jirametrics 2.22 → 2.30

data/lib/jirametrics/download_config.rb

@@ -25,10 +25,25 @@ class DownloadConfig
     @no_earlier_than
   end
 
+  def github_repos
+    @github_repos ||= []
+  end
+
+  def github_repo *repos
+    github_repos.concat(repos.map { |r| normalize_github_repo(r) })
+  end
+
   def start_date today:
     date = today.to_date - @rolling_date_count if @rolling_date_count
     date = [date, @no_earlier_than].max if date && @no_earlier_than
     date = @no_earlier_than if date.nil? && @no_earlier_than
     date
   end
+
+  private
+
+  def normalize_github_repo repo
+    match = repo.match(%r{github\.com/([^/]+/[^/]+?)/?$})
+    match ? match[1] : repo
+  end
 end

data/lib/jirametrics/downloader.rb

@@ -33,21 +33,23 @@ class Downloader
   # For testing only
   attr_reader :start_date_in_query, :board_id_to_filter_id
 
-  def self.create download_config:, file_system:, jira_gateway:
+  def self.create download_config:, file_system:, jira_gateway:, github_pr_cache: {}
     is_cloud = jira_gateway.settings['jira_cloud'] || jira_gateway.cloud?
     (is_cloud ? DownloaderForCloud : DownloaderForDataCenter).new(
       download_config: download_config,
       file_system: file_system,
-      jira_gateway: jira_gateway
+      jira_gateway: jira_gateway,
+      github_pr_cache: github_pr_cache
     )
   end
 
-  def initialize download_config:, file_system:, jira_gateway:
+  def initialize download_config:, file_system:, jira_gateway:, github_pr_cache: {}
     @metadata = {}
     @download_config = download_config
     @target_path = @download_config.project_config.target_path
     @file_system = file_system
     @jira_gateway = jira_gateway
+    @github_pr_cache = github_pr_cache
     @board_id_to_filter_id = {}
 
     @issue_keys_downloaded_in_current_run = []
@@ -72,17 +74,36 @@ class Downloader
     download_statuses
     find_board_ids.each do |id|
       board = download_board_configuration board_id: id
+      board.project_config = @download_config.project_config
       download_issues board: board
     end
     download_users
 
     save_metadata
+    download_github_prs if @download_config.github_repos.any?
   end
 
   def log text, both: false
     @file_system.log text, also_write_to_stderr: both
   end
 
+  def log_start text
+    @file_system.log_start text
+  end
+
+  def start_progress
+    @file_system.start_progress
+  end
+
+  def progress_dot message = nil
+    @file_system.log message if message
+    @file_system.progress_dot
+  end
+
+  def end_progress
+    @file_system.end_progress
+  end
+
   def find_board_ids
     ids = @download_config.project_config.board_configs.collect(&:id)
     raise 'Board ids must be specified' if ids.empty?
@@ -165,11 +186,28 @@ class Downloader
     # actually look at the returned json.
     @board_id_to_filter_id[board_id] = json['filter']['id'].to_i
 
+    if json['type'] == 'simple'
+      features_json = download_features board_id: board_id
+      if BoardFeature.from_raw(features_json).any? { |f| f.name == 'jsw.agility.sprints' && f.enabled? }
+        download_sprints board_id: board_id
+      end
+    end
     download_sprints board_id: board_id if json['type'] == 'scrum'
     # TODO: Should be passing actual statuses, not empty list
     Board.new raw: json, possible_statuses: StatusCollection.new
   end
 
+  def download_features board_id:
+    log "  Downloading features for board #{board_id}", both: true
+    json = @jira_gateway.call_url relative_url: "/rest/agile/1.0/board/#{board_id}/features"
+
+    @file_system.save_json(
+      json: json,
+      filename: File.join(@target_path, "#{file_prefix}_board_#{board_id}_features.json")
+    )
+    json
+  end
+
   def download_sprints board_id:
     log "  Downloading sprints for board #{board_id}", both: true
     max_results = 100
@@ -211,19 +249,29 @@ class Downloader
         value = Date.parse(value) if value.is_a?(String) && value =~ /^\d{4}-\d{2}-\d{2}$/
         @metadata[key] = value
       end
+
     end
 
     # Even if this is the old format, we want to obey this one tag
     @metadata['no-download'] = hash['no-download'] if hash['no-download']
   end
 
