Merge pull request #1 from dropbox/master

Merge upstream

Author: ptallada

README.rst

@@ -1,31 +1,31 @@
-================================
-Project is currently unsupported
-================================
+========================================================
+PyHive project has been donated to  Apache Kyuubi
+========================================================
 
+You can follow it's development and report any issues you are experiencing here: https://github.com/apache/kyuubi/tree/master/python/pyhive 
 
 
 
-.. image:: https://travis-ci.org/dropbox/PyHive.svg?branch=master
-    :target: https://travis-ci.org/dropbox/PyHive
-.. image:: https://img.shields.io/codecov/c/github/dropbox/PyHive.svg
+Legacy notes / instructions
+===========================
 
-======
 PyHive
-======
+**********
+
 
 PyHive is a collection of Python `DB-API <http://www.python.org/dev/peps/pep-0249/>`_ and
-`SQLAlchemy <http://www.sqlalchemy.org/>`_ interfaces for `Presto <http://prestodb.io/>`_ and
-`Hive <http://hive.apache.org/>`_.
+`SQLAlchemy <http://www.sqlalchemy.org/>`_ interfaces for `Presto <http://prestodb.io/>`_ ,
+`Hive <http://hive.apache.org/>`_ and `Trino <https://trino.io/>`_.
 
 Usage
-=====
+**********
 
 DB-API
 ------
 .. code-block:: python
 
     from pyhive import presto  # or import hive or import trino
-    cursor = presto.connect('localhost').cursor()
+    cursor = presto.connect('localhost').cursor()  # or use hive.connect or use trino.connect
     cursor.execute('SELECT * FROM my_awesome_data LIMIT 10')
     print cursor.fetchone()
     print cursor.fetchall()
@@ -61,7 +61,7 @@ In Python 3.7 `async` became a keyword; you can use `async_` instead:
 
 SQLAlchemy
 ----------
-First install this package to register it with SQLAlchemy (see ``setup.py``).
+First install this package to register it with SQLAlchemy, see ``entry_points`` in ``setup.py``.
 
 .. code-block:: python
 
@@ -71,9 +71,11 @@ First install this package to register it with SQLAlchemy (see ``setup.py``).
     # Presto
     engine = create_engine('presto://localhost:8080/hive/default')
     # Trino
-    engine = create_engine('trino://localhost:8080/hive/default')
+    engine = create_engine('trino+pyhive://localhost:8080/hive/default')
     # Hive
     engine = create_engine('hive://localhost:10000/default')
+
+    # SQLAlchemy < 2.0
     logs = Table('my_awesome_data', MetaData(bind=engine), autoload=True)
     print select([func.count('*')], from_obj=logs).scalar()
 
@@ -82,6 +84,20 @@ First install this package to register it with SQLAlchemy (see ``setup.py``).
     logs = Table('my_awesome_data', MetaData(bind=engine), autoload=True)
     print select([func.count('*')], from_obj=logs).scalar()
 
+    # SQLAlchemy >= 2.0
+    metadata_obj = MetaData()
+    books = Table("books", metadata_obj, Column("id", Integer), Column("title", String), Column("primary_author", String))
+    metadata_obj.create_all(engine)
+    inspector = inspect(engine)
+    inspector.get_columns('books')
+
+    with engine.connect() as con:
+        data = [{ "id": 1, "title": "The Hobbit", "primary_author": "Tolkien" }, 
+                { "id": 2, "title": "The Silmarillion", "primary_author": "Tolkien" }]
+        con.execute(books.insert(), data[0])
+        result = con.execute(text("select * from books"))
+        print(result.fetchall())
+
 Note: query generation functionality is not exhaustive or fully tested, but there should be no
 problem with raw SQL.
 
