lv_img_conv_py: minimal python port of node module

.devcontainer/Dockerfile

@@ -11,6 +11,7 @@ RUN apt-get update -qq \
       make \
       python3 \
       python3-pip \
+      python3-pil \
       tar \
       unzip \
       wget \ 

.github/workflows/main.yml

@@ -31,6 +31,10 @@ jobs:
         uses: actions/checkout@v3
         with:
           submodules: recursive
+      - name: Install resource build dependencies
+        run:  |
+          apt-get update
+          apt-get -y install --no-install-recommends python3-pil
       - name: Build
         shell: bash
         run: /opt/build.sh all

doc/buildAndProgram.md

@@ -42,7 +42,7 @@ CMake configures the project according to variables you specify the command line
 **NRF5_SDK_PATH**|path to the NRF52 SDK|`-DNRF5_SDK_PATH=/home/jf/nrf52/Pinetime/sdk`|
 **CMAKE_BUILD_TYPE (\*)**| Build type (Release or Debug). Release is applied by default if this variable is not specified.|`-DCMAKE_BUILD_TYPE=Debug`
 **BUILD_DFU (\*\*)**|Build DFU files while building (needs [adafruit-nrfutil](https://github.com/adafruit/Adafruit_nRF52_nrfutil)).|`-DBUILD_DFU=1`
-**BUILD_RESOURCES (\*\*)**| Generate external resource while building (needs [lv_font_conv](https://github.com/lvgl/lv_font_conv) and [lv_img_conv](https://github.com/lvgl/lv_img_conv). |`-DBUILD_RESOURCES=1`
+**BUILD_RESOURCES (\*\*)**| Generate external resource while building (needs [lv_font_conv](https://github.com/lvgl/lv_font_conv) and [python3-pil/pillow](https://pillow.readthedocs.io) module). |`-DBUILD_RESOURCES=1`
 **TARGET_DEVICE**|Target device, used for hardware configuration. Allowed: `PINETIME, MOY-TFK5, MOY-TIN5, MOY-TON5, MOY-UNK`|`-DTARGET_DEVICE=PINETIME` (Default)
 
 #### (\*) Note about **CMAKE_BUILD_TYPE**

docker/Dockerfile

@@ -11,6 +11,7 @@ RUN apt-get update -qq \
       make \
       python3 \
       python3-pip \
+      python3-pil \
       python-is-python3 \
       tar \
       unzip \
@@ -39,10 +40,6 @@ RUN pip3 install -Iv cryptography==3.3
 RUN pip3 install cbor
 RUN npm i lv_font_conv@1.5.2 -g
 
-RUN npm i ts-node@10.9.1 -g
-RUN npm i @swc/core -g
-RUN npm i lv_img_conv@0.3.0 -g
-
 # build.sh knows how to compile
 COPY build.sh /opt/
 

src/resources/CMakeLists.txt

@@ -3,8 +3,8 @@ find_program(LV_FONT_CONV "lv_font_conv" NO_CACHE REQUIRED
     HINTS "${CMAKE_SOURCE_DIR}/node_modules/.bin")
 message(STATUS "Using ${LV_FONT_CONV} to generate font files")
 
-find_program(LV_IMG_CONV "lv_img_conv" NO_CACHE REQUIRED
-    HINTS "${CMAKE_SOURCE_DIR}/node_modules/.bin")
+find_program(LV_IMG_CONV "lv_img_conv.py" NO_CACHE REQUIRED
+    HINTS "${CMAKE_CURRENT_SOURCE_DIR}")
 message(STATUS "Using ${LV_IMG_CONV} to generate font files")
 
 if(CMAKE_VERSION VERSION_GREATER_EQUAL 3.12)

src/resources/generate-img.py

@@ -11,6 +11,9 @@
 
 def gen_lvconv_line(lv_img_conv: str, dest: str, color_format: str, output_format: str, binary_format: str, sources: str):
     args = [lv_img_conv, sources, '--force', '--output-file', dest, '--color-format', color_format, '--output-format', output_format, '--binary-format', binary_format]
+    if lv_img_conv.endswith(".py"):
+        # lv_img_conv is a python script, call with current python executable
+        args = [sys.executable] + args
 
