Skip to main content

PyPI version

Pygfried

pygfried is a CPython extension that brings siegfried - a powerful, signature-based file format identification tool written in Go - into the Python ecosystem.

Identifying a file with pygfried

Instead of reimplementing siegfried's logic, pygfried embeds the original Go code directly, making core siegfried functionality available to Python users without any changes to the underlying detection engine.

No animals were harmed in the making of this extension.

Usage

$ pip install pygfried
$ python -q
>>> import pygfried
>>> pygfried.version()
'1.11.9'
>>> pygfried.identify("example.png")
'fmt/13'
>>> pygfried.identify("example.png", detailed=True)
{'siegfried': '1.11.9', 'scandate': '2026-09-30T17:35:59+02:00', 'signature': 'default.sig', 'created': '2026-09-27T12:04:02+10:00', 'identifiers': [{'name': 'pronom', 'details': 'DROID_SignatureFile_V125.xml; container-signature-20260119.xml'}], 'files': [{'filename': 'example.png', 'filesize': 676214, 'modified': '2025-06-13T11:53:28+02:00', 'errors': '', 'matches': [{'ns': 'pronom', 'id': 'fmt/13', 'format': 'Portable Network Graphics', 'version': '1.2', 'mime': 'image/png', 'class': 'Image (Raster)', 'basis': 'extension match png; byte match at [[0 16] [2962 4] [676202 12]]', 'warning': ''}]}]}
>>> pygfried.identify_many(["example.png", "README.md"], workers=2)
{'siegfried': '1.11.9', ...}
>>> pygfried.identify_dir("samples", recursive=True, workers=2)
{'siegfried': '1.11.9', ...}

When identifying an empty file, identify(path) raises pygfried.EmptySourceError, a subclass of pygfried.GoError and OSError. This also applies to Scanner.identify(path). Existing handlers for GoError continue to catch this exception.

With detailed=True, or when using identify_many or identify_dir, empty files are reported in the results with errors: "empty source" and any extension-based matches. Other files in the scan are still identified.

Custom scanners

The module-level functions use Siegfried's embedded default.sig. To select another complete Siegfried signature database, create a Scanner with its path:

import pygfried

scanner = pygfried.Scanner(signature="/usr/share/siegfried/custom.sig")
scanner.identify("example.png")
scanner.identify_many(["example.png", "README.md"], workers=2)
scanner.identify_dir("samples", recursive=True, workers=2)

External databases are validated and loaded into memory when the scanner is constructed, so changing or deleting the file afterward does not affect that scanner.

As a convenience, a scanner can instead select a bundled signature profile by name:

scanner = pygfried.Scanner(profile="archivematica")
scanner.identify("disk-image.ad1")
# 'archivematica-fmt/2'

Available bundled profiles are:

  • default, which is also selected by Scanner() and matches the behavior of the module-level functions.
  • archivematica, which contains the complete PRONOM database plus Archivematica's extended identifiers for AD1, encrypted AD1, raw disk images, and AFF.

The Archivematica database is copied from the exact Siegfried module version used to compile pygfried. It is embedded in the extension, so it does not need to be installed separately.

profile and signature are keyword-only and mutually exclusive. The read-only scanner.profile property reports the bundled profile, or None for an external database. The read-only scanner.signature property reports the database's filename without its directory path.

Batch and directory scans

Use identify_many when you already have a list of paths, or identify_dir when you want pygfried to scan a directory for you. Both functions return the same detailed result shape as identify(..., detailed=True).

>>> from pathlib import Path
>>> paths = [str(path) for path in Path("samples").rglob("*.png")]
>>> pygfried.identify_many(paths, workers=4)
{'siegfried': '1.11.9', ...}
>>> pygfried.identify_dir("samples", recursive=True, workers=4)
{'siegfried': '1.11.9', ...}

Batch and directory scans avoid repeated Python-to-Go calls. The workers argument controls Go-side concurrency and defaults to 1. Higher values allow multiple files to be identified in parallel. To tune performance, start with the number of CPU cores available to your process, then measure with your own files.

By default identify_dir skips symlinks. Use follow_symlinks=True to identify file symlinks and descend symlinked directories; directory cycles are skipped, and repeated links to the same directory are scanned once.

Concurrency

Both the module-level functions and Scanner methods are safe to call from multiple threads. The module-level API shares one immutable default scanner; each custom scanner owns an independent, immutable Siegfried engine.

The extension holds Python's GIL during identification, so Python threads do not run scans in parallel. Use workers for parallel file identification within a batch.

Each identify_many or identify_dir call controls its own Go-side concurrency with workers, whether it is called on the module or on a scanner. There is no process-wide worker limit beyond the per-call range of 1 to 1024.

For process-based parallelism, use the spawn start method explicitly (multiprocessing.get_context("spawn")). Import pygfried and create any custom scanners inside each worker process. Do not pass scanner instances between processes. Other start methods have not been verified.

Limitations

Go libraries can clash

This project uses Go's -buildmode=c-shared to provide its Python extension. Loading multiple Go-based shared libraries in the same process has known issues and may result in panics or crashes due to conflicts between separate Go runtimes.

