Resolve "Create result class"

Closes #6 (closed) Adds a result class to handle intermediate output instead of managing everything in a giant dictionary structure. Instances of result class will be added to the new Run class to manage data better.

chipathlon/db.py

@@ -14,6 +14,9 @@ import hashlib
 class MongoDB(object):
 
     def __init__(self, host, username, password):
+        self.host = host
+        self.username = username
+        self.password = password
         self.client = MongoClient(host)
         self.db = self.client.chipseq
         try:
@@ -39,33 +42,44 @@ class MongoDB(object):
             print "result_id %s doesn't exist." % (result_id,)
         return
 
-    def check_result(self, file_name, control_sample_ids, experiment_sample_ids, ref_genome, result_type, params, debug=False):
+    def _get_result_query(self, result, genome):
+        query = {
+            "result_type": result.file_type,
+            "assembly": genome.assembly,
+            "timestamp": {"$exists": True},
+            "file_name": result.full_name
+        }
+        # In the case that there are 0 samples we just want to check for existence.
+        control_sample_accessions = result.get_accessions("control")
+        signal_sample_accessions = result.get_accessions("signal")
+        query["control_sample_accessions"] = {"$all": control_sample_accessions} if (len(control_sample_accessions) > 0) else {"$exists": True}
+        query["signal_sample_accessions"] = {"$all": signal_sample_accessions} if (len(signal_sample_accessions) > 0) else {"$exists": True}
+        for job in result.all_jobs:
+            job_args = job.get_db_arguments()
+            arg_keys = job_args.keys()
+            if len(arg_keys) == 0:
+                query[job.job_name] = {"$exists": True}
+            else:
+                for arg_name in arg_keys:
+                    query[job.job_name + "." + arg_name] = job_args[arg_name]
+        return query
+
+    def result_exists(self, result, genome):
         try:
-            query = {
-                "result_type": result_type,
-                "ref_genome": ref_genome,
-                "timestamp": {"$exists": True},
-                "file_name": file_name
-            }
-            # In the case that there are 0 samples we just want to check for existence.
-            query["control_sample_ids"] = {"$all": control_sample_ids} if (len(control_sample_ids) > 0) else {"$exists": True}
-            query["experiment_sample_ids"] = {"$all": experiment_sample_ids} if (len(experiment_sample_ids) > 0) else {"$exists": True}
-            for job_name in params:
-                param_keys = params[job_name].keys()
-                if len(param_keys) == 0:
-                    query[job_name] = {"$exists": True}
-                else:
-                    for param_name in param_keys:
-                        query[job_name + "." + param_name] = params[job_name][param_name]
-            cursor = self.db.results.find(query)
-            if debug:
-                print "[DB.check_result] Query = %s" % (query,)
-                print "[DB.check_result] count = %s" % (cursor.count(),)
+            cursor = self.db.results.find(self._get_result_query(result, genome))
+            return cursor.count() > 0
+        except pymongo.errors.OperationFailure as e:
+            print "Error checking result [%s]: %s" % (file_name, e)
+        return False
+
+
+    def get_result(self, result, genome):
+        try:
+            cursor = self.db.results.find(self._get_result_query(result, genome))
             if cursor.count() > 0:
-                # Use the latest if available
                 return cursor.sort("timestamp", pymongo.DESCENDING).next()
         except pymongo.errors.OperationFailure as e:
-            print "Error with experiment_id %s: %s" % (experiment_id, e)
+            print "Error checking result [%s]: %s" % (file_name, e)
         return None
 
     def create_result(self, output_file, control_sample_ids, experiment_sample_ids, result_type, additional_data = {}, gfs_attributes = {}):