First Steps

A handful of commands are enough to get started using byebug. The following session illustrates these commands. Take the following sample file:

#
# The n'th triangle number: triangle(n) = n*(n+1)/2 = 1 + 2 + ... + n
#
def triangle(n)
  tri = 0

  0.upto(n) { |i| tri += i }

  tri
end

t = triangle(3)
puts t

Let's debug it.

$ byebug /path/to/triangle.rb

[1, 10] in /path/to/triangle.rb
    1: #
    2: # The n'th triangle number: triangle(n) = n*(n+1)/2 = 1 + 2 + ... + n
    3: #
=>  4: def triangle(n)
    5:   tri = 0
    6:
    7:   0.upto(n) { |i| tri += i }
    8:
    9:   tri
   10: end
(byebug)

We are currently stopped before the first executable line of the program: line 4 of triangle.rb. If you are used to less dynamic languages and have used debuggers for more statically compiled languages like C, C++, or Java, it may seem odd to be stopped before a function definition but in Ruby line 4 is executed.

Byebug's prompt is (byebug). If the program has died and you are in post-mortem debugging, (byebug:post-mortem) is used instead. If the program has terminated normally and the --no-quit option has been specified in the command line, the prompt will be (byebug:ctrl) instead. The commands available change depending on the program's state.

Byebug automatically lists 10 lines of code centered around the current line every time it is stopped. The current line is marked with =>. If the range would overflow the beggining or the end of the file, byebug will move it accordingly so that only actual real lines of code are displayed.

Now let us step through the program.

(byebug) step

[5, 14] in /path/to/triangle.rb
    5:   tri = 0
    6:
    7:   0.upto(n) { |i| tri += i }
    9:   end
   10:
   11:   tri
   12: end
   13:
=> 14: triangle(3)
(byebug) <RET> # hit enter

[1, 10] in /path/to/triangle.rb
    1: #
    2: # The n'th triangle number: triangle(n) = n*(n+1)/2 = 1 + 2 + ... + n
    3: #
    4: def triangle(n)
=>  5:   tri = 0
    6:
    7:   0.upto(n) { |i| tri += i }
    8:
    9:   tri
   10: end
(byebug) p tri
nil
(byebug) step

[2, 11] in /path/to/triangle.rb
    2: # The n'th triangle number: triangle(n) = n*(n+1)/2 = 1 + 2 + ... + n
    3: #
    4: def triangle(n)
    5:   tri = 0
    6:
=>  7:   0.upto(n) { |i| tri += i }
    8:
    9:   tri
   10: end
   11:
(byebug) p tri
0

The first step command runs the script one executable unit. The second command we entered was just hitting the return key: byebug remembers the last command you entered was step and runs it again.

One way to print the values of variables is p (there are other ways). When we look at the value of tri the first time, we see it is nil. Again we are stopped before the assignment on line 5, and this variable hadn't been set previously. However after issuing another step command we see that the value is 0 as expected. If every time we stop we want to see the value of tri to see how things are going, there is a better way by setting a display expression:

(byebug) display tri
1: tri = 0

Now let us run the program until right before we return from the function. We'll want to see which lines get run, so we turn on line tracing. If we don't want whole paths to be displayed when tracing, we can turn on basename.

(byebug) set linetrace
linetrace is on
(byebug) set basename
basename is on
(byebug) finish 0
Tracing: triangle.rb:7   0.upto(n) { |i| tri += i }
1: tri = 0
Tracing: triangle.rb:7   0.upto(n) { |i| tri += i }
1: tri = 0
Tracing: triangle.rb:7   0.upto(n) { |i| tri += i }
1: tri = 1
Tracing: triangle.rb:7   0.upto(n) { |i| tri += i }
1: tri = 3
Tracing: triangle.rb:9   tri
1: tri = 6
1: tri = 6

[4, 13] in /home/davidr/Proyectos/byebug/triangle.rb
    4: def triangle(n)
    5:   tri = 0
    6:
    7:   0.upto(n) { |i| tri += i }
    8:
    9:   tri
=> 10: end
   11:
   12: t = triangle(3)
   13: puts t
(byebug) quit
Really quit? (y/n)
y

So far, so good. As you can see from the above to get out of byebug, one can issue a quit command (q and exit are just as good). If you want to quit without being prompted, suffix the command with an exclamation mark, e.g., q!.

Second Sample Session: Delving Deeper

In this section we'll introduce breakpoints, the call stack and restarting. Below we will debug a simple Ruby program to solve the classic Towers of Hanoi puzzle. It is augmented by the bane of programming: some command-parameter processing with error checking.

#
# Solves the classic Towers of Hanoi puzzle.
#
def hanoi(n, a, b, c)
  hanoi(n - 1, a, c, b) if n - 1 > 0

  puts "Move disk #{a} to #{b}"

  hanoi(n - 1, c, b, a) if n - 1 > 0
end

n_args = $ARGV.length

fail('*** Need number of disks or no parameter') if n_args > 1

Recall in the first section it was stated that before the def is run, the method it names is undefined. Let's check that out. First let's see what private methods we can call before running def hanoi.

$ byebug path/to/hanoi.rb

    1: #
    2: # Solves the classic Towers of Hanoi puzzle.
    3: #
    4: def hanoi(n, a, b, c)
    5:   hanoi(n - 1, a, c, b) if n - 1 > 0
    6:
    7:   puts "Move disk #{a} to #{b}"
    8:
    9:   hanoi(n - 1, c, b, a) if n - 1 > 0
   10: end
(byebug) private_methods
[:public, :private, :include, :using, :define_method, :default_src_encoding, ...

private_methods is not a byebug command but a Ruby feature. By default, when byebug doesn't understand a command, it will evaluate it as if it was a Ruby command. If you don't want this behaviour, you can use set noautoeval or even drop it in your .byebugrc file if you want that behaviour permanently. The output of private_methods, thought, is unwieldy for our purpose: check whether hanoi method is in the list. Fortunately, byebug has nice formatting features: we can sort the output and put it into columns list using the print command ps. It also has a width setting that let's us adapt the width of the output so that it nicely fits our screen.

