try to run convert, but fall back on gs if needed

IbnNafis007 · Mar 3, 2019 · 29b0886 · 29b0886
1 parent ea282c2
commit 29b0886
Showing 1 changed file with 33 additions and 19 deletions.
diff --git a/src/paperless_tesseract/parsers.py b/src/paperless_tesseract/parsers.py
@@ -46,25 +46,39 @@ def get_thumbnail(self):
         """
 
         out_path = os.path.join(self.tempdir, "convert.png")
-        gs_out_path = os.path.join(self.tempdir, "gs_out.png")
-
-        # Extract the first PDF page as a PNG using Ghostscript  
-        # https://github.com/danielquinn/paperless/issues/447
-        # call gs first
-        cmd = [self.GHOSTSCRIPT, 
-               "-q",
-               "-sDEVICE=pngalpha",
-               "-o", gs_out_path,
-               self.document_path]
-        if not subprocess.Popen(cmd).wait() == 0:
-            raise ParseError("Thumbnail (gs) failed at {}".format(cmd))
-        # then run convert on the output from gs
-        run_convert(
-            self.CONVERT,
-            "-scale", "500x5000",
-            "-alpha", "remove",
-            gs_out_path,
-            out_path
+
+        # Run convert to get a decent thumbnail
+        try:
+            run_convert(
+                self.CONVERT,
+                "-scale", "500x5000",
+                "-alpha", "remove",
+                "{}[0]".format(self.document_path),
+                out_path
+            )
+        except ParseError:
+            # if convert fails, fall back to extracting
+            # the first PDF page as a PNG using Ghostscript
+            self.log(
+                "warning",
+                "Thumbnail generation with ImageMagick failed, "
+                "falling back to Ghostscript."
+            )
+            gs_out_path = os.path.join(self.tempdir, "gs_out.png")
+            cmd = [self.GHOSTSCRIPT,
+                   "-q",
+                   "-sDEVICE=pngalpha",
+                   "-o", gs_out_path,
+                   self.document_path]
+            if not subprocess.Popen(cmd).wait() == 0:
+                raise ParseError("Thumbnail (gs) failed at {}".format(cmd))
+            # then run convert on the output from gs
+            run_convert(
+                self.CONVERT,
+                "-scale", "500x5000",
+                "-alpha", "remove",
+                gs_out_path,
+                out_path
         )
 
         return out_path