2017-07-12 11:06:35 -04:00
|
|
|
# Gitaly note: JV: needs RPC for Gitlab::Git::Diff.between.
|
|
|
|
|
2017-01-04 13:43:06 -05:00
|
|
|
# Gitlab::Git::Diff is a wrapper around native Rugged::Diff object
|
|
|
|
module Gitlab
|
|
|
|
module Git
|
|
|
|
class Diff
|
2017-03-01 06:00:37 -05:00
|
|
|
TimeoutError = Class.new(StandardError)
|
2017-06-01 17:21:14 -04:00
|
|
|
include Gitlab::EncodingHelper
|
2017-01-04 13:43:06 -05:00
|
|
|
|
|
|
|
# Diff properties
|
|
|
|
attr_accessor :old_path, :new_path, :a_mode, :b_mode, :diff
|
|
|
|
|
|
|
|
# Stats properties
|
|
|
|
attr_accessor :new_file, :renamed_file, :deleted_file
|
|
|
|
|
2017-05-15 13:10:29 -04:00
|
|
|
alias_method :new_file?, :new_file
|
|
|
|
alias_method :deleted_file?, :deleted_file
|
|
|
|
alias_method :renamed_file?, :renamed_file
|
|
|
|
|
2017-05-26 19:27:30 -04:00
|
|
|
attr_accessor :expanded
|
2017-06-09 07:48:25 -04:00
|
|
|
attr_writer :too_large
|
2017-01-04 13:43:06 -05:00
|
|
|
|
2017-06-06 17:28:06 -04:00
|
|
|
alias_method :expanded?, :expanded
|
|
|
|
|
2017-06-09 07:48:25 -04:00
|
|
|
SERIALIZE_KEYS = %i(diff new_path old_path a_mode b_mode new_file renamed_file deleted_file too_large).freeze
|
2017-05-30 15:50:02 -04:00
|
|
|
|
2017-10-02 08:51:51 -04:00
|
|
|
# The maximum size of a diff to display.
|
|
|
|
SIZE_LIMIT = 100.kilobytes
|
2017-06-09 08:12:27 -04:00
|
|
|
|
2017-10-02 08:51:51 -04:00
|
|
|
# The maximum size before a diff is collapsed.
|
|
|
|
COLLAPSE_LIMIT = 10.kilobytes
|
2017-01-04 13:43:06 -05:00
|
|
|
|
2017-10-02 08:51:51 -04:00
|
|
|
class << self
|
2017-01-04 13:43:06 -05:00
|
|
|
def between(repo, head, base, options = {}, *paths)
|
|
|
|
straight = options.delete(:straight) || false
|
|
|
|
|
|
|
|
common_commit = if straight
|
|
|
|
base
|
|
|
|
else
|
|
|
|
# Only show what is new in the source branch
|
|
|
|
# compared to the target branch, not the other way
|
|
|
|
# around. The linex below with merge_base is
|
|
|
|
# equivalent to diff with three dots (git diff
|
|
|
|
# branch1...branch2) From the git documentation:
|
|
|
|
# "git diff A...B" is equivalent to "git diff
|
|
|
|
# $(git-merge-base A B) B"
|
|
|
|
repo.merge_base_commit(head, base)
|
|
|
|
end
|
|
|
|
|
|
|
|
options ||= {}
|
|
|
|
actual_options = filter_diff_options(options)
|
|
|
|
repo.diff(common_commit, head, actual_options, *paths)
