Added support for changing color conversion strategy

src/ocrmypdf/_exec/ghostscript.py

@@ -28,6 +28,17 @@ except AttributeError:
     # Pillow 9 shim
     Transpose = Image  # type: ignore
 
+
+COLOR_CONVERSION_STRATEGIES = frozenset(
+    [
+        'CMYK',
+        'Gray',
+        'LeaveColorUnchanged',
+        'RGB',
+        'UseDeviceIndependentColor',
+    ]
+)
+
 log = logging.getLogger(__name__)
 
 # Most reliable way to get the bitness of Python interpreter, according to Python docs
@@ -160,6 +171,7 @@ def generate_pdfa(
     output_file: os.PathLike,
     *,
     compression: str,
+    color_conversion_strategy: str,
     pdf_version: str = '1.5',
     pdfa_part: str = '2',
     progressbar_class=None,
@@ -209,16 +221,16 @@ def generate_pdfa(
             "-dBATCH",
             "-dNOPAUSE",
             "-dSAFER",
-            "-dCompatibilityLevel=" + str(pdf_version),
+            f"-dCompatibilityLevel={str(pdf_version)}",
             "-sDEVICE=pdfwrite",
             "-dAutoRotatePages=/None",
-            "-sColorConversionStrategy=" + strategy,
+            f"-sColorConversionStrategy={color_conversion_strategy}",
         ]
         + (['-dPDFSTOPONERROR'] if stop_on_error else [])
         + compression_args
         + [
             "-dJPEGQ=95",
-            "-dPDFA=" + pdfa_part,
+            f"-dPDFA={pdfa_part}",
             "-dPDFACompatibilityPolicy=1",
             "-o",
             "-",

src/ocrmypdf/_pipeline.py

@@ -793,6 +793,7 @@ def convert_to_pdfa(input_pdf: Path, input_ps_stub: Path, context: PdfContext) -
         pdfmark=input_ps_stub,
         output_file=output_file,
         compression=options.pdfa_image_compression,
+        color_conversion_strategy=options.color_conversion_strategy,
         pdfa_part=options.output_type[-1],  # is pdfa-1, pdfa-2, or pdfa-3
         progressbar_class=(
             context.plugin_manager.hook.get_progressbar_class()

src/ocrmypdf/builtin_plugins/ghostscript.py

@@ -14,6 +14,20 @@ from ocrmypdf.subprocess import check_external_program
 log = logging.getLogger(__name__)
 
 
+@hookimpl
+def add_options(parser):
+    gs = parser.add_argument_group("Ghostscript", "Advanced control of Ghostscript")
+    gs.add_argument(
+        '--color-conversion-strategy',
+        action='store',
+        type=str,
+        metavar='STRATEGY',
+        choices=ghostscript.COLOR_CONVERSION_STRATEGIES,
+        default='LeaveColorUnchanged',
+        help="Set Ghostscript color conversion strategy",
+    )
+
+
 @hookimpl
 def check_options(options):
     """Check that the options are valid for this plugin."""
@@ -33,6 +47,10 @@ def check_options(options):
 
     if options.output_type == 'pdfa':
         options.output_type = 'pdfa-2'
+    if options.color_conversion_strategy not in ghostscript.COLOR_CONVERSION_STRATEGIES:
+        raise ValueError(
+            f"Invalid color conversion strategy: {options.color_conversion_strategy}"
+        )
 
 
 @hookimpl
@@ -68,6 +86,7 @@ def generate_pdfa(
     pdfmark,
     output_file,
     compression,
+    color_conversion_strategy,
     pdf_version,
     pdfa_part,
     progressbar_class,
@@ -78,6 +97,7 @@ def generate_pdfa(
         pdf_pages=[*pdf_pages, pdfmark],
         output_file=output_file,
         compression=compression,
+        color_conversion_strategy=color_conversion_strategy,
         pdf_version=pdf_version,
         pdfa_part=pdfa_part,
         progressbar_class=progressbar_class,

src/ocrmypdf/pluginspec.py

@@ -467,6 +467,7 @@ def generate_pdfa(
     pdfmark: Path,
     output_file: Path,
     compression: str,
+    color_conversion_strategy: str,
     pdf_version: str,
     pdfa_part: str,
     progressbar_class,
@@ -489,6 +490,8 @@ def generate_pdfa(
             possible. For lossless, all images should be converted to FlateEncode
             (lossless PNG). If an empty string, the PDF generator should make its
             own decisions about how to encode images.
+        color_conversion_strategy: A valid Ghostscript color conversion strategy
+            name.
         pdf_version: The minimum PDF version that the output file should be.
             At its own discretion, the PDF/A generator may raise the version,
             but should not lower it.