Upgrade to Gremlin-Python 3.5 and Jupyter Notebook 6.x (#209)

* Remove dependency on Tornado and use the Gremlin 3.5.x replacement * HashableDict is part of Gremlin 3.5 no longer need to backport * Merge GREMLIN35-v2 into GREMLIN35 (#208) * Pin ipython and ipykernel dependency versions (#207) * Pin ipython and ipykernel dependency versions * Update Changelog Co-authored-by: Michael Chin <chnmch@amazon.com> * Apply nest_asyncio for Gremlin client requests, Update dependencies and install docs * Switch to python script for starting jupyter * Allow specifying notebooks directory on startup * Pinning nbconvert version to resolve ipython-contrib/jupyter_contrib_nbextensions#1529 * Fix OC query smart indent bug * Fix failure to overwrite notebook.json Co-authored-by: Michael Chin <chnmch@amazon.com> * ChangeLog.md Co-authored-by: Kelvin Lawrence <gfxman@yahoo.com> Co-authored-by: Michael Chin <chnmch@amazon.com>
aws · Oct 21, 2021 · 371efdf · 371efdf
1 parent 0c35457
commit 371efdf
Show file tree

Hide file tree

Showing 6 changed files with 67 additions and 21 deletions.
diff --git a/ChangeLog.md b/ChangeLog.md
@@ -7,7 +7,8 @@ Starting with v1.31.6, this file will contain a record of major features and upd
 - Allow `%%neptune_ml` to accept JSON blob as parameter input for most phases ([Link to PR](https://github.com/aws/graph-notebook/pull/202))
 - Added `--silent` option for suppressing query output ([PR #1](https://github.com/aws/graph-notebook/pull/201)) ([PR #2](https://github.com/aws/graph-notebook/pull/203))
 - Added all `parserConfiguration` options to `%load` ([Link to PR](https://github.com/aws/graph-notebook/pull/205))
-- Pinned `ipython` and `ipykernel` dependency versions ([Link to PR](https://github.com/aws/graph-notebook/pull/207))
+- Upgraded to Gremlin-Python 3.5 and Jupyter Notebook 6.x ([Link to PR](https://github.com/aws/graph-notebook/pull/209))
+- Resolved smart indent bug in openCypher magic cells ([Link to PR](https://github.com/aws/graph-notebook/pull/209))
 
 ## Release 3.0.6 (September 20, 2021)
 - Added a new `%stream_viewer` magic that allows interactive exploration of the Neptune CDC stream (if enabled). ([Link to PR](https://github.com/aws/graph-notebook/pull/191))

diff --git a/README.md b/README.md
@@ -89,9 +89,7 @@ It is recommended to check the [ChangeLog.md](ChangeLog.md) file periodically to
 
 You will need:
 
-* [Python](https://www.python.org/downloads/) 3.6.1-3.6.12
-* [Jupyter Notebook](https://jupyter.org/install) 5.7.13
-* [Tornado](https://pypi.org/project/tornado/) 4.5.3
+* [Python](https://www.python.org/downloads/) 3.6.13-3.9.7
 * [RDFLib](https://pypi.org/project/rdflib/) 5.0.0
 * A graph database that provides one or more of:
   *  A SPARQL 1.1 endpoint 
@@ -102,10 +100,6 @@ You will need:
 
 ```
 # pin specific versions of required dependencies
-pip install tornado==4.5.3
-pip install ipykernel==5.3.4
-pip install "ipython>=7.16.1,<=7.19.0"
-pip install notebook==5.7.13
 pip install rdflib==5.0.0
 
 # install the package
@@ -123,7 +117,7 @@ python -m graph_notebook.nbextensions.install
 python -m graph_notebook.notebooks.install --destination ~/notebook/destination/dir  
 
 # start jupyter
-jupyter notebook ~/notebook/destination/dir
+python -m graph_notebook.start_notebook --notebooks-dir ~/notebook/destination/dir
 ```
 
 ## Connecting to a graph database

diff --git a/requirements.txt b/requirements.txt
@@ -1,20 +1,20 @@
-tornado==4.5.3
 SPARQLWrapper==1.8.4
 networkx==2.4
 Jinja2==2.11.3
 jupyter
-notebook==5.7.13
+notebook>=6.1.5
 ipywidgets==7.5.1
 jupyter-contrib-nbextensions
 widgetsnbextension
-gremlinpython<=3.4.*
+gremlinpython>=3.5.1
 requests==2.24.0
-ipython>=7.16.1,<=7.19.0
+ipython>=7.16.1
 ipykernel==5.3.4
 neo4j==4.2.1
 rdflib~=5.0.0
 traitlets~=4.3.3
 setuptools~=40.6.2
+nbconvert<6
 