|
|
|
|
end
|
|
|
|
|
|
|
|
# Return a copy of the +options+ hash containing only keys that can be
|
|
|
|
# passed to Rugged. Allowed options are:
|
|
|
|
#
|
|
|
|
# :ignore_whitespace_change ::
|
|
|
|
# If true, changes in amount of whitespace will be ignored.
|
|
|
|
#
|
|
|
|
# :disable_pathspec_match ::
|
|
|
|
# If true, the given +*paths+ will be applied as exact matches,
|
|
|
|
# instead of as fnmatch patterns.
|
|
|
|
#
|
|
|
|
def filter_diff_options(options, default_options = {})
|
2017-07-12 11:06:35 -04:00
|
|
|
allowed_options = [:ignore_whitespace_change,
|
|
|
|
:disable_pathspec_match, :paths,
|
2017-05-31 15:41:25 -04:00
|
|
|
:max_files, :max_lines, :limits, :expanded]
|
2017-01-04 13:43:06 -05:00
|
|
|
|
|
|
|
if default_options
|
|
|
|
actual_defaults = default_options.dup
|
|
|
|
actual_defaults.keep_if do |key|
|
|
|
|
allowed_options.include?(key)
|
|
|
|
end
|
|
|
|
else
|
|
|
|
actual_defaults = {}
|
|
|
|
end
|
|
|
|
|
|
|
|
if options
|
|
|
|
filtered_opts = options.dup
|
|
|
|
filtered_opts.keep_if do |key|
|
|
|
|
allowed_options.include?(key)
|
|
|
|
end
|
|
|
|
filtered_opts = actual_defaults.merge(filtered_opts)
|
|
|
|
else
|
|
|
|
filtered_opts = actual_defaults
|
|
|
|
end
|
|
|
|
|
|
|
|
filtered_opts
|
|
|
|
end
|
2017-09-03 07:45:44 -04:00
|
|
|
|
|
|
|
# Return a binary diff message like:
|
2017-09-04 15:32:57 -04:00
|
|
|
#
|
2017-09-03 07:45:44 -04:00
|
|
|
# "Binary files a/file/path and b/file/path differ\n"
|
2017-09-05 13:16:08 -04:00
|
|
|
# This is used when we detect that a diff is binary
|
|
|
|
# using CharlockHolmes when Rugged treats it as text.
|
2017-09-03 07:45:44 -04:00
|
|
|
def binary_message(old_path, new_path)
|
|
|
|
"Binary files #{old_path} and #{new_path} differ\n"
|
|
|
|
end
|
2017-01-04 13:43:06 -05:00
|
|
|
end
|
|
|
|
|
2017-05-26 19:27:30 -04:00
|
|
|
def initialize(raw_diff, expanded: true)
|
|
|
|
@expanded = expanded
|
|
|
|
|
2017-01-04 13:43:06 -05:00
|
|
|
case raw_diff
|
|
|
|
when Hash
|
2017-02-24 10:53:44 -05:00
|
|
|
init_from_hash(raw_diff)
|
2017-05-26 19:27:30 -04:00
|
|
|
prune_diff_if_eligible
|
2017-01-04 13:43:06 -05:00
|
|
|
when Rugged::Patch, Rugged::Diff::Delta
|
2017-05-26 19:27:30 -04:00
|
|
|
init_from_rugged(raw_diff)
|
2017-05-30 15:30:05 -04:00
|
|
|
when Gitlab::GitalyClient::Diff
|
2017-02-24 10:53:44 -05:00
|
|
|
init_from_gitaly(raw_diff)
|
2017-05-26 19:27:30 -04:00
|
|
|
prune_diff_if_eligible
|
2017-05-05 10:55:12 -04:00
|
|
|
when Gitaly::CommitDelta
|
|
|
|
init_from_gitaly(raw_diff)
|
2017-01-04 13:43:06 -05:00
|
|
|
when nil
|
|
|
|
raise "Nil as raw diff passed"
|
|
|
|
else
|
|
|
|
raise "Invalid raw diff type: #{raw_diff.class}"
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
def to_hash
|
|
|
|
hash = {}
|
|
|
|
|
2017-06-09 07:48:25 -04:00
|
|
|
SERIALIZE_KEYS.each do |key|
|
2017-08-03 22:20:34 -04:00
|
|
|
hash[key] = send(key) # rubocop:disable GitlabSecurity/PublicSend
|
2017-01-04 13:43:06 -05:00
|
|
|
end
|
|
|
|
|
|
|
|
hash
|
|
|
|
end
|
|
|
|
|
2017-05-15 13:10:29 -04:00
|
|
|
def mode_changed?
|
|
|
|
a_mode && b_mode && a_mode != b_mode
|
|
|
|
end
|
|
|
|
|
2017-01-04 13:43:06 -05:00
|
|
|
def submodule?
|
|
|
|
a_mode == '160000' || b_mode == '160000'
|
|
|
|
end
|
|
|
|
|
|
|
|
def line_count
|
|
|
|
@line_count ||= Util.count_lines(@diff)
|
|
|
|
end
|
|
|
|
|
|
|
|
def too_large?
