Merge pull request #425 from mboudet/dev

Manage file processing in background

CHANGELOG.md

@@ -14,9 +14,11 @@ This changelog was started for release 4.2.0.
 - "askomics:instancesLabel" predicate can be defined at the entity-level, to set a specific attribute URI as the 'label' (ie, visible by default)
   - Play the same role as 'askomics:instancesHaveNoLabels', except a specific attribute is visible instead of the URI attribute.
 - Added the *TIMEOUT* env variable, which will set the web workers tiemout value. Default 300s
+- Added the 'skip_rdf_preview' config option. This will skip loading RDF files in memory to get the location and remote graph. Warning: This means you must enter the values at the integration step yourself.
 
 ### Changed
 
+- Now pre-process files to store integration parameters (column, entities, etc...) in DB, to avoid re-processing every time.
 - Rewrote the ontology part. Instead of specifying 'children of' and other values, users can tick the 'recursive' button to customize the query. While this is less intuitive, this change is more flexible for the various types of ontological relations
 
 ## [4.5.0] - 2023-10-20

askomics/api/file.py

@@ -130,6 +130,8 @@ def upload_chunk():
         }), 400
 
     data = request.get_json()
+
+    skip_preview = data.get('skip_preview', False)
     if not (data and all([key in data for key in ["first", "last", "size", "name", "type", "size", "chunk"]])):
         return jsonify({
             "path": '',
@@ -146,7 +148,7 @@ def upload_chunk():
 
     try:
         files = FilesHandler(current_app, session)
-        path = files.persist_chunk(data)
+        path = files.persist_chunk(data, skip_preview)
     except Exception as e:
         traceback.print_exc(file=sys.stdout)
         return jsonify({
@@ -243,6 +245,9 @@ def get_preview():
 
         results = []
         for file in files_handler.files:
+            if file.status == "error":
+                continue
+
             file.set_preview()
             res = file.get_preview()
             results.append(res)
@@ -254,11 +259,19 @@ def get_preview():
             'errorMessage': str(e)
         }), 500
 
+    errorMessage = ''
+    error = False
+    errorCode = 200
+    if not results:
+        errorMessage = "None of the selected files are in an integrable state"
+        error = True
+        errorCode = 400
+
     return jsonify({
         'previewFiles': results,
-        'error': False,
-        'errorMessage': ''
-    })
+        'error': error,
+        'errorMessage': errorMessage
+    }), errorCode
 
 
 @file_bp.route('/api/files/delete', methods=['POST'])
@@ -332,6 +345,9 @@ def integrate():
 
         for file in files_handler.files:
 
+            if file.status == "error":
+                continue
+
             data["externalEndpoint"] = data["externalEndpoint"] if (data.get("externalEndpoint") and isinstance(file, RdfFile)) else None
             data["externalGraph"] = data["externalGraph"] if (data.get("externalGraph") and isinstance(file, RdfFile)) else None
             data["customUri"] = data["customUri"] if (data.get("customUri") and not isinstance(file, RdfFile)) else None

askomics/libaskomics/BedFile.py

@@ -41,6 +41,11 @@ def __init__(self, app, session, file_info, host_url=None, external_endpoint=Non
 
     def set_preview(self):
         """Set entity name preview"""
+
+        if self.preview:
+            self.entity_name = self.preview['entity_name']
+            return
+
         try:
             BedTool(self.path).count()
             self.entity_name = self.human_name
@@ -49,6 +54,18 @@ def set_preview(self):
             self.error_message = "Malformated BED ({})".format(str(e))
             traceback.print_exc(file=sys.stdout)
 
+    def save_preview(self):
+        """Save location and endpoint in preview"""
+        data = None
+        error = None
+        self.set_preview()
+
+        if self.error:
+            error = self.error_message
+        else:
+            data = {'entity_name': self.entity_name}
+        self.save_preview_in_db(data, error)
+
     def get_preview(self):
         """Get file preview
 

askomics/libaskomics/CsvFile.py

@@ -59,6 +59,12 @@ def set_preview(self):
         self.set_preview_and_header()
         self.set_columns_type()
 
+    def save_preview(self):
+        """Save location and endpoint in preview"""
+        data = None
+        error = None
+        self.save_preview_in_db(data, error)
+
     def get_preview(self):
         """Get a preview of the file
 

askomics/libaskomics/Database.py

@@ -419,6 +419,28 @@ def create_files_table(self):
         except Exception:
             pass
 