+  def timezone_offset
+    @download_config.project_config.exporter.timezone_offset
+  end
+
+  def today_in_project_timezone
+    Time.now.getlocal(timezone_offset).to_date
+  end
+
   def save_metadata
     @metadata['version'] = CURRENT_METADATA_VERSION
+    @metadata['rolling_date_count'] = @download_config.rolling_date_count
     @metadata['date_start_from_last_query'] = @start_date_in_query if @start_date_in_query
 
     if @download_date_range.nil?
       log "Making up a date range in meta since one wasn't specified. You'll want to change that.", both: true
-      today = Date.today
+      today = today_in_project_timezone
       @download_date_range = (today - 7)..today
     end
 
@@ -258,7 +306,8 @@ class Downloader
     end
   end
 
-  def make_jql filter_id:, today: Date.today
+  def make_jql filter_id:, today: nil
+    today ||= today_in_project_timezone
     segments = []
     segments << "filter=#{filter_id}"
 
@@ -283,6 +332,39 @@ class Downloader
     segments.join ' AND '
   end
 
+  def download_github_prs
+    project_keys = extract_project_keys_from_downloaded_issues
+    if project_keys.empty?
+      log '  No project keys found in downloaded issues, skipping GitHub PR download', both: true
+      return
+    end
+
+    prs = @download_config.github_repos.flat_map do |repo|
+      GithubGateway.new(
+        repo: repo,
+        project_keys: project_keys,
+        file_system: @file_system,
+        raw_pr_cache: @github_pr_cache
+      ).fetch_pull_requests(since: @download_date_range&.begin)
+    end
+
+    @file_system.save_json(
+      json: prs.map(&:raw),
+      filename: File.join(@target_path, "#{file_prefix}_github_prs.json")
+    )
+  end
+
+  def extract_project_keys_from_downloaded_issues
+    path = File.join(@target_path, "#{file_prefix}_issues")
+    return [] unless @file_system.dir_exist?(path)
+
+    keys = []
+    @file_system.foreach(path) do |filename|
+      keys << filename.split('-').first if filename.match?(/^[A-Z][A-Z_0-9]+-\d+-\d+\.json$/)
+    end
+    keys.uniq
+  end
+
   def file_prefix
     @download_config.project_config.get_file_prefix
   end

data/lib/jirametrics/downloader_for_cloud.rb

@@ -5,6 +5,45 @@ class DownloaderForCloud < Downloader
     'Jira Cloud'
   end
 
+  def run
+    super
+    download_fix_versions
+  end
+
+  def download_board_configuration board_id:
+    board = super
+    location = board.raw['location']
+    @project_key ||= location['key'] if location&.[]('type') == 'project'
+    board
+  end
+
+  def download_fix_versions
+    return unless @project_key
+
+    log "  Downloading fix versions for project #{@project_key}", both: true
+    max_results = 50
+    start_at = 0
+    all_versions = []
+
+    loop do
+      json = @jira_gateway.call_url(
+        relative_url: "/rest/api/3/project/#{@project_key}/version?" \
+          "startAt=#{start_at}&maxResults=#{max_results}"
+      )
+
+      values = json['values'] || []
+      all_versions.concat(values)
+      break if json['isLast'] || values.empty?
+
+      start_at += values.size
+    end
+
+    @file_system.save_json(
+      json: all_versions,
+      filename: File.join(@target_path, "#{file_prefix}_fix_versions.json")
+    )
+  end
+
   def search_for_issues jql:, board_id:, path:
     log "  JQL: #{jql}"
     escaped_jql = CGI.escape jql
@@ -14,6 +53,7 @@ class DownloaderForCloud < Downloader
     next_page_token = nil
     issue_count = 0
 
+    start_progress
     loop do
       relative_url = +''
       relative_url << '/rest/api/3/search/jql'
@@ -36,11 +76,12 @@ class DownloaderForCloud < Downloader
         issue_count += 1
       end
 
-      message = "    Found #{issue_count} issues"
-      log message, both: true
+      progress_dot "    Found #{issue_count} issues"
 
       break unless next_page_token
     end
+    end_progress
+
     hash
   end
 
@@ -49,7 +90,7 @@ class DownloaderForCloud < Downloader
     # that only returns the "recent" changes, not all of them. So now we get the issue
     # without changes and then make a second call for that changes. Then we insert it
     # into the raw issue as if it had been there all along.
-    log "  Downloading #{issue_datas.size} issues", both: true
+    log "  Downloading #{issue_datas.size} issues"
     payload = {
       'fields' => ['*all'],
       'issueIdsOrKeys' => issue_datas.collect(&:key)
@@ -67,11 +108,24 @@ class DownloaderForCloud < Downloader
       }
       issue = Issue.new(raw: issue_json, board: board)
       data = issue_datas.find { |d| d.key == issue.key }
+      unless data
+        log "  Skipping #{issue.key}: returned by Jira but key not in request (issue may have been moved)"
+        next
+      end
       data.up_to_date = true
       data.last_modified = issue.updated
       data.issue = issue
     end
 
