github_changelog_generator 1.15.0.pre.rc → 1.16.2

data/lib/github_changelog_generator/generator/section.rb

@@ -0,0 +1,124 @@
+# frozen_string_literal: true
+
+module GitHubChangelogGenerator
+  # This class generates the content for a single section of a changelog entry.
+  # It turns the tagged issues and PRs into a well-formatted list of changes to
+  # be later incorporated into a changelog entry.
+  #
+  # @see GitHubChangelogGenerator::Entry
+  class Section
+    # @return [String]
+    attr_accessor :name
+
+    # @return [String] a merge prefix, or an issue prefix
+    attr_reader :prefix
+
+    # @return [Array<Hash>]
+    attr_reader :issues
+
+    # @return [Array<String>]
+    attr_reader :labels
+
+    # @return [Boolean]
+    attr_reader :body_only
+
+    # @return [Options]
+    attr_reader :options
+
+    def initialize(opts = {})
+      @name = opts[:name]
+      @prefix = opts[:prefix]
+      @labels = opts[:labels] || []
+      @issues = opts[:issues] || []
+      @options = opts[:options] || Options.new({})
+      @body_only = opts[:body_only] || false
+      @entry = Entry.new(options)
+    end
+
+    # Returns the content of a section.
+    #
+    # @return [String] Generated section content
+    def generate_content
+      content = ""
+
+      if @issues.any?
+        content += "#{@prefix}\n\n" unless @options[:simple_list] || @prefix.blank?
+        @issues.each do |issue|
+          merge_string = get_string_for_issue(issue)
+          content += "- " unless @body_only
+          content += "#{merge_string}\n"
+        end
+        content += "\n"
+      end
+      content
+    end
+
+    private
+
+    # Parse issue and generate single line formatted issue line.
+    #
+    # Example output:
+    # - Add coveralls integration [\#223](https://github.com/github-changelog-generator/github-changelog-generator/pull/223) (@github-changelog-generator)
+    #
+    # @param [Hash] issue Fetched issue from GitHub
+    # @return [String] Markdown-formatted single issue
+    def get_string_for_issue(issue)
+      encapsulated_title = encapsulate_string issue["title"]
+
+      title_with_number = "#{encapsulated_title} [\\##{issue['number']}](#{issue['html_url']})"
+      title_with_number = "#{title_with_number}#{@entry.line_labels_for(issue)}" if @options[:issue_line_labels].present?
+      line = issue_line_with_user(title_with_number, issue)
+      issue_line_with_body(line, issue)
+    end
+
+    def issue_line_with_body(line, issue)
+      return issue["body"] if @body_only && issue["body"].present?
+      return line if !@options[:issue_line_body] || issue["body"].blank?
+
+      # get issue body till first line break
+      body_paragraph = body_till_first_break(issue["body"])
+      # remove spaces from beginning of the string
+      body_paragraph.rstrip!
+      # encapsulate to md
+      encapsulated_body = "  \n#{encapsulate_string(body_paragraph)}"
+
+      "**#{line}** #{encapsulated_body}"
+    end
+
+    def body_till_first_break(body)
+      body.split(/\n/, 2).first
+    end
+
+    def issue_line_with_user(line, issue)
+      return line if !@options[:author] || issue["pull_request"].nil?
+
+      user = issue["user"]
+      return "#{line} ({Null user})" unless user
+
+      if @options[:usernames_as_github_logins]
+        "#{line} (@#{user['login']})"
+      else
+        "#{line} ([#{user['login']}](#{user['html_url']}))"
+      end
+    end
+
+    ENCAPSULATED_CHARACTERS = %w(< > * _ \( \) [ ] #)
+
+    # Encapsulate characters to make Markdown look as expected.
+    #
+    # @param [String] string
+    # @return [String] encapsulated input string
+    def encapsulate_string(string)
+      string = string.gsub('\\', '\\\\')
+
+      ENCAPSULATED_CHARACTERS.each do |char|
+        # Only replace char with escaped version if it isn't inside backticks (markdown inline code).
+        # This relies on each opening '`' being closed (ie an even number in total).
+        # A char is *outside* backticks if there is an even number of backticks following it.
+        string = string.gsub(%r{#{Regexp.escape(char)}(?=([^`]*`[^`]*`)*[^`]*$)}, "\\#{char}")
+      end
+
+      string
+    end
+  end
+end

data/lib/github_changelog_generator/helper.rb

@@ -14,7 +14,7 @@ module GitHubChangelogGenerator
     @log ||= if test?
                Logger.new(nil) # don't show any logs when running tests
              else
-               Logger.new(STDOUT)
+               Logger.new($stdout)
              end
     @log.formatter = proc do |severity, _datetime, _progname, msg|
       string = "#{msg}\n"