+        query = '''
+        ALTER TABLE files
+        ADD preview text NULL
+        DEFAULT(NULL)
+        '''
+
+        try:
+            self.execute_sql_query(query)
+        except Exception:
+            pass
+
+        query = '''
+        ALTER TABLE files
+        ADD preview_error text NULL
+        DEFAULT(NULL)
+        '''
+
+        try:
+            self.execute_sql_query(query)
+        except Exception:
+            pass
+
     def create_abstraction_table(self):
         """Create abstraction table"""
         query = """

askomics/libaskomics/File.py

@@ -1,5 +1,6 @@
 import datetime
 import os
+import json
 import time
 from dateutil import parser
 from urllib.parse import quote
@@ -92,6 +93,8 @@ def __init__(self, app, session, file_info, host_url=None, external_endpoint=Non
         self.path = file_info['path']
         self.type = file_info['type']
         self.size = file_info['size']
+        self.status = file_info['status']
+        self.preview = json.loads(file_info['preview']) if file_info['preview'] else None
         self.id = file_info['id']
         self.public = False
         self.ntriples = 0
@@ -536,3 +539,20 @@ def convert_type(self, value, try_date=False):
                     return value
 
         return value
+
+    def save_preview_in_db(self, preview, error):
+        database = Database(self.app, self.session)
+
+        status = "available" if not error else "error"
+        data = json.dumps(preview) if preview else None
+
+        query = '''
+        UPDATE files SET
+        preview = ?,
+        preview_error = ?,
+        status = ?
+        WHERE id= ?
+        '''
+
+        variables = [data, error, status, self.id]
+        database.execute_sql_query(query, tuple(variables))

askomics/libaskomics/FilesHandler.py

@@ -90,7 +90,7 @@ def get_files_infos(self, files_id=None, files_path=None, return_path=False):
             subquery_str = '(' + ' OR '.join(['id = ?'] * len(files_id)) + ')'
 
             query = '''
-            SELECT id, name, type, size, path, date, status
+            SELECT id, name, type, size, path, date, status, preview, preview_error
             FROM files
             WHERE user_id = ?
             AND {}
@@ -102,7 +102,7 @@ def get_files_infos(self, files_id=None, files_path=None, return_path=False):
             subquery_str = '(' + ' OR '.join(['path = ?'] * len(files_path)) + ')'
 
             query = '''
-            SELECT id, name, type, size, path, date, status
+            SELECT id, name, type, size, path, date, status, preview, preview_error
             FROM files
             WHERE user_id = ?
             AND {}
@@ -113,7 +113,7 @@ def get_files_infos(self, files_id=None, files_path=None, return_path=False):
         else:
 
             query = '''
-            SELECT id, name, type, size, path, date, status
+            SELECT id, name, type, size, path, date, status, preview, preview_error
             FROM files
             WHERE user_id = ?
             '''
@@ -128,7 +128,9 @@ def get_files_infos(self, files_id=None, files_path=None, return_path=False):
                 'type': row[2],
                 'size': row[3],
                 'date': row[5],
-                'status': row[6]
+                'status': row[6],
+                'preview': row[7],
+                'preview_error': row[8]
             }
             if return_path:
                 file['path'] = row[4]
@@ -206,7 +208,7 @@ def write_data_into_file(self, data, file_name, mode, should_exist=False):
         with open(file_path, mode) as file:
             file.write(data)
 
-    def store_file_info_in_db(self, name, filetype, file_name, size, status="available", task_id=None):
+    def store_file_info_in_db(self, name, filetype, file_name, size, status="available", task_id=None, skip_preview=False):
         """Store the file info in the database
 
         Parameters
@@ -239,10 +241,15 @@ def store_file_info_in_db(self, name, filetype, file_name, size, status="availab
             ?,
             ?,
             ?,
+            ?,
+            ?,
             ?
         )
         '''
 
+        if not skip_preview:
+            status = 'processing'
+
         # Type
         if filetype in ('text/tab-separated-values', 'tabular'):
             filetype = 'csv/tsv'
@@ -259,9 +266,13 @@ def store_file_info_in_db(self, name, filetype, file_name, size, status="availab
 
         self.date = int(time.time())
 
-        return database.execute_sql_query(query, (self.session['user']['id'], name, filetype, file_path, size, self.date, status, task_id), get_id=True)
+        id = database.execute_sql_query(query, (self.session['user']['id'], name, filetype, file_path, size, self.date, status, task_id, None, None), get_id=True)
+
+        if not skip_preview:
+            self.app.celery.send_task('save_preview', ({"user": self.session["user"]}, id))
+        return id
 
