Writing Your Own Debugger¶

Environment Variables

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

  sub DB::DB {}

It can easily be defined like this:

  $ PERL5DB="sub DB::DB {}" perl -d your-script

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

  sub DB::DB {print ++$i; scalar <STDIN>}

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, " $sub\n"; &$sub}
  }

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

When it starts, the debugger reads your rc file (./.perldb or ~/.perldb under Unix), which can set important options. (A subroutine (&afterinit) 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 "o ..." debugger command (q.v. in "Configurable Options" in perldebug).

Debugger Internal Variables

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

• @DB::dbline is an alias for "@{"::_<current_file"}", which holds the lines of the currently-selected file (compiled by Perl), either explicitly chosen with the debugger's "f" 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.

• %DB::dbline is an alias for "%{"::_<current_file"}", which contains breakpoints and actions keyed by line number in the currently-selected file, either explicitly chosen with the debugger's "f" 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 perl5db.pl have the form "$break_condition\0$action".

Debugger Customization Functions

Some functions are provided to simplify customization.

• See "Configurable Options" in perldebug for a description of options parsed by "DB::parse_options(string)".
• "DB::dump_trace(skip[,count])" skips the specified number of frames and returns a list containing information about the calling frames (all of them, if "count" is missing). Each entry is reference to a hash with keys "context" (either ".", "$", or "@"), "sub" (subroutine name, or info about "eval"), "args" ("undef" or a reference to an array), "file", and "line".
• "DB::print_trace(FH, skip[, count[, short]])" prints formatted info about caller frames. The last two functions may be convenient as arguments to "<", "<<" commands.

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

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 <d>(14)
    14: CURLYX[0] {1,32767}(28)
    16:   OPEN1(18)
    18:     EXACT <e>(20)
    20:     STAR(23)
    21:       EXACT <f>(0)
    23:     EXACT <g>(25)
    25:   CLOSE1(27)
    27:   WHILEM[1/1](0)
    28: NOTHING(29)
    29: EXACT <h>(31)
    31: ANYOF[ij](42)
    42: EXACT <k>(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 $` $& $' 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+"size"*8. The next line shows the label id 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 "de" at offset 1, plus substring "gh" at some offset between 3 and infinity. Moreover, when checking for these substrings (to abandon impossible matches quickly), Perl will check for the substring "gh" before checking for the substring "de". The optimizer may also use the knowledge that the match starts (at the "first" id) 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

"anchored" STRING "at" POS

"floating" STRING "at" POS1..POS2

See above.

"matching floating/anchored"

Which substring to check first.

"minlen"

The minimal length of the match.

"stclass" TYPE

Type of first matching node.

"noscan"

Don't scan for the found substrings.

"isall"

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.

"GPOS"

Set if the pattern contains "\G".

"plus"

Set if the pattern starts with a repeated char (as in "x+y").

"implicit"

Set if the pattern starts with ".*".

"with eval"

Set if the pattern contain eval-groups, such as "(?{ code })" and "(??{ code })".

"anchored(TYPE)"

If the pattern may match only at a handful of places, with "TYPE" being "BOL", "MBOL", or "GPOS". See the table below.

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

The optimizer-specific information is used to avoid entering (a slow) regex engine on strings that will not definitely match. If the "isall" 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 nodes of the compiled form of the regex. Each line has format

" "id: TYPE OPTIONAL-INFO (next-id)

Types of Nodes¶

Here are the possible types, with short descriptions:

