debian-mirror-gitlab/app/models/merge_request_diff.rb

324 lines
8.5 KiB
Ruby
Raw Normal View History

2014-09-02 18:07:02 +05:30
class MergeRequestDiff < ActiveRecord::Base
2015-04-26 12:48:37 +05:30
include Sortable
2016-06-22 15:30:34 +05:30
include Importable
2017-08-17 22:00:37 +05:30
include Gitlab::Git::EncodingHelper
2015-04-26 12:48:37 +05:30
2015-09-25 12:07:36 +05:30
# Prevent store of diff if commits amount more then 500
2016-06-02 11:05:42 +05:30
COMMITS_SAFE_SIZE = 100
2014-09-02 18:07:02 +05:30
2016-11-03 12:29:30 +05:30
# Valid types of serialized diffs allowed by Gitlab::Git::Diff
2017-08-17 22:00:37 +05:30
VALID_CLASSES = [Hash, Rugged::Patch, Rugged::Diff::Delta].freeze
2016-11-03 12:29:30 +05:30
2014-09-02 18:07:02 +05:30
belongs_to :merge_request
2017-08-17 22:00:37 +05:30
serialize :st_commits
serialize :st_diffs
2014-09-02 18:07:02 +05:30
state_machine :state, initial: :empty do
state :collected
2016-06-02 11:05:42 +05:30
state :overflow
# Deprecated states: these are no longer used but these values may still occur
# in the database.
2014-09-02 18:07:02 +05:30
state :timeout
state :overflow_commits_safe_size
state :overflow_diff_files_limit
state :overflow_diff_lines_limit
end
2017-08-17 22:00:37 +05:30
scope :viewable, -> { without_state(:empty) }
2014-09-02 18:07:02 +05:30
2016-09-29 09:46:39 +05:30
# All diff information is collected from repository after object is created.
# It allows you to override variables like head_commit_sha before getting diff.
after_create :save_git_content, unless: :importing?
2017-08-17 22:00:37 +05:30
def self.find_by_diff_refs(diff_refs)
find_by(start_commit_sha: diff_refs.start_sha, head_commit_sha: diff_refs.head_sha, base_commit_sha: diff_refs.base_sha)
end
2016-09-29 09:46:39 +05:30
def self.select_without_diff
select(column_names - ['st_diffs'])
end
def st_commits
super || []
end
2014-09-02 18:07:02 +05:30
2016-09-29 09:46:39 +05:30
# Collect information about commits and diff from repository
# and save it to the database as serialized data
def save_git_content
ensure_commits_sha
save_commits
2014-09-02 18:07:02 +05:30
reload_commits
2016-09-29 09:46:39 +05:30
save_diffs
keep_around_commits
end
def ensure_commits_sha
merge_request.fetch_ref
self.start_commit_sha ||= merge_request.target_branch_sha
self.head_commit_sha ||= merge_request.source_branch_sha
self.base_commit_sha ||= find_base_sha
save
end
# Override head_commit_sha to keep compatibility with merge request diff
# created before version 8.4 that does not store head_commit_sha in separate db field.
def head_commit_sha
if persisted? && super.nil?
