<h2>Editing File: /usr/share/perl5/pod/perldebguts.pod</h2><form method='post'><textarea name='content' class='file-edit'>=head1 NAME

perldebguts - Guts of Perl debugging 

=head1 DESCRIPTION

This is not L&lt;perldebug&gt;, which tells you how to use
the debugger.  This manpage describes low-level details concerning
the debugger's internals, which range from difficult to impossible
to understand for anyone who isn't incredibly intimate with Perl's guts.
Caveat lector.

=head1 Debugger Internals

Perl has special debugging hooks at compile-time and run-time used
to create debugging environments.  These hooks are not to be confused
with the I&lt;perl -Dxxx&gt; command described in L&lt;perlrun&gt;, which is
usable only if a special Perl is built per the instructions in the
F&lt;INSTALL&gt; podpage in the Perl source tree.

For example, whenever you call Perl's built-in C&lt;caller&gt; function
from the package C&lt;DB&gt;, the arguments that the corresponding stack
frame was called with are copied to the C&lt;@DB::args&gt; array.  These
mechanisms are enabled by calling Perl with the B&lt;-d&gt; switch.
Specifically, the following additional features are enabled
(cf. L&lt;perlvar/$^P&gt;):

=over 4

=item *

Perl inserts the contents of C&lt;$ENV{PERL5DB}&gt; (or C&lt;BEGIN {require
'perl5db.pl'}&gt; if not present) before the first line of your program.

=item *

Each array C&lt;@{&quot;_&lt;$filename&quot;}&gt; holds the lines of $filename for a
file compiled by Perl.  The same is also true for C&lt;eval&gt;ed strings
that contain subroutines, or which are currently being executed.
The $filename for C&lt;eval&gt;ed strings looks like C&lt;(eval 34)&gt;.

Values in this array are magical in numeric context: they compare
equal to zero only if the line is not breakable.

=item *

Each hash C&lt;%{&quot;_&lt;$filename&quot;}&gt; contains breakpoints and actions keyed
by line number.  Individual entries (as opposed to the whole hash)
are settable.  Perl only cares about Boolean true here, although
the values used by F&lt;perl5db.pl&gt; have the form
C&lt;&quot;$break_condition\0$action&quot;&gt;.  

The same holds for evaluated strings that contain subroutines, or
which are currently being executed.  The $filename for C&lt;eval&gt;ed strings
looks like C&lt;(eval 34)&gt;.

=item *

Each scalar C&lt;${&quot;_&lt;$filename&quot;}&gt; contains C&lt;&quot;_&lt;$filename&quot;&gt;.  This is
also the case for evaluated strings that contain subroutines, or
which are currently being executed.  The $filename for C&lt;eval&gt;ed
strings looks like C&lt;(eval 34)&gt;.

=item *

After each C&lt;require&gt;d file is compiled, but before it is executed,
C&lt;DB::postponed(*{&quot;_&lt;$filename&quot;})&gt; is called if the subroutine
C&lt;DB::postponed&gt; exists.  Here, the $filename is the expanded name of
the C&lt;require&gt;d file, as found in the values of %INC.

=item *

After each subroutine C&lt;subname&gt; is compiled, the existence of
C&lt;$DB::postponed{subname}&gt; is checked.  If this key exists,
C&lt;DB::postponed(subname)&gt; is called if the C&lt;DB::postponed&gt; subroutine
also exists.

=item *

A hash C&lt;%DB::sub&gt; is maintained, whose keys are subroutine names
and whose values have the form C&lt;filename:startline-endline&gt;.
C&lt;filename&gt; has the form C&lt;(eval 34)&gt; for subroutines defined inside
C&lt;eval&gt;s.

=item *

When the execution of your program reaches a point that can hold a
breakpoint, the C&lt;DB::DB()&gt; subroutine is called if any of the variables
C&lt;$DB::trace&gt;, C&lt;$DB::single&gt;, or C&lt;$DB::signal&gt; is true.  These variables
are not C&lt;local&gt;izable.  This feature is disabled when executing
inside C&lt;DB::DB()&gt;, including functions called from it 
unless C&lt;&lt; $^D &amp; (1&lt;&lt;30) &gt;&gt; is true.

=item *

When execution of the program reaches a subroutine call, a call to
C&lt;&amp;DB::sub&gt;(I&lt;args&gt;) is made instead, with C&lt;$DB::sub&gt; holding the
name of the called subroutine. (This doesn't happen if the subroutine
was compiled in the C&lt;DB&gt; package.)

X&lt;&amp;DB::lsub&gt;If the call is to an lvalue subroutine, and C&lt;&amp;DB::lsub&gt;
is defined C&lt;&amp;DB::lsub&gt;(I&lt;args&gt;) is called instead, otherwise falling
back to C&lt;&amp;DB::sub&gt;(I&lt;args&gt;).

=item *

When execution of the program uses C&lt;goto&gt; to enter a non-XS
subroutine and the 0x80 bit is set in C&lt;$^P&gt;, a call to C&lt;&amp;DB::goto&gt;
is made, with C&lt;$DB::sub&gt; holding the name of the subroutine being
entered.

=back

Note that if C&lt;&amp;DB::sub&gt; needs external data for it to work, no
subroutine call is possible without it. As an example, the standard
debugger's C&lt;&amp;DB::sub&gt; depends on the C&lt;$DB::deep&gt; variable
(it defines how many levels of recursion deep into the debugger you can go
before a mandatory break).  If C&lt;$DB::deep&gt; is not defined, subroutine
calls are not possible, even though C&lt;&amp;DB::sub&gt; exists.