+    # Mark any unmatched requests as up_to_date to prevent infinite re-fetching.
+    # This happens when Jira returns a different key (moved issue) leaving the original unmatched.
+    issue_datas.each do |data|
+      next if data.up_to_date
+
+      log "  Skipping #{data.key}: not returned by Jira (issue may have been deleted or moved)"
+      data.up_to_date = true
+    end
+
     issue_datas
   end
 
@@ -127,16 +181,20 @@ class DownloaderForCloud < Downloader
 
     issue_data_hash = search_for_issues jql: jql, board_id: board.id, path: path
 
+    checked_for_related = Set.new
+    in_related_phase = false
+
     loop do
       related_issue_keys = Set.new
-      issue_data_hash
-        .values
-        .reject { |data| data.up_to_date }
-        .each_slice(100) do |slice|
-          slice = bulk_fetch_issues(
-            issue_datas: slice, board: board, in_initial_query: true
-          )
+      stale = issue_data_hash.values.reject { |data| data.up_to_date }
+      unless stale.empty?
+        log_start '  Downloading more issues ' unless in_related_phase
+        stale.each_slice(100) do |slice|
+          slice = bulk_fetch_issues(issue_datas: slice, board: board, in_initial_query: true)
+          progress_dot
           slice.each do |data|
+            next unless data.issue
+
             @file_system.save_json(
               json: data.issue.raw, filename: data.cache_path
             )
@@ -144,20 +202,25 @@ class DownloaderForCloud < Downloader
             # to parse the file just to find the timestamp
             @file_system.utime time: data.issue.updated, file: data.cache_path
 
-            issue = data.issue
-            next unless issue
-
-            parent_key = issue.parent_key(project_config: @download_config.project_config)
-            related_issue_keys << parent_key if parent_key
-
-            # Sub-tasks
-            issue.raw['fields']['subtasks']&.each do |raw_subtask|
-              related_issue_keys << raw_subtask['key']
-            end
+            collect_related_issue_keys issue: data.issue, related_issue_keys: related_issue_keys
+            checked_for_related << data.key
           end
         end
+        end_progress unless in_related_phase
+      end
+
+      # Also scan up-to-date cached issues we haven't checked yet — they may reference
+      # related issues that are not in the primary query result.
+      issue_data_hash.each_value do |data|
+        next if checked_for_related.include?(data.key)
+        next unless @file_system.file_exist?(data.cache_path)
+
+        checked_for_related << data.key
+        raw = @file_system.load_json(data.cache_path)
+        collect_related_issue_keys issue: Issue.new(raw: raw, board: board), related_issue_keys: related_issue_keys
+      end
 
-      # Remove all the ones we already downloaded
+      # Remove all the ones we already have
       related_issue_keys.reject! { |key| issue_data_hash[key] }
 
       related_issue_keys.each do |key|
@@ -169,9 +232,15 @@ class DownloaderForCloud < Downloader
       end
       break if related_issue_keys.empty?
 
-      log "  Downloading linked issues for board #{board.id}", both: true
+      unless in_related_phase
+        in_related_phase = true
+        log "  Identifying related issues (parents, subtasks, links) for board #{board.id}", both: true
+        log_start '  Downloading more issues '
+      end
     end
 
+    end_progress if in_related_phase
+
     delete_issues_from_cache_that_are_not_in_server(
       issue_data_hash: issue_data_hash, path: path
     )
@@ -196,6 +265,22 @@ class DownloaderForCloud < Downloader
     end
   end
 
+  def collect_related_issue_keys issue:, related_issue_keys:
+    parent_key = issue.parent_key(project_config: @download_config.project_config)
+    related_issue_keys << parent_key if parent_key
+
+    issue.raw['fields']['subtasks']&.each do |raw_subtask|
+      related_issue_keys << raw_subtask['key']
+    end
+
+    issue.raw['fields']['issuelinks']&.each do |link|
+      next if link['type']['name'] == 'Cloners'
+
+      linked = link['inwardIssue'] || link['outwardIssue']
+      related_issue_keys << linked['key'] if linked
+    end
+  end
+
   def last_modified filename:
     File.mtime(filename) if File.exist?(filename)
   end

data/lib/jirametrics/downloader_for_data_center.rb

@@ -25,7 +25,7 @@ class DownloaderForDataCenter < Downloader
       keys_to_request = @issue_keys_pending_download[0..99]
       @issue_keys_pending_download.reject! { |key| keys_to_request.include? key }
       jql = "key in (#{keys_to_request.join(', ')})"
-      jira_search_by_jql(jql: jql, initial_query: false, board: board, path: path)
+      jira_search_by_jql(jql: jql, initial_query: true, board: board, path: path)
     end
   end
 
