Sha256: 84f5099f76d9a8ffaa6b45ab6822a9c834191ce609398713a427368d9a3f7e6a
Contents?: true
Size: 1.48 KB
Versions: 28
Compression:
Stored size: 1.48 KB
Contents
#!/usr/bin/ruby -I../lib
require 'gratr/import'
require 'gratr/dot'
# This program gives an example of dynamic analysis of a program's call stack,
# that exports to dot and creates a jpg visualization of the call diagram.
class GraphSelf
# Setup some data to call Dijkstra's Algorithm
def initialize
@d = Digraph[ [:a,:b] => 9, [:a,:e] => 3,
[:b,:c] => 2, [:b,:e] => 6,
[:c,:d] => 1,
[:d,:c] => 2,
[:e,:b] => 2, [:e,:f] => 1,
[:f,:c] => 2, [:f,:d] => 7, [:f,:e] => 2 ]
@call_stack = []
@call_graph = Digraph.new
end
# Get the call graph variable
def call_graph() @call_graph; end
# Turn capturing of call graph on
def capture_func
Proc.new do |event, f, l, id, b, klass|
# Only interested in the GRATR library itself
if ( klass.to_s =~ /GRATR/ )
case event.to_s
when /call/
method = "#{klass.to_s.split('::')[1]}.#{id}" # Removes GRATR::
@call_graph.add_edge!(@call_stack[-1],method) if @call_stack[-1]
@call_stack.push(method)
when /return/ : @call_stack.pop
end
end
end
end
# Run a capture of the call graph for Dijkstra's algorithm
def run
set_trace_func capture_func
@d.dijkstras_algorithm(:a)
set_trace_func nil
self
end
end
# Run a capture and generate the resulting jpg file
GraphSelf.new.run.call_graph.write_to_graphic_file('png','self_graph')
Version data entries
28 entries across 28 versions & 3 rubygems