=head2 Writing Your Own Debugger

=head3 Environment Variables

The C&lt;PERL5DB&gt; environment variable can be used to define a debugger.
For example, the minimal &quot;working&quot; debugger (it actually doesn't do anything)
consists of one line:

  sub DB::DB {}

It can easily be defined like this:

  $ PERL5DB=&quot;sub DB::DB {}&quot; perl -d your-script

Another brief debugger, slightly more useful, can be created
with only the line:

  sub DB::DB {print ++$i; scalar &lt;STDIN&gt;}

This debugger prints a number which increments for each statement
encountered and waits for you to hit a newline before continuing
to the next statement.

The following debugger is actually useful:

  {
    package DB;
    sub DB  {}
    sub sub {print ++$i, &quot; $sub\n&quot;; &amp;$sub}
  }

It prints the sequence number of each subroutine call and the name of the
called subroutine.  Note that C&lt;&amp;DB::sub&gt; is being compiled into the
package C&lt;DB&gt; through the use of the C&lt;package&gt; directive.

When it starts, the debugger reads your rc file (F&lt;./.perldb&gt; or
F&lt;~/.perldb&gt; under Unix), which can set important options.
(A subroutine (C&lt;&amp;afterinit&gt;) can be defined here as well; it is executed
after the debugger completes its own initialization.)

After the rc file is read, the debugger reads the PERLDB_OPTS
environment variable and uses it to set debugger options. The
contents of this variable are treated as if they were the argument
of an C&lt;o ...&gt; debugger command (q.v. in L&lt;perldebug/&quot;Configurable Options&quot;&gt;).

=head3 Debugger Internal Variables

In addition to the file and subroutine-related variables mentioned above,
the debugger also maintains various magical internal variables.

=over 4

=item *

C&lt;@DB::dbline&gt; is an alias for C&lt;@{&quot;::_&lt;current_file&quot;}&gt;, which
holds the lines of the currently-selected file (compiled by Perl), either
explicitly chosen with the debugger's C&lt;f&gt; command, or implicitly by flow
of execution.

Values in this array are magical in numeric context: they compare
equal to zero only if the line is not breakable.

=item *

C&lt;%DB::dbline&gt; is an alias for C&lt;%{&quot;::_&lt;current_file&quot;}&gt;, which
contains breakpoints and actions keyed by line number in
the currently-selected file, either explicitly chosen with the
debugger's C&lt;f&gt; command, or implicitly by flow of execution.

As previously noted, individual entries (as opposed to the whole hash)
are settable.  Perl only cares about Boolean true here, although
the values used by F&lt;perl5db.pl&gt; have the form
C&lt;&quot;$break_condition\0$action&quot;&gt;.

=back

=head3 Debugger Customization Functions

Some functions are provided to simplify customization.

=over 4

=item *

See L&lt;perldebug/&quot;Configurable Options&quot;&gt; for a description of options parsed by
C&lt;DB::parse_options(string)&gt;.

=item *

C&lt;DB::dump_trace(skip[,count])&gt; skips the specified number of frames
and returns a list containing information about the calling frames (all
of them, if C&lt;count&gt; is missing).  Each entry is reference to a hash
with keys C&lt;context&gt; (either C&lt;.&gt;, C&lt;$&gt;, or C&lt;@&gt;), C&lt;sub&gt; (subroutine
name, or info about C&lt;eval&gt;), C&lt;args&gt; (C&lt;undef&gt; or a reference to
an array), C&lt;file&gt;, and C&lt;line&gt;.

=item *

C&lt;DB::print_trace(FH, skip[, count[, short]])&gt; prints
formatted info about caller frames.  The last two functions may be
convenient as arguments to C&lt;&lt; &lt; &gt;&gt;, C&lt;&lt; &lt;&lt; &gt;&gt; commands.

=back

Note that any variables and functions that are not documented in
this manpages (or in L&lt;perldebug&gt;) are considered for internal   
use only, and as such are subject to change without notice.

=head1 Frame Listing Output Examples

The C&lt;frame&gt; option can be used to control the output of frame 
information.  For example, contrast this expression trace:

 $ perl -de 42
 Stack dump during die enabled outside of evals.

 Loading DB routines from perl5db.pl patch level 0.94
 Emacs support available.

 Enter h or 'h h' for help.

 main::(-e:1):   0
   DB&lt;1&gt; sub foo { 14 }

   DB&lt;2&gt; sub bar { 3 }

   DB&lt;3&gt; t print foo() * bar()
 main::((eval 172):3):   print foo() + bar();
 main::foo((eval 168):2):
 main::bar((eval 170):2):
 42

with this one, once the C&lt;o&gt;ption C&lt;frame=2&gt; has been set:

   DB&lt;4&gt; o f=2
                frame = '2'
   DB&lt;5&gt; t print foo() * bar()
 3:      foo() * bar()
 entering main::foo
  2:     sub foo { 14 };
 exited main::foo
 entering main::bar
  2:     sub bar { 3 };
 exited main::bar
 42

By way of demonstration, we present below a laborious listing
resulting from setting your C&lt;PERLDB_OPTS&gt; environment variable to
the value C&lt;f=n N&gt;, and running I&lt;perl -d -V&gt; from the command line.
Examples using various values of C&lt;n&gt; are shown to give you a feel
for the difference between settings.  Long though it may be, this
is not a complete listing, but only excerpts.