@@ -63,7 +63,8 @@ class DownloaderForDataCenter < Downloader
     end
   end
 
-  def make_jql filter_id:, today: Date.today
+  def make_jql filter_id:, today: nil
+    today ||= today_in_project_timezone
     segments = []
     segments << "filter=#{filter_id}"
 

data/lib/jirametrics/estimate_accuracy_chart.rb

@@ -5,7 +5,7 @@ class EstimateAccuracyChart < ChartBase
     super()
 
     header_text 'Estimate Accuracy'
-    description_text <<-HTML
+    description_text <<~HTML
       <div class="p">
         This chart graphs estimates against actual recorded cycle times. Since
         estimates can change over time, we're graphing the estimate at the time that the story started.
@@ -20,8 +20,18 @@ class EstimateAccuracyChart < ChartBase
           far to the right then you know you have a problem.
         <% end %>
       </div>
+      <% if @correlation_coefficient %>
+        <div class="p">
+          The completed items here have a correlation coefficient of <b><%= @correlation_coefficient.round(3) %></b>.
+          The closer it is to +1, the stronger the positive correlation. The closer it is to -1,
+          the stronger the negative collalation. Zero would mean no correlation at all.
+        </div>
+      <% end %>
     HTML
 
+    @x_axis_title = 'Cycletime (days)'
+    @y_axis_title = 'Estimate'
+
     @y_axis_type = 'linear'
     @y_axis_block = ->(issue, start_time) { estimate_at(issue: issue, start_time: start_time)&.to_f }
     @y_axis_sort_order = nil
@@ -30,9 +40,9 @@ class EstimateAccuracyChart < ChartBase
   end
 
   def run
-    if @y_axis_label.nil?
+    if @y_axis_title.nil?
       text = current_board.estimation_configuration.units == :story_points ? 'Story Points' : 'Days'
-      @y_axis_label = "Estimated #{text}"
+      @y_axis_title = "Estimated #{text}"
     end
     data_sets = scan_issues
 
@@ -43,7 +53,7 @@ class EstimateAccuracyChart < ChartBase
 
   def scan_issues
     completed_hash, aging_hash = split_into_completed_and_aging issues: issues
-
+    @correlation_coefficient = correlation_coefficient(completed_hash) unless completed_hash.empty?
     estimation_units = current_board.estimation_configuration.units
     @has_aging_data = !aging_hash.empty?
 
@@ -170,4 +180,32 @@ class EstimateAccuracyChart < ChartBase
     end
     @y_axis_block = block
   end
+
+  # Correlation coefficient is calculated using the Pearson Correlation Coefficient
+  # r = Σ((xi - x̄)(yi - ȳ)) / sqrt(Σ(xi - x̄)² · Σ(yi - ȳ)²)
+  def correlation_coefficient completed_hash
+    list1 = []
+    list2 = []
+    completed_hash.each do |(estimate, cycle_time), issues|
+      issues.size.times do
+        list1 << estimate
+        list2 << cycle_time
+      end
+    end
+
+    n = list1.size
+    return nil if n < 2
+
+    mean1 = list1.sum.to_f / n
+    mean2 = list2.sum.to_f / n
+
+    numerator = list1.zip(list2).sum { |x, y| (x - mean1) * (y - mean2) }
+    sum_sq1 = list1.sum { |x| (x - mean1)**2 }
+    sum_sq2 = list2.sum { |y| (y - mean2)**2 }
+
+    denominator = Math.sqrt(sum_sq1 * sum_sq2)
+    return nil if denominator.zero?
+
+    numerator / denominator
+  end
 end

data/lib/jirametrics/examples/aggregated_project.rb

@@ -10,9 +10,9 @@
 class Exporter
   def aggregated_project name:, project_names:, settings: {}
     project name: name do
-      puts name
+      file_system.log name
       file_prefix name
-      self.settings.merge! settings
+      self.settings.merge! stringify_keys(settings)
 