This limitation should only affect you if you're using pygfried together with another Python library that also uses a Go extension (built with the same c-shared mechanism) in the same process. If you're just using pygfried on its own, you don't need to worry - everything should work as expected.

Credits

pygfried is powered by the original siegfried project, which is distributed under the Apache License, Version 2.0. All core file format identification logic and signatures are provided by siegfried. We gratefully acknowledge the work of the siegfried project and its contributors.

Metadata

Release files for pygfried 0.19.1

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

Source distribution (sdist)

Source distribution for pygfried 0.19.1
File Size Uploaded
pygfried-0.19.1.tar.gz 887.9 kB Details

Built distributions (wheels)

Table of built distributions (wheels) for pygfried 0.19.1
File
pygfried-0.19.1-cp310-abi3-win_amd64.whl CPython 3.10 abi3 Windows x86-64 Details
pygfried-0.19.1-cp310-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl CPython 3.10 abi3 Linux glibc 2.17+ ARM64 Details
pygfried-0.19.1-cp310-abi3-manylinux1_x86_64.manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_5_x86_64.whl CPython 3.10 abi3 Linux glibc 2.17+ x86-64, Linux glibc 2.5+ x86-64 Details
pygfried-0.19.1-cp310-abi3-macosx_11_0_universal2.whl CPython 3.10 abi3 macOS 11.0+ universal2 (ARM64, x86-64) Details

Total release size: 12.4 MB

Release files / pygfried-0.19.1.tar.gz

Download URL pygfried-0.19.1.tar.gz
Size 887.9 kB
Tags Source
SHA-256 checksum
How to use checksums
069cad95a3a31e30595c4599af60bc38e582439e51948b80db763de88b7a532d
BLAKE2b-256 checksum
How to use checksums
16ffaaae1c51c0450de71994b87f39abf8f94471d6b6e470c6942e9de709bba2
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 Oct 1, 2026.

Transparency log

Release files / pygfried-0.19.1-cp310-abi3-win_amd64.whl

Download URL pygfried-0.19.1-cp310-abi3-win_amd64.whl
Size 3.0 MB
Tags CPython 3.10 Windows x86-64 abi3
SHA-256 checksum
How to use checksums
e7e9d1f3aa7edb31ecc63ff8bffa613cfde13cc8b02d12761e3722c955d7c344
BLAKE2b-256 checksum
How to use checksums
c2159e648058760c2a9244f2d26c9760a7e303cb1d6b37740a78b319738f342e
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 Oct 1, 2026.

Transparency log

Release files / pygfried-0.19.1-cp310-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl

Download URL pygfried-0.19.1-cp310-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
Size 2.8 MB
Tags CPython 3.10 Linux glibc 2.17+ ARM64 abi3
SHA-256 checksum
How to use checksums
bc1944765fae157cf60a7f26ab3966ba40190547f01362f5dcc73fe16fa1b6db
BLAKE2b-256 checksum
How to use checksums
08caab9926c43a0b3d8f360d7d5c50edd6f953edd19d5dac07dc8627771cd2a1
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 Oct 1, 2026.

Transparency log

Release files / pygfried-0.19.1-cp310-abi3-manylinux1_x86_64.manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_5_x86_64.whl

Download URL pygfried-0.19.1-cp310-abi3-manylinux1_x86_64.manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_5_x86_64.whl
Size 3.0 MB
Tags CPython 3.10 Linux glibc 2.17+ x86-64 Linux glibc 2.5+ x86-64 abi3
SHA-256 checksum
How to use checksums
942f333dfa5d5b6675dc10fdd86439dba1498d5bb20e82efbfedbf786ae88ce9
BLAKE2b-256 checksum
How to use checksums
7f1c1d43a60445fbff25b982a54017d2e759fb07424f53be18e12aff8364552e
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 Oct 1, 2026.

Transparency log

Release files / pygfried-0.19.1-cp310-abi3-macosx_11_0_universal2.whl

Download URL pygfried-0.19.1-cp310-abi3-macosx_11_0_universal2.whl
Size 2.7 MB
Tags CPython 3.10 abi3 macOS 11.0+ universal2 (ARM64, x86-64)
SHA-256 checksum
How to use checksums
6b370f113ed4a56a4c42d6a49b3d564221542f9101c11d2da2504636df11582d
BLAKE2b-256 checksum
How to use checksums
b4654d2e4d028a232aac9b517851ac8bd7f2a84c5399ea0e5e39f9757e8c319a
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 Oct 1, 2026.

Transparency log

Release history Release notifications | RSS feed

0.20.0

5 release files

This release

0.19.1 This release

5 release files

0.19.0

5 release files

0.18.1

5 release files

0.18.0

5 release files

0.17.0

5 release files

0.16.1

5 release files

0.16.0

5 release files

0.15.1

5 release files

0.15.0

5 release files

0.14.1

5 release files

0.12.0

8 release files

0.11.0

8 release files

0.10.0

4 release files

0.9.1

3 release files

0.9.0

2 release files

0.8.0

2 release files

0.7.0

3 release files

0.6.0

3 release files

0.5.0

4 release files

0.4.0

4 release files

0.3.0

4 release files

0.2.0

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