Counting 3,742 Big Data & Machine Learning Frameworks, Toolsets, and Examples...
Suggestion? Feedback? Tweet @stkim1

Last Commit
Apr. 21, 2019
Jan. 2, 2015

smart_open — utils for streaming large files in Python

License Travis


smart_open is a Python 2 & Python 3 library for efficient streaming of very large files from/to S3, HDFS, WebHDFS, HTTP, or local storage. It supports transparent, on-the-fly (de-)compression for a variety of different formats.

smart_open is a drop-in replacement for Python's built-in open(): it can do anything open can (100% compatible, falls back to native open wherever possible), plus lots of nifty extra stuff on top.

smart_open is well-tested, well-documented, and has a simple, Pythonic API:

>>> from smart_open import open
>>> # stream lines from an S3 object
>>> for line in open('s3://commoncrawl/robots.txt'):
...    print(repr(line))
...    break
'User-Agent: *\n'

>>> # stream from/to compressed files, with transparent (de)compression:
>>> for line in open('smart_open/tests/test_data/1984.txt.gz', encoding='utf-8'):
...    print(repr(line))
'It was a bright cold day in April, and the clocks were striking thirteen.\n'
'Winston Smith, his chin nuzzled into his breast in an effort to escape the vile\n'
'wind, slipped quickly through the glass doors of Victory Mansions, though not\n'
'quickly enough to prevent a swirl of gritty dust from entering along with him.\n'

>>> # can use context managers too:
>>> with open('smart_open/tests/test_data/1984.txt.gz') as fin:
...    with open('smart_open/tests/test_data/1984.txt.bz2', 'w') as fout:
...        for line in fin:
...           fout.write(line)

>>> # can use any IOBase operations, like seek
>>> with open('s3://commoncrawl/robots.txt', 'rb') as fin:
...     for line in fin:
...         print(repr(line.decode('utf-8')))
...         break
...     offset =  # seek to the beginning
...     print(
'User-Agent: *\n'

>>> # stream from HTTP
>>> for line in open(''):
...     print(repr(line))
...     break
'<!doctype html>\n'

Other examples of URLs that smart_open accepts:


For detailed API info, see the online help:


or click here to view the help in your browser.

More examples:

>>> import boto3
>>> # stream content *into* S3 (write mode) using a custom session
>>> url = 's3://smart-open-py37-benchmark-results/test.txt'
>>> lines = [b'first line\n', b'second line\n', b'third line\n']
>>> transport_params = {'session': boto3.Session(profile_name='smart_open')}
>>> with open(url, 'wb', transport_params=transport_params) as fout:
...     for line in lines:
...         bytes_written = fout.write(line)
# stream from HDFS
for line in open('hdfs://user/hadoop/my_file.txt', encoding='utf8'):

# stream from WebHDFS
for line in open('webhdfs://host:port/user/hadoop/my_file.txt'):

# stream content *into* HDFS (write mode):
with open('hdfs://host:port/user/hadoop/my_file.txt', 'wb') as fout:
    fout.write(b'hello world')

# stream content *into* WebHDFS (write mode):
with open('webhdfs://host:port/user/hadoop/my_file.txt', 'wb') as fout:
    fout.write(b'hello world')

# stream from a completely custom s3 server, like s3proxy:
for line in open('s3u://user:secret@host:port@mybucket/mykey.txt'):

# Stream to Digital Ocean Spaces bucket providing credentials from boto profile
transport_params = {
    'session': boto3.Session(profile_name='digitalocean'),
    'resource_kwargs': {
        'endpoint_url': '',
with open('s3://bucket/key.txt', 'wb', transport_params=transport_params) as fout:
    fout.write(b'here we stand')


Working with large S3 files using Amazon's default Python library, boto and boto3, is a pain. Its key.set_contents_from_string() and key.get_contents_as_string() methods only work for small files (loaded in RAM, no streaming). There are nasty hidden gotchas when using boto's multipart upload functionality that is needed for large files, and a lot of boilerplate.

smart_open shields you from that. It builds on boto3 but offers a cleaner, Pythonic API. The result is less code for you to write and fewer bugs to make.


pip install smart_open

Or, if you prefer to install from the source tar.gz:

python test  # run unit tests
python install

To run the unit tests (optional), you'll also need to install mock , moto and responses (pip install mock moto responses). The tests are also run automatically with Travis CI on every commit push & pull request.