-    def update_file_info(self, file_id, size=None, status="", task_id=""):
+    def update_file_info(self, file_id, size=None, status="", task_id="", error=None):
         """Update file size and status
 
         Parameters
@@ -285,33 +296,39 @@ def update_file_info(self, file_id, size=None, status="", task_id=""):
         size_query = ""
         status_query = ""
         task_query = ""
+        error_query = ""
 
         # Should be a cleaner way of doing this...
         if size is not None:
-            size_query = "size=?," if (status or task_id) else "size=?"
+            size_query = "size=?," if (status or task_id or error) else "size=?"
             query_vars.append(size)
 
         if status:
-            status_query = "status=?," if task_id else "status=?"
+            status_query = "status=?," if (task_id or error) else "status=?"
             query_vars.append(status)
 
         if task_id:
-            task_query = "task_id=?"
+            task_query = "task_id=?," if error else "task_id=?"
             query_vars.append(task_id)
 
+        if error:
+            error_query = "preview_error=?"
+            query_vars.append(error)
+
         query_vars.append(file_id)
 
         query = '''
         UPDATE files SET
         {}
         {}
         {}
+        {}
         WHERE id=?
-        '''.format(size_query, status_query, task_query)
+        '''.format(size_query, status_query, task_query, error_query)
 
         database.execute_sql_query(query, tuple(query_vars))
 
-    def persist_chunk(self, chunk_info):
+    def persist_chunk(self, chunk_info, skip_preview=False):
         """Persist a file by chunk. Store info in db if the chunk is the last
 
         Parameters
@@ -324,14 +341,15 @@ def persist_chunk(self, chunk_info):
         str
             local filename
         """
+
         try:
             # 1 chunk file
             if chunk_info["first"] and chunk_info["last"]:
                 # Write data into file
                 file_name = self.get_file_name()
                 self.write_data_into_file(chunk_info["chunk"], file_name, "w")
                 # store file info in db
-                self.store_file_info_in_db(chunk_info["name"], chunk_info["type"], file_name, chunk_info["size"])
+                self.store_file_info_in_db(chunk_info["name"], chunk_info["type"], file_name, chunk_info["size"], skip_preview=skip_preview)
             # first chunk of large file
             elif chunk_info["first"]:
                 file_name = self.get_file_name()
@@ -340,7 +358,7 @@ def persist_chunk(self, chunk_info):
             elif chunk_info["last"]:
                 file_name = chunk_info["path"]
                 self.write_data_into_file(chunk_info["chunk"], file_name, "a")
-                self.store_file_info_in_db(chunk_info["name"], chunk_info["type"], file_name, chunk_info["size"])
+                self.store_file_info_in_db(chunk_info["name"], chunk_info["type"], file_name, chunk_info["size"], skip_preview=skip_preview)
             # chunk of large file
             else:
                 file_name = chunk_info["path"]
@@ -395,8 +413,8 @@ def download_url(self, url, task_id):
             # Update final value
             self.update_file_info(file_id, size=os.path.getsize(path), status="available")
 
-        except Exception:
-            self.update_file_info(file_id, size=os.path.getsize(path), status="error")
+        except Exception as e:
+            self.update_file_info(file_id, size=os.path.getsize(path), status="error", error=str(e))
 
     def get_type(self, file_ext):
         """Get files type, based on extension

askomics/libaskomics/GffFile.py

@@ -54,6 +54,11 @@ def __init__(self, app, session, file_info, host_url=None, external_endpoint=Non
 
     def set_preview(self):
         """Summary"""
+
+        if self.preview:
+            self.entities = self.preview['entities']
+            return
+
         try:
             exam = GFFExaminer()
             handle = open(self.path, encoding="utf-8", errors="ignore")
@@ -67,6 +72,18 @@ def set_preview(self):
             self.error_message = "Malformated GFF ({})".format(str(e))
             traceback.print_exc(file=sys.stdout)
 
+    def save_preview(self):
+        """Save location and endpoint in preview"""
+        data = None
+        error = None
+        self.set_preview()
+
+        if self.error:
+            error = self.error_message
+        else:
+            data = {'entities': self.entities}
+        self.save_preview_in_db(data, error)
+
     def get_preview(self):
         """Get gff file preview (list of entities)