Introducing optional env vars for ImageMagick

paperless.conf.example

@@ -32,7 +32,32 @@ PAPERLESS_PASSPHRASE="secret"
 # have a shared secret here.
 PAPERLESS_SHARED_SECRET=""
 
+#
+# The following values use sensible defaults for modern systems, but if you're
+# running Paperless on a low-resource machine (like a Rasberry Pi), modifying
+# some of these values may be necessary.
+#
+
 # By default, Paperless will attempt to use all available CPU cores to process
 # a document, but if you would like to limit that, you can set this value to
 # an integer:
 #PAPERLESS_OCR_THREADS=1
+
+# On smaller systems, or even in the case of Very Large Documents, the consumer
+# may explode, complaining about how it's "unable to extent pixel cache".  In
+# such cases, try setting this to a reasonably low value, like 32000000.  The
+# default is to use whatever is necessary to do everything without writing to
+# disk, and units are in megabytes.
+#
+# For more information on how to use this value, you should probably search
+# the web for "MAGICK_MEMORY_LIMIT".
+#PAPERLESS_CONVERT_MEMORY_LIMIT=0
+
+# Similar to the memory limit, if you've got a small system and your OS mounts
+# /tmp as tmpfs, you should set this to a path that's on a physical disk, like
+# /home/your_user/tmp or something.  ImageMagick will use this as scratch space
+# when crunching through very large documents.
+#
+# For more information on how to use this value, you should probably search
+# the web for "MAGICK_TMPDIR".
+#PAPERLESS_CONVERT_TMPDIR=/var/tmp/paperless

src/documents/consumer.py

@@ -129,10 +129,13 @@ class Consumer(object):
 
         # Convert PDF to multiple PNMs
         pnm = os.path.join(tempdir, "convert-%04d.pnm")
-        subprocess.Popen((
-            self.CONVERT, "-density", "300", "-depth", "8",
-            "-type", "grayscale", doc, pnm
-        )).wait()
+        run_convert(
+            self.CONVERT,
+            "-density", "300",
+            "-depth", "8",
+            "-type", "grayscale",
+            doc, pnm,
+        )
 
         # Get a list of converted images
         pnms = []
@@ -159,13 +162,14 @@ class Consumer(object):
 
         self.log("info", "Generating the thumbnail")
 
-        subprocess.Popen((
+        run_convert(
             self.CONVERT,
             "-scale", "500x5000",
             "-alpha", "remove",
+            "-limit", "memory", "20MiB",
             doc,
             os.path.join(tempdir, "convert-%04d.png")
-        )).wait()
+        )
 
         return os.path.join(tempdir, "convert-0000.png")
 
@@ -334,6 +338,16 @@ def image_to_string(args):
 
 def run_unpaper(args):
     unpaper, pnm = args
-    subprocess.Popen((
-        unpaper, pnm, pnm.replace(".pnm", ".unpaper.pnm")
-    )).wait()
+    subprocess.Popen(
+        (unpaper, pnm, pnm.replace(".pnm", ".unpaper.pnm"))).wait()
+
+
+def run_convert(*args):
+
+    environment = {}
+    if settings.CONVERT_MEMORY_LIMIT:
+        environment["MAGICK_MEMORY_LIMIT"] = settings.CONVERT_MEMORY_LIMIT
+    if settings.CONVERT_TMPDIR:
+        environment["MAGICK_TMPDIR"] = settings.CONVERT_TMPDIR
+
+    subprocess.Popen(args, env=environment).wait()

src/paperless/settings.py

@@ -189,6 +189,8 @@ GNUPG_HOME = os.getenv("HOME", "/tmp")
 
 # Convert is part of the ImageMagick package
 CONVERT_BINARY = os.getenv("PAPERLESS_CONVERT_BINARY")
+CONVERT_TMPDIR = os.getenv("PAPERLESS_CONVERT_TMPDIR")
+CONVERT_MEMORY_LIMIT = os.getenv("PAPERLESS_CONVERT_MEMORY_LIMIT")
 
 # Unpaper
 UNPAPER_BINARY = os.getenv("PAPERLESS_UNPAPER_BINARY", "unpaper")
@@ -226,7 +228,7 @@ PASSPHRASE = os.getenv("PAPERLESS_PASSPHRASE")
 SHARED_SECRET = os.getenv("PAPERLESS_SHARED_SECRET", "")
 
 #
-# TODO: Remove after 1.2
+# TODO: Remove after 0.2
 #
 # This logic is here to address issue #44, wherein we were using inconsistent
 # constant names vs. environment variables.  If you're using Paperless for the