Module: YfAsDataframe::CurlImpersonateIntegration

Defined in:
lib/yf_as_dataframe/curl_impersonate_integration.rb

Class Attribute Summary collapse

Class Method Summary collapse

Class Attribute Details

.curl_impersonate_connect_timeoutObject

Returns the value of attribute curl_impersonate_connect_timeout.



18
19
20
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 18

def curl_impersonate_connect_timeout
  @curl_impersonate_connect_timeout
end

.curl_impersonate_enabledObject

Returns the value of attribute curl_impersonate_enabled.



18
19
20
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 18

def curl_impersonate_enabled
  @curl_impersonate_enabled
end

.curl_impersonate_fallbackObject

Returns the value of attribute curl_impersonate_fallback.



18
19
20
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 18

def curl_impersonate_fallback
  @curl_impersonate_fallback
end

.curl_impersonate_process_timeoutObject

Returns the value of attribute curl_impersonate_process_timeout.



18
19
20
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 18

def curl_impersonate_process_timeout
  @curl_impersonate_process_timeout
end

.curl_impersonate_retriesObject

Returns the value of attribute curl_impersonate_retries.



18
19
20
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 18

def curl_impersonate_retries
  @curl_impersonate_retries
end

.curl_impersonate_retry_delayObject

Returns the value of attribute curl_impersonate_retry_delay.



18
19
20
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 18

def curl_impersonate_retry_delay
  @curl_impersonate_retry_delay
end

.curl_impersonate_timeoutObject

Returns the value of attribute curl_impersonate_timeout.



18
19
20
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 18

def curl_impersonate_timeout
  @curl_impersonate_timeout
end

Class Method Details

.available_executablesObject

Find available curl-impersonate executables



30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 30

def self.available_executables
  @available_executables ||= begin
    executables = []
    Dir.glob(File.join(executable_directory, "curl_*")).each do |path|
      executable = File.basename(path)
      if executable.start_with?('curl_')
        browser_type = case executable
                      when /^curl_chrome/ then :chrome
                      when /^curl_ff/ then :firefox
                      when /^curl_edge/ then :edge
                      when /^curl_safari/ then :safari
                      else :unknown
                      end
        executables << { path: path, executable: executable, browser: browser_type }
      end
    end
    executables
  end
end

.executable_directoryObject

Get the curl-impersonate executable directory from environment variable or default



25
26
27
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 25

def self.executable_directory
  ENV['CURL_IMPERSONATE_DIR'] || '/usr/local/bin'
end

.get_random_executableObject

Get a random executable



51
52
53
54
55
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 51

def self.get_random_executable
  available = available_executables
  return nil if available.empty?
  available.sample
end

.make_request(url, headers: {}, params: {}, timeout: nil, retries: nil) ⇒ Object

Make a curl-impersonate request with improved timeout handling



58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
# File 'lib/yf_as_dataframe/curl_impersonate_integration.rb', line 58

def self.make_request(url, headers: {}, params: {}, timeout: nil, retries: nil)
  executable_info = get_random_executable
  return nil unless executable_info

  timeout ||= @curl_impersonate_timeout
  retries ||= @curl_impersonate_retries

  cmd = [
    executable_info[:path],
    "--max-time", timeout.to_s,
    "--connect-timeout", @curl_impersonate_connect_timeout.to_s,
    "--retry", retries.to_s,
    "--retry-delay", @curl_impersonate_retry_delay.to_s,
    "--retry-max-time", (timeout * 2).to_s,
    "--fail",
    "--silent",
    "--show-error"
  ]
  headers.each { |key, value| cmd.concat(["-H", "#{key}: #{value}"]) }
  unless params.empty?
    query_string = params.map { |k, v| "#{k}=#{v}" }.join('&')
    separator = url.include?('?') ? '&' : '?'
    url = "#{url}#{separator}#{query_string}"
  end
  cmd << url

  # puts "DEBUG: curl-impersonate command: #{cmd.join(' ')}"
  # puts "DEBUG: curl-impersonate timeout: #{timeout} seconds"

  begin
    stdout_str = ''
    stderr_str = ''
    status = nil
    Open3.popen3(*cmd) do |stdin, stdout, stderr, wait_thr|
      stdin.close
      pid = wait_thr.pid
      done = false
      monitor = Thread.new do
        sleep(timeout + 10)
        unless done
          # puts "DEBUG: Killing curl-impersonate PID \\#{pid} after timeout"
          Process.kill('TERM', pid) rescue nil
          sleep(1)
          Process.kill('KILL', pid) rescue nil if wait_thr.alive?
        end
      end
      stdout_str = stdout.read
      stderr_str = stderr.read
      status = wait_thr.value
      done = true
      monitor.kill
    end
    # puts "DEBUG: curl-impersonate stdout: #{stdout_str[0..200]}..." if stdout_str && !stdout_str.empty?
    # puts "DEBUG: curl-impersonate stderr: #{stderr_str}" if stderr_str && !stderr_str.empty?
    # puts "DEBUG: curl-impersonate status: #{status.exitstatus}"
    if status.success?
      response = OpenStruct.new
        response.body = stdout_str
      response.code = 200
      response.define_singleton_method(:success?) { true }
        response.parsed_response = parse_json_if_possible(stdout_str)
      response
    else
      # puts "DEBUG: curl-impersonate failed with error: \\#{error_message}"
      error_message = "curl failed with code \\#{status.exitstatus}: \\#{stderr_str}"
      nil
    end
  rescue => e
    # puts "DEBUG: curl-impersonate exception: \\#{e.message}"
    nil
  end
end