add pyexasol datasource, ensure that integer dont overflow in javascr…

…ipt (#4378)

redash/query_runner/exasol.py

@@ -0,0 +1,142 @@
+import datetime
+
+from redash.query_runner import *
+from redash.utils import json_dumps
+
+def _exasol_type_mapper(val, data_type):
+    if val is None:
+        return None
+    elif data_type['type'] == 'DECIMAL':
+        if data_type['scale'] == 0 and data_type['precision'] < 16:
+            return int(val)
+        elif data_type['scale'] == 0 and data_type['precision'] >= 16:
+            return val
+        else:
+            return float(val)
+    elif data_type['type'] == 'DATE':
+        return datetime.date(int(val[0:4]), int(val[5:7]), int(val[8:10]))
+    elif data_type['type'] == 'TIMESTAMP':
+        return datetime.datetime(int(val[0:4]), int(val[5:7]), int(val[8:10]),           # year, month, day
+                                 int(val[11:13]), int(val[14:16]), int(val[17:19]),      # hour, minute, second
+                                 int(val[20:26].ljust(6, '0')) if len(val) > 20 else 0)  # microseconds (if available)
+    else:
+        return val
+
+def _type_mapper(data_type):
+    if data_type['type'] == 'DECIMAL':
+        if data_type['scale'] == 0 and data_type['precision'] < 16:
+            return TYPE_INTEGER
+        elif data_type['scale'] == 0 and data_type['precision'] >= 16:
+            return TYPE_STRING
+        else:
+            return TYPE_FLOAT
+    elif data_type['type'] == 'DATE':
+        return TYPE_DATE
+    elif data_type['type'] == 'TIMESTAMP':
+        return TYPE_DATETIME
+    else:
+        return TYPE_STRING
+
+try:
+    import pyexasol
+    enabled = True
+except ImportError:
+    enabled = False
+
+
+class Exasol(BaseQueryRunner):
+    noop_query = 'SELECT 1 FROM DUAL'
+
+    @classmethod
+    def configuration_schema(cls):
+        return {
+            'type': 'object',
+            'properties': {
+                'user': {
+                    'type': 'string'
+                },
+                'password': {
+                    'type': 'string'
+                },
+                'host': {
+                    'type': 'string'
+                },
+                'port': {
+                    'type': 'number',
+                    'default': 8563
+                },
+            },
+            'required': ['host', 'port', 'user', 'password'],
+            'order': ['host', 'port', 'user', 'password'],
+            'secret': ['password']
+        }
+
+    def _get_connection(self):
+        exahost = "%s:%s" % (self.configuration.get('host', None), self.configuration.get('port', 8563))
+        return pyexasol.connect(
+            dsn=exahost,
+            user=self.configuration.get('user', None) ,
+            password=self.configuration.get('password', None),
+            compression=True,
+            json_lib='rapidjson',
+            fetch_mapper=_exasol_type_mapper,
+        )
+
+    def run_query(self, query, user):
+        connection = self._get_connection()
+        statement = None
+        error = None
+        try:
+            statement = connection.execute(query)
+            columns = [{'name': n, 'friendly_name': n,'type': _type_mapper(t)} for (n, t) in statement.columns().items()]
+            cnames = statement.column_names()
+
+            rows = [dict(zip(cnames, row)) for row in statement]
+            data = {'columns': columns, 'rows': rows}
+            json_data = json_dumps(data)
+        finally:
+            if statement is not None:
+                statement.close()
+
+            connection.close()
+
+        return json_data, error
+
+
+
+    def get_schema(self, get_stats=False):
+        query = """
+        SELECT
+            COLUMN_SCHEMA,
+            COLUMN_TABLE,
+            COLUMN_NAME
+        FROM EXA_ALL_COLUMNS
+        """
+
+        connection = self._get_connection()
+        statement = None
+        try:
+            statement = connection.execute(query)
+            result = {}
+
+            for (schema, table_name, column) in statement:
+                table_name_with_schema = '%s.%s' % (schema,table_name)
+
+                if table_name_with_schema not in result:
+                    result[table_name_with_schema] = {'name': table_name_with_schema, 'columns': []}
+
+                result[table_name_with_schema]['columns'].append(column)
+        finally:
+            if statement is not None:
+                statement.close()
+
+            connection.close()
+
+        return result.values()
+
+    @classmethod
+    def enabled(cls):
+        return enabled
+
+
+register(Exasol)

redash/settings/__init__.py

@@ -303,6 +303,7 @@ def email_server_is_configured():
     'redash.query_runner.cass',
     'redash.query_runner.dgraph',
     'redash.query_runner.azure_kusto',
+    'redash.query_runner.exasol',
     'redash.query_runner.cloudwatch',
     'redash.query_runner.cloudwatch_insights',
 ]

requirements_all_ds.txt

@@ -32,3 +32,5 @@ phoenixdb==0.7
 certifi>=2019.9.11
 pydgraph==2.0.2
 azure-kusto-data==0.0.35
+pyexasol==0.9.1
+python-rapidjson==0.8.0