fast-export Git Command Guide

The git fast-export command is used to export Git data in a format suitable for piping into git fast-import. It dumps revisions in a human-readable format that can be used as bundle replacement or for history rewrites.

git fast-export Syntax:

git fast-export [<options>]

Options:

Option	Description
—progress=	Insert progress statements every n objects
—signed-tags=(action)	Handle signed tags (verbatim
—signed-commits=(action)	Handle signed commits
-M, -C	Perform move and copy detection
—export-marks=	Dump marks table to file
—import-marks=	Load marks from file
—mark-tags	Label tags with mark IDs
—fake-missing-tagger	Fake tagger for old repos
—use-done-feature	Use done stanza
—no-data	Refer to blobs by SHA-1 hash
—full-tree	Output deleteall + full file list
—anonymize	Anonymize repository contents
—tag-of-filtered-object=(action)	Handle filtered tags

Parameters:

Parameter	Description
(none)	Operates on current repository
[…]	Optional commit ranges to export

Caution

With —no-data option, output can only be imported into repositories that already contain the required blob objects. Signed tags and commits may become invalid after transformations.

git fast-export Command Samples:

Basic export and import

git fast-export | git fast-import

Exports current repository and imports it back (creates a copy).

Export all branches with progress

git fast-export --progress=100 | git fast-import

Shows progress every 100 objects during the import process.

Incremental export with marks

git fast-export --export-marks=marks.txt HEAD | git fast-import --import-marks=marks.txt

Uses marks for incremental exports, allowing bidirectional syncing.

Export with move and copy detection

git fast-export -M -C HEAD~10..HEAD | git fast-import

Enables rename and copy detection in the exported stream.

Anonymize repository data

git fast-export --anonymize | git fast-import

Removes identifying information while preserving repository structure.

Export to bundle-like format

git fast-export --no-data HEAD | gzip > repo.bundle.gz

Creates a compressed bundle with blob references instead of content.

Tip

Use marks files (—export-marks, —import-marks) for incremental bidirectional repository transformations and syncing.

How do I export the entire repository?

To export the entire repository, run:

git fast-export HEAD | git fast-import

How can I handle incremental exports?

To handle incremental exports, use marks files:

git fast-export --export-marks=marks.txt HEAD
git fast-export --import-marks=marks.txt --export-marks=marks.txt NEW_COMMIT

How do I anonymize repository data?

To anonymize repository data, use the —anonymize option:

git fast-export --anonymize HEAD | git fast-import

What is the output format of fast-export?

git fast-export outputs a text stream format designed for git fast-import, containing commit, blob, and tree objects in a structured format suitable for editing and transformation.

How do I handle signed commits and tags?

To handle signed commits and tags, use:

git fast-export --signed-tags=warn-strip --signed-commits=warn-strip HEAD

How can I show progress during export/import?

To show progress during export/import operations, use —progress option with fast-import (fast-export doesn’t have built-in progress):

git fast-export HEAD | git fast-import --progress=50

Applications of the git fast-export command

1. Repository migration between Git hosts
2. Historical data rewriting and filtering
3. Creating anonymized repository copies for testing
4. Implementing custom Git transformation pipelines
5. Backup and archival of repository in text format