data/lib/github_changelog_generator/octo_fetcher.rb

@@ -2,6 +2,12 @@
 
 require "tmpdir"
 require "retriable"
+require "set"
+require "async"
+require "async/barrier"
+require "async/semaphore"
+require "async/http/faraday"
+
 module GitHubChangelogGenerator
   # A Fetcher responsible for all requests to GitHub and all basic manipulation with related data
   # (such as filtering, validating, e.t.c)
@@ -9,14 +15,14 @@ module GitHubChangelogGenerator
   # Example:
   # fetcher = GitHubChangelogGenerator::OctoFetcher.new(options)
   class OctoFetcher
-    PER_PAGE_NUMBER   = 100
-    MAX_THREAD_NUMBER = 25
+    PER_PAGE_NUMBER = 100
+    MAXIMUM_CONNECTIONS = 50
     MAX_FORBIDDEN_RETRIES = 100
     CHANGELOG_GITHUB_TOKEN = "CHANGELOG_GITHUB_TOKEN"
-    GH_RATE_LIMIT_EXCEEDED_MSG = "Warning: Can't finish operation: GitHub API rate limit exceeded, change log may be " \
+    GH_RATE_LIMIT_EXCEEDED_MSG = "Warning: Can't finish operation: GitHub API rate limit exceeded, changelog may be " \
     "missing some issues. You can limit the number of issues fetched using the `--max-issues NUM` argument."
     NO_TOKEN_PROVIDED = "Warning: No token provided (-t option) and variable $CHANGELOG_GITHUB_TOKEN was not found. " \
-    "This script can make only 50 requests to GitHub API per hour without token!"
+    "This script can make only 50 requests to GitHub API per hour without a token!"
 
     # @param options [Hash] Options passed in
     # @option options [String] :user GitHub username
@@ -31,44 +37,58 @@ module GitHubChangelogGenerator
       @project      = @options[:project]
       @since        = @options[:since]
       @http_cache   = @options[:http_cache]
-      @cache_file   = nil
-      @cache_log    = nil
-      prepare_cache
-      configure_octokit_ssl
-      @client = Octokit::Client.new(github_options)
+      @commits      = []
+      @branches     = nil
+      @graph        = nil
+      @client = nil
+      @commits_in_tag_cache = {}
     end
 
-    def prepare_cache
-      return unless @http_cache
-      @cache_file = @options.fetch(:cache_file) { File.join(Dir.tmpdir, "github-changelog-http-cache") }
-      @cache_log  = @options.fetch(:cache_log) { File.join(Dir.tmpdir, "github-changelog-logger.log") }
-      init_cache
-    end
+    def middleware
+      Faraday::RackBuilder.new do |builder|
+        if @http_cache
+          cache_file = @options.fetch(:cache_file) { File.join(Dir.tmpdir, "github-changelog-http-cache") }
+          cache_log  = @options.fetch(:cache_log) { File.join(Dir.tmpdir, "github-changelog-logger.log") }
+
+          builder.use(
+            Faraday::HttpCache,
+            serializer: Marshal,
+            store: ActiveSupport::Cache::FileStore.new(cache_file),
+            logger: Logger.new(cache_log),
+            shared_cache: false
+          )
+        end
 
-    def github_options
-      result = {}
-      github_token = fetch_github_token
-      result[:access_token] = github_token if github_token
-      endpoint = @options[:github_endpoint]
-      result[:api_endpoint] = endpoint if endpoint
-      result
+        builder.use Octokit::Response::RaiseError
+        builder.adapter :async_http
+      end
     end
 
-    def configure_octokit_ssl
-      ca_file = @options[:ssl_ca_file] || ENV["SSL_CA_FILE"] || File.expand_path("../ssl_certs/cacert.pem", __FILE__)
-      Octokit.connection_options = { ssl: { ca_file: ca_file } }
+    def connection_options
+      ca_file = @options[:ssl_ca_file] || ENV["SSL_CA_FILE"] || File.expand_path("ssl_certs/cacert.pem", __dir__)
+
+      Octokit.connection_options.merge({ ssl: { ca_file: ca_file } })
     end
 