(byebug) set width 80
Maximum width of byebug's output is 80
(byebug) ps private_methods
Array             default_src_encoding  open                        sleep      
Complex           define_method         p                           spawn      
Digest            eval                  pp                          sprintf    
Float             exec                  print                       srand      
Hash              exit                  printf                      syscall    
Integer           exit!                 private                     system     
Pathname          fail                  proc                        test       
Rational          fork                  public                      throw      
String            format                putc                        timeout    
URI               gem_original_require  puts                        trace_var  
__callee__        gets                  raise                       trap       
__dir__           global_variables      rand                        untrace_var
__method__        include               readline                    using      
`                 initialize            readlines                   warn       
abort             initialize_clone      require                     y          
at_exit           initialize_copy       require_relative          
autoload          initialize_dup        respond_to_missing?       
autoload?         iterator?             rubygems_require          
binding           lambda                select                    
block_given?      load                  set_trace_func            
caller            local_variables       singleton_method_added    
caller_locations  loop                  singleton_method_removed  
catch             method_missing        singleton_method_undefined
(byebug)

Now let's see what happens after stepping:

(byebug) private_methods.member?(:hanoi)
false
(byebug) step

[5, 14] in /path/to/hanoi.rb
    5:   hanoi(n - 1, a, c, b) if n - 1 > 0
    6:
    7:   puts "Move disk #{a} to #{b}"
    8:
    9:   hanoi(n - 1, c, b, a) if n - 1 > 0
   10: end
   11:
=> 12: n_args = $ARGV.length
   13:
   14: fail('*** Need number of disks or no parameter') if n_args > 1
(byebug) private_methods.member?(:hanoi)
true
(byebug)

Okay, lets go on and talk about program arguments.

(byebug) $ARGV
[]

Oops. We forgot to specify any parameters to this program. Let's try again. We can use the restart command here.

(byebug) restart 3
Re exec'ing:
  /path/to/bin/byebug /path/to/hanoi.rb 3

[1, 10] in /path/to/hanoi.rb
    1: #
    2: # Solves the classic Towers of Hanoi puzzle.
    3: #
=>  4: def hanoi(n, a, b, c)
    5:   hanoi(n - 1, a, c, b) if n - 1 > 0
    6:
    7:   puts "Move disk #{a} to #{b}"
    8:
    9:   hanoi(n - 1, c, b, a) if n - 1 > 0
   10: end
(byebug) break 5
Created breakpoint 1 at /path/to/hanoi.rb:5
(byebug) continue
Stopped by breakpoint 1 at /path/to/hanoi.rb:5

[1, 10] in /path/to/hanoi.rb
    1: #
    2: # Solves the classic Towers of Hanoi puzzle.
    3: #
    4: def hanoi(n, a, b, c)
=>  5:   hanoi(n - 1, a, c, b) if n - 1 > 0
    6:
    7:   puts "Move disk #{a} to #{b}"
    8:
    9:   hanoi(n - 1, c, b, a) if n - 1 > 0
   10: end
(byebug) display n
1: n = 3
(byebug) display a
2: a = :a
(byebug) display b
3: b = :b
(byebug) undisplay 3
(byebug) continue
Stopped by breakpoint 1 at /path/to/hanoi.rb:5
1: n = 2
2: a = :a
[1, 10] in /path/to/hanoi.rb
    1: #
    2: # Solves the classic Towers of Hanoi puzzle.
    3: #
    4: def hanoi(n, a, b, c)
=>  5:   hanoi(n - 1, a, c, b) if n - 1 > 0
    6:
    7:   puts "Move disk #{a} to #{b}"
    8:
    9:   hanoi(n - 1, c, b, a) if n - 1 > 0
   10: end

(byebug) c
Stopped by breakpoint 1 at /path/to/hanoi.rb:5
1: n = 1
2: a = :a

[1, 10] in /path/to/hanoi.rb
    1: #
    2: # Solves the classic Towers of Hanoi puzzle.
    3: #
    4: def hanoi(n, a, b, c)
=>  5:   hanoi(n - 1, a, c, b) if n - 1 > 0
    6:
    7:   puts "Move disk #{a} to #{b}"
    8:
    9:   hanoi(n - 1, c, b, a) if n - 1 > 0
   10: end
(byebug) set nofullpath
fullpath is off
(byebug) where
--> #0  Object.hanoi(n#Fixnum, a#Symbol, b#Symbol, c#Symbol) at .../shortpath/to/hanoi.rb:5
    #1  Object.hanoi(n#Fixnum, a#Symbol, b#Symbol, c#Symbol) at .../shortpath/to/hanoi.rb:5
    #2  <top (required)> at .../Proyectos/byebug/hanoi.rb:28
(byebug)

In the above we added new commands: break (see breakpoints), which indicates to stop just before that line of code is run, and continue, which resumes execution. To remove a display expression undisplay is used. If we give a display number, just that display expression is removed.

We also used a new command where(see backtrace) to show the callstack. In the above situation, starting from the bottom line we see we called the hanoi method from line 28 of the file hanoi.rb and the hanoi method called itself two more times at line 5.