@@ -101,7 +117,7 @@ Passing session configuration
                       'session_props': {'query_max_run_time': '1234m'}}
     )
     create_engine(
-        'trino://user@host:443/hive',
+        'trino+pyhive://user@host:443/hive',
         connect_args={'protocol': 'https',
                       'session_props': {'query_max_run_time': '1234m'}}
     )
@@ -116,27 +132,30 @@ Passing session configuration
     )
 
 Requirements
-============
+************
 
 Install using
 
-- ``pip install 'pyhive[hive]'`` for the Hive interface and
-- ``pip install 'pyhive[presto]'`` for the Presto interface.
+- ``pip install 'pyhive[hive]'`` or ``pip install 'pyhive[hive_pure_sasl]'`` for the Hive interface
+- ``pip install 'pyhive[presto]'`` for the Presto interface
 - ``pip install 'pyhive[trino]'`` for the Trino interface
 
+Note: ``'pyhive[hive]'`` extras uses `sasl <https://pypi.org/project/sasl/>`_ that doesn't support Python 3.11, See `github issue <https://github.com/cloudera/python-sasl/issues/30>`_.
+Hence PyHive also supports `pure-sasl <https://pypi.org/project/pure-sasl/>`_ via additional extras ``'pyhive[hive_pure_sasl]'`` which support Python 3.11.
+
 PyHive works with
 
 - Python 2.7 / Python 3
-- For Presto: Presto install
-- For Trino: Trino install
+- For Presto: `Presto installation <https://prestodb.io/docs/current/installation.html>`_
+- For Trino: `Trino installation <https://trino.io/docs/current/installation.html>`_
 - For Hive: `HiveServer2 <https://cwiki.apache.org/confluence/display/Hive/Setting+up+HiveServer2>`_ daemon
 
 Changelog
-=========
+*********
 See https://github.com/dropbox/PyHive/releases.
 
 Contributing
-============
+************
 - Please fill out the Dropbox Contributor License Agreement at https://opensource.dropbox.com/cla/ and note this in your pull request.
 - Changes must come with tests, with the exception of trivial things like fixing comments. See .travis.yml for the test environment setup.
 - Notes on project scope:
@@ -146,8 +165,28 @@ Contributing
   - We prefer having a small number of generic features over a large number of specialized, inflexible features.
     For example, the Presto code takes an arbitrary ``requests_session`` argument for customizing HTTP calls, as opposed to having a separate parameter/branch for each ``requests`` option.
 
+Tips for test environment setup
+****************************************
+You can setup test environment by following ``.travis.yaml`` in this repository. It uses `Cloudera's CDH 5 <https://docs.cloudera.com/documentation/enterprise/release-notes/topics/cdh_vd_cdh_download_510.html>`_ which requires username and password for download.
+It may not be feasible for everyone to get those credentials. Hence below are alternative instructions to setup test environment.
+
+You can clone `this repository <https://github.com/big-data-europe/docker-hive/blob/master/docker-compose.yml>`_ which has Docker Compose setup for Presto and Hive.
+You can add below lines to its docker-compose.yaml to start Trino in same environment::
+ 
+    trino:
+        image: trinodb/trino:351    
+        ports:     
+            - "18080:18080"    
+        volumes:    
+            - ./trino:/etc/trino
+
+Note: ``./trino`` for docker volume defined above is `trino config from PyHive repository <https://github.com/dropbox/PyHive/tree/master/scripts/travis-conf/trino>`_
+
+Then run::
+    docker-compose up -d
+
 Testing
-=======
+*******
 .. image:: https://travis-ci.org/dropbox/PyHive.svg
     :target: https://travis-ci.org/dropbox/PyHive
 .. image:: http://codecov.io/github/dropbox/PyHive/coverage.svg?branch=master
@@ -166,7 +205,7 @@ WARNING: This drops/creates tables named ``one_row``, ``one_row_complex``, and `
 database called ``pyhive_test_database``.
 