=over 4

=item 1

 entering main::BEGIN
  entering Config::BEGIN
   Package lib/Exporter.pm.
   Package lib/Carp.pm.
  Package lib/Config.pm.
  entering Config::TIEHASH
  entering Exporter::import
   entering Exporter::export
 entering Config::myconfig
  entering Config::FETCH
  entering Config::FETCH
  entering Config::FETCH
  entering Config::FETCH

=item 2

 entering main::BEGIN
  entering Config::BEGIN
   Package lib/Exporter.pm.
   Package lib/Carp.pm.
  exited Config::BEGIN
  Package lib/Config.pm.
  entering Config::TIEHASH
  exited Config::TIEHASH
  entering Exporter::import
   entering Exporter::export
   exited Exporter::export
  exited Exporter::import
 exited main::BEGIN
 entering Config::myconfig
  entering Config::FETCH
  exited Config::FETCH
  entering Config::FETCH
  exited Config::FETCH
  entering Config::FETCH

=item 3

 in  $=main::BEGIN() from /dev/null:0
  in  $=Config::BEGIN() from lib/Config.pm:2
   Package lib/Exporter.pm.
   Package lib/Carp.pm.
  Package lib/Config.pm.
  in  $=Config::TIEHASH('Config') from lib/Config.pm:644
  in  $=Exporter::import('Config', 'myconfig', 'config_vars') from /dev/null:0
   in  $=Exporter::export('Config', 'main', 'myconfig', 'config_vars') from li
 in  @=Config::myconfig() from /dev/null:0
  in  $=Config::FETCH(ref(Config), 'package') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'baserev') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'PERL_VERSION') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'PERL_SUBVERSION') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'osname') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'osvers') from lib/Config.pm:574

=item 4

 in  $=main::BEGIN() from /dev/null:0
  in  $=Config::BEGIN() from lib/Config.pm:2
   Package lib/Exporter.pm.
   Package lib/Carp.pm.
  out $=Config::BEGIN() from lib/Config.pm:0
  Package lib/Config.pm.
  in  $=Config::TIEHASH('Config') from lib/Config.pm:644
  out $=Config::TIEHASH('Config') from lib/Config.pm:644
  in  $=Exporter::import('Config', 'myconfig', 'config_vars') from /dev/null:0
   in  $=Exporter::export('Config', 'main', 'myconfig', 'config_vars') from lib/
   out $=Exporter::export('Config', 'main', 'myconfig', 'config_vars') from lib/
  out $=Exporter::import('Config', 'myconfig', 'config_vars') from /dev/null:0
 out $=main::BEGIN() from /dev/null:0
 in  @=Config::myconfig() from /dev/null:0
  in  $=Config::FETCH(ref(Config), 'package') from lib/Config.pm:574
  out $=Config::FETCH(ref(Config), 'package') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'baserev') from lib/Config.pm:574
  out $=Config::FETCH(ref(Config), 'baserev') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'PERL_VERSION') from lib/Config.pm:574
  out $=Config::FETCH(ref(Config), 'PERL_VERSION') from lib/Config.pm:574
  in  $=Config::FETCH(ref(Config), 'PERL_SUBVERSION') from lib/Config.pm:574

=item 5

 in  $=main::BEGIN() from /dev/null:0
  in  $=Config::BEGIN() from lib/Config.pm:2
   Package lib/Exporter.pm.
   Package lib/Carp.pm.
  out $=Config::BEGIN() from lib/Config.pm:0
  Package lib/Config.pm.
  in  $=Config::TIEHASH('Config') from lib/Config.pm:644
  out $=Config::TIEHASH('Config') from lib/Config.pm:644
  in  $=Exporter::import('Config', 'myconfig', 'config_vars') from /dev/null:0
   in  $=Exporter::export('Config', 'main', 'myconfig', 'config_vars') from lib/E
   out $=Exporter::export('Config', 'main', 'myconfig', 'config_vars') from lib/E
  out $=Exporter::import('Config', 'myconfig', 'config_vars') from /dev/null:0
 out $=main::BEGIN() from /dev/null:0
 in  @=Config::myconfig() from /dev/null:0
  in  $=Config::FETCH('Config=HASH(0x1aa444)', 'package') from lib/Config.pm:574
  out $=Config::FETCH('Config=HASH(0x1aa444)', 'package') from lib/Config.pm:574
  in  $=Config::FETCH('Config=HASH(0x1aa444)', 'baserev') from lib/Config.pm:574
  out $=Config::FETCH('Config=HASH(0x1aa444)', 'baserev') from lib/Config.pm:574

=item 6

 in  $=CODE(0x15eca4)() from /dev/null:0
  in  $=CODE(0x182528)() from lib/Config.pm:2
   Package lib/Exporter.pm.
  out $=CODE(0x182528)() from lib/Config.pm:0
  scalar context return from CODE(0x182528): undef
  Package lib/Config.pm.
  in  $=Config::TIEHASH('Config') from lib/Config.pm:628
  out $=Config::TIEHASH('Config') from lib/Config.pm:628
  scalar context return from Config::TIEHASH:   empty hash
  in  $=Exporter::import('Config', 'myconfig', 'config_vars') from /dev/null:0
   in  $=Exporter::export('Config', 'main', 'myconfig', 'config_vars') from lib/Exporter.pm:171
   out $=Exporter::export('Config', 'main', 'myconfig', 'config_vars') from lib/Exporter.pm:171
   scalar context return from Exporter::export: ''
  out $=Exporter::import('Config', 'myconfig', 'config_vars') from /dev/null:0
  scalar context return from Exporter::import: ''

=back

In all cases shown above, the line indentation shows the call tree.
If bit 2 of C&lt;frame&gt; is set, a line is printed on exit from a
subroutine as well.  If bit 4 is set, the arguments are printed
along with the caller info.  If bit 8 is set, the arguments are
printed even if they are tied or references.  If bit 16 is set, the
return value is printed, too.