 # TYPE arg-description [num-args] [longjump-len] DESCRIPTION
 # Exit points
 END   no        End of program.
 SUCCEED   no        Return from a subroutine, basically.
 # Anchors:
 BOL        no      Match "" at beginning of line.
 MBOL       no      Same, assuming multiline.
 SBOL       no      Same, assuming singleline.
 EOS        no      Match "" at end of string.
 EOL        no      Match "" at end of line.
 MEOL       no      Same, assuming multiline.
 SEOL       no      Same, assuming singleline.
 BOUND      no      Match "" at any word boundary using native charset
                    semantics for non-utf8
 BOUNDL     no      Match "" at any locale word boundary
 BOUNDU     no      Match "" at any word boundary using Unicode semantics
 BOUNDA     no      Match "" at any word boundary using ASCII semantics
 NBOUND     no      Match "" at any word non-boundary using native charset
                    semantics for non-utf8
 NBOUNDL    no      Match "" at any locale word non-boundary
 NBOUNDU    no      Match "" at any word non-boundary using Unicode semantics
 NBOUNDA    no      Match "" at any word non-boundary using ASCII semantics
 GPOS       no      Matches where last m//g left off.
 # [Special] alternatives:
 REG_ANY    no      Match any one character (except newline).
 SANY       no      Match any one character.
 CANY       no      Match any one byte.
 ANYOF      sv      Match character in (or not in) this class, single char
                    match only
 ANYOFV     sv      Match character in (or not in) this class, can
                    match-multiple chars
 ALNUM      no      Match any alphanumeric character using native charset
                    semantics for non-utf8
 ALNUML     no      Match any alphanumeric char in locale
 ALNUMU     no      Match any alphanumeric char using Unicode semantics
 ALNUMA     no      Match [A-Za-z_0-9]
 NALNUM     no      Match any non-alphanumeric character using native charset
                    semantics for non-utf8
 NALNUML    no      Match any non-alphanumeric char in locale
 NALNUMU    no      Match any non-alphanumeric char using Unicode semantics
 NALNUMA    no      Match [^A-Za-z_0-9]
 SPACE      no      Match any whitespace character using native charset
                    semantics for non-utf8
 SPACEL     no      Match any whitespace char in locale
 SPACEU     no      Match any whitespace char using Unicode semantics
 SPACEA     no      Match [ \t\n\f\r]
 NSPACE     no      Match any non-whitespace character using native charset
                    semantics for non-utf8
 NSPACEL    no      Match any non-whitespace char in locale
 NSPACEU    no      Match any non-whitespace char using Unicode semantics
 NSPACEA    no      Match [^ \t\n\f\r]
 DIGIT      no      Match any numeric character using native charset semantics
                    for non-utf8
 DIGITL     no      Match any numeric character in locale
 DIGITA     no      Match [0-9]
 NDIGIT     no      Match any non-numeric character using native charset
 i                  semantics for non-utf8
 NDIGITL    no      Match any non-numeric character in locale
 NDIGITA    no      Match [^0-9]
 CLUMP      no      Match any extended grapheme cluster sequence
 # Alternation
 # BRANCH        The set of branches constituting a single choice are hooked
 #               together with their "next" pointers, since precedence prevents
 #               anything being concatenated to any individual branch.  The
 #               "next" pointer of the last BRANCH in a choice points to the
 #               thing following the whole choice.  This is also where the
 #               final "next" 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...
 # Back pointer
 # BACK          Normal "next" pointers all implicitly point forward; BACK
 #               exists to make loop structures possible.
 # not used
 BACK       no      Match "", "next" ptr points backward.
 # Literals
 EXACT      str     Match this string (preceded by length).
 EXACTF     str     Match this string, folded, native charset semantics for
                    non-utf8 (prec. by length).
 EXACTFL    str     Match this string, folded in locale (w/len).
 EXACTFU    str     Match this string, folded, Unicode semantics for non-utf8
                    (prec. by length).
 EXACTFA    str     Match this string, folded, Unicode semantics for non-utf8,
                    but no ASCII-range character matches outside ASCII (prec.
                    by length),.
 # 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 using BACK.  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
                    semantics for non-utf8
 REFFL      num 1   Match already matched string, folded in loc.
 REFFU      num 1   Match already matched string, folded using unicode
                    semantics for non-utf8
 REFFA      num 1   Match already matched string, folded using unicode
                    semantics 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
                    semantics for non-utf8
 NREFFL     no-sv 1 Match already matched string, folded in loc.
 NREFFU     num   1 Match already matched string, folded using unicode
                    semantics for non-utf8
 NREFFA     num   1 Match already matched string, folded using unicode
                    semantics for non-utf8, no mixing ASCII, non-ASCII
 IFMATCH    off 1 2 Succeeds if the following matches.
 UNLESSM    off 1 2 Fails if the following matches.
 SUSPEND    off 1 1 "Independent" sub-RE.
 IFTHEN     off 1 1 Switch, should be preceded by switcher.
 GROUPP     num 1   Whether the group matched.
 # Support for long RE
 LONGJMP    off 1 1 Jump far away.
 BRANCHJ    off 1 1 BRANCH with long offset.
 # The heavy worker
 EVAL       evl 1   Execute some Perl code.
 # 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.
 # NOTE: the relative order of the TRIE-like regops  is significant
 TRIE       trie 1    Match many EXACT(F[ALU]?)? at once. flags==type
 TRIEC      charclass Same as TRIE, but with embedded charclass data
 # For start classes, contains an added fail table.
 AHOCORASICK trie 1   Aho Corasick stclass. flags==type
 AHOCORASICKC charclass Same as AHOCORASICK, but with embedded charclass data
 # Regex Subroutines
 GOSUB      num/ofs 2L recurse to paren arg1 at (signed) ofs arg2
 GOSTART    no         recurse to start of pattern
 # 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     none      Same as (?!)
 ACCEPT     parno 1   Accepts the current matched string.
 # 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 $&.
 KEEPS      no        $& begins here.
 # New charclass like patterns
 LNBREAK    none      generic newline pattern
 VERTWS     none      vertical whitespace         (Perl 6)
 NVERTWS    none      not vertical whitespace     (Perl 6)
 HORIZWS    none      horizontal whitespace       (Perl 6)
 NHORIZWS   none      not horizontal whitespace   (Perl 6)
 FOLDCHAR   codepoint 1 codepoint with tricky case folding properties.
 # SPECIAL  REGOPS
 # This is not really a node, but an optimized away piece of a "long" 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 "unoptimizable" and PSEUDO mean
 # "not seen anything to optimize yet".
 PSEUDO     off       Pseudo opcode for internal use.