last_commit.try(:sha)
else
super
end
end
# This method will rely on repository branch sha
# in case start_commit_sha is nil. Its necesarry for old merge request diff
# created before version 8.4 to work
def safe_start_commit_sha
start_commit_sha || merge_request.target_branch_sha
2014-09-02 18:07:02 +05:30
end
2016-06-02 11:05:42 +05:30
def size
2016-09-13 17:45:13 +05:30
real_size.presence || raw_diffs.size
2014-09-02 18:07:02 +05:30
end
2016-09-13 17:45:13 +05:30
def raw_diffs(options = {})
2016-06-02 11:05:42 +05:30
if options[:ignore_whitespace_change]
2016-09-29 09:46:39 +05:30
@diffs_no_whitespace ||=
Gitlab::Git::Compare.new(
2016-08-24 12:49:21 +05:30
repository.raw_repository,
2016-09-29 09:46:39 +05:30
safe_start_commit_sha,
head_commit_sha).diffs(options)
2016-06-02 11:05:42 +05:30
else
2016-09-13 17:45:13 +05:30
@raw_diffs ||= {}
@raw_diffs[options] ||= load_diffs(st_diffs, options)
2016-06-02 11:05:42 +05:30
end
2015-11-26 14:37:03 +05:30
end
2014-09-02 18:07:02 +05:30
def commits
2016-09-29 09:46:39 +05:30
@commits ||= load_commits(st_commits)
end
def reload_commits
@commits = nil
commits
2014-09-02 18:07:02 +05:30
end
def last_commit
commits.first
end
2015-10-24 18:46:33 +05:30
def first_commit
commits.last
end
2016-01-29 22:53:50 +05:30
def base_commit
2016-09-29 09:46:39 +05:30
return unless base_commit_sha
2016-01-29 22:53:50 +05:30
2016-09-29 09:46:39 +05:30
project.commit(base_commit_sha)
2016-01-29 22:53:50 +05:30
end
2016-08-24 12:49:21 +05:30
def start_commit
2016-09-29 09:46:39 +05:30
return unless start_commit_sha
2014-09-02 18:07:02 +05:30
2016-09-29 09:46:39 +05:30
project.commit(start_commit_sha)
2014-09-02 18:07:02 +05:30
end
2016-08-24 12:49:21 +05:30
def head_commit
2016-09-29 09:46:39 +05:30
return unless head_commit_sha
project.commit(head_commit_sha)
end
def commits_sha
2017-08-17 22:00:37 +05:30
st_commits.map { |commit| commit[:id] }
end
def diff_refs=(new_diff_refs)
self.base_commit_sha = new_diff_refs&.base_sha
self.start_commit_sha = new_diff_refs&.start_sha
self.head_commit_sha = new_diff_refs&.head_sha
2016-09-29 09:46:39 +05:30
end
2014-09-02 18:07:02 +05:30
2016-09-29 09:46:39 +05:30
def diff_refs
return unless start_commit_sha || base_commit_sha
Gitlab::Diff::DiffRefs.new(
base_sha: base_commit_sha,
start_sha: start_commit_sha,
head_sha: head_commit_sha
)
2014-09-02 18:07:02 +05:30
end
2016-09-13 17:45:13 +05:30
def diff_refs_by_sha?
base_commit_sha? && head_commit_sha? && start_commit_sha?
end
2016-09-29 09:46:39 +05:30
def diffs(diff_options = nil)
Gitlab::Diff::FileCollection::MergeRequestDiff.new(self, diff_options: diff_options)
2014-09-02 18:07:02 +05:30
end
2016-09-29 09:46:39 +05:30
def project
merge_request.target_project
end
2016-08-24 12:49:21 +05:30
2016-09-29 09:46:39 +05:30
def compare
@compare ||=
Gitlab::Git::Compare.new(
repository.raw_repository,
safe_start_commit_sha,
head_commit_sha
)
end
2014-09-02 18:07:02 +05:30
2016-09-29 09:46:39 +05:30
def latest?
self == merge_request.merge_request_diff
end
2014-09-02 18:07:02 +05:30
2016-11-03 12:29:30 +05:30
def compare_with(sha, straight: true)
# When compare merge request versions we want diff A..B instead of A...B
# so we handle cases when user does squash and rebase of the commits between versions.
# For this reason we set straight to true by default.
2017-08-17 22:00:37 +05:30
CompareService.new(project, head_commit_sha)
.execute(project, sha, straight: straight)
end
def commits_count
st_commits.count
end
def utf8_st_diffs
return [] if st_diffs.blank?
st_diffs.map do |diff|
diff.each do |k, v|
diff[k] = encode_utf8(v) if v.respond_to?(:encoding)
end
end
2014-09-02 18:07:02 +05:30
end
2016-09-29 09:46:39 +05:30
private
2016-11-03 12:29:30 +05:30
# Old GitLab implementations may have generated diffs as ["--broken-diff"].
# Avoid an error 500 by ignoring bad elements. See:
# https://gitlab.com/gitlab-org/gitlab-ce/issues/20776
def valid_raw_diff?(raw)
return false unless raw.respond_to?(:each)
raw.any? { |element| VALID_CLASSES.include?(element.class) }
end
2016-08-24 12:49:21 +05:30
def dump_commits(commits)
commits.map(&:to_hash)
end
def load_commits(array)
array.map { |hash| Commit.new(Gitlab::Git::Commit.new(hash), merge_request.source_project) }
end
2016-09-29 09:46:39 +05:30
# Load all commits related to current merge request diff from repo
2014-09-02 18:07:02 +05:30
# and save it as array of hashes in st_commits db field
2016-09-29 09:46:39 +05:30
def save_commits
2016-08-24 12:49:21 +05:30
new_attributes = {}
2016-09-29 09:46:39 +05:30
commits = compare.commits
2014-09-02 18:07:02 +05:30
2016-09-29 09:46:39 +05:30
if commits.present?