In the callstack we show a current frame mark, the frame number, the method being called, the names of the parameters, the types those parameters currently have and the file-line position. Remember it's possible that when the program was called the parameters had different types, since the types of variables can change dynamically. You can alter the style of what to show in the trace (see callstyle).

Now let's move around the callstack.

(byebug) undisplay
Clear all expressions? (y/n) y
(byebug) n_args
NameError Exception: undefined local variable or method `n_args' for main:Object
(byebug) frame 2

[19, 28] in /path/to/hanoi.rb
   19:   begin
   20:     n = $ARGV[0].to_i
   21:   rescue ValueError
   22:     raise("** Expecting an integer, got: #{$ARGV[0]}")
   23:   end
   24: end
   25:
   26: fail('*** Number of disks should be between 1 and 100') if n < 1 || n > 100
   27:
=> 28: hanoi(n, :a, :b, :c)
(byebug) n_args
0
(byebug) p n
3
(byebug) down 2

[1, 10] in /path/to/hanoi.rb
    1: #
    2: # Solves the classic Towers of Hanoi puzzle.
    3: #
    4: def hanoi(n, a, b, c)
=>  5:   hanoi(n - 1, a, c, b) if n - 1 > 0
    6:
    7:   puts "Move disk #{a} to #{b}"
    8:
    9:   hanoi(n - 1, c, b, a) if n - 1 > 0
   10: end
(byebug) p n
2

Notice in the above to get the value of variable n we had to use a print command like p n. If we entered just n, that would be taken to mean byebug command next. In the current scope, variable n_args is not defined. However I can change to the top-most frame by using the frame 2 command. Notice that inside frame #2, the value of n_args can be shown. Also note that the value of variable n is different.

Attaching to a running program with `byebug`

In the previous sessions we've been calling byebug right at the outset, but there is another mode of operation you might use. If there's a lot of code that needs to be run before the part you want to inspect, it might not be efficient or convenient to run byebug from the outset.

In this section we'll show how to enter the code in the middle of your program, while delving more into byebug's operation. We will also use unit testing. Using unit tests will greatly reduce the amount of debugging needed, while at the same time, will increase the quality of your program.

What we'll do is take the triangle code from the first session and write a unit test for that. In a sense we did write a tiny test for the program which was basically the last line where we printed the value of triangle(3). This test however wasn't automated: the expectation is that someone would look at the output and verify that what was printed is what was expected.

Before we can turn that into something that can be required, we probably want to remove that output. However I like to keep in that line so that when I look at the file, I have an example of how to run it. Therefore we will conditionally run this line if that file is invoked directly, but skip it if it is not. NOTE: byebug resets $0 to try to make things like this work.

if __FILE__ == $PROGRAM_NAME
  t = triangle(3)
  puts t
end

Okay, we're now ready to write our unit test and we'll use the minitest framework for that. Here's the test code, it should be placed in the same directory as triangle.rb.

require 'minitest/autorun'
require_relative 'triangle.rb'

class TestTriangle < Minitest::Test
  def test_basic
    solutions = []

    0.upto(5) { |i| solutions << triangle(i) }

    assert_equal([0, 1, 3, 6, 10, 15], solutions, 'First 5 triangle numbers')
  end
end

Let's say we want to stop before the first statement in our test method, we'll add the following:

...
def test_basic
  byebug
  solutions = []
...

Now we run the program, requiring byebug

$ ruby -rbyebug test_triangle.rb
Run options: --seed 31679

# Running:


[2, 11] in test_triangle.rb
    2: require_relative 'triangle.rb'
    3: 
    4: class TestTriangle < Minitest::Test
    5:   def test_basic
    6:     byebug
=>  7:     solutions = []
    8:
    9:     0.upto(5) { |i| solutions << triangle(i) }
   10:
   11:     assert_equal([0, 1, 3, 6, 10, 15], solutions, 'First 5 triangle numbers')
(byebug)

and we see that we are stopped at line 7 just before the initialization of the list solutions.

Now let's see where we are...

(byebug) set nofullpath
Displaying frame's full file names is off.
(byebug) bt
--> #0  TestTriangle.test_basic at .../Proyectos/byebug/test_triangle.rb:7
    #1  block (3 levels) in Minitest::Test.run at .../lib/minitest/test.rb:108
    #2  Minitest::Test.capture_exceptions at .../lib/minitest/test.rb:206
    #3  block (2 levels) in Minitest::Test.run at .../lib/minitest/test.rb:105
    #4  Minitest::Test.time_it at .../lib/minitest/test.rb:258
    #5  block in Minitest::Test.run at .../lib/minitest/test.rb:104
    #6  #<Class:Minitest::Runnable>.on_signal(name#String, action#Proc) at .../minitest-5.5.0/lib/minitest.rb:321
    #7  Minitest::Test.with_info_handler(&block#Proc) at .../lib/minitest/test.rb:278
    #8  Minitest::Test.run at .../lib/minitest/test.rb:103
    #9  #<Class:Minitest>.run_one_method(klass#Class, method_name#String) at .../minitest-5.5.0/lib/minitest.rb:768
    #10 #<Class:Minitest::Runnable>.run_one_method(klass#Class, method_name#String, reporter#Minitest::CompositeReporter) at .../minitest-5.5.0/lib/minitest.rb:295
    #11 block (2 levels) in #<Class:Minitest::Runnable>.run(reporter#Minitest::CompositeReporter, options#Hash) at .../minitest-5.5.0/lib/minitest.rb:289
    ͱ-- #12 Array.each at .../minitest-5.5.0/lib/minitest.rb:288
    #13 block in #<Class:Minitest::Runnable>.run(reporter#Minitest::CompositeReporter, options#Hash) at .../minitest-5.5.0/lib/minitest.rb:288
    #14 #<Class:Minitest::Runnable>.on_signal(name#String, action#Proc) at .../minitest-5.5.0/lib/minitest.rb:321
    #15 #<Class:Minitest::Runnable>.with_info_handler(reporter#Minitest::CompositeReporter, &block#Proc) at .../minitest-5.5.0/lib/minitest.rb:308
    #16 #<Class:Minitest::Runnable>.run(reporter#Minitest::CompositeReporter, options#Hash) at .../minitest-5.5.0/lib/minitest.rb:287
    #17 block in #<Class:Minitest>.__run(reporter#Minitest::CompositeReporter, options#Hash) at .../minitest-5.5.0/lib/minitest.rb:150
    ͱ-- #18 Array.map at .../minitest-5.5.0/lib/minitest.rb:150
    #19 #<Class:Minitest>.__run(reporter#Minitest::CompositeReporter, options#Hash) at .../minitest-5.5.0/lib/minitest.rb:150
    #20 #<Class:Minitest>.run(args#Array) at .../minitest-5.5.0/lib/minitest.rb:127
    #21 block in #<Class:Minitest>.autorun at .../minitest-5.5.0/lib/minitest.rb:56
(byebug)