|
2017-05-30 22:48:30 -04:00
|
|
|
if @too_large.nil?
|
2017-10-02 08:51:51 -04:00
|
|
|
@too_large = @diff.bytesize >= SIZE_LIMIT
|
2017-05-30 22:48:30 -04:00
|
|
|
else
|
|
|
|
@too_large
|
|
|
|
end
|
2017-01-04 13:43:06 -05:00
|
|
|
end
|
|
|
|
|
2017-06-09 07:48:25 -04:00
|
|
|
# This is used by `to_hash` and `init_from_hash`.
|
|
|
|
alias_method :too_large, :too_large?
|
|
|
|
|
2017-05-26 19:27:30 -04:00
|
|
|
def too_large!
|
2017-01-04 13:43:06 -05:00
|
|
|
@diff = ''
|
|
|
|
@line_count = 0
|
|
|
|
@too_large = true
|
|
|
|
end
|
|
|
|
|
|
|
|
def collapsed?
|
|
|
|
return @collapsed if defined?(@collapsed)
|
2017-05-26 19:27:30 -04:00
|
|
|
|
2017-10-02 08:51:51 -04:00
|
|
|
@collapsed = !expanded && @diff.bytesize >= COLLAPSE_LIMIT
|
2017-01-04 13:43:06 -05:00
|
|
|
end
|
|
|
|
|
2017-05-26 19:27:30 -04:00
|
|
|
def collapse!
|
2017-01-04 13:43:06 -05:00
|
|
|
@diff = ''
|
|
|
|
@line_count = 0
|
|
|
|
@collapsed = true
|
|
|
|
end
|
|
|
|
|
2017-09-04 13:34:15 -04:00
|
|
|
def json_safe_diff
|
2017-09-05 13:16:08 -04:00
|
|
|
return @diff unless detect_binary?(@diff)
|
2017-09-04 13:34:15 -04:00
|
|
|
|
|
|
|
# the diff is binary, let's make a message for it
|
2017-09-04 15:32:57 -04:00
|
|
|
Diff.binary_message(@old_path, @new_path)
|
2017-09-04 13:34:15 -04:00
|
|
|
end
|
|
|
|
|
2017-09-11 08:52:27 -04:00
|
|
|
def has_binary_notice?
|
|
|
|
@diff.start_with?('Binary')
|
|
|
|
end
|
|
|
|
|
2017-01-04 13:43:06 -05:00
|
|
|
private
|
|
|
|
|
2017-05-26 19:27:30 -04:00
|
|
|
def init_from_rugged(rugged)
|
2017-01-04 13:43:06 -05:00
|
|
|
if rugged.is_a?(Rugged::Patch)
|
2017-05-26 19:27:30 -04:00
|
|
|
init_from_rugged_patch(rugged)
|
2017-01-04 13:43:06 -05:00
|
|
|
d = rugged.delta
|
|
|
|
else
|
|
|
|
d = rugged
|
|
|
|
end
|
|
|
|
|
|
|
|
@new_path = encode!(d.new_file[:path])
|
|
|
|
@old_path = encode!(d.old_file[:path])
|
|
|
|
@a_mode = d.old_file[:mode].to_s(8)
|
|
|
|
@b_mode = d.new_file[:mode].to_s(8)
|
|
|
|
@new_file = d.added?
|
|
|
|
@renamed_file = d.renamed?
|
|
|
|
@deleted_file = d.deleted?
|
|
|
|
end
|
|
|
|
|
2017-05-26 19:27:30 -04:00
|
|
|
def init_from_rugged_patch(patch)