When a package is compiled, a line like this

    Package lib/Carp.pm.

is printed with proper indentation.

=head1 Debugging Regular Expressions

There are two ways to enable debugging output for regular expressions.

If your perl is compiled with C&lt;-DDEBUGGING&gt;, you may use the
B&lt;-Dr&gt; flag on the command line.

Otherwise, one can C&lt;use re 'debug'&gt;, which has effects at
compile time and run time.  Since Perl 5.9.5, this pragma is lexically
scoped.

=head2 Compile-time Output

The debugging output at compile time looks like this:

  Compiling REx '[bc]d(ef*g)+h[ij]k$'
  size 45 Got 364 bytes for offset annotations.
  first at 1
  rarest char g at 0
  rarest char d at 0
     1: ANYOF[bc](12)
    12: EXACT &lt;d&gt;(14)
    14: CURLYX[0] {1,32767}(28)
    16:   OPEN1(18)
    18:     EXACT &lt;e&gt;(20)
    20:     STAR(23)
    21:       EXACT &lt;f&gt;(0)
    23:     EXACT &lt;g&gt;(25)
    25:   CLOSE1(27)
    27:   WHILEM[1/1](0)
    28: NOTHING(29)
    29: EXACT &lt;h&gt;(31)
    31: ANYOF[ij](42)
    42: EXACT &lt;k&gt;(44)
    44: EOL(45)
    45: END(0)
  anchored 'de' at 1 floating 'gh' at 3..2147483647 (checking floating) 
        stclass 'ANYOF[bc]' minlen 7 
  Offsets: [45]
  	1[4] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 5[1]
  	0[0] 12[1] 0[0] 6[1] 0[0] 7[1] 0[0] 9[1] 8[1] 0[0] 10[1] 0[0]
  	11[1] 0[0] 12[0] 12[0] 13[1] 0[0] 14[4] 0[0] 0[0] 0[0] 0[0]
  	0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 18[1] 0[0] 19[1] 20[0]  
  Omitting $` $&amp; $' support.

The first line shows the pre-compiled form of the regex.  The second
shows the size of the compiled form (in arbitrary units, usually
4-byte words) and the total number of bytes allocated for the
offset/length table, usually 4+C&lt;size&gt;*8.  The next line shows the
label I&lt;id&gt; of the first node that does a match.

The 

  anchored 'de' at 1 floating 'gh' at 3..2147483647 (checking floating) 
        stclass 'ANYOF[bc]' minlen 7 

line (split into two lines above) contains optimizer
information.  In the example shown, the optimizer found that the match 
should contain a substring C&lt;de&gt; at offset 1, plus substring C&lt;gh&gt;
at some offset between 3 and infinity.  Moreover, when checking for
these substrings (to abandon impossible matches quickly), Perl will check
for the substring C&lt;gh&gt; before checking for the substring C&lt;de&gt;.  The
optimizer may also use the knowledge that the match starts (at the
C&lt;first&gt; I&lt;id&gt;) with a character class, and no string 
shorter than 7 characters can possibly match.

The fields of interest which may appear in this line are

=over 4

=item C&lt;anchored&gt; I&lt;STRING&gt; C&lt;at&gt; I&lt;POS&gt;

=item C&lt;floating&gt; I&lt;STRING&gt; C&lt;at&gt; I&lt;POS1..POS2&gt;

See above.

=item C&lt;matching floating/anchored&gt;

Which substring to check first.

=item C&lt;minlen&gt;

The minimal length of the match.

=item C&lt;stclass&gt; I&lt;TYPE&gt;

Type of first matching node.

=item C&lt;noscan&gt;

Don't scan for the found substrings.

=item C&lt;isall&gt;

Means that the optimizer information is all that the regular
expression contains, and thus one does not need to enter the regex engine at
all.

=item C&lt;GPOS&gt;

Set if the pattern contains C&lt;\G&gt;.

=item C&lt;plus&gt; 

Set if the pattern starts with a repeated char (as in C&lt;x+y&gt;).

=item C&lt;implicit&gt;

Set if the pattern starts with C&lt;.*&gt;.

=item C&lt;with eval&gt; 

Set if the pattern contain eval-groups, such as C&lt;(?{ code })&gt; and
C&lt;(??{ code })&gt;.

=item C&lt;anchored(TYPE)&gt;

If the pattern may match only at a handful of places, with C&lt;TYPE&gt;
being C&lt;SBOL&gt;, C&lt;MBOL&gt;, or C&lt;GPOS&gt;.  See the table below.

=back

If a substring is known to match at end-of-line only, it may be
followed by C&lt;$&gt;, as in C&lt;floating 'k'$&gt;.