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 "offset[length]". Entries are numbered starting with 1, so entry #1 here is "1[4]" and entry #12 is "5[1]". "1[4]" indicates that the node labeled "1:" (the "1: ANYOF[bc]") begins at character position 1 in the pre-compiled form of the regex, and has a length of 4 characters. "5[1]" in position 12 indicates that the node labeled "12:" (the "12: EXACT <d>") begins at character position 5 in the pre-compiled form of the regex, and has a length of 1 character. "12[1]" in position 14 indicates that the node labeled "14:" (the "14: CURLYX[0] {1,32767}") 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 "+" symbol in the precompiled regex.

"0[0]" items indicate that there is no corresponding node.

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 <ab> <cdefg__gh_>    |  1: ANYOF
     3 <abc> <defg__gh_>    | 11: EXACT <d>
     4 <abcd> <efg__gh_>    | 13: CURLYX {1,32767}
     4 <abcd> <efg__gh_>    | 26:   WHILEM
                                0 out of 1..32767  cc=effff31c
     4 <abcd> <efg__gh_>    | 15:     OPEN1
     4 <abcd> <efg__gh_>    | 17:     EXACT <e>
     5 <abcde> <fg__gh_>    | 19:     STAR
                             EXACT <f> can match 1 times out of 32767...
    Setting an EVAL scope, savestack=3
     6 <bcdef> <g__gh__>    | 22:       EXACT <g>
     7 <bcdefg> <__gh__>    | 24:       CLOSE1
     7 <bcdefg> <__gh__>    | 26:       WHILEM
                                    1 out of 1..32767  cc=effff31c
    Setting an EVAL scope, savestack=12
     7 <bcdefg> <__gh__>    | 15:         OPEN1
     7 <bcdefg> <__gh__>    | 17:         EXACT <e>
       restoring \1 to 4(4)..7
                                    failed, try continuation...
     7 <bcdefg> <__gh__>    | 27:         NOTHING
     7 <bcdefg> <__gh__>    | 28:         EXACT <h>
                                    failed...
                                failed...

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

" "STRING-OFFSET <PRE-STRING> <POST-STRING> |ID: TYPE

The TYPE info is indented with respect to the backtracking level. Other incidental information appears interspersed within.

Using $ENV{PERL_DEBUG_MSTATS}¶

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 "$ENV{PERL_DEBUG_MSTATS} > 1", and before termination of the program when "$ENV{PERL_DEBUG_MSTATS} >= 1". The report format is similar to the following example:

  $ PERL_DEBUG_MSTATS=2 perl -e "require Carp"
  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:

"buckets SMALLEST(APPROX)..GREATEST(APPROX)"

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 "APPROX" field above.

Free/Used

The 1 or 2 rows of numbers following that correspond to the number of buckets of each size between "SMALLEST" and "GREATEST". 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 "above".

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-"DEBUGGING" perl, the buckets starting from 128 have a 4-byte overhead, and thus an 8192-long bucket may take up to 8188-byte allocations.

"Total sbrk(): SBRKed/SBRKs:CONTINUOUS"

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.

"pad: 0"

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

"heads: 2192"

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.

"chain: 0"

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.

"tail: 6144"

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.