OCR Handler

CHANGES.rst

@@ -5,5 +5,7 @@ Changelog
 1.0a1 (unreleased)
 ------------------
 
+- Publish Redis messages on file change
+  [kreafox]
 - Initial release.
   [MrTango]

setup.py

@@ -56,6 +56,7 @@
         "z3c.jbot",
         "plone.api>=1.8.4",
         "plone.app.dexterity",
+        "redis",
     ],
     extras_require={
         "test": [

src/collective/ocrmypdf/config.py

@@ -0,0 +1,5 @@
+import os
+
+REDIS_HOST = os.getenv("REDIS_HOST", "redis")
+REDIS_PORT = int(os.getenv("REDIS_PORT", 6379))
+REDIS_CHANNEL = os.getenv("REDIS_CHANNEL", "ocr_queue")

src/collective/ocrmypdf/datamanager.py

@@ -0,0 +1,89 @@
+import logging
+
+import transaction
+
+logger = logging.getLogger("collective.ocrmypdf")
+
+
+class CallbacksDataManager(object):
+    """Transaction aware data manager for calling callbacks at commit time"""
+
+    def __init__(self):
+        self.sp = 0
+        self.callbacks = []
+        self.txn = None
+
+    def tpc_begin(self, txn):
+        self.txn = txn
+
+    def tpc_finish(self, txn):
+        self.callbacks = []
+
+    def tpc_vote(self, txn):
+        pass
+
+    def tpc_abort(self, txn):
+        self._checkTransaction(txn)
+
+        if self.txn is not None:
+            self.txn = None
+
+        self.callbacks = []
+
+    def abort(self, txn):
+        self.callbacks = []
+
+    def commit(self, txn):
+        self._checkTransaction(txn)
+
+        for callback in self.callbacks:
+            try:
+                callback()
+            except Exception:
+                logger.exception("Error executing callback.")
+
+        self.txn = None
+        self.callbacks = []
+
+    def savepoint(self):
+        self.sp += 1
+
+        return Savepoint(self)
+
+    def sortKey(self):
+        return self.__class__.__name__
+
+    def add(self, callback):
+        logger.info("Add callback to queue %s", callback)
+        self.callbacks.append(callback)
+
+    def _checkTransaction(self, txn):
+        if txn is not self.txn and self.txn is not None:
+            raise TypeError("Transaction missmatch", txn, self.txn)
+
+
+class Savepoint(object):
+    """Savepoint implementation to allow rollback of queued callbacks"""
+
+    def __init__(self, dm):
+        self.dm = dm
+        self.sp = dm.sp
+        self.callbacks = dm.callbacks[:]
+        self.txn = dm.txn
+
+    def rollback(self):
+        if self.txn is not self.dm.txn:
+            raise TypeError("Attempt to rollback stale rollback")
+
+        if self.dm.sp < self.sp:
+            raise TypeError(
+                "Attempt to roll back to invalid save point", self.sp, self.dm.sp
+            )
+        self.dm.sp = self.sp
+        self.dm.callbacks = self.callbacks[:]
+
+
+def queue_callback(callback):
+    cdm = CallbacksDataManager()
+    transaction.get().join(cdm)
+    cdm.add(callback)

src/collective/ocrmypdf/subscribers/configure.zcml

@@ -2,12 +2,16 @@
     xmlns="http://namespaces.zope.org/zope"
     i18n_domain="collective.ocrmypdf">
 
- <!-- -*- extra stuff goes here -*- -->
+   <subscriber for="plone.app.contenttypes.interfaces.IFile
+                   zope.lifecycleevent.interfaces.IObjectAddedEvent"
+              handler=".file_modified_ocr.handler"
+   />
+
 
-  <subscriber for="plone.dexterity.interfaces.IDexterityContent
+  <subscriber for="plone.app.contenttypes.interfaces.IFile
                    zope.lifecycleevent.interfaces.IObjectModifiedEvent"
               handler=".file_modified_ocr.handler"
-              />
+  />
 
 
 </configure>

src/collective/ocrmypdf/subscribers/file_modified_ocr.py

@@ -1,6 +1,30 @@
 # -*- coding: utf-8 -*-
+import json
+import redis
+from zope.globalrequest import getRequest
+from collective.ocrmypdf.datamanager import queue_callback
+from collective.ocrmypdf.config import REDIS_HOST, REDIS_PORT, REDIS_CHANNEL
+
+redis_client = redis.StrictRedis(host=REDIS_HOST, port=REDIS_PORT, db=0)
 
 
 def handler(obj, event):
-    """Event handler"""
-    print("{0} on object {1}".format(event.__class__, obj.absolute_url()))
+    """Publish UID of PDF files to the Redis channel."""
+
+    if obj.content_type() == "application/pdf":
+        request = getRequest()
+        # Check the 'ocr' flag to prevent re-triggering the handler
+        # when the processed file is uploaded by the consumer.
+        ocr_flag = request.form.get("ocr")
+
+        if ocr_flag == "1":
+            return
+
+        uid = obj.UID()
+        message = {"uid": uid}
+        data = json.dumps(message)
+
+        def callback():
+            redis_client.publish(REDIS_CHANNEL, data)
+
+        queue_callback(callback)