We get the same result as if we had run byebug from the outset.

Debugging Oddities: How debugging Ruby may be different from other languages

If you are used to debugging in other languages like C, C++, Perl, Java or even Bash (see bashdb), there may be a number of things that seem or feel a little bit different and may confuse you. A number of these things aren't oddities of the debugger per se but differences in how Ruby works compared to those other languages. Because Ruby works a little differently from those other languages, writing a debugger has to also be a little different as well if it is to be useful. In this respect, using Byebug may help you understand Ruby better.

We've already seen one such difference: the fact that we stop on method definitions or def's and that is because these are in fact executable statements. In other compiled languages this would not happen because that's already been done when you compile the program (or in Perl when it scans in the program). In this section we'll consider some other things that might throw off new users to Ruby who are familiar with other languages and debugging in them.

Bouncing Around in Blocks (iterators)
No Parameter Values in a Call Stack
Lines You Can Stop At

Bouncing Around in Blocks (iterators)

When debugging languages with coroutines like Python and Ruby, a method call may not necessarily go to the first statement after the method header. It's possible that the call will continue after a yield statement from a prior call.

#
# Enumerator for primes
#
class SievePrime
  def initialize
    @odd_primes = []
  end

  def next_prime
    candidate = 2
    yield candidate
    not_prime = false
    candidate += 1

    loop do
      @odd_primes.each do |p|
        not_prime = (0 == (candidate % p))
        break if not_prime
      end

      unless not_prime
        @odd_primes << candidate
        yield candidate
      end

      candidate += 2
    end
  end
end

SievePrime.new.next_prime do |prime|
  puts prime
  break if prime > 10
end

$ byebug primes.rb
[1, 10] in /path/to/primes.rb
    1: #
    2: # Enumerator for primes
    3: #
=>  4: class SievePrime
    5:   def initialize
    6:     @odd_primes = []
    7:   end
    8:
    9:   def self.next_prime(&block)
   10:    candidate = 2
(byebug) set tracing
line tracing is on.
(byebug) set basename
basename in on.
(byebug) step 9
Tracing: primes.rb:5   def initialize
Tracing: primes.rb:9   def next_prime
Tracing: primes.rb:31 SievePrime.new.next_prime do |prime|
Tracing: primes.rb:6     @odd_primes = []
Tracing: primes.rb:10     candidate = 2
Tracing: primes.rb:11     yield candidate
Tracing: primes.rb:32   puts prime
2
Tracing: primes.rb:33   break if prime > 10
Tracing: primes.rb:12     not_prime = false

[7, 16] in /path/to/primes.rb
    7:   end
    8:
    9:   def next_prime
   10:     candidate = 2
   11:     yield candidate
=> 12:     not_prime = false
   13:     candidate += 1
   14:
   15:     loop do
   16:       @odd_primes.each do |p|
   17:         not_prime = (0 == (candidate % p))
(byebug)

The loop between lines 31-34 gets interleaved between those of SievePrime#next_prime, lines 9-28 above.

No Parameter Values in a Call Stack

In traditional debuggers, in a call stack you can generally see the names of the parameters and the values that were passed in.

Ruby is a very dynamic language and it tries to be efficient within the confines of the language definition. Values generally aren't taken out of a variable or expression and pushed onto a stack. Instead a new scope is created and the parameters are given initial values. Parameter passing is by reference not by value as it is say Algol, C, or Perl. During the execution of a method, parameter values can change (and often do). In fact even the class of the object can change.

So at present, the name of the parameter is shown. The call-style setting (callstyle) can be used to set whether the name is shown or the name and the current class of the object.

Lines You Can Stop At

Consider the following little Ruby program.

'Yes it does' =~ /
(Yes) \s+
it  \s+
does
/ix
puts $1

The stopping points that Ruby records are the last two lines, lines 5 and 6.

Inside byebug you can get a list of stoppable lines for a file using the info file command.

To be continued...

Threading Support.
More complex examples with objects, pretty printing and irb.
Line tracing and non-interactive tracing.
Post-mortem debugging.

Getting in & out

Starting byebug

There is a wrapper script called byebug which basically require's the gem then loads byebug before its argument (the program to be debugged) is started. If you don't need to pass dash options to your program, which might be confused with byebug options, then you don't need to add the --. To get a brief list of options and descriptions, use the --help option.

