read-protobuf
Small library to read serialized protobuf(s) directly into Pandas Dataframe.
This is meant to be a simple shortcut to getting from serialized protobuf bytes / files directly to a dataframe.
Runs in Python 2.7+ and Python 3
Note: This currently only supports basic proto3. I have not yet tested it with proto2, though I believe that should work. I plan to expand the utility of this library with time and need.
Install
Available via pip:
$ pip install read-protobuf
Usage
Run the demo-notebook for an interactive demo.
import demo_pb2 # compiled protobuf message module
from read_protobuf import read_protobuf
MessageType = demo_pb2.MessageType() # instantiate a new message type
df = read_protobuf(b'\x00\x00', MessageType) # create a dataframe from serialized protobuf bytes
df = read_protobuf([b'\x00\x00', b'x00\x00'] MessageType) # read multiple protobuf bytes
df = read_protobuf('demo.pb', MessageType) # use file instead of bytes
df = read_protobuf(['demo.pb', 'demo2.pb'], MessageType) # read multiple files
# options
df = read_protobuf('demo.pb', MessageType, flatten=False) # don't flatten pb messages
df = read_protobuf('demo.pb', MessageType, prefix_nested=True) # prefix nested messages with parent keys (like pandas.io.json.json_normalize)
To compile a protobuf Message class from python, use:
$ protoc --python_out="." demo.proto
Alternatives
protobuf-to-dict
https://github.com/benhodgson/protobuf-to-dict
This library was developed earlier to convert protobufs to JSON via a dict.
JSON
The google protobuf library comes with a utility to convert messages to JSON. Then the JSON objects could be loaded into pands via pd.read_json()
.
from google.protobuf.json_format import MessageToJson
In my brief tests, the read_protobuf
package is about twice as fast as converting a protobuf to a dataframe using MessageToJson
.
Develop
Currently developed for Python 3 using the anaconda python distribution. To install a development version of the package, run from the root directory:
$ pip install -e .
- To install development dependencies, use pip on the
dev-requirements.txt
file:
$ pip install -r dev-requirements.txt
Lint
Uses pylint
to lint application.
$ pylint read_protobuf
Configuration options are specified in .pylintrc
Test
Uses pytest
to run unit tests. From the root of the repository, run:
$ pytest
$ pytest -k "TestRead::test_read_bytes" # specify test
Configuration options are specified in setup.cfg
Code Coverage
We use coverage
to monitor code coverage during tests. To record coverage while running tests, run:
$ coverage run -m pytest # watch files while testing
$ coverage report # will display coverage report
UnLicense
This is free and unencumbered software released into the public domain.
Anyone is free to copy, modify, publish, use, compile, sell, or distribute this software, either in source code form or as a compiled binary, for any purpose, commercial or non-commercial, and by any means.
In jurisdictions that recognize copyright laws, the author or authors of this software dedicate any and all copyright interest in the software to the public domain. We make this dedication for the benefit of the public at large and to the detriment of our heirs and successors. We intend this dedication to be an overt act of relinquishment in perpetuity of all present and future rights to this software under copyright law.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
For more information, please refer to http://unlicense.org/