       aggregate do
         project_names.each do |project_name|

data/lib/jirametrics/examples/standard_project.rb

@@ -6,14 +6,14 @@ class Exporter
   def standard_project name:, file_prefix:, ignore_issues: nil, starting_status: nil, boards: {},
       default_board: nil, anonymize: false, settings: {}, status_category_mappings: {},
       rolling_date_count: 90, no_earlier_than: nil, ignore_types: %w[Sub-task Subtask Epic],
-      show_experimental_charts: false
-
+      show_experimental_charts: false, github_repos: nil
+    exporter = self
     project name: name do
-      puts name
+      file_system.log name, also_write_to_stderr: true
       file_prefix file_prefix
 
       self.anonymize if anonymize
-      self.settings.merge! settings
+      self.settings.merge! stringify_keys(settings)
 
       boards.each_key do |board_id|
         block = boards[board_id]
@@ -35,19 +35,20 @@ class Exporter
       download do
         self.rolling_date_count(rolling_date_count) if rolling_date_count
         self.no_earlier_than(no_earlier_than) if no_earlier_than
+        github_repo *github_repos if github_repos
       end
 
       issues.reject! do |issue|
         ignore_types.include? issue.type
       end
 
+      exporter.filter_issues issues, ignore_issues
+
       discard_changes_before status_becomes: (starting_status || :backlog) # rubocop:disable Style/RedundantParentheses
 
       file do
         file_suffix '.html'
 
-        issues.reject! { |issue| ignore_issues.include? issue.key } if ignore_issues
-
         html_report do
           board_id default_board if default_board
 
@@ -57,37 +58,39 @@ class Exporter
             html "<div><a href='#{board.url}'>#{id} #{board.name}</a> (#{board.board_type})</div>",
                  type: :header
           end
-
           daily_view
-
+          cumulative_flow_diagram
           cycletime_scatterplot do
             show_trend_lines
           end
           cycletime_histogram
 
           throughput_chart do
-            description_text '<h2>Number of items completed, grouped by issue type</h2>'
+            description_text <<~TEXT
+              <div>Throughput data is very useful for#{' '}
+                <a href="https://blog.mikebowler.ca/2024/06/02/probabilistic-forecasting/">probabilistic forecasting</a>,
+                to determine when we'll be done. Try it now with the
+                <a href="<%= throughput_forecaster_url %>" target="_blank" rel="noopener noreferrer">
+                Focused Objective throughput forecaster,</a> to see how long it would take to complete all of the
+                <%= @not_started_count %> items you currently have in your backlog.
+              </div>
+              <h2>Number of items completed, grouped by issue type</h2>'
+            TEXT
           end
-          throughput_chart do
-            header_text nil
+          throughput_by_completed_resolution_chart do
             description_text '<h2>Number of items completed, grouped by completion status and resolution</h2>'
-            grouping_rules do |issue, rules|
-              if issue.resolution
-                rules.label = "#{issue.status.name}:#{issue.resolution}"
-              else
-                rules.label = issue.status.name
-              end
-            end
           end
 
           aging_work_in_progress_chart
+          wip_by_column_chart do
+            show_recommendations
+          end
           aging_work_bar_chart
           aging_work_table
           daily_wip_by_age_chart
           daily_wip_by_blocked_stalled_chart
           daily_wip_by_parent_chart
           flow_efficiency_scatterplot if show_experimental_charts
-          expedited_chart
           sprint_burndown
           estimate_accuracy_chart
           dependency_chart
@@ -95,4 +98,14 @@ class Exporter
       end
     end
   end
+
+  # Extracted as a separate method so it can be tested independently, without needing to invoke
+  # the full standard_project DSL setup.
+  def filter_issues issues, ignore_issues
+    return unless ignore_issues
+
+    issues.reject! do |issue|
+      ignore_issues.is_a?(Proc) ? ignore_issues.call(issue) : ignore_issues.include?(issue.key)
+    end
+  end
 end

data/lib/jirametrics/expedited_chart.rb

@@ -38,6 +38,8 @@ class ExpeditedChart < ChartBase
       </div>
       #{describe_non_working_days}
     HTML
+    @x_axis_title = 'Date'
+    @y_axis_title = 'Age in days'
 
     instance_eval(&block)
   end
@@ -48,7 +50,7 @@ class ExpeditedChart < ChartBase
     end
 
     if data_sets.empty?
-      '<h1 class="foldable">Expedited work</h1><p>There is no expedited work in this time period.</p>'
+      '<h1 class="foldable">Expedited work</h1><div>There is no expedited work in this time period.</div>'
     else
       wrap_and_render(binding, __FILE__)
     end