$ byebug --help

  byebug 3.5.1

  Usage: byebug [options] <script.rb> -- <script.rb parameters>

    -d, --debug               Set $DEBUG=true
    -I, --include list        Add to paths to $LOAD_PATH
    -m, --[no-]post-mortem    Use post-mortem mode
    -q, --[no-]quit           Quit when script finishes
    -x, --[no-]rc             Run byebug initialization file
    -s, --[no-]stop           Stop when script is loaded
    -r, --require file        Require library before script
    -R, --remote [host:]port  Remote debug [host:]port
    -t, --[no-]trace          Turn on line tracing
    -v, --version             Print program version
    -h, --help                Display this message

Many options appear as a long option name, such as --help and a short one letter option name, such as -h. The list of options is detailed below:

-h | --help

It causes byebug to print some basic help and exit

-v | --version

It causes byebug to print its version number and exit.

-d | --debug

Sets $DEBUG to true. Compatible with Ruby's flag.

-I | --include

Adds path to load path. path can be a single path or a colon separated path list.

-m | --post-mortem

If your program raises an exception that isn't caught you can enter byebug for inspection of what went wrong. You may also want to use this option in conjunction with --no-stop. See also Post-Mortem Debugging.

--no-quit

Keep inside byebug after your program terminates normally.

--no-stop

Normally byebug stops before executing the first statement. If instead you want it to start running initially and perhaps break it later in the execution, use this option.

-r | --require

Requires the library before executing the script. This option is compatible with Ruby's.

-t | --trace

Turns on line tracing. Running byebug --trace <rubyscript>.rb is pretty much like running ruby -rtracer <rubyscript>.rb. If all you want to do however is get a line trace, tracer is most likely faster than byebug.

$ time byebug --trace --no-stop hanoi.rb > /dev/null

real	0m0.743s
user	0m0.668s
sys	0m0.068s
$ time ruby -rtracer hanoi.rb > /dev/null

real	0m0.077s
user	0m0.072s
sys	0m0.004s

Byebug default options

Byebug has many command-line options,; it seems that some people want to set them differently from the defaults. For example, some people may want --no-quit to be the default behavior. One could write a wrapper script or set a shell alias to handle this.

Command Files

A command file is a file of lines that are byebug commands. Comments (lines starting with #) may also be included. An empty line in a command file does nothing; it does not mean to repeat the last command, as it would from the terminal.

When you start byebug, it automatically executes commands from its init file, called .byebugrc. During startup, byebug does the following:

Processes command line options and operands. Reads the init file in your current directory, if any, and then checks your home directory. The home directory is the directory named in the $HOME or $HOMEPATH environment variable. Thus, you can have more than one init file, one generic in your home directory, and another, specific to the program you are debugging, in the directory where you invoke byebug.

You can also request the execution of a command file with the source command (see Source).

Quitting byebug

To exit byebug, use the quit command (abbreviated q and aliased exit). Normally if you are in an interactive session, this command will prompt to ask if you really want to quit. If you don't want any questions asked, enter quit unconditionally (abbreviated q!). Another way to terminate byebug is to use the kill command. This does the more forceful kill -9. It can be used in cases where quit doesn't work (I haven't seen those yet).

Calling byebug from inside your program

Running a program from byebug adds a bit of overhead and slows it down a little. Furthermore, by necessity, debuggers change the operation of the program they are debugging. And this can lead to unexpected and unwanted differences. It has happened so often that the term Heisenbugs was coined to describe the situation where using a debugger (among other possibilities) changes the behavior of the program so that the bug doesn't manifest itself anymore.

There is another way to get into byebug which adds no overhead or slowdown until you reach the point at which you want to start debugging. However here you must change the script and make an explicit call to byebug. Because byebug isn't involved before the first call, there is no overhead and the script will run at the same speed as if there were no byebug.

To enter byebug this way, just drop byebug in whichever line you want to start debugging at. You also have to require byebug somehow. If using bundler, it will take care of that for you, otherwise you can use the ruby -r flag or add require 'byebug' in the line previous to the byebug call.

If speed is crucial, you may want to start and stop this around certain sections of code, using Byebug.start and Byebug.stop. Alternatively, instead of issuing an explicit Byebug.stop you can add a block to the Byebug.start and debugging is turned on for that block. If the block of code raises an uncaught exception that would cause the block to terminate, the stop will occur. See Byebug.start with a block.

When byebugis run, .byebugrc is read.

You may want to enter byebug at several points in the program where there is a problem you want to investigate. And since byebug is just a method call it's possible to enclose it in a conditional expression, for example

byebug if 'bar' == foo and 20 == iter_count

Restarting Byebug

You can restart the program using restart [program args]. This is a re-exec - all byebug state is lost. If command arguments are passed, those are used. Otherwise program arguments from the last invocation are used.

You won't be able to restart your program in all cases. First, the program should have been invoked at the outset rather than having been called from inside your program or invoked as a result of post-mortem handling.

Also, since this relies on the OS exec call, this command is available only if your OS supports exec.

Debugging remote programs

It is possible to set up debugging so that you can issue byebug commands from outside the process running the Ruby code. In fact, you might even be on a different computer than the one running the Ruby program.

To setup remote debugging, drop the following somewhere before the point in the program that you want to debug (In Rails, the config/environments/development.rb could be a good candidate).

  require 'byebug'
  Byebug.wait_connection = true
  Byebug.start_server('localhost', <port>)

Once this piece gets executed, you can connect to the remote debugger from your local machine, by running: byebug -R localhost:<port>.

Next, at a place of program execution which gets run just before the code you want to debug, add a call to byebug as was done without remote execution:

   # work, work, work...
   byebug
   some ruby code  # byebug will stop before this line is run

Byebug Command Reference

Command Syntax

Usually a command is put on a single line. There is no limit on how long it can be. It starts with a command name, which is followed by arguments whose meaning depends on the command name. For example, the command step accepts an argument which is the number of times to step, as in step 5. You can also use the step command with no arguments. Some commands do not allow any arguments.