|
2017-01-04 13:43:06 -05:00
|
|
|
# Don't bother initializing diffs that are too large. If a diff is
|
|
|
|
# binary we're not going to display anything so we skip the size check.
|
2017-05-26 19:27:30 -04:00
|
|
|
return if !patch.delta.binary? && prune_large_patch(patch)
|
2017-01-04 13:43:06 -05:00
|
|
|
|
2017-09-04 13:34:15 -04:00
|
|
|
@diff = encode!(strip_diff_headers(patch.to_s))
|
2017-01-04 13:43:06 -05:00
|
|
|
end
|
|
|
|
|
2017-02-24 10:53:44 -05:00
|
|
|
def init_from_hash(hash)
|
2017-01-04 13:43:06 -05:00
|
|
|
raw_diff = hash.symbolize_keys
|
|
|
|
|
2017-06-09 07:48:25 -04:00
|
|
|
SERIALIZE_KEYS.each do |key|
|
2017-08-03 22:20:34 -04:00
|
|
|
send(:"#{key}=", raw_diff[key.to_sym]) # rubocop:disable GitlabSecurity/PublicSend
|
2017-01-04 13:43:06 -05:00
|
|
|
end
|
2017-02-24 10:53:44 -05:00
|
|
|
end
|
|
|
|
|
2017-05-30 15:30:05 -04:00
|
|
|
def init_from_gitaly(diff)
|
2017-06-21 19:51:46 -04:00
|
|
|
@diff = encode!(diff.patch) if diff.respond_to?(:patch)
|
2017-05-30 15:30:05 -04:00
|
|
|
@new_path = encode!(diff.to_path.dup)
|
|
|
|
@old_path = encode!(diff.from_path.dup)
|
|
|
|
@a_mode = diff.old_mode.to_s(8)
|
|
|
|
@b_mode = diff.new_mode.to_s(8)
|
|
|
|
@new_file = diff.from_id == BLANK_SHA
|
|
|
|
@renamed_file = diff.from_path != diff.to_path
|
|
|
|
@deleted_file = diff.to_id == BLANK_SHA
|
2017-07-13 18:22:09 -04:00
|
|
|
|
|
|
|
collapse! if diff.respond_to?(:collapsed) && diff.collapsed
|
2017-02-24 10:53:44 -05:00
|
|
|
end
|
2017-01-04 13:43:06 -05:00
|
|
|
|
2017-05-26 19:27:30 -04:00
|
|
|
def prune_diff_if_eligible
|
|
|
|
if too_large?
|
|
|
|
too_large!
|
|
|
|
elsif collapsed?
|
|
|
|
collapse!
|
|
|
|
end
|
2017-01-04 13:43:06 -05:00
|
|
|
end
|
|
|
|
|
|
|
|
# If the patch surpasses any of the diff limits it calls the appropiate
|
|
|
|
# prune method and returns true. Otherwise returns false.
|
2017-05-26 19:27:30 -04:00
|
|
|
def prune_large_patch(patch)
|
2017-01-04 13:43:06 -05:00
|
|
|
size = 0
|
|
|
|
|
|
|
|
patch.each_hunk do |hunk|
|
|
|
|
hunk.each_line do |line|
|
|
|
|
size += line.content.bytesize
|
|
|
|
|
2017-10-02 08:51:51 -04:00
|
|
|
if size >= SIZE_LIMIT
|
2017-05-26 19:27:30 -04:00
|
|
|
too_large!
|
2017-01-04 13:43:06 -05:00
|
|
|
return true
|
|
|
|
end
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
2017-10-02 08:51:51 -04:00
|
|
|
if !expanded && size >= COLLAPSE_LIMIT
|
2017-05-26 19:27:30 -04:00
|
|
|
collapse!
|
2017-01-04 13:43:06 -05:00
|
|
|
return true
|
|
|
|
end
|
|
|
|
|
|
|
|
false
|
|
|
|
end
|
|
|
|
|
|
|
|
# Strip out the information at the beginning of the patch's text to match
|
|
|
|
# Grit's output
|
|
|
|
def strip_diff_headers(diff_text)
|
|
|
|
# Delete everything up to the first line that starts with '---' or
|
|
|
|
# 'Binary'
|
|
|
|
diff_text.sub!(/\A.*?^(---|Binary)/m, '\1')
|
|
|
|
|
|
|
|
if diff_text.start_with?('---', 'Binary')
|
|
|
|
diff_text
|
|
|
|
else
|
|
|
|
# If the diff_text did not contain a line starting with '---' or
|
|
|
|
# 'Binary', return the empty string. No idea why; we are just
|
|
|
|
# preserving behavior from before the refactor.
|
|
|
|
''
|
|
|
|
end
|
|
|
|
end
|
|
|
|
end
|
|
|
|
end
|
|
|
|
end
|