 Updating TCLIService
-====================
+********************
 
 The TCLIService module is autogenerated using a ``TCLIService.thrift`` file. To update it, the
 ``generate.py`` file can be used: ``python generate.py <TCLIServiceURL>``. When left blank, the

dev_requirements.txt

@@ -12,6 +12,8 @@ pytest-timeout==1.2.0
 requests>=1.0.0
 requests_kerberos>=0.12.0
 sasl>=0.2.1
+pure-sasl>=0.6.2
+kerberos>=1.3.0
 thrift>=0.10.0
 #thrift_sasl>=0.1.0
 git+https://github.com/cloudera/thrift_sasl  # Using master branch in order to get Python 3 SASL patches

pyhive/__init__.py

@@ -1,3 +1,3 @@
 from __future__ import absolute_import
 from __future__ import unicode_literals
-__version__ = '0.6.3'
+__version__ = '0.7.0'

pyhive/common.py

@@ -18,6 +18,11 @@
 from future.utils import with_metaclass
 from itertools import islice
 
+try:
+    from collections.abc import Iterable
+except ImportError:
+    from collections import Iterable
+
 
 class DBAPICursor(with_metaclass(abc.ABCMeta, object)):
     """Base class for some common DB-API logic"""
@@ -245,7 +250,7 @@ def escape_item(self, item):
             return self.escape_number(item)
         elif isinstance(item, basestring):
             return self.escape_string(item)
-        elif isinstance(item, collections.Iterable):
+        elif isinstance(item, Iterable):
             return self.escape_sequence(item)
         elif isinstance(item, datetime.datetime):
             return self.escape_datetime(item, self._DATETIME_FORMAT)

pyhive/hive.py

@@ -54,6 +54,45 @@
 }
 
 
+def get_sasl_client(host, sasl_auth, service=None, username=None, password=None):
+    import sasl
+    sasl_client = sasl.Client()
+    sasl_client.setAttr('host', host)
+
+    if sasl_auth == 'GSSAPI':
+        sasl_client.setAttr('service', service)
+    elif sasl_auth == 'PLAIN':
+        sasl_client.setAttr('username', username)
+        sasl_client.setAttr('password', password)
+    else:
+        raise ValueError("sasl_auth only supports GSSAPI and PLAIN")
+
+    sasl_client.init()
+    return sasl_client
+
+
+def get_pure_sasl_client(host, sasl_auth, service=None, username=None, password=None):
+    from pyhive.sasl_compat import PureSASLClient
+
+    if sasl_auth == 'GSSAPI':
+        sasl_kwargs = {'service': service}
+    elif sasl_auth == 'PLAIN':
+        sasl_kwargs = {'username': username, 'password': password}
+    else:
+        raise ValueError("sasl_auth only supports GSSAPI and PLAIN")
+
+    return PureSASLClient(host=host, **sasl_kwargs)
+
+
+def get_installed_sasl(host, sasl_auth, service=None, username=None, password=None):
+    try:
+        return get_sasl_client(host=host, sasl_auth=sasl_auth, service=service, username=username, password=password)
+        # The sasl library is available
+    except ImportError:
+        # Fallback to pure-sasl library
+        return get_pure_sasl_client(host=host, sasl_auth=sasl_auth, service=service, username=username, password=password)
+    
+
 def _parse_timestamp(value):
     if value:
         match = _TIMESTAMP_PATTERN.match(value)
@@ -232,7 +271,6 @@ def __init__(
                 self._transport = thrift.transport.TTransport.TBufferedTransport(socket)
             elif auth in ('LDAP', 'KERBEROS', 'NONE', 'CUSTOM'):
                 # Defer import so package dependency is optional
-                import sasl
                 import thrift_sasl
 
                 if auth == 'KERBEROS':
@@ -243,20 +281,8 @@ def __init__(
                     if password is None:
                         # Password doesn't matter in NONE mode, just needs to be nonempty.
                         password = 'x'
-
-                def sasl_factory():
-                    sasl_client = sasl.Client()
-                    sasl_client.setAttr('host', host)
-                    if sasl_auth == 'GSSAPI':
-                        sasl_client.setAttr('service', kerberos_service_name)
-                    elif sasl_auth == 'PLAIN':
-                        sasl_client.setAttr('username', username)
-                        sasl_client.setAttr('password', password)
-                    else:
-                        raise AssertionError
-                    sasl_client.init()
-                    return sasl_client
-                self._transport = thrift_sasl.TSaslClientTransport(sasl_factory, sasl_auth, socket)
+                
+                self._transport = thrift_sasl.TSaslClientTransport(lambda: get_installed_sasl(host=host, sasl_auth=sasl_auth, service=kerberos_service_name, username=username, password=password), sasl_auth, socket)
             else:
                 # All HS2 config options:
                 # https://cwiki.apache.org/confluence/display/Hive/Setting+Up+HiveServer2#SettingUpHiveServer2-Configuration

pyhive/presto.py

@@ -9,6 +9,8 @@
 from __future__ import unicode_literals
 
 from builtins import object
+from decimal import Decimal
+
 from pyhive import common
 from pyhive.common import DBAPITypeObject
 # Make all exceptions visible in this module per DB-API
@@ -34,6 +36,11 @@
 
