Skip to main content

🚪 Doorway

Essential data wrangling utilities.

license python versions pypi version tests status code coverage

Contributions are welcome!


Table Of Contents


Overview

Doorway is a common library for interacting with files and data wranging, with zero deps over the common utils.

Get started with doorway by installing it with $pip install doorway or cloning this repository.

Experimental features need pip install "doorway[extras]"

Features

Doorway includes the following features:

  • Partial "fast" hashing of files
  • Stale file detection
  • Atomic file writing and overwriting via a separate temporary file that is moved into place
  • Path name modification
  • Context managers
  • Environment Vars & Validation
  • String Colors
  • String Formatting
  • Downloading with a progress bar
  • Deterministic dataset partitioning using

As well as experimental:

  • Arbitrary URI interop and handling
  • Proxy downloader -- multithreading, auto-retry, auto proxy retrieval and rotation.

Atomic IO

Write or overwrite a file without worrying about it being half-written or corrupted, and left in a bad state.

from doorway import AtomicOpen, AtomicPath

# write to a temporary file, then move it into place once the context closes
with AtomicOpen("my/file.txt", "w") as f:
    f.write("Hello, World!")

# generate a temporary path and raise an exception if the file does not exist when the context closes
with AtomicPath("my/file.txt") as path:
    with open(path, "w") as f:  # `doorway.AtomicPath` + open ~= `doorway.AtomicOpen`
        f.write("Hello, World!")

Fast Hashing

from doorway import hash_file

# normal hashing
hast_full = hash_file("my/large.file", hash_mode="full", hash_algo="md5")
# fast hashing, uses start, middle and end bytes
hash_fast = hash_file("my/large.file", hash_mode="fast", hash_algo="md5")

# `doorway.hash_bytes`, `doorway.hash_str`, `doorway.hash_file_is_valid` are also available

Stale Files

from doorway import stalefile_generate, stalefile_is_stale

# generate a file if it is stale (doesn't exist or the hash is different)
stalefile_generate(
    make_file_fn=lambda path: open(path, "w").close(),
    path="my/file.txt",
    hash="some-long-hash-fffffffffffffffffffffffffff",
)

# check if a file is stale (doesn't exist or the hash is different)
exists = stalefile_is_stale(
    path="my/file.txt",
    hash="some-long-hash-fffffffffffffffffffffffffff",
)

Path Modification

from doorway import path_basename_modify

# `basename` is the full last part of the path e.g. "file.1.2.txt"
# `name` is the basename without the `extension` e.g. "file.1.2" or "file" (depending on the `name_contains_sep` parameter)
path = "my/path/to/file.1.2.txt"

p0 = path_basename_modify(
    path,
    ext="csv",
    name_prefix="a_",
    name_suffix="_b",
    name_contains_sep=True,
)
p1 = path_basename_modify(
    path,
    ext="csv",
    name_prefix="a_",
    name_suffix="_b",
    name_contains_sep=False,
)

assert p0 == "my/path/to/a_file.1.2_b.csv"
assert p1 == "my/path/to/a_file_b.csv"

# see `doorway.basename_modify` for the non-path version of this
# see `doorway.basename_split_ext` for splitting the name from the extension

File Downloading

Easy file downloading wrapping AtomicIO with a progress bar if tqdm is installed.

from doorway import io_download

# download a file with a progress bar
io_download(
    src_url="https://example.com/file.txt",
    dst_path="my/file.txt",
    exists_mode="error",  # options are: "error" (default), "overwrite", "skip"
    progress=True,  # (default)
)

Context Managers

Context managers for various tasks.

import os
from doorway import ctx_temp_environ

# temporarily set an environment variable
with ctx_temp_environ(TEMP_VAR="temp_value"):
    assert os.environ.get("TEMP_VAR", None) == "temp_value"
assert os.environ.get("TEMP_VAR", None) is None

# most of the other context managers are similar
# all gracefully handle exceptions and undo their changes
# - ctx_temp_wd:       temporarily update the working directory
# - ctx_temp_attr:     temporarily update an attribute on an object
# - ctx_temp_sys_args: temporarily update sys.argv
# - ctx_no_stdout:     temporarily suppress stdout
# - ctx_no_stderr:     temporarily suppress stderr
# - ctx_do_undo:       do something (run "do" fn), then undo it (run "undo" fn)

Environment Variables

Parse environment variables and validate their values.

from doorway import EnvVar, ctx_temp_environ

MY_VAR = EnvVar.env_int(
    "MY_VAR",
    default=42,
    validator=EnvVar.validator_min_max(0, 100),
)

# get the value of the environment variable (and validate)
with ctx_temp_environ(MY_VAR="7"):
    assert MY_VAR.get() == 7

# get a default value if set (and validate)
# - if no default was set, an exception is raised
assert MY_VAR.get() == 42

# override the default value for this call (and validate)
with ctx_temp_environ(MY_VAR="8"):
    assert MY_VAR.get(default=43) == 8
assert MY_VAR.get(default=43) == 43

# force a value (and validate)
with ctx_temp_environ(MY_VAR="9"):
    assert MY_VAR.get(override=99) == 99
assert MY_VAR.get(override=99) == 99

ANSI Colors

from doorway import c

print(f"{c.lRED}This is red {c.RST}This is reset to defaults")