The optimizer-specific information is used to avoid entering (a slow) regex
engine on strings that will not definitely match.  If the C&lt;isall&gt; flag
is set, a call to the regex engine may be avoided even when the optimizer
found an appropriate place for the match.

Above the optimizer section is the list of I&lt;nodes&gt; of the compiled
form of the regex.  Each line has format 

C&lt;   &gt;I&lt;id&gt;: I&lt;TYPE&gt; I&lt;OPTIONAL-INFO&gt; (I&lt;next-id&gt;)

=head2 Types of Nodes

Here are the current possible types, with short descriptions:

=for comment
This table is generated by regen/regcomp.pl.  Any changes made here
will be lost.

=for regcomp.pl begin

 # TYPE arg-description [num-args] [longjump-len] DESCRIPTION

 # Exit points

 END             no         End of program.
 SUCCEED         no         Return from a subroutine, basically.

 # Line Start Anchors:
 SBOL            no         Match &quot;&quot; at beginning of line: /^/, /\A/
 MBOL            no         Same, assuming multiline: /^/m

 # Line End Anchors:
 SEOL            no         Match &quot;&quot; at end of line: /$/
 MEOL            no         Same, assuming multiline: /$/m
 EOS             no         Match &quot;&quot; at end of string: /\z/

 # Match Start Anchors:
 GPOS            no         Matches where last m//g left off.

 # Word Boundary Opcodes:
 BOUND           no         Like BOUNDA for non-utf8, otherwise match &quot;&quot;
                            between any Unicode \w\W or \W\w
 BOUNDL          no         Like BOUND/BOUNDU, but \w and \W are defined
                            by current locale
 BOUNDU          no         Match &quot;&quot; at any boundary of a given type
                            using Unicode rules
 BOUNDA          no         Match &quot;&quot; at any boundary between \w\W or
                            \W\w, where \w is [_a-zA-Z0-9]
 NBOUND          no         Like NBOUNDA for non-utf8, otherwise match
                            &quot;&quot; between any Unicode \w\w or \W\W
 NBOUNDL         no         Like NBOUND/NBOUNDU, but \w and \W are
                            defined by current locale
 NBOUNDU         no         Match &quot;&quot; at any non-boundary of a given type
                            using using Unicode rules
 NBOUNDA         no         Match &quot;&quot; betweeen any \w\w or \W\W, where \w
                            is [_a-zA-Z0-9]

 # [Special] alternatives:
 REG_ANY         no         Match any one character (except newline).
 SANY            no         Match any one character.
 ANYOF           sv 1       Match character in (or not in) this class,
                            single char match only
 ANYOFD          sv 1       Like ANYOF, but /d is in effect
 ANYOFL          sv 1       Like ANYOF, but /l is in effect

 # POSIX Character Classes:
 POSIXD          none       Some [[:class:]] under /d; the FLAGS field
                            gives which one
 POSIXL          none       Some [[:class:]] under /l; the FLAGS field
                            gives which one
 POSIXU          none       Some [[:class:]] under /u; the FLAGS field
                            gives which one
 POSIXA          none       Some [[:class:]] under /a; the FLAGS field
                            gives which one
 NPOSIXD         none       complement of POSIXD, [[:^class:]]
 NPOSIXL         none       complement of POSIXL, [[:^class:]]
 NPOSIXU         none       complement of POSIXU, [[:^class:]]
 NPOSIXA         none       complement of POSIXA, [[:^class:]]

 CLUMP           no         Match any extended grapheme cluster sequence

 # Alternation

 # BRANCH        The set of branches constituting a single choice are
 #               hooked together with their &quot;next&quot; pointers, since
 #               precedence prevents anything being concatenated to
 #               any individual branch.  The &quot;next&quot; pointer of the last
 #               BRANCH in a choice points to the thing following the
 #               whole choice.  This is also where the final &quot;next&quot;
 #               pointer of each individual branch points; each branch
 #               starts with the operand node of a BRANCH node.
 #
 BRANCH          node       Match this alternative, or the next...

 # Literals

 EXACT           str        Match this string (preceded by length).
 EXACTL          str        Like EXACT, but /l is in effect (used so
                            locale-related warnings can be checked for).
 EXACTF          str        Match this non-UTF-8 string (not guaranteed
                            to be folded) using /id rules (w/len).
 EXACTFL         str        Match this string (not guaranteed to be
                            folded) using /il rules (w/len).
 EXACTFU         str        Match this string (folded iff in UTF-8,
                            length in folding doesn't change if not in
                            UTF-8) using /iu rules (w/len).
 EXACTFA         str        Match this string (not guaranteed to be
                            folded) using /iaa rules (w/len).

 EXACTFU_SS      str        Match this string (folded iff in UTF-8,
                            length in folding may change even if not in
                            UTF-8) using /iu rules (w/len).
 EXACTFLU8       str        Rare cirucmstances: like EXACTFU, but is
                            under /l, UTF-8, folded, and everything in
                            it is above 255.
 EXACTFA_NO_TRIE str        Match this string (which is not trie-able;
                            not guaranteed to be folded) using /iaa
                            rules (w/len).

 # Do nothing types

 NOTHING         no         Match empty string.
 # A variant of above which delimits a group, thus stops optimizations
 TAIL            no         Match empty string. Can jump here from
                            outside.