 _logger = logging.getLogger(__name__)
 
+TYPES_CONVERTER = {
+    "decimal": Decimal,
+    # As of Presto 0.69, binary data is returned as the varbinary type in base64 format
+    "varbinary": base64.b64decode
+}
 
 class PrestoParamEscaper(common.ParamEscaper):
     def escape_datetime(self, item, format):
@@ -307,14 +314,13 @@ def _fetch_more(self):
         """Fetch the next URI and update state"""
         self._process_response(self._requests_session.get(self._nextUri, **self._requests_kwargs))
 
-    def _decode_binary(self, rows):
-        # As of Presto 0.69, binary data is returned as the varbinary type in base64 format
-        # This function decodes base64 data in place
+    def _process_data(self, rows):
         for i, col in enumerate(self.description):
-            if col[1] == 'varbinary':
+            col_type = col[1].split("(")[0].lower()
+            if col_type in TYPES_CONVERTER:
                 for row in rows:
                     if row[i] is not None:
-                        row[i] = base64.b64decode(row[i])
+                        row[i] = TYPES_CONVERTER[col_type](row[i])
 
     def _process_response(self, response):
         """Given the JSON response from Presto's REST API, update the internal state with the next
@@ -341,7 +347,7 @@ def _process_response(self, response):
         if 'data' in response_json:
             assert self._columns
             new_data = response_json['data']
-            self._decode_binary(new_data)
+            self._process_data(new_data)
             self._data += map(tuple, new_data)
         if 'nextUri' not in response_json:
             self._state = self._STATE_FINISHED

pyhive/sasl_compat.py

@@ -0,0 +1,56 @@
+# Original source of this file is https://github.com/cloudera/impyla/blob/master/impala/sasl_compat.py 
+# which uses Apache-2.0 license as of 21 May 2023.
+# This code was added to Impyla in 2016 as a compatibility layer to allow use of either python-sasl or pure-sasl 
+# via PR https://github.com/cloudera/impyla/pull/179
+# Even though thrift_sasl lists pure-sasl as dependency here https://github.com/cloudera/thrift_sasl/blob/master/setup.py#L34 
+# but it still calls functions native to python-sasl in this file https://github.com/cloudera/thrift_sasl/blob/master/thrift_sasl/__init__.py#L82
+# Hence this code is required for the fallback to work.
+ 
+
+from puresasl.client import SASLClient, SASLError
+from contextlib import contextmanager
+
+@contextmanager
+def error_catcher(self, Exc = Exception):
+    try:
+        self.error = None
+        yield
+    except Exc as e:
+        self.error = str(e)
+
+
+class PureSASLClient(SASLClient):
+    def __init__(self, *args, **kwargs):
+        self.error = None
+        super(PureSASLClient, self).__init__(*args, **kwargs)
+
+    def start(self, mechanism):
+        with error_catcher(self, SASLError):
+            if isinstance(mechanism, list):
+                self.choose_mechanism(mechanism)
+            else:
+                self.choose_mechanism([mechanism])
+            return True, self.mechanism, self.process()
+        # else
+        return False, mechanism, None
+
+    def encode(self, incoming):
+        with error_catcher(self):
+            return True, self.unwrap(incoming)
+        # else
+        return False, None
+
+    def decode(self, outgoing):
+        with error_catcher(self):
+            return True, self.wrap(outgoing)
+        # else
+        return False, None
+
+    def step(self, challenge=None):
+        with error_catcher(self):
+            return True, self.process(challenge)
+        # else
+        return False, None
+
+    def getError(self):
+        return self.error