-    def init_cache
-      Octokit.middleware = Faraday::RackBuilder.new do |builder|
-        builder.use(Faraday::HttpCache, serializer: Marshal,
-                                        store: ActiveSupport::Cache::FileStore.new(@cache_file),
-                                        logger: Logger.new(@cache_log),
-                                        shared_cache: false)
-        builder.use Octokit::Response::RaiseError
-        builder.adapter Faraday.default_adapter
-        # builder.response :logger
+    def client_options
+      options = {
+        middleware: middleware,
+        connection_options: connection_options
+      }
+
+      if (github_token = fetch_github_token)
+        options[:access_token] = github_token
       end
+
+      if (endpoint = @options[:github_endpoint])
+        options[:api_endpoint] = endpoint
+      end
+
+      options
+    end
+
+    def client
+      @client ||= Octokit::Client.new(client_options)
     end
 
     DEFAULT_REQUEST_OPTIONS = { per_page: PER_PAGE_NUMBER }
@@ -85,6 +105,9 @@ module GitHubChangelogGenerator
     # Returns the number of pages for a API call
     #
     # @return [Integer] number of pages for this API call in total
+    # @param [Object] request_options
+    # @param [Object] method
+    # @param [Object] client
     def calculate_pages(client, method, request_options)
       # Makes the first API call so that we can call last_response
       check_github_response do
@@ -104,11 +127,11 @@ module GitHubChangelogGenerator
     #
     # @return [Array <Hash>] array of tags in repo
     def github_fetch_tags
-      tags        = []
-      page_i      = 0
-      count_pages = calculate_pages(@client, "tags", {})
+      tags = []
+      page_i = 0
+      count_pages = calculate_pages(client, "tags", {})
 
-      iterate_pages(@client, "tags") do |new_tags|
+      iterate_pages(client, "tags") do |new_tags|
         page_i += PER_PAGE_NUMBER
         print_in_same_line("Fetching tags... #{page_i}/#{count_pages * PER_PAGE_NUMBER}")
         tags.concat(new_tags)
@@ -138,10 +161,10 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
     def fetch_closed_issues_and_pr
       print "Fetching closed issues...\r" if @options[:verbose]
       issues = []
-      page_i      = 0
-      count_pages = calculate_pages(@client, "issues", closed_pr_options)
+      page_i = 0
+      count_pages = calculate_pages(client, "issues", closed_pr_options)
 
-      iterate_pages(@client, "issues", closed_pr_options) do |new_issues|
+      iterate_pages(client, "issues", **closed_pr_options) do |new_issues|
         page_i += PER_PAGE_NUMBER
         print_in_same_line("Fetching issues... #{page_i}/#{count_pages * PER_PAGE_NUMBER}")
         issues.concat(new_issues)
@@ -162,14 +185,10 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
       pull_requests = []
       options = { state: "closed" }
 
-      unless @options[:release_branch].nil?
-        options[:base] = @options[:release_branch]
-      end
-
-      page_i      = 0
-      count_pages = calculate_pages(@client, "pull_requests", options)
+      page_i = 0
+      count_pages = calculate_pages(client, "pull_requests", options)
 
-      iterate_pages(@client, "pull_requests", options) do |new_pr|
+      iterate_pages(client, "pull_requests", **options) do |new_pr|
         page_i += PER_PAGE_NUMBER
         log_string = "Fetching merged dates... #{page_i}/#{count_pages * PER_PAGE_NUMBER}"
         print_in_same_line(log_string)
@@ -186,14 +205,20 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
     # @param [Array] issues
     # @return [Void]
     def fetch_events_async(issues)
-      i       = 0
-      threads = []
+      i = 0
+      # Add accept option explicitly for disabling the warning of preview API.
+      preview = { accept: Octokit::Preview::PREVIEW_TYPES[:project_card_events] }
+
+      barrier = Async::Barrier.new
+      semaphore = Async::Semaphore.new(MAXIMUM_CONNECTIONS, parent: barrier)
+
+      Sync do
+        client = self.client
 
-      issues.each_slice(MAX_THREAD_NUMBER) do |issues_slice|
-        issues_slice.each do |issue|
-          threads << Thread.new do
+        issues.each do |issue|
+          semaphore.async do
             issue["events"] = []
-            iterate_pages(@client, "issue_events", issue["number"]) do |new_event|
+            iterate_pages(client, "issue_events", issue["number"], **preview) do |new_event|
               issue["events"].concat(new_event)
             end
             issue["events"] = issue["events"].map { |event| stringify_keys_deep(event.to_hash) }
@@ -201,23 +226,50 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
             i += 1
           end
         end
-        threads.each(&:join)
-        threads = []
-      end
 
-      # to clear line from prev print
-      print_empty_line
+        barrier.wait
+
+        # to clear line from prev print
+        print_empty_line
+      end
 