 # requirements for testing
 botocore~=1.18.18

diff --git a/setup.py b/setup.py
@@ -66,14 +66,13 @@ def get_version():
     package_dir={'': 'src'},
     include_package_data=True,
     install_requires=[
-        'gremlinpython<3.5.*',
+        'gremlinpython>=3.5.1',
         'SPARQLWrapper==1.8.4',
-        'tornado==4.5.3',
         'requests',
         'ipywidgets',
         'networkx==2.4',
         'Jinja2==2.11.3',
-        'notebook',
+        'notebook>=6.1.5',
         'jupyter-contrib-nbextensions',
         'widgetsnbextension',
         'jupyter>=1.0.0',
@@ -83,7 +82,8 @@ def get_version():
         'ipython>=7.16.1,<=7.19.0',
         'neo4j==4.3.2',
         'rdflib==5.0.0',
-        'ipykernel==5.3.4'
+        'ipykernel==5.3.4',
+        'nbconvert==5.6.1'
     ],
     package_data={
         'graph_notebook': ['graph_notebook/widgets/nbextensions/static/*.js',

diff --git a/src/graph_notebook/neptune/client.py b/src/graph_notebook/neptune/client.py
@@ -14,9 +14,11 @@
 from botocore.awsrequest import AWSRequest
 from gremlin_python.driver import client
 from neo4j import GraphDatabase
-from tornado import httpclient
+import nest_asyncio
 
-import graph_notebook.neptune.gremlin.graphsonV3d0_MapType_objectify_patch  # noqa F401
+# This patch is no longer needed when graph_notebook is using the a Gremlin Python
+# client >= 3.5.0 as the HashableDict is now part of that client driver.
+# import graph_notebook.neptune.gremlin.graphsonV3d0_MapType_objectify_patch  # noqa F401
 
 DEFAULT_SPARQL_CONTENT_TYPE = 'application/x-www-form-urlencoded'
 DEFAULT_PORT = 8182
@@ -158,12 +160,13 @@ def sparql_cancel(self, query_id: str, silent: bool = False):
         return self._query_status('sparql', query_id=query_id, silent=silent, cancelQuery=True)
 
     def get_gremlin_connection(self) -> client.Client:
+        nest_asyncio.apply()
+
         uri = f'{self._http_protocol}://{self.host}:{self.port}/gremlin'
         request = self._prepare_request('GET', uri)
 
         ws_url = f'{self._ws_protocol}://{self.host}:{self.port}/gremlin'
-        ws_request = httpclient.HTTPRequest(ws_url, headers=dict(request.headers))
-        return client.Client(ws_request, 'g')
+        return client.Client(ws_url, 'g', headers=dict(request.headers))
 
     def gremlin_query(self, query, bindings=None):
         c = self.get_gremlin_connection()

diff --git a/src/graph_notebook/start_notebook.py b/src/graph_notebook/start_notebook.py
@@ -0,0 +1,48 @@
+"""
+Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+SPDX-License-Identifier: Apache-2.0
+"""
+
+import os
+import argparse
+import json
+
+HOME_PATH = os.path.expanduser("~")
+NOTEBOOK_CFG_PATH = HOME_PATH + '/.jupyter/nbconfig/notebook.json'
+
+
+def patch_cm_cypher_config():
+    cypher_cfg = {
+        "cm_config": {
+            "smartIndent": False,
+            "mode": "cypher"
+        }
+    }
+
+    try:
+        with open(NOTEBOOK_CFG_PATH, 'r') as file:
+            notebook_cfg = json.load(file)
+    except (json.decoder.JSONDecodeError, FileNotFoundError) as e:
+        notebook_cfg = {}
+
+    notebook_cfg["CodeCell"] = cypher_cfg
+
+    with open(NOTEBOOK_CFG_PATH, 'w') as file:
+        json.dump(notebook_cfg, file)
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--notebooks-dir', default='', type=str, help='The directory to start Jupyter from.')
+
+    args = parser.parse_args()
+
+    patch_cm_cypher_config()
+
+    kernel_manager_option = "--NotebookApp.kernel_manager_class=notebook.services.kernels.kernelmanager.AsyncMappingKernelManager"
+    notebooks_dir = '~/notebook/destination/dir' if args.notebooks_dir == '' else args.notebooks_dir
+    os.system(f'''jupyter notebook {kernel_manager_option} {notebooks_dir}''')
+
+
+if __name__ == '__main__':
+    main()