_width, 57)

image1_dr = image1.get_selected_drawables()

Gimp.edit_paste(image1_dr[0], True)

In this code, I select and copy an area of size (width, 57) from coordinate (0,1479). Then, I scale the image to a new width. I then try and paste the selected area at coordinates (0,1479) and then at (original_width, 1479). So the whole of the bottom of the new image should have the original pattern. However, only a region on the bottom right of the image is covered, and not at the specified regions. In an example, if width was originally 25378, and then scaled to 31234, the selected area was only pasted onto the region from (15617,1479) to (31234,1479). The other area where I pasted first from (0,1479) to (15617,1479) has the distorted pattern after scaling and not the original pattern that was pasted onto it.

Can you please explain why this occurs, and suggest a way to properly copy and paste the original pattern onto the scaled image?

Thank you for your help!

Endianess of pixel data in a Python region for gimp 2.10.

Mon, 14 Jul 2025 14:49:47 +0000

For gimp 2.10, is my understanding correct?

Given a region's bytes from a layer ‘l’:

r = l.get_pixel_rgn(0, 0, w, h)
rb = r[:, :]

I think rb will hold the bytes of the channel's values in the machine's native endianess because:

- a region maps onto tiles,
- getting the tile's pixels uses a simple memcpy(), and
- a tile's pixels are in the native endianess.

I'm aware tile data in an XCF file is big-endian from
https://testing.developer.gimp.org/core/...ganization
But I'm interested in what's in memory and guaranteed to be seen through slicing a Python region into a string of bytes.