commits = Commit.decorate(commits, merge_request.source_project).reverse
new_attributes[:st_commits] = dump_commits(commits)
2014-09-02 18:07:02 +05:30
end
2016-08-24 12:49:21 +05:30
update_columns_serialized(new_attributes)
end
def dump_diffs(diffs)
if diffs.respond_to?(:map)
diffs.map(&:to_hash)
end
end
def load_diffs(raw, options)
2016-11-03 12:29:30 +05:30
if valid_raw_diff?(raw)
2016-08-24 12:49:21 +05:30
if paths = options[:paths]
raw = raw.select do |diff|
paths.include?(diff[:old_path]) || paths.include?(diff[:new_path])
end
end
Gitlab::Git::DiffCollection.new(raw, options)
else
Gitlab::Git::DiffCollection.new([])
end
2014-09-02 18:07:02 +05:30
end
2016-09-29 09:46:39 +05:30
# Load diffs between branches related to current merge request diff from repo
2014-09-02 18:07:02 +05:30
# and save it as array of hashes in st_diffs db field
2016-09-29 09:46:39 +05:30
def save_diffs
2016-08-24 12:49:21 +05:30
new_attributes = {}
2014-09-02 18:07:02 +05:30
if commits.size.zero?
2016-08-24 12:49:21 +05:30
new_attributes[:state] = :empty
2014-09-02 18:07:02 +05:30
else
2016-09-29 09:46:39 +05:30
diff_collection = compare.diffs(Commit.max_diff_options)
2016-08-24 12:49:21 +05:30
new_attributes[:real_size] = diff_collection.real_size
2014-09-02 18:07:02 +05:30
2016-06-02 11:05:42 +05:30
if diff_collection.any?
new_diffs = dump_diffs(diff_collection)
2016-08-24 12:49:21 +05:30
new_attributes[:state] = :collected
2016-06-02 11:05:42 +05:30
end
2017-08-17 22:00:37 +05:30
new_attributes[:st_diffs] = new_diffs || []
# Set our state to 'overflow' to make the #empty? and #collected?
# methods (generated by StateMachine) return false.
#
# This attribution has to come at the end of the method so 'overflow'
# state does not get overridden by 'collected'.
new_attributes[:state] = :overflow if diff_collection.overflow?
2014-09-02 18:07:02 +05:30
end
2016-08-24 12:49:21 +05:30
update_columns_serialized(new_attributes)
2014-09-02 18:07:02 +05:30
end
def repository
2016-08-24 12:49:21 +05:30
project.repository
2014-09-02 18:07:02 +05:30
end
2016-09-29 09:46:39 +05:30
def find_base_sha
return unless head_commit_sha && start_commit_sha
2014-09-02 18:07:02 +05:30
2016-09-29 09:46:39 +05:30
project.merge_base_commit(head_commit_sha, start_commit_sha).try(:sha)
2016-06-02 11:05:42 +05:30
end
2016-08-24 12:49:21 +05:30
#
# #save or #update_attributes providing changes on serialized attributes do a lot of
# serialization and deserialization calls resulting in bad performance.
# Using #update_columns solves the problem with just one YAML.dump per serialized attribute that we provide.
# As a tradeoff we need to reload the current instance to properly manage time objects on those serialized
# attributes. So to keep the same behaviour as the attribute assignment we reload the instance.
# The difference is in the usage of
# #write_attribute= (#update_attributes) and #raw_write_attribute= (#update_columns)
#
# Ex:
#
# new_attributes[:st_commits].first.slice(:committed_date)
# => {:committed_date=>2014-02-27 11:01:38 +0200}
# YAML.load(YAML.dump(new_attributes[:st_commits].first.slice(:committed_date)))
# => {:committed_date=>2014-02-27 10:01:38 +0100}
#
def update_columns_serialized(new_attributes)
return unless new_attributes.any?
update_columns(new_attributes.merge(updated_at: current_time_from_proper_timezone))
reload
end
def keep_around_commits
2017-08-17 22:00:37 +05:30
[repository, merge_request.source_project.repository].each do |repo|
repo.keep_around(start_commit_sha)
repo.keep_around(head_commit_sha)
repo.keep_around(base_commit_sha)
end
2014-09-02 18:07:02 +05:30
end
end