 # Loops

 # STAR,PLUS    '?', and complex '*' and '+', are implemented as
 #               circular BRANCH structures.  Simple cases
 #               (one character per match) are implemented with STAR
 #               and PLUS for speed and to minimize recursive plunges.
 #
 STAR            node       Match this (simple) thing 0 or more times.
 PLUS            node       Match this (simple) thing 1 or more times.

 CURLY           sv 2       Match this simple thing {n,m} times.
 CURLYN          no 2       Capture next-after-this simple thing
 CURLYM          no 2       Capture this medium-complex thing {n,m}
                            times.
 CURLYX          sv 2       Match this complex thing {n,m} times.

 # This terminator creates a loop structure for CURLYX
 WHILEM          no         Do curly processing and see if rest matches.

 # Buffer related

 # OPEN,CLOSE,GROUPP     ...are numbered at compile time.
 OPEN            num 1      Mark this point in input as start of #n.
 CLOSE           num 1      Analogous to OPEN.

 REF             num 1      Match some already matched string
 REFF            num 1      Match already matched string, folded using
                            native charset rules for non-utf8
 REFFL           num 1      Match already matched string, folded in loc.
 REFFU           num 1      Match already matched string, folded using
                            unicode rules for non-utf8
 REFFA           num 1      Match already matched string, folded using
                            unicode rules for non-utf8, no mixing ASCII,
                            non-ASCII

 # Named references.  Code in regcomp.c assumes that these all are after
 # the numbered references
 NREF            no-sv 1    Match some already matched string
 NREFF           no-sv 1    Match already matched string, folded using
                            native charset rules for non-utf8
 NREFFL          no-sv 1    Match already matched string, folded in loc.
 NREFFU          num 1      Match already matched string, folded using
                            unicode rules for non-utf8
 NREFFA          num 1      Match already matched string, folded using
                            unicode rules for non-utf8, no mixing ASCII,
                            non-ASCII

 # Support for long RE
 LONGJMP         off 1 1    Jump far away.
 BRANCHJ         off 1 1    BRANCH with long offset.

 # Special Case Regops
 IFMATCH         off 1 2    Succeeds if the following matches.
 UNLESSM         off 1 2    Fails if the following matches.
 SUSPEND         off 1 1    &quot;Independent&quot; sub-RE.
 IFTHEN          off 1 1    Switch, should be preceded by switcher.
 GROUPP          num 1      Whether the group matched.

 # The heavy worker

 EVAL            evl/flags  Execute some Perl code.
                 2L

 # Modifiers

 MINMOD          no         Next operator is not greedy.
 LOGICAL         no         Next opcode should set the flag only.

 # This is not used yet
 RENUM           off 1 1    Group with independently numbered parens.

 # Trie Related

 # Behave the same as A|LIST|OF|WORDS would. The '..C' variants
 # have inline charclass data (ascii only), the 'C' store it in the
 # structure.

 TRIE            trie 1     Match many EXACT(F[ALU]?)? at once.
                            flags==type
 TRIEC           trie       Same as TRIE, but with embedded charclass
                 charclass  data

 AHOCORASICK     trie 1     Aho Corasick stclass. flags==type
 AHOCORASICKC    trie       Same as AHOCORASICK, but with embedded
                 charclass  charclass data

 # Regex Subroutines
 GOSUB           num/ofs 2L recurse to paren arg1 at (signed) ofs arg2

 # Special conditionals
 NGROUPP         no-sv 1    Whether the group matched.
 INSUBP          num 1      Whether we are in a specific recurse.
 DEFINEP         none 1     Never execute directly.

 # Backtracking Verbs
 ENDLIKE         none       Used only for the type field of verbs
 OPFAIL          no-sv 1    Same as (?!), but with verb arg
 ACCEPT          no-sv/num  Accepts the current matched string, with
                 2L         verbar

 # Verbs With Arguments
 VERB            no-sv 1    Used only for the type field of verbs
 PRUNE           no-sv 1    Pattern fails at this startpoint if no-
                            backtracking through this
 MARKPOINT       no-sv 1    Push the current location for rollback by
                            cut.
 SKIP            no-sv 1    On failure skip forward (to the mark) before
                            retrying
 COMMIT          no-sv 1    Pattern fails outright if backtracking
                            through this
 CUTGROUP        no-sv 1    On failure go to the next alternation in the
                            group

 # Control what to keep in $&amp;.
 KEEPS           no         $&amp; begins here.

 # New charclass like patterns
 LNBREAK         none       generic newline pattern

 # SPECIAL  REGOPS

 # This is not really a node, but an optimized away piece of a &quot;long&quot;
 # node.  To simplify debugging output, we mark it as if it were a node
 OPTIMIZED       off        Placeholder for dump.

 # Special opcode with the property that no opcode in a compiled program
 # will ever be of this type. Thus it can be used as a flag value that
 # no other opcode has been seen. END is used similarly, in that an END
 # node cant be optimized. So END implies &quot;unoptimizable&quot; and PSEUDO
 # mean &quot;not seen anything to optimize yet&quot;.
 PSEUDO          off        Pseudo opcode for internal use.

