Get pluggy to work with forking workers

src/ocrmypdf/_jobcontext.py

@@ -19,6 +19,9 @@ import logging
 import os
 import shutil
 import sys
+from functools import partial
+
+from ocrmypdf._plugin_manager import get_plugin_manager
 
 
 class PDFContext:
@@ -59,10 +62,22 @@ class PageContext:
         self.name = pdf_context.name
         self.pageno = pageno
         self.pageinfo = pdf_context.pdfinfo[pageno]
+        self.plugin_manager = pdf_context.plugin_manager
 
     def get_path(self, name):
         return os.path.join(self.work_folder, "%06d_%s" % (self.pageno + 1, name))
 
+    def __getstate__(self):
+        state = self.__dict__.copy()
+        del state['plugin_manager']
+        state['construct_plugin_manager'] = partial(get_plugin_manager, self.options)
+        return state
+
+    def __setstate__(self, state):
+        self.__dict__.update(state)
+        self.plugin_manager = self.__dict__['construct_plugin_manager']()
+        del self.__dict__['construct_plugin_manager']
+
 
 def cleanup_working_files(work_folder, options):
     if options.keep_temporary_files:

src/ocrmypdf/_pipeline.py

@@ -521,6 +521,9 @@ def create_ocr_image(image, page_context):
             im = pix.topil()
 
         del draw
+
+        im = page_context.plugin_manager.hook.filter_ocr_image(image=im)
+
         # Pillow requires integer DPI
         dpi = tuple(round(coord) for coord in im.info['dpi'])
         im.save(output_file, dpi=dpi)

src/ocrmypdf/_plugin_manager.py

@@ -0,0 +1,15 @@
+import importlib
+
+import pluggy
+
+from ocrmypdf import pluginspec
+
+
+def get_plugin_manager(options):
+    pm = pluggy.PluginManager('ocrmypdf')
+    pm.add_hookspecs(pluginspec)
+
+    for name in options.plugins:
+        module = importlib.import_module(name)
+        pm.register(module)
+    return pm

src/ocrmypdf/_sync.py

@@ -29,7 +29,7 @@ from tempfile import mkdtemp
 import PIL
 import pluggy
 
-from ocrmypdf import _pluginspec
+from ocrmypdf import pluginspec
 from ocrmypdf._concurrent import exec_progress_pool
 from ocrmypdf._graft import OcrGrafter
 from ocrmypdf._jobcontext import PDFContext, cleanup_working_files
@@ -59,6 +59,7 @@ from ocrmypdf._pipeline import (
     triage,
     validate_pdfinfo_options,
 )
+from ocrmypdf._plugin_manager import get_plugin_manager
 from ocrmypdf._validation import (
     check_requested_output_file,
     create_input_file,
@@ -292,16 +293,6 @@ def configure_debug_logging(log_filename, prefix=''):
     return log_file_handler
 
 
-def get_plugin_manager(options):
-    pm = pluggy.PluginManager('ocrmypdf')
-    pm.add_hookspecs(_pluginspec)
-
-    for name in options.plugins:
-        module = importlib.import_module(name)
-        pm.register(module)
-    return pm
-
-
 def run_pipeline(options, api=False):
     # Any changes to options will not take effect for options that are already
     # bound to function parameters in the pipeline. (For example

src/ocrmypdf/example.py

@@ -1,6 +1,11 @@
-import ocrmypdf
+from ocrmypdf import hookimpl
 
 
-@ocrmypdf.hookimpl
+@hookimpl
 def prepare(options):
-    raise ValueError('foo')
+    pass
+
+
+@hookimpl
+def filter_ocr_image(image):
+    return image