Class: GithubDailyDigest::GithubService

Inherits:
Object
  • Object
show all
Defined in:
lib/github_service.rb

Constant Summary collapse

MAX_RETRIES =

Local retry specifically for rate limits within a method call

3

Instance Method Summary collapse

Constructor Details

#initialize(token:, logger:, config:) ⇒ GithubService

Returns a new instance of GithubService.



10
11
12
13
14
15
16
17
18
19
20
21
22
 
# File 'lib/github_service.rb', line 10

def initialize(token:, logger:, config:)
  @logger = logger
  @config = config
  @client = Octokit::Client.new(access_token: token)
  @client.auto_paginate = true # Essential for members/repos
  verify_authentication
rescue Octokit::Unauthorized => e
  @logger.fatal("GitHub authentication failed. Check GITHUB_TOKEN. Error: #{e.message}")
  raise # Re-raise to stop execution
rescue => e
  @logger.fatal("Failed to initialize GitHub client: #{e.message}")
  raise
end

Instance Method Details

#fetch_active_repos(org_name, since_time) ⇒ Object

Fetches all repositories with activity since a given time



110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
 
# File 'lib/github_service.rb', line 110

def fetch_active_repos(org_name, since_time)
  @logger.info("Fetching active repositories for organization: #{org_name} since #{since_time}")
  repos = fetch_org_repos(org_name)
  
  @logger.info("Checking #{repos.size} repositories for activity")
  active_repos = {}
  
  repos.each_with_index do |repo, index|
    repo_full_name = repo.full_name
    @logger.info("Checking for activity in #{repo_full_name} since #{since_time} [#{index+1}/#{repos.size}]")
    
    begin
      # Get all branches for this repository
      branches = handle_api_errors(catch_conflicts: true) do
        @client.branches(repo_full_name)
      end
      
      if branches.nil? || branches.empty?
        @logger.debug("No branches found in #{repo_full_name}")
        next
      end
      
      @logger.info("Found #{branches.count} branches in #{repo_full_name}")
      
      # Find branches with recent activity
      active_branches = []
      all_commits = []
      
      # We'll check each branch in parallel
      branches.each do |branch|
        branch_name = branch.name
        
        # Get latest commit for the branch
        latest_commit = branch.commit
        if latest_commit
          commit_date = nil
          
          # Get full commit details to check date
          commit_details = handle_api_errors(catch_conflicts: true) do
            @client.commit(repo_full_name, latest_commit.sha)
          end
          
          if commit_details && commit_details.commit && commit_details.commit.author
            commit_date = commit_details.commit.author.date
          end
          
          # If this branch has commits since our cutoff date, flag it as active
          if commit_date && Time.parse(commit_date.to_s) >= Time.parse(since_time.to_s)
            active_branches << branch_name
          end
        end
      end
      
      if active_branches.any?
        @logger.info("Found #{active_branches.size} active branches in #{repo_full_name}: #{active_branches.join(', ')}")
        
        # Now get commits for each active branch
        active_branches.each do |branch_name|
          branch_commits = handle_api_errors(catch_conflicts: true) do
            @client.commits(repo_full_name, { sha: branch_name, since: since_time })
          end
          
          if branch_commits && branch_commits.any?
            @logger.debug("Found #{branch_commits.count} commits in branch #{branch_name} of #{repo_full_name}")
            
            # Add branch information to each commit
            branch_commits.each do |commit|
              commit.branch = branch_name
            end
            
            all_commits.concat(branch_commits)
          end
        end
        
        # Remove duplicate commits (same SHA across multiple branches)
        # But preserve branch information
        unique_commits = {}
        all_commits.each do |commit|
          if unique_commits[commit.sha]
            # If we've seen this commit before, add branch to its branches list
            unique_commits[commit.sha].branches ||= []
            unique_commits[commit.sha].branches << commit.branch unless unique_commits[commit.sha].branches.include?(commit.branch)
          else
            # First time seeing this commit
            commit.branches = [commit.branch]
            unique_commits[commit.sha] = commit
          end
        end
        
        commits = unique_commits.values
        
        if commits.any?
          @logger.info("Found #{commits.count} unique commits across active branches in #{repo_full_name}")
          active_repos[repo_full_name] = commits
        end
      else
        @logger.debug("No active branches found in #{repo_full_name} since #{since_time}")
      end
      