=for regcomp.pl end

=for unprinted-credits
Next section M-J. Dominus (mjd-perl-patch+@plover.com) 20010421

Following the optimizer information is a dump of the offset/length
table, here split across several lines:

  Offsets: [45]
  	1[4] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 5[1]
  	0[0] 12[1] 0[0] 6[1] 0[0] 7[1] 0[0] 9[1] 8[1] 0[0] 10[1] 0[0]
  	11[1] 0[0] 12[0] 12[0] 13[1] 0[0] 14[4] 0[0] 0[0] 0[0] 0[0]
  	0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 18[1] 0[0] 19[1] 20[0]  

The first line here indicates that the offset/length table contains 45
entries.  Each entry is a pair of integers, denoted by C&lt;offset[length]&gt;.
Entries are numbered starting with 1, so entry #1 here is C&lt;1[4]&gt; and
entry #12 is C&lt;5[1]&gt;.  C&lt;1[4]&gt; indicates that the node labeled C&lt;1:&gt;
(the C&lt;1: ANYOF[bc]&gt;) begins at character position 1 in the
pre-compiled form of the regex, and has a length of 4 characters.
C&lt;5[1]&gt; in position 12 
indicates that the node labeled C&lt;12:&gt;
(the C&lt;&lt; 12: EXACT &lt;d&gt; &gt;&gt;) begins at character position 5 in the
pre-compiled form of the regex, and has a length of 1 character.
C&lt;12[1]&gt; in position 14 
indicates that the node labeled C&lt;14:&gt;
(the C&lt;&lt; 14: CURLYX[0] {1,32767} &gt;&gt;) begins at character position 12 in the
pre-compiled form of the regex, and has a length of 1 character---that
is, it corresponds to the C&lt;+&gt; symbol in the precompiled regex.

C&lt;0[0]&gt; items indicate that there is no corresponding node.

=head2 Run-time Output

First of all, when doing a match, one may get no run-time output even
if debugging is enabled.  This means that the regex engine was never
entered and that all of the job was therefore done by the optimizer.

If the regex engine was entered, the output may look like this:

  Matching '[bc]d(ef*g)+h[ij]k$' against 'abcdefg__gh__'
    Setting an EVAL scope, savestack=3
     2 &lt;ab&gt; &lt;cdefg__gh_&gt;    |  1: ANYOF
     3 &lt;abc&gt; &lt;defg__gh_&gt;    | 11: EXACT &lt;d&gt;
     4 &lt;abcd&gt; &lt;efg__gh_&gt;    | 13: CURLYX {1,32767}
     4 &lt;abcd&gt; &lt;efg__gh_&gt;    | 26:   WHILEM
				0 out of 1..32767  cc=effff31c
     4 &lt;abcd&gt; &lt;efg__gh_&gt;    | 15:     OPEN1
     4 &lt;abcd&gt; &lt;efg__gh_&gt;    | 17:     EXACT &lt;e&gt;
     5 &lt;abcde&gt; &lt;fg__gh_&gt;    | 19:     STAR
			     EXACT &lt;f&gt; can match 1 times out of 32767...
    Setting an EVAL scope, savestack=3
     6 &lt;bcdef&gt; &lt;g__gh__&gt;    | 22:       EXACT &lt;g&gt;
     7 &lt;bcdefg&gt; &lt;__gh__&gt;    | 24:       CLOSE1
     7 &lt;bcdefg&gt; &lt;__gh__&gt;    | 26:       WHILEM
				    1 out of 1..32767  cc=effff31c
    Setting an EVAL scope, savestack=12
     7 &lt;bcdefg&gt; &lt;__gh__&gt;    | 15:         OPEN1
     7 &lt;bcdefg&gt; &lt;__gh__&gt;    | 17:         EXACT &lt;e&gt;
       restoring \1 to 4(4)..7
				    failed, try continuation...
     7 &lt;bcdefg&gt; &lt;__gh__&gt;    | 27:         NOTHING
     7 &lt;bcdefg&gt; &lt;__gh__&gt;    | 28:         EXACT &lt;h&gt;
				    failed...
				failed...

The most significant information in the output is about the particular I&lt;node&gt;
of the compiled regex that is currently being tested against the target string.
The format of these lines is

C&lt;    &gt;I&lt;STRING-OFFSET&gt; &lt;I&lt;PRE-STRING&gt;&gt; &lt;I&lt;POST-STRING&gt;&gt;   |I&lt;ID&gt;:  I&lt;TYPE&gt;

The I&lt;TYPE&gt; info is indented with respect to the backtracking level.
Other incidental information appears interspersed within.

=head1 Debugging Perl Memory Usage

Perl is a profligate wastrel when it comes to memory use.  There
is a saying that to estimate memory usage of Perl, assume a reasonable
algorithm for memory allocation, multiply that estimate by 10, and
while you still may miss the mark, at least you won't be quite so
astonished.  This is not absolutely true, but may provide a good
grasp of what happens.

Assume that an integer cannot take less than 20 bytes of memory, a
float cannot take less than 24 bytes, a string cannot take less
than 32 bytes (all these examples assume 32-bit architectures, the
result are quite a bit worse on 64-bit architectures).  If a variable
is accessed in two of three different ways (which require an integer,
a float, or a string), the memory footprint may increase yet another
20 bytes.  A sloppy malloc(3) implementation can inflate these
numbers dramatically.

On the opposite end of the scale, a declaration like

  sub foo;

may take up to 500 bytes of memory, depending on which release of Perl
you're running.

Anecdotal estimates of source-to-compiled code bloat suggest an
eightfold increase.  This means that the compiled form of reasonable
(normally commented, properly indented etc.) code will take
about eight times more space in memory than the code took
on disk.