     return args
 

src/resources/lv_img_conv.py

@@ -0,0 +1,193 @@
+#!/usr/bin/env python3
+import argparse
+import pathlib
+import sys
+import decimal
+from PIL import Image
+
+
+def classify_pixel(value, bits):
+    def round_half_up(v):
+        """python3 implements "propper" "banker's rounding" by rounding to the nearest
+        even number. Javascript rounds to the nearest integer.
+        To have the same output as the original JavaScript implementation add a custom
+        rounding function, which does "school" rounding (to the nearest integer).
+
+        see: https://stackoverflow.com/questions/43851273/how-to-round-float-0-5-up-to-1-0-while-still-rounding-0-45-to-0-0-as-the-usual
+        """
+        return int(decimal.Decimal(v).quantize(decimal.Decimal('1'), rounding=decimal.ROUND_HALF_UP))
+    tmp = 1 << (8 - bits)
+    val = round_half_up(value / tmp) * tmp
+    if val < 0:
+        val = 0
+    return val
+
+
+def test_classify_pixel():
+    # test difference between round() and round_half_up()
+    assert classify_pixel(18, 5) == 16
+    # school rounding 4.5 to 5, but banker's rounding 4.5 to 4
+    assert classify_pixel(18, 6) == 20
+
+
+def main():
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument("img",
+        help="Path to image to convert to C header file")
+    parser.add_argument("-o", "--output-file",
+        help="output file path (for single-image conversion)",
+        required=True)
+    parser.add_argument("-f", "--force",
+        help="allow overwriting the output file",
+        action="store_true")
+    parser.add_argument("-i", "--image-name",
+        help="name of image structure (not implemented)")
+    parser.add_argument("-c", "--color-format",
+        help="color format of image",
+        default="CF_TRUE_COLOR_ALPHA",
+        choices=[
+            "CF_ALPHA_1_BIT", "CF_ALPHA_2_BIT", "CF_ALPHA_4_BIT",
+            "CF_ALPHA_8_BIT", "CF_INDEXED_1_BIT", "CF_INDEXED_2_BIT", "CF_INDEXED_4_BIT",
+            "CF_INDEXED_8_BIT", "CF_RAW", "CF_RAW_CHROMA", "CF_RAW_ALPHA",
+            "CF_TRUE_COLOR", "CF_TRUE_COLOR_ALPHA", "CF_TRUE_COLOR_CHROMA", "CF_RGB565A8",
+        ],
+        required=True)
+    parser.add_argument("-t", "--output-format",
+        help="output format of image",
+        default="bin", # default in original is 'c'
+        choices=["c", "bin"])
+    parser.add_argument("--binary-format",
+        help="binary color format (needed if output-format is binary)",
+        default="ARGB8565_RBSWAP",
+        choices=["ARGB8332", "ARGB8565", "ARGB8565_RBSWAP", "ARGB8888"])
+    parser.add_argument("-s", "--swap-endian",
+        help="swap endian of image (not implemented)",
+        action="store_true")
+    parser.add_argument("-d", "--dither",
+        help="enable dither (not implemented)",
+        action="store_true")
+    args = parser.parse_args()
+
+    img_path = pathlib.Path(args.img)
+    out = pathlib.Path(args.output_file)
+    if not img_path.is_file():
+        print(f"Input file is missing: '{args.img}'")
+        return 1
+    print(f"Beginning conversion of {args.img}")
+    if out.exists():
+        if args.force:
+            print(f"overwriting {args.output_file}")
+        else:
+            pritn(f"Error: refusing to overwrite {args.output_file} without -f specified.")
+            return 1
+    out.touch()
+
+    # only implemented the bare minimum, everything else is not implemented
+    if args.color_format not in ["CF_INDEXED_1_BIT", "CF_TRUE_COLOR_ALPHA"]:
+        raise NotImplementedError(f"argument --color-format '{args.color_format}' not implemented")
+    if args.output_format != "bin":
+        raise NotImplementedError(f"argument --output-format '{args.output_format}' not implemented")
+    if args.binary_format not in ["ARGB8565_RBSWAP", "ARGB8888"]:
+        raise NotImplementedError(f"argument --binary-format '{args.binary_format}' not implemented")
+    if args.image_name:
