Diff

A diff shows the changes between trees, an index or the working dir.

Repository.diff(a=None, b=None, cached=False, flags=0, context_lines=3, interhunk_lines=0)

Show changes between the working tree and the index or a tree, changes between the index and a tree, changes between two trees, or changes between two blobs.

Keyword arguments:

a

None, a str (that refers to an Object, see revparse_single()) or a Reference object. If None, b must be None, too. In this case the working directory is compared with the index. Otherwise the referred object is compared to ‘b’.

b

None, a str (that refers to an Object, see revparse_single()) or a Reference object. If None, the working directory is compared to ‘a’. (except ‘cached’ is True, in which case the index is compared to ‘a’). Otherwise the referred object is compared to ‘a’

cached

If ‘b’ is None, by default the working directory is compared to ‘a’. If ‘cached’ is set to True, the index/staging area is used for comparing.

flag

A combination of GIT_DIFF_* constants. For a list of the constants, with a description, see git_diff_option_t in https://github.com/libgit2/libgit2/blob/master/include/git2/diff.h

context_lines

The number of unchanged lines that define the boundary of a hunk (and to display before and after)

interhunk_lines

The maximum number of unchanged lines between hunk boundaries before the hunks will be merged into a one

Examples:

# Changes in the working tree not yet staged for the next commit
>>> diff()

# Changes between the index and your last commit
>>> diff(cached=True)

# Changes in the working tree since your last commit
>>> diff('HEAD')

# Changes between commits
>>> t0 = revparse_single('HEAD')
>>> t1 = revparse_single('HEAD^')
>>> diff(t0, t1)
>>> diff('HEAD', 'HEAD^') # equivalent

If you want to diff a tree against an empty tree, use the low level API (Tree.diff_to_tree()) directly.

Examples

# Changes between commits
>>> t0 = revparse_single('HEAD')
>>> t1 = revparse_single('HEAD^')
>>> repo.diff(t0, t1)
>>> t0.diff(t1)           # equivalent
>>> repo.diff('HEAD', 'HEAD^') # equivalent

# Get all patches for a diff
>>> diff = repo.diff('HEAD^', 'HEAD~3')
>>> patches = [p for p in diff]

# Get the stats for a diff
>>> diff = repo.diff('HEAD^', 'HEAD~3')
>>> diff.stats

# Diffing the empty tree
>>> tree = revparse_single('HEAD').tree
>>> tree.diff_to_tree()

# Diff empty tree to a tree
>>> tree = revparse_single('HEAD').tree
>>> tree.diff_to_tree(swap=True)

The Diff type

class pygit2.Diff

Diff objects.

__iter__()

Returns an iterator over the deltas/patches in this diff.

__len__()

Returns the number of deltas/patches in this diff.

deltas

Iterate over the diff deltas.

find_similar(flags: int = GIT_DIFF_FIND_BY_CONFIG, rename_threshold: int = 50, copy_threshold: int = 50, rename_from_rewrite_threshold: int = 50, break_rewrite_threshold: int = 60, rename_limit: int = 1000)

Transform a diff marking file renames, copies, etc.

This modifies a diff in place, replacing old entries that look like renames or copies with new entries reflecting those changes. This also will, if requested, break modified files into add/remove pairs if the amount of change is above a threshold.

flags - Combination of GIT_DIFF_FIND_* and GIT_DIFF_BREAK_* constants.

merge(diff: Diff)

Merge one diff into another.

static parse_diff(git_diff: str | bytes) Diff

Parses a git unified diff into a diff object without a repository

patch

Patch diff string. Can be None in some cases, such as empty commits.

patchid

Corresponding patchid.

stats

Accumulate diff statistics for all patches.

The Patch type

Attributes:

class pygit2.Patch

Diff patch object.

static create_from()

Create a patch from blobs, buffers, or a blob and a buffer

data

The raw bytes of the patch’s contents.

delta

Get the delta associated with a patch.

hunks
line_stats

Get line counts of each type in a patch (context, additions, deletions).

text

Patch diff string. Can be None in some cases, such as empty commits. Note that this decodes the content to Unicode assuming UTF-8 encoding. For non-UTF-8 content that can lead be a lossy, non-reversible process. To access the raw, un-decoded patch, use patch.data.

The DiffDelta type

class pygit2.DiffDelta

DiffDelta object.

flags

Combination of GIT_DIFF_FLAG_* flags.

is_binary

True if binary data, False if text, None if not (yet) known.

new_file

“to” side of the diff.

nfiles

Number of files in the delta.

old_file

“from” side of the diff.

similarity

For renamed and copied.

status

A GIT_DELTA_* constant.

status_char() str

Return the single character abbreviation for a delta status code.

The DiffFile type

class pygit2.DiffFile

DiffFile object.

flags

Combination of GIT_DIFF_FLAG_* flags.

static from_c()

Method exposed for _checkout_notify_cb to hook into

id

Oid of the item.

mode

Mode of the entry.

path

Path to the entry.

raw_path

Path to the entry (bytes).

size

Size of the entry.

The DiffHunk type

class pygit2.DiffHunk

DiffHunk object.

header

Header.

lines

Lines.

new_lines

New lines.

new_start

New start.

old_lines

Old lines.

old_start

Old start.

The DiffStats type

class pygit2.DiffStats

DiffStats object.

deletions

Total number of deletions

files_changed

Total number of files changed

format(format: int, width: int) str

Format the stats as a string.

Returns: str.

Parameters:

format

The format to use. A pygit2.GIT_DIFF_STATS_* constant.

width

The width of the output. The output will be scaled to fit.

insertions

Total number of insertions

The DiffLine type

class pygit2.DiffLine

DiffLine object.

content

Content of the diff line

content_offset

Offset in the original file to the content

new_lineno

Line number in new file or -1 for deleted line

num_lines

Number of newline characters in content

old_lineno

Line number in old file or -1 for added line

origin

Type of the diff line

raw_content

Content of the diff line (byte string)