      # Avoid hitting rate limits
      sleep(0.1) if index > 0 && index % 10 == 0
    rescue => e
      @logger.error("Error checking repo #{repo_full_name}: #{e.message}")
      @logger.error(e.backtrace.join("\n"))
      next
    end
  end
  
  if active_repos.empty?
    @logger.warn("No active repositories found with commits since #{since_time}")
  else
    @logger.info("Found #{active_repos.size} active repositories with commits out of #{repos.size} total repos")
    active_repos.keys.each do |repo_name|
      @logger.info("Active repo: #{repo_name} with #{active_repos[repo_name].size} commits")
    end
  end
  
  active_repos
end

#fetch_members(org_name) ⇒ Object 



24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
 
# File 'lib/github_service.rb', line 24

def fetch_members(org_name)
  @logger.info("Fetching members for organization: #{org_name}")
  members = handle_api_errors { @client.organization_members(org_name) }
  if members
    logins = members.map(&:login)
    @logger.info("Found #{logins.count} members.")
    logins
  else
    @logger.error("Could not fetch members for #{org_name}.")
    [] # Return empty array on failure after retries
  end
rescue Octokit::NotFound => e
  @logger.error("Organization '#{org_name}' not found or token lacks permission. Error: #{e.message}")
  []
end

#fetch_org_repos(org_name) ⇒ Object 



57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
 
# File 'lib/github_service.rb', line 57

def fetch_org_repos(org_name)
  @logger.info("Fetching repositories for organization: #{org_name}")

  # First try with type: 'all'
  repos = handle_api_errors do
    @client.organization_repositories(org_name, { type: 'all', per_page: 100 })
  end

  if repos
    @logger.info("Found #{repos.count} repositories.")

    # Log some details about the first few repos for debugging
    if repos.any?
      sample_repos = repos.take(3)
      sample_repos.each do |repo|
        @logger.info("Sample repo: #{repo.full_name}, Private: #{repo.private}, Fork: #{repo.fork}")
      end
    end

    repos # Return the array of Sawyer::Resource objects
  else
    @logger.error("Could not fetch repositories for #{org_name}.")
    []
  end
end

#fetch_user_commits_in_repo(repo_full_name, username, since_time) ⇒ Object

Fetches commits for a specific user in a specific repo since a given time



84
85
86
87
88
89
90
91
92
93
94
95
96
97
 
# File 'lib/github_service.rb', line 84

def fetch_user_commits_in_repo(repo_full_name, username, since_time)
  @logger.debug("Fetching commits by #{username} in #{repo_full_name} since #{since_time}")
  options = { author: username, since: since_time }
  commits = handle_api_errors(catch_conflicts: true) do
    @client.commits_since(repo_full_name, since_time, options)
    # Alternative if above doesn't work reliably with author filter:
    # @client.commits(repo_full_name, since: since_time).select { |c| c.author&.login == username }
  end
  commits || [] # Return empty array on failure
rescue Octokit::Conflict, Octokit::NotFound => e
  # Repo might be empty, disabled issues/wiki, or inaccessible
  @logger.warn("Skipping repo #{repo_full_name} for user #{username}. Reason: #{e.message}")
  []
end

#get_current_user ⇒ Object

Get information about the authenticated user



41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
 
# File 'lib/github_service.rb', line 41

def get_current_user
  handle_api_errors do
    user = @client.user
    {
      login: user.,
      name: user.name,
      email: user.email,
      avatar_url: user.avatar_url,
      scopes: @client.scopes
    }
  end
rescue => e
  @logger.error("Failed to get current user information: #{e.message}")
  nil
end

#map_commits_to_users(active_repos) ⇒ Object

Maps commits to users for efficient activity tracking



231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
 
# File 'lib/github_service.rb', line 231

def map_commits_to_users(active_repos)
  @logger.info("Mapping commits to users")
  user_commits = {}

  active_repos.each do |repo_full_name, commits|
    commits.each do |commit|
      author = commit.author&.
      next unless author # Skip commits without a valid GitHub author

      # Fetch commit details to get line changes
      commit_details = handle_api_errors do
        @client.commit(repo_full_name, commit.sha)
      end

      user_commits[author] ||= []

      if commit_details
        # Add commit with line changes information if available
        user_commits[author] << format_commit(commit, repo_full_name, commit_details)
      else
        # Fallback to basic commit info if details couldn't be fetched
        user_commits[author] << format_commit(commit, repo_full_name)
      end
    end
  end

  @logger.info("Found commits from #{user_commits.size} users")
  user_commits
end

#search_user_reviews(username, org_name, since_time) ⇒ Object

Searches for PRs reviewed by the user



100
101
102
103
104
105
106
107
 
# File 'lib/github_service.rb', line 100

def search_user_reviews(username, org_name, since_time)
  @logger.debug("Searching PR reviews for user: #{username} since #{since_time}")
  query = "is:pr reviewed-by:#{username} org:#{org_name} updated:>#{since_time}"
  results = handle_api_errors { @client.search_issues(query, per_page: 1) } # Fetch 1 to get total_count efficiently
  count = results ? results.total_count : 0
  @logger.debug("Found #{count} PRs reviewed by #{username} via search.")
  count
end