Multiple commands can be put on a line by separating each with a semicolon ;. You can disable the meaning of a semicolon to separate commands by escaping it with a backslash.

For example, if you have autoeval set, which is the default, you might want to enter the following code to compute the 5th Fibonacci number.

(byebug) fib1=0; fib2=1; 5.times {|temp| temp=fib1; fib1=fib2; fib2 += temp }
0
1
SyntaxError Exception: /home/davidr/Proyectos/sample_app/trace.rb:1: syntax
error, unexpected end-of-input, expecting '}'
 5.times { |temp| temp=fib1
                           ^
nil
1
SyntaxError Exception: /home/davidr/Proyectos/sample_app/trace.rb:1: syntax
error, unexpected tSTRING_DEND, expecting end-of-input
 fib2 += temp }
               ^
nil
(byebug) fib1=0\; fib2=1\; 5.times {|temp| temp=fib1\; fib1=fib2\; fib2 += temp }
5
(byebug) fib2
8

You might also consider using the irb or pry commands and then you won't have to escape semicolons.

A blank line as input (typing just <RET>) means to repeat the previous command.

Byebug uses readline, which handles line editing and retrieval of previous commands. Up arrow, for example, moves to the previous byebug command; down arrow moves to the next more recent command (provided you are not already at the last command). Command history is saved in file .byebug_hist. A limit is put on the history size. You can see this with the show history size command. See history for history parameters.

Command Output

In the command-line interface, when byebug is waiting for input it presents a prompt of the form (byebug). If the program has terminated normally the prompt will be (byebug:ctrl) and in post-mortem debugging it will be (byebug:post-mortem).

Whenever byebug gives an error message such as for an invalid command or an invalid location position, it will generally preface the message with ***.

Command Help

Once inside byebug you can always ask it for information on its commands using the help command. You can use help (abbreviated h) with no arguments to display a short list of named classes of commands

(byebug) help
Type "help <command-name>" for help on a specific command

Available commands:
backtrace  delete   enable  help  method  ps        save       step       where
break      disable  eval    info  next    putl      set        trace      catch
display    exit     irb     p     quit    show      undisplay  condition  down
finish     kill     pp      skip  up      continue  edit       frame      list
pry        restart  source  var

With a command name as help argument, byebug displays short information on how to use that command.

(byebug) help list
l[ist]    list forward
l[ist] -  list backward
l[ist] =  list current line
l[ist] nn-mm  list given lines
* NOTE - to turn on autolist, use 'set autolist'
(byebug)

A number of commands, namely info, set, show, enable and disable, have many sub-parameters or subcommands. When you ask for help for one of these commands, you will get help for all of the subcommands that command offers. Sometimes you may want help only on a subcommand and to do this just follow the command with its subcommand name. For example, help info breakpointswill just give help about the info breakpoints command. Furthermore it will give longer help than the summary information that appears when you ask for help. You don't need to list the full subcommand name, just enough of the letters to make that subcommand distinct from others will do. For example, help info b is the same as help info breakpoints.

Some examples follow.

(byebug) help info
info[ subcommand]

Generic command for showing things about the program being debugged.

--
List of "info" subcommands:
--
info args        -- Argument variables of current stack frame
info breakpoints -- Status of user-settable breakpoints
info catch       -- Exceptions that can be caught in the current stack frame
info display     -- Expressions to display when program stops
info file        -- Info about a particular file read in
info files       -- File names and timestamps of files read in
info line        -- Line number and filename of current position in source file
info program     -- Execution status of the program

(byebug) help info breakpoints
Status of user-settable breakpoints.
Without argument, list info about all breakpoints.
With an integer argument, list info on that breakpoint.

(byebug) help info b
Status of user-settable breakpoints.
Without argument, list info about all breakpoints.
With an integer argument, list info on that breakpoint.

Control Commands: quit, restart, source

Quit

To exit byebug, type quit (abbreviated q and aliased exit). Normally if you are in an interactive session, this command will prompt you to confirm you really want to quit. If you don't want any questions asked, enter quit unconditionally (abbreviated q!).

Restart

To restart the program, use the restart|r command. This is a re-exec - all byebug state is lost. If command arguments are passed, those are used. Otherwise program arguments from the last invocation are used.

You won't be able to restart your program in all cases. First, the program should have been invoked at the outset rather than having been called from inside your program or invoked as a result of post-mortem handling.

Source

You can run byebug commands inside a file, using the command source <file>. The lines in a command file are executed sequentially. They are not printed as they are executed. If there is an error, execution proceeds to the next command in the file. For information about command files that get run automatically on startup see Command Files.

Display Commands: display, undisplay

Display

If you find that you want to print the value of an expression frequently (to see how it changes), you might want to add it to the automatic display list* so that byebug evaluates it each time your program stops or after a line is printed if line tracing is enabled. Each expression added to the list is given a number to identify it; to remove an expression from the list, you specify that number. The automatic display looks like this:

(byebug) display n
1: n = 3

This display shows item numbers, expressions and their current values. If the expression is undefined or illegal the expression will be printed but no value will appear.

(byebug) display undefined_variable
2: undefined_variable =
(byebug) display 1/0
3: 1/0 =

If you use display with no argument, byebug will display the current values of the expressions in the list, just as it is done when your program stops. Using info display has the same effect.

Undisplay

To remove an item from the list, use undisplay followed by the number identifying the expression you want to remove. undisplay does not repeat if you press <RET>after using it (otherwise you would just get the error No display number n)

You can also temporarily disable or enable display expressions, so that the will not be printed but they won't be forgotten either, so you can toggle them again later. To do that, use disable display or enable display followed by the expression number.

