Stricter parameter checking for many public functions

src/ocrmypdf/_exec/ghostscript.py

@@ -166,6 +166,7 @@ class GhostscriptFollower:
 def generate_pdfa(
     pdf_pages,
     output_file: os.PathLike,
+    *,
     compression: str,
     pdf_version: str = '1.5',
     pdfa_part: str = '2',

src/ocrmypdf/_exec/tesseract.py

@@ -221,6 +221,7 @@ def _generate_null_hocr(output_hocr, output_text, image):
 
 
 def generate_hocr(
+    *,
     input_file: Path,
     output_hocr: Path,
     output_text: Path,

src/ocrmypdf/_exec/unpaper.py

@@ -69,7 +69,7 @@ def _setup_unpaper_io(tmpdir: Path, input_file: Path) -> Tuple[Path, Path]:
 
 
 def run(
-    input_file: Path, output_file: Path, dpi: DecFloat, mode_args: List[str]
+    input_file: Path, output_file: Path, *, dpi: DecFloat, mode_args: List[str]
 ) -> None:
     args_unpaper = ['unpaper', '-v', '--dpi', str(round(dpi, 6))] + mode_args
 
@@ -114,6 +114,7 @@ def validate_custom_args(args: str) -> List[str]:
 def clean(
     input_file: Path,
     output_file: Path,
+    *,
     dpi: DecFloat,
     unpaper_args: Optional[List[str]] = None,
 ):
@@ -130,4 +131,4 @@ def clean(
     ]
     if not unpaper_args:
         unpaper_args = default_args
-    run(input_file, output_file, dpi, unpaper_args)
+    run(input_file, output_file, dpi=dpi, mode_args=unpaper_args)

src/ocrmypdf/_pipeline.py

@@ -480,7 +480,12 @@ def preprocess_deskew(input_file: Path, page_context: PageContext):
 def preprocess_clean(input_file: Path, page_context: PageContext):
     output_file = page_context.get_path('pp_clean.png')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    unpaper.clean(input_file, output_file, dpi.x, page_context.options.unpaper_args)
+    unpaper.clean(
+        input_file,
+        output_file,
+        dpi=dpi.x,
+        unpaper_args=page_context.options.unpaper_args,
+    )
     return output_file
 
 
@@ -616,9 +621,9 @@ def render_hocr_page(hocr: Path, page_context: PageContext):
     dpi = get_page_square_dpi(page_context.pageinfo, options)
     debug_mode = options.pdf_renderer == 'hocrdebug'
 
-    hocrtransform = HocrTransform(hocr, dpi.x)  # square
+    hocrtransform = HocrTransform(hocr_filename=hocr, dpi=dpi.x)  # square
     hocrtransform.to_pdf(
-        output_file,
+        out_filename=output_file,
         image_filename=None,
         show_bounding_boxes=False if not debug_mode else True,
         invisible_text=True if not debug_mode else False,

src/ocrmypdf/api.py

@@ -45,6 +45,7 @@ class Verbosity(IntEnum):
 
 def configure_logging(
     verbosity: Verbosity,
+    *,
     progress_bar_friendly: bool = True,
     manage_root_logger: bool = False,
 ):

src/ocrmypdf/hocrtransform.py

@@ -77,7 +77,7 @@ class HocrTransform:
         {'ff': 'ff', 'ffi': 'f‌f‌i', 'ffl': 'f‌f‌l', 'fi': 'fi', 'fl': 'fl'}
     )
 
-    def __init__(self, hocr_filename: Union[str, Path], dpi: float):
+    def __init__(self, *, hocr_filename: Union[str, Path], dpi: float):
         self.dpi = dpi
         self.hocr = ElementTree.parse(os.fspath(hocr_filename))
 
@@ -182,6 +182,7 @@ class HocrTransform:
 
     def to_pdf(
         self,
+        *,
         out_filename: Path,
         image_filename: Optional[Path] = None,
         show_bounding_boxes: bool = False,
@@ -433,10 +434,10 @@ if __name__ == "__main__":
     parser.add_argument('outputfile', help='Path to the PDF file to be generated')
     args = parser.parse_args()
 
-    hocr = HocrTransform(args.hocrfile, args.resolution)
+    hocr = HocrTransform(hocr_filename=args.hocrfile, dpi=args.resolution)
     hocr.to_pdf(
-        args.outputfile,
-        args.image,
-        args.boundingboxes,
+        out_filename=args.outputfile,
+        image_filename=args.image,
+        show_bounding_boxes=args.boundingboxes,
         interword_spaces=args.interword_spaces,
     )

tests/test_hocrtransform.py

@@ -53,8 +53,10 @@ def test_mono_image(blank_hocr, outdir):
         im.putpixel((n, n), 1)
     im.save(outdir / 'mono.tif', format='TIFF')
 
-    hocr = hocrtransform.HocrTransform(str(blank_hocr), 300)
-    hocr.to_pdf(str(outdir / 'mono.pdf'), image_filename=str(outdir / 'mono.tif'))
+    hocr = hocrtransform.HocrTransform(hocr_filename=str(blank_hocr), dpi=300)
+    hocr.to_pdf(
+        out_filename=str(outdir / 'mono.pdf'), image_filename=str(outdir / 'mono.tif')
+    )
 
     check_pdf(str(outdir / 'mono.pdf'))
 

tests/test_optimize.py

@@ -143,7 +143,7 @@ def test_multiple_pngs(resources, outdir):
             outputstream=inpdf,
         )
 
-    def mockquant(input_file, output_file, _quality_min, _quality_max):
+    def mockquant(input_file, output_file, *args):
         with Image.open(input_file) as im:
             draw = ImageDraw.Draw(im)
             draw.rectangle((0, 0, im.width, im.height), fill=128)