+        raise NotImplementedError(f"argument --image-name not implemented")
+    if args.swap_endian:
+        raise NotImplementedError(f"argument --swap-endian not implemented")
+    if args.dither:
+        raise NotImplementedError(f"argument --dither not implemented")
+
+    # open image using Pillow
+    img = Image.open(img_path)
+    img_height = img.height
+    img_width = img.width
+    if args.color_format == "CF_TRUE_COLOR_ALPHA" and args.binary_format == "ARGB8888":
+        buf = bytearray(img_height*img_width*4) # 4 bytes (32 bit) per pixel
+        for y in range(img_height):
+            for x in range(img_width):
+                i = (y*img_width + x)*4 # buffer-index
+                pixel = img.getpixel((x,y))
+                r, g, b, a = pixel
+                buf[i + 0] = r
+                buf[i + 1] = g
+                buf[i + 2] = b
+                buf[i + 3] = a
+
+    elif args.color_format == "CF_TRUE_COLOR_ALPHA" and args.binary_format == "ARGB8565_RBSWAP":
+        buf = bytearray(img_height*img_width*3) # 3 bytes (24 bit) per pixel
+        for y in range(img_height):
+            for x in range(img_width):
+                i = (y*img_width + x)*3 # buffer-index
+                pixel = img.getpixel((x,y))
+                r_act = classify_pixel(pixel[0], 5)
+                g_act = classify_pixel(pixel[1], 6)
+                b_act = classify_pixel(pixel[2], 5)
+                a = pixel[3]
+                r_act = min(r_act, 0xF8)
+                g_act = min(g_act, 0xFC)
+                b_act = min(b_act, 0xF8)
+                c16 = ((r_act) << 8) | ((g_act) << 3) | ((b_act) >> 3) # RGR565
+                buf[i + 0] = (c16 >> 8) & 0xFF
+                buf[i + 1] = c16 & 0xFF
+                buf[i + 2] = a
+
+    elif args.color_format == "CF_INDEXED_1_BIT": # ignore binary format, use color format as binary format
+        w = img_width >> 3
+        if img_width & 0x07:
+            w+=1
+        max_p = w * (img_height-1) + ((img_width-1) >> 3) + 8  # +8 for the palette
+        buf = bytearray(max_p+1)
+
+        for y in range(img_height):
+            for x in range(img_width):
+                c, a = img.getpixel((x,y))
+                p = w * y + (x >> 3) + 8  # +8 for the palette
+                buf[p] |= (c & 0x1) << (7 - (x & 0x7))
+        # write palette information, for indexed-1-bit we need palette with two values
+        # write 8 palette bytes
+        buf[0] = 0
+        buf[1] = 0
+        buf[2] = 0
+        buf[3] = 0
+        # Normally there is much math behind this, but for the current use case this is close enough
+        # only needs to be more complicated if we have more than 2 colors in the palette
+        buf[4] = 255
+        buf[5] = 255
+        buf[6] = 255
+        buf[7] = 255
+    else:
+        # raise just to be sure
+        raise NotImplementedError(f"args.color_format '{args.color_format}' with args.binary_format '{args.binary_format}' not implemented")
+
+    # write header
+    match args.color_format:
+        case "CF_TRUE_COLOR_ALPHA":
+            lv_cf = 5
+        case "CF_INDEXED_1_BIT":
+            lv_cf = 7
+        case _:
+            # raise just to be sure
+            raise NotImplementedError(f"args.color_format '{args.color_format}' not implemented")
+    header_32bit = lv_cf | (img_width << 10) | (img_height << 21)
+    buf_out = bytearray(4 + len(buf))
+    buf_out[0] = header_32bit & 0xFF
+    buf_out[1] = (header_32bit & 0xFF00) >> 8
+    buf_out[2] = (header_32bit & 0xFF0000) >> 16
+    buf_out[3] = (header_32bit & 0xFF000000) >> 24
+    buf_out[4:] = buf
+
+    # write byte buffer to file
+    with open(out, "wb") as f:
+        f.write(buf_out)
+    return 0
+
+
+if __name__ == '__main__':
+    if "--test" in sys.argv:
+        # run small set of tests and exit
+        print("running tests")
+        test_classify_pixel()
+        print("success!")
+        sys.exit(0)
+    # run normal program
+    sys.exit(main())