Print Commands

One way to examine and change data in your script is with the eval command (abbreviated p). byebug by default evaluates any input that is not recognized as a command, so in most situations eval is not necessary and byebug will work like a REPL. One case where it's necessary could be when trying to print a variable called n. In this case, you have no choice because typing just n will execute byebug's command next.

A similar command to eval|p is pp which tries to pretty print the result.

If the value you want to print is an array, sometimes a columnized list looks nicer. Use putl for that. Notice however that entries are sorted to run down first rather than across. If the value is not an array putl will just call pretty-print.

Sometimes you may want to print the array not only columnized, but sorted as well. The list of byebug help commands appears this way, and so does the output of the method commands. Use ps for that. If the value is not an array ps will just call pretty-print.

(byebug) Kernel.instance_methods
[:nil?, :===, :=~, :!~, :eql?, :hash, :<=>, :class, :singleton_class, :clone,
:dup, :taint, :tainted?, :untaint, :untrust, :untrusted?, :trust, :freeze,
:frozen?, :to_s, :inspect, :methods, :singleton_methods, :protected_methods,
:private_methods, :public_methods, :instance_variables, :instance_variable_get,
:instance_variable_set, :instance_variable_defined?, :remove_instance_variable,
:instance_of?, :kind_of?, :is_a?, :tap, :send, :public_send, :respond_to?,
:extend, :display, :method, :public_method, :define_singleton_method,
:object_id, :to_enum, :enum_for, :gem, :pretty_inspect, :byebug]
(byebug) p Kernel.instance_methods
[:nil?, :===, :=~, :!~, :eql?, :hash, :<=>, :class, :singleton_class, :clone,
:dup, :taint, :tainted?, :untaint, :untrust, :untrusted?, :trust, :freeze,
:frozen?, :to_s, :inspect, :methods, :singleton_methods, :protected_methods,
:private_methods, :public_methods, :instance_variables, :instance_variable_get,
:instance_variable_set, :instance_variable_defined?, :remove_instance_variable,
:instance_of?, :kind_of?, :is_a?, :tap, :send, :public_send, :respond_to?,
:extend, :display, :method, :public_method, :define_singleton_method,
:object_id, :to_enum, :enum_for, :gem, :pretty_inspect, :byebug]
(byebug) pp Kernel.instance_methods
[:nil?,
 :===,
 :=~,
 :!~,
 :eql?,
 :hash,
 :<=>,
 :class,
 :singleton_class,
 :clone,
 :dup,
 :taint,
 :tainted?,
 :untaint,
 :untrust,
 :untrusted?,
 :trust,
 :freeze,
 :frozen?,
 :to_s,
 :inspect,
 :methods,
 :singleton_methods,
 :protected_methods,
 :private_methods,
 :public_methods,
 :instance_variables,
 :instance_variable_get,
 :instance_variable_set,
 :instance_variable_defined?,
 :remove_instance_variable,
 :instance_of?,
 :kind_of?,
 :is_a?,
 :tap,
 :send,
 :public_send,
 :respond_to?,
 :extend,
 :display,
 :method,
 :public_method,
 :define_singleton_method,
 :object_id,
 :to_enum,
 :enum_for,
 :gem,
 :pretty_inspect,
 :byebug]
(byebug) putl Kernel.instance_methods
nil?  <=>              tainted?    frozen?            private_methods             remove_instance_variable  public_send    define_singleton_method  byebug
===   class            untaint     to_s               public_methods              instance_of?              respond_to?    object_id
=~    singleton_class  untrust     inspect            instance_variables          kind_of?                  extend         to_enum
!~    clone            untrusted?  methods            instance_variable_get       is_a?                     display        enum_for
eql?  dup              trust       singleton_methods  instance_variable_set       tap                       method         gem
hash  taint            freeze      protected_methods  instance_variable_defined?  send                      public_method  pretty_inspect
(byebug) ps Kernel.instance_methods
!~      clone                    extend   instance_of?                kind_of?        private_methods           respond_to?        tap      untrusted?
<=>     define_singleton_method  freeze   instance_variable_defined?  method          protected_methods         send               to_enum
===     display                  frozen?  instance_variable_get       methods         public_method             singleton_class    to_s   
=~      dup                      gem      instance_variable_set       nil?            public_methods            singleton_methods  trust  
byebug  enum_for                 hash     instance_variables          object_id       public_send               taint              untaint
class   eql?                     inspect  is_a?                       pretty_inspect  remove_instance_variable  tainted?           untrust

Finally, if you need more advanced functionality from REPL's, you can enter irb or pry using irb or pry commands. The bindings environment will be set to the current state in the program. When you leave the repl and go back to byebug's command prompt we show the file, line and text position of the program. If you issue a list without location information, the default location used is the current line rather than the current position that may have got updated via a prior list command.

