Add image conversion backends

camelot/backends/__init__.py

@@ -0,0 +1,3 @@
+# -*- coding: utf-8 -*-
+
+from .image_conversion import ImageConversionBackend

camelot/backends/ghostscript_backend.py

@@ -0,0 +1,17 @@
+# -*- coding: utf-8 -*-
+
+import ghostscript
+
+
+class GhostscriptBackend(object):
+    def convert(self, pdf_path, png_path, resolution=300):
+        gs_args = [
+            "gs",
+            "-q",
+            "-sDEVICE=png16m",
+            "-o",
+            png_path,
+            f"-r{resolution}",
+            pdf_path,
+        ]
+        ghostscript.Ghostscript(*gs_args)

camelot/backends/image_conversion.py

@@ -0,0 +1,15 @@
+# -*- coding: utf-8 -*-
+
+from .poppler_backend import PopplerBackend
+from .ghostscript_backend import GhostscriptBackend
+
+backends = {"poppler": PopplerBackend, "ghostscript": GhostscriptBackend}
+
+
+class ImageConversionBackend(object):
+    def __init__(self, backend="poppler"):
+        self.backend = backend
+
+    def convert(self, pdf_path, png_path):
+        converter = backends[self.backend]()
+        converter.convert(pdf_path, png_path)

camelot/backends/poppler_backend.py

@@ -0,0 +1,8 @@
+# -*- coding: utf-8 -*-
+
+from pdftopng import pdftopng
+
+
+class PopplerBackend(object):
+    def convert(self, pdf_path, png_path):
+        pdftopng.convert(pdf_path, png_path)

camelot/parsers/lattice.py

@@ -29,6 +29,7 @@ from ..image_processing import (
     find_contours,
     find_joints,
 )
+from ..backends import ImageConversionBackend
 
 
 logger = logging.getLogger("camelot")
@@ -111,7 +112,7 @@ class Lattice(BaseParser):
         threshold_constant=-2,
         iterations=0,
         resolution=300,
-        **kwargs
+        **kwargs,
     ):
         self.table_regions = table_regions
         self.table_areas = table_areas
@@ -208,9 +209,8 @@ class Lattice(BaseParser):
         return t
 
     def _generate_image(self):
-        from pdftopng import pdftopng
-
-        pdftopng.convert(pdf_path=self.filename, png_path=self.imagename)
+        converter = ImageConversionBackend()
+        converter.convert(self.filename, self.imagename)
 
     def _generate_table_bbox(self):
         def scale_areas(areas):