Human Readable Byte Conversion

from doorway import fmt_bytes_to_human

assert fmt_bytes_to_human(1025**0, base=1024) == "1.000 B"
assert fmt_bytes_to_human(1025**0, base=1000) == "1.000 B"

assert fmt_bytes_to_human(1025**1, base=1024) == "1.001 KiB"
assert fmt_bytes_to_human(1025**1, base=1000) == "1.025 kB"  # lowercase is more correct for kB in this case

assert fmt_bytes_to_human(1025**2, base=1024) == "1.002 MiB"
assert fmt_bytes_to_human(1025**2, base=1000) == "1.051 MB"

assert fmt_bytes_to_human(1025**3, base=1024) == "1.003 GiB"
assert fmt_bytes_to_human(1025**3, base=1000) == "1.077 GB"

assert fmt_bytes_to_human(1025**4, base=1024) == "1.004 TiB"
assert fmt_bytes_to_human(1025**4, base=1000) == "1.104 TB"

Deterministic Dataset Partitioning

Split datapoints deterministically but pseudo-randomly into buckets.

This is useful when you have datasets that are increasing or decreasing in size, but you want to keep adding the same datapoints to the same splits/buckets/shards. (A standard shuffle with seed CANNOT do this)

The example below only has 9 datapoints, which doesn't show the full power of this function, this is useful for much larger datasets when assignment to buckets follows the expected statistical distribution.

from doorway import sharded, sharded_weighted

# create 3 buckets and assign values to them pseudo-randomly
shards = sharded(
    values=[1, 2, 3, 4, 5, 6, 7, 8, 9],
    num_shards=3,
    shard_key=lambda x: str(x),
)
print(shards)  # [[2, 5, 6, 7], [1, 3], [4, 8, 9]]

# create 3 buckets with relative weights 9:1:1 and assign values to them pseudo-randomly
train, test, split = sharded_weighted(
    values=[1, 2, 3, 4, 5, 6, 7, 8, 9],
    shard_weights=[8, 3, 1],
    shard_key=str,
)
print(train, test, split)  # [2, 6, 9, 7, 8, 1, 3] [4, 5] []

Proxy Downloader

# EXPERIMENTAL, API may change -- `pip install doorway[extras]`
from doorway.x import ProxyDownloader, proxy_download


# collect a default list of proxies and cache them
# - see: `proxies_register_scraper` for registering custom proxies, or pass in a list of proxies directly to this
downloader = ProxyDownloader()

# download a single file with a randomly chosen proxy
path = downloader.download(
    "https://example.com/file.txt",  # src
    "my/file.txt",  # dst
    exists_mode="error",  # options are: "error" (default), "overwrite", "skip"
)

# download multiple files in parallel while rotating proxies and retrying if any fail
failed = downloader.download_threaded(
    [
        ("https://example.com/file1.txt", "my/file1.txt"),
        ("https://example.com/file2.txt", "my/file2.txt"),
        ("https://example.com/file3.txt", "my/file3.txt"),
    ],
    exists_mode="error",  # options are: "error" (default), "overwrite", "skip"
    verbose=True,  # (default is False)
    ignore_failures=True,  # (default is False)
)

Proxy Issues?

The scrape logic used to obtain the proxy list will probably go out of date. You can override the default scrape logic by registering a new scrape function.

from doorway.x import proxies_register_scraper


@proxies_register_scraper(name="my_proxy_source", is_default=True)
def custom_proxy_scraper(proxy_type):
    # you should respect this setting
    assert proxy_type in ("all", "http", "https")
    # proxies is a list of dictionaries, where each dictionary only has one entry:
    # - the key is the protocol
    # - the value is the matching full url
    return [
        {"HTTP": "http://<my-http-proxy>.com"},
        {"HTTPS": "https://<my-https-proxy>.com"},
    ]

URI Handling

Interop between URIs from different locations, e.g. S3, Local, HTTP, etc.

# TODO: W.I.P

TODO

  • More Docs
  • More Examples
  • More APIs

Release files for doorway 0.7.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for doorway 0.7.1
File Size Uploaded
doorway-0.7.1.tar.gz 79.2 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for doorway 0.7.1
File Interpreter ABI Platform
doorway-0.7.1-py3-none-any.whl Python 3 none any Details

Total release size: 128.8 kB

Release files / doorway-0.7.1.tar.gz

Download URL doorway-0.7.1.tar.gz
Size 79.2 kB
Tags Source
SHA-256 checksum
How to use checksums
88e4018ab7731c4f29b1bcc2ed3d23d413c9aa349915857bd4db8659516bcb56
BLAKE2b-256 checksum
How to use checksums
f5ebab8df4bf7c447fcee78b2df7758b729aaa46adaed3ccc6907d54586d9432
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 13, 2026.

Transparency log

Release files / doorway-0.7.1-py3-none-any.whl

Download URL doorway-0.7.1-py3-none-any.whl
Size 49.6 kB
Tags Python 3
SHA-256 checksum
How to use checksums
18725234c556e059d4eac7a8582a475bd3aad146cff21e1c1c061d835d2b546d
BLAKE2b-256 checksum
How to use checksums
9149bfa7908079bf08b64e1ed3b210e0f5d6cd72b12f2421a1a7d1b62dd18b86
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 13, 2026.

Transparency log

Release history Release notifications | RSS feed

0.7.2

2 release files

This release

0.7.1 This release

2 release files

0.7.0

2 release files

0.6.3

2 release files

0.6.2

2 release files

0.6.1

2 release files

0.6.0

2 release files

0.5.0

2 release files

0.4.0

2 release files

0.3.0

2 release files

0.2.0

2 release files

0.1.1

2 release files

0.1.0

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page