$ byebug triangle.rb
[1, 10] in /path/to/triangle.rb
    1: # Compute the n'th triangle number, the hard way: triangle(n) == (n*(n+1))/2
=>  2: def triangle(n)
    3:   tri = 0
    4:   0.upto(n) do |i|
    5:     tri += i
    6:   end
    7:   tri
    8: end
    9:
   10: if __FILE__ == $0
(byebug) irb
2.0.0-p247 :001 > (0..6).inject{|sum, i| sum +=i}
 => 21
2.0.0-p247 :002 > exit
/home/davidr/Proyectos/byebug/old_doc/triangle.rb @ 2
def triangle(n)
(byebug) list # same line range as before going into irb
[1, 10] in /path/to/triangle.rb
    1: # Compute the n'th triangle number, the hard way: triangle(n) == (n*(n+1))/2
=>  2: def triangle(n)
    3:   tri = 0
    4:   0.upto(n) do |i|
    5:     tri += i
    6:   end
    7:   tri
    8: end
    9:
   10: if __FILE__ == $0
(byebug)

Printing variables

Byebug can print many different information about variables. Such as

var const <object>. Show the constants of <object>. This is basically listing variables and their values in <object>.constant.
var instance <object>. Show the instance variables of <object>. This is basically listing <object>.instance_variables.
var instance. Show instance_variables of self.
var local. Show local variables.
var global. Show global variables.
var all. Show local, global and instance and class variables of self.
method instance <object>. Show methods of <object>. Basically this is the same as running ps <object>.instance_methods(false).
method <class-or-module>. Show methods of the class or module <class-or-module>. Basically this is the same as running ps <class-or-module>.methods.

Examining Program Source Files: list

byebug can print parts of your script's source. When your script stops, byebug spontaneously lists the source code around the line where it stopped that line. It does that when you change the current stack frame as well. Implicitly there is a default line location. Each time a list command is run that implicit location is updated, so that running several list commands in succession shows a contiguous block of program text.

If you don't need code context displayed every time, you can issue the set noautolist command. Now whenever you want code listed, you can explicitly issue the list command or its abbreviation l. Notice that when a second listing is displayed, we continue listing from the place we last left off. When the beginning or end of the file is reached, the line range to be shown is adjusted so "it doesn't overflow". You can set the noautolist option by default by dropping set noautolist in byebug's startup file .byebugrc.

If you want to set how many lines to be printed by default rather than use the initial number of lines, 10, use the set listsize command ([listsize()). To see the entire program in one shot, give an explicit starting and ending line number. You can print other portions of source files by giving explicit position as a parameter to the list command.

There are several ways to specify what part of the file you want to print. list nnn prints lines centered around line number nnn in the current source file. l prints more lines, following the last lines printed. list - prints lines just before the lines last printed. list nnn-mmm prints lines between nnn and mmm inclusive. list = prints lines centered around where the script is stopped. Repeating a list command with RET discards the argument, so it is equivalent to typing just list. This is more useful than listing the same lines again. An exception is made for an argument of -: that argument is preserved in repetition so that each repetition moves up in the source file.

Editing Source files: edit

To edit a source file, use the edit command. The editor of your choice is invoked with the current line set to the active line in the program. Alternatively, you can give a line specification to specify what part of the file you want to edit.

You can customize byebug to use any editor you want by using the EDITOR environment variable. The only restriction is that your editor (say ex) recognizes the following command-line syntax:

ex +nnn file

The optional numeric value +nnn specifies the line number in the file where you want to start editing. For example, to configure byebug to use the vi editor, you could use these commands with the sh shell:

EDITOR=/usr/bin/vi
export EDITOR
byebug ...

or in the csh shell,

setenv EDITOR /usr/bin/vi
byebug ...

The stack trace

When your script has stopped, one thing you'll probably want to know is where it stopped and some idea of how it got there.

Each time your script calls a method or enters a block, information about this action is saved. This information is what we call a stack frame or just a frame. The set of all frames at a certain point in the program's execution is called the stack trace or just the stack. Each frame contains a line number and the source-file name that the line refers to. If the frame is the beginning of a method it also contains the method name.

When your script is started, the stack has only one frame, that of the main method. This is called the initial frame or the outermost frame. Each time a method is called, a new frame is added to the stack trace. Each time a method returns, the frame for that method invocation is removed. If a method is recursive, there can be many frames for the same method. The frame for the method in which execution is actually occurring is called the innermost frame. This is the most recently created of all the stack frames that still exist.

Every time the debugger stops, one entry in the stack is selected as the current frame. Many byebug commands refer implicitly to the selected block. In particular, whenever you ask Byebug to list lines without giving a line number or location the value is found in the selected frame. There are special commands to select whichever frame you're interested in, such as up, down and frame.

After switching frames, when you issue a list command without any position information, the position used is the location in the frame that you just switched between, rather than a location that got updated via a prior list command.

Byebug assigns numbers to all existing stack frames, starting with zero for the innermost frame, one for the frame that called it, and so on upward. These numbers do not really exist in your script, they are assigned by Byebug to give you a way of designating stack frames in commands.

Printing the Stack: `where` command

The command where, aliased to bt or backtrace prints the call stack., It shows one line per frame, for many frames, starting with the place that you are stopped at (frame zero), followed by its caller (frame one), and on up the stack. Each frame is numbered and can be referred to in the frame command. The position of the current frame is marked with -->.

The are some special frames generated for methods that are implemented in C. One such method is each. They are marked differently in the call stack to indicate that we cannot switch to those frames. This is because they have no source code in Ruby, so we can not debug them using Byebug.

(byebug) where
--> #0 Object.gcd(a#Fixnum, b#Fixnum) at line gcd.rb:6
    #1 at line gcd.rb:19

Selecting a frame: `up`, `down` and `frame` commands

up <n>: Move n frames up the stack, towards the outermost frame (higher frame numbers, frames that have existed longer). n defaults to one.
down <n>: Move n frames down the stack, towards the innermost frame (lower frame numbers, frames that were created more recently). n defaults to one.
frame <n>: Allows you to move to an arbitrary frame. n is the stack frame number or 0 if no frame number is given. frame 0 will show the current and most recent stack frame. If a negative number is given, counting is from the other end of the stack frame, so frame -1 shows the least-recent, outermost stack frame. Without an argument, frame prints the current stack frame.

Files

GUIDE.md

Latest commit

History