Supported Compression Formats

smart_open allows reading and writing gzip and bzip2 files. They are transparently handled over HTTP, S3, and other protocols, too, based on the extension of the file being opened. You can easily add support for other file extensions and compression formats. For example, to open xz-compressed files:

>>> import lzma, os
>>> from smart_open import open, register_compressor

>>> def _handle_xz(file_obj, mode):
...      return lzma.LZMAFile(filename=file_obj, mode=mode, format=lzma.FORMAT_XZ)

>>> register_compressor('.xz', _handle_xz)

>>> with open('smart_open/tests/test_data/crime-and-punishment.txt.xz') as fin:
...     text =
>>> print(len(text))

lzma is in the standard library in Python 3.3 and greater. For 2.7, use backports.lzma.

Transport-specific Options

smart_open supports a wide range of transport options out of the box, including:

  • S3
  • HTTP, HTTPS (read-only)
  • SSH, SCP and SFTP
  • WebHDFS

Each option involves setting up its own set of parameters. For example, for accessing S3, you often need to set up authentication, like API keys or a profile name. smart_open's open function accepts a keyword argument transport_params which accepts additional parameters for the transport layer. Here are some examples of using this parameter:

>>> import boto3
>>> fin = open('s3://commoncrawl/robots.txt', transport_params=dict(session=boto3.Session()))
>>> fin = open('s3://commoncrawl/robots.txt', transport_params=dict(buffer_size=1024))

For the full list of keyword arguments supported by each transport option, see the documentation:


S3 Credentials

smart_open uses the boto3 library to talk to S3. boto3 has several mechanisms for determining the credentials to use. By default, smart_open will defer to boto3 and let the latter take care of the credentials. There are several ways to override this behavior.

The first is to pass a boto3.Session object as a transport parameter to the open function. You can customize the credentials when constructing the session. smart_open will then use the session when talking to S3.

session = boto3.Session(
fin = open('s3://bucket/key', transport_params=dict(session=session), ...)

Your second option is to specify the credentials within the S3 URL itself:

fin = open('s3://aws_access_key_id:aws_secret_access_key@bucket/key', ...)

Important: The two methods above are mutually exclusive. If you pass an AWS session and the URL contains credentials, smart_open will ignore the latter.

Iterating Over an S3 Bucket's Contents

Since going over all (or select) keys in an S3 bucket is a very common operation, there's also an extra function smart_open.s3_iter_bucket() that does this efficiently, processing the bucket keys in parallel (using multiprocessing):

>>> from smart_open import s3_iter_bucket
>>> # get data corresponding to 2010 and later under "silo-open-data/annual/monthly_rain"
>>> # we use workers=1 for reproducibility; you should use as many workers as you have cores
>>> bucket = 'silo-open-data'
>>> prefix = 'annual/monthly_rain/'
>>> for key, content in s3_iter_bucket(bucket, prefix=prefix, accept_key=lambda key: '/201' in key, workers=1, key_limit=3):
...     print(key, round(len(content) / 2**20))
annual/monthly_rain/ 14
annual/monthly_rain/ 14
annual/monthly_rain/ 14

Comments, bug reports

smart_open lives on Github. You can file issues or pull requests there. Suggestions, pull requests and improvements welcome!

smart_open is open source software released under the MIT license. Copyright (c) 2015-now Radim Řehůřek.

Latest Releases
 Apr. 17 2019
 Apr. 7 2019
 Jan. 17 2019
 Sep. 19 2018
 Sep. 18 2018