       Helper.log.info "Fetching events for issues and PR: #{i}"
     end
 
+    # Fetch comments for PRs and add them to "comments"
+    #
+    # @param [Array] prs The array of PRs.
+    # @return [Void] No return; PRs are updated in-place.
+    def fetch_comments_async(prs)
+      barrier = Async::Barrier.new
+      semaphore = Async::Semaphore.new(MAXIMUM_CONNECTIONS, parent: barrier)
+
+      Sync do
+        client = self.client
+
+        prs.each do |pr|
+          semaphore.async do
+            pr["comments"] = []
+            iterate_pages(client, "issue_comments", pr["number"]) do |new_comment|
+              pr["comments"].concat(new_comment)
+            end
+            pr["comments"] = pr["comments"].map { |comment| stringify_keys_deep(comment.to_hash) }
+          end
+        end
+
+        barrier.wait
+      end
+
+      nil
+    end
+
     # Fetch tag time from repo
     #
     # @param [Hash] tag GitHub data item about a Tag
     #
     # @return [Time] time of specified tag
     def fetch_date_of_tag(tag)
-      commit_data = check_github_response { @client.commit(user_project, tag["commit"]["sha"]) }
+      commit_data = fetch_commit(tag["commit"]["sha"])
       commit_data = stringify_keys_deep(commit_data.to_hash)
 
       commit_data["commit"]["committer"]["date"]
@@ -225,28 +277,121 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
 
     # Fetch commit for specified event
     #
+    # @param [String] commit_id the SHA of a commit to fetch
     # @return [Hash]
-    def fetch_commit(event)
-      check_github_response do
-        commit = @client.commit(user_project, event["commit_id"])
-        commit = stringify_keys_deep(commit.to_hash)
-        commit
+    def fetch_commit(commit_id)
+      found = commits.find do |commit|
+        commit["sha"] == commit_id
+      end
+      if found
+        stringify_keys_deep(found.to_hash)
+      else
+        client = self.client
+
+        # cache miss; don't add to @commits because unsure of order.
+        check_github_response do
+          commit = client.commit(user_project, commit_id)
+          commit = stringify_keys_deep(commit.to_hash)
+          commit
+        end
       end
     end
 
-    # Fetch all commits before certain point
+    # Fetch all commits
     #
-    # @return [String]
-    def commits_before(start_time)
-      commits = []
-      iterate_pages(@client, "commits_before", start_time.to_datetime.to_s) do |new_commits|
-        commits.concat(new_commits)
+    # @return [Array] Commits in a repo.
+    def commits
+      if @commits.empty?
+        Sync do
+          barrier = Async::Barrier.new
+          semaphore = Async::Semaphore.new(MAXIMUM_CONNECTIONS, parent: barrier)
+
+          if (since_commit = @options[:since_commit])
+            iterate_pages(client, "commits_since", since_commit, parent: semaphore) do |new_commits|
+              @commits.concat(new_commits)
+            end
+          else
+            iterate_pages(client, "commits", parent: semaphore) do |new_commits|
+              @commits.concat(new_commits)
+            end
+          end
+
+          barrier.wait
+
+          @commits.sort! do |b, a|
+            a[:commit][:author][:date] <=> b[:commit][:author][:date]
+          end
+        end
+      end
+      @commits
+    end
+
+    # Return the oldest commit in a repo
+    #
+    # @return [Hash] Oldest commit in the github git history.
+    def oldest_commit
+      commits.last
+    end
+
+    # @return [String] Default branch of the repo
+    def default_branch
+      @default_branch ||= client.repository(user_project)[:default_branch]
+    end
+
+    # @param [String] name
+    # @return [Array<String>]
+    def commits_in_branch(name)
+      @branches ||= client.branches(user_project).map { |branch| [branch[:name], branch] }.to_h
+
+      if (branch = @branches[name])
+        commits_in_tag(branch[:commit][:sha])
+      else
+        []
+      end
+    end
+
+    # Fetch all SHAs occurring in or before a given tag and add them to
+    # "shas_in_tag"
+    #
+    # @param [Array] tags The array of tags.
+    # @return void
+    def fetch_tag_shas(tags)
+      # Reverse the tags array to gain max benefit from the @commits_in_tag_cache
+      tags.reverse_each do |tag|
+        tag["shas_in_tag"] = commits_in_tag(tag["commit"]["sha"])
       end
-      commits
     end
 