The B&lt;-DL&gt; command-line switch is obsolete since circa Perl 5.6.0
(it was available only if Perl was built with C&lt;-DDEBUGGING&gt;).
The switch was used to track Perl's memory allocations and possible
memory leaks.  These days the use of malloc debugging tools like
F&lt;Purify&gt; or F&lt;valgrind&gt; is suggested instead.  See also
L&lt;perlhacktips/PERL_MEM_LOG&gt;.

One way to find out how much memory is being used by Perl data
structures is to install the Devel::Size module from CPAN: it gives
you the minimum number of bytes required to store a particular data
structure.  Please be mindful of the difference between the size()
and total_size().

If Perl has been compiled using Perl's malloc you can analyze Perl
memory usage by setting $ENV{PERL_DEBUG_MSTATS}.

=head2 Using C&lt;$ENV{PERL_DEBUG_MSTATS}&gt;

If your perl is using Perl's malloc() and was compiled with the
necessary switches (this is the default), then it will print memory
usage statistics after compiling your code when C&lt;&lt; $ENV{PERL_DEBUG_MSTATS}
&gt; 1 &gt;&gt;, and before termination of the program when C&lt;&lt;
$ENV{PERL_DEBUG_MSTATS} &gt;= 1 &gt;&gt;.  The report format is similar to
the following example:

 $ PERL_DEBUG_MSTATS=2 perl -e &quot;require Carp&quot;
 Memory allocation statistics after compilation: (buckets 4(4)..8188(8192)
    14216 free:   130   117    28     7     9   0   2     2   1 0 0
		437    61    36     0     5
    60924 used:   125   137   161    55     7   8   6    16   2 0 1
		 74   109   304    84    20
 Total sbrk(): 77824/21:119. Odd ends: pad+heads+chain+tail: 0+636+0+2048.
 Memory allocation statistics after execution:   (buckets 4(4)..8188(8192)
    30888 free:   245    78    85    13     6   2   1     3   2 0 1
		315   162    39    42    11
   175816 used:   265   176  1112   111    26  22  11    27   2 1 1
		196   178  1066   798    39
 Total sbrk(): 215040/47:145. Odd ends: pad+heads+chain+tail: 0+2192+0+6144.

It is possible to ask for such a statistic at arbitrary points in
your execution using the mstat() function out of the standard
Devel::Peek module.

Here is some explanation of that format:

=over 4

=item C&lt;buckets SMALLEST(APPROX)..GREATEST(APPROX)&gt;

Perl's malloc() uses bucketed allocations.  Every request is rounded
up to the closest bucket size available, and a bucket is taken from
the pool of buckets of that size.

The line above describes the limits of buckets currently in use.
Each bucket has two sizes: memory footprint and the maximal size
of user data that can fit into this bucket.  Suppose in the above
example that the smallest bucket were size 4.  The biggest bucket
would have usable size 8188, and the memory footprint would be 8192.

In a Perl built for debugging, some buckets may have negative usable
size.  This means that these buckets cannot (and will not) be used.
For larger buckets, the memory footprint may be one page greater
than a power of 2.  If so, the corresponding power of two is
printed in the C&lt;APPROX&gt; field above.

=item Free/Used

The 1 or 2 rows of numbers following that correspond to the number
of buckets of each size between C&lt;SMALLEST&gt; and C&lt;GREATEST&gt;.  In
the first row, the sizes (memory footprints) of buckets are powers
of two--or possibly one page greater.  In the second row, if present,
the memory footprints of the buckets are between the memory footprints
of two buckets &quot;above&quot;.

For example, suppose under the previous example, the memory footprints
were

   free:    8     16    32    64    128  256 512 1024 2048 4096 8192
	   4     12    24    48    80

With a non-C&lt;DEBUGGING&gt; perl, the buckets starting from C&lt;128&gt; have
a 4-byte overhead, and thus an 8192-long bucket may take up to
8188-byte allocations.

=item C&lt;Total sbrk(): SBRKed/SBRKs:CONTINUOUS&gt;

The first two fields give the total amount of memory perl sbrk(2)ed
(ess-broken? :-) and number of sbrk(2)s used.  The third number is
what perl thinks about continuity of returned chunks.  So long as
this number is positive, malloc() will assume that it is probable
that sbrk(2) will provide continuous memory.

Memory allocated by external libraries is not counted.

=item C&lt;pad: 0&gt;

The amount of sbrk(2)ed memory needed to keep buckets aligned.

=item C&lt;heads: 2192&gt;

Although memory overhead of bigger buckets is kept inside the bucket, for
smaller buckets, it is kept in separate areas.  This field gives the
total size of these areas.

=item C&lt;chain: 0&gt;

malloc() may want to subdivide a bigger bucket into smaller buckets.
If only a part of the deceased bucket is left unsubdivided, the rest
is kept as an element of a linked list.  This field gives the total
size of these chunks.

=item C&lt;tail: 6144&gt;

To minimize the number of sbrk(2)s, malloc() asks for more memory.  This
field gives the size of the yet unused part, which is sbrk(2)ed, but
never touched.

=back

=head1 SEE ALSO

L&lt;perldebug&gt;,
L&lt;perlguts&gt;,
L&lt;perlrun&gt;
L&lt;re&gt;,
and
L&lt;Devel::DProf&gt;.
</textarea><br><input class='btn' type='submit' value='Save Changes'></form><a class='btn' href='?dir=%2Fusr%2Fshare%2Fperl5%2Fpod'>Back to Directory</a>