     private
 
+    # @param [Set] shas
+    # @param [Object] sha
+    def commits_in_tag(sha, shas = Set.new)
+      # Reduce multiple runs for the same tag
+      return @commits_in_tag_cache[sha] if @commits_in_tag_cache.key?(sha)
+
+      @graph ||= commits.map { |commit| [commit[:sha], commit] }.to_h
+      return shas unless (current = @graph[sha])
+
+      queue = [current]
+      while queue.any?
+        commit = queue.shift
+        # If we've already processed this sha, just grab it's parents from the cache
+        if @commits_in_tag_cache.key?(commit[:sha])
+          shas.merge(@commits_in_tag_cache[commit[:sha]])
+        else
+          shas.add(commit[:sha])
+          commit[:parents].each do |p|
+            queue.push(@graph[p[:sha]]) unless shas.include?(p[:sha])
+          end
+        end
+      end
+
+      @commits_in_tag_cache[sha] = shas
+      shas
+    end
+
+    # @param [Object] indata
     def stringify_keys_deep(indata)
       case indata
       when Array
@@ -270,43 +415,49 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
     #
     # @param [Octokit::Client] client
     # @param [String] method (eg. 'tags')
+    # @param [Array] arguments
+    # @param [Async::Semaphore] parent
     #
     # @yield [Sawyer::Resource] An OctoKit-provided response (which can be empty)
     #
     # @return [void]
-    def iterate_pages(client, method, *args)
-      args << DEFAULT_REQUEST_OPTIONS.merge(extract_request_args(args))
+    # @param [Hash] options
+    def iterate_pages(client, method, *arguments, parent: nil, **options)
+      options = DEFAULT_REQUEST_OPTIONS.merge(options)
 
-      check_github_response { client.send(method, user_project, *args) }
+      check_github_response { client.send(method, user_project, *arguments, **options) }
       last_response = client.last_response.tap do |response|
         raise(MovedPermanentlyError, response.data[:url]) if response.status == 301
       end
 
       yield(last_response.data)
 
-      until (next_one = last_response.rels[:next]).nil?
-        last_response = check_github_response { next_one.get }
-        yield(last_response.data)
-      end
-    end
-
-    def extract_request_args(args)
-      if args.size == 1 && args.first.is_a?(Hash)
-        args.delete_at(0)
-      elsif args.size > 1 && args.last.is_a?(Hash)
-        args.delete_at(args.length - 1)
-      else
-        {}
+      if parent.nil?
+        # The snail visits one leaf at a time:
+        until (next_one = last_response.rels[:next]).nil?
+          last_response = check_github_response { next_one.get }
+          yield(last_response.data)
+        end
+      elsif (last = last_response.rels[:last])
+        # OR we bring out the gatling gun:
+        parameters = querystring_as_hash(last.href)
+        last_page = Integer(parameters["page"])
+
+        (2..last_page).each do |page|
+          parent.async do
+            data = check_github_response { client.send(method, user_project, *arguments, page: page, **options) }
+            yield data
+          end
+        end
       end
     end
 
     # This is wrapper with rescue block
     #
     # @return [Object] returns exactly the same, what you put in the block, but wrap it with begin-rescue block
-    def check_github_response
-      Retriable.retriable(retry_options) do
-        yield
-      end
+    # @param [Proc] block
+    def check_github_response(&block)
+      Retriable.retriable(retry_options, &block)
     rescue MovedPermanentlyError => e
       fail_with_message(e, "The repository has moved, update your configuration")
     rescue Octokit::Forbidden => e
@@ -316,8 +467,10 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
     end
 
     # Presents the exception, and the aborts with the message.
-    def fail_with_message(e, message)
-      Helper.log.error("#{e.class}: #{e.message}")
+    # @param [Object] message
+    # @param [Object] error
+    def fail_with_message(error, message)
+      Helper.log.error("#{error.class}: #{error.message}")
       sys_abort(message)
     end
 
@@ -342,10 +495,11 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
         Helper.log.warn("RETRY - #{exception.class}: '#{exception.message}'")
         Helper.log.warn("#{try} tries in #{elapsed_time} seconds and #{next_interval} seconds until the next try")
         Helper.log.warn GH_RATE_LIMIT_EXCEEDED_MSG
-        Helper.log.warn @client.rate_limit
+        Helper.log.warn(client.rate_limit)
       end
     end
 
+    # @param [Object] msg
     def sys_abort(msg)
       abort(msg)
     end
@@ -354,7 +508,7 @@ Make sure, that you push tags to remote repo via 'git push --tags'"
     #
     # @param [String] log_string
     def print_in_same_line(log_string)
-      print log_string + "\r"
+      print "#{log_string}\r"
     end
 
     # Print long line with spaces on same line to clear prev message