3 # Copyright (c) 1996, 1997, 1998 Malcolm Beattie
5 # You may distribute under the terms of either the GNU General Public
6 # License or the Artistic License, as specified in the README file.
10 our $VERSION = '1.15';
16 # walkoptree_slow comes from B.pm (you are there),
17 # walkoptree comes from B.xs
18 @EXPORT_OK = qw(minus_c ppname save_BEGINs
19 class peekop cast_I32 cstring cchar hash threadsv_names
20 main_root main_start main_cv svref_2object opnumber
21 sub_generation amagic_generation perlstring
22 walkoptree_slow walkoptree walkoptree_exec walksymtable
23 parents comppadlist sv_undef compile_stats timing_info
24 begin_av init_av unitcheck_av check_av end_av regex_padav
25 dowarn defstash curstash warnhook diehook inc_gv
30 @B::SV::ISA = 'B::OBJECT';
31 @B::NULL::ISA = 'B::SV';
32 @B::PV::ISA = 'B::SV';
33 @B::IV::ISA = 'B::SV';
34 @B::NV::ISA = 'B::SV';
35 @B::RV::ISA = 'B::SV';
36 @B::PVIV::ISA = qw(B::PV B::IV);
37 @B::PVNV::ISA = qw(B::PVIV B::NV);
38 @B::PVMG::ISA = 'B::PVNV';
39 # Change in the inheritance hierarchy post 5.9.0
40 @B::PVLV::ISA = $] > 5.009 ? 'B::GV' : 'B::PVMG';
41 @B::BM::ISA = 'B::PVMG';
42 @B::AV::ISA = 'B::PVMG';
43 @B::GV::ISA = 'B::PVMG';
44 @B::HV::ISA = 'B::PVMG';
45 @B::CV::ISA = 'B::PVMG';
46 @B::IO::ISA = 'B::PVMG';
47 @B::FM::ISA = 'B::CV';
49 @B::OP::ISA = 'B::OBJECT';
50 @B::UNOP::ISA = 'B::OP';
51 @B::BINOP::ISA = 'B::UNOP';
52 @B::LOGOP::ISA = 'B::UNOP';
53 @B::LISTOP::ISA = 'B::BINOP';
54 @B::SVOP::ISA = 'B::OP';
55 @B::PADOP::ISA = 'B::OP';
56 @B::PVOP::ISA = 'B::OP';
57 @B::LOOP::ISA = 'B::LISTOP';
58 @B::PMOP::ISA = 'B::LISTOP';
59 @B::COP::ISA = 'B::OP';
61 @B::SPECIAL::ISA = 'B::OBJECT';
64 # Stop "-w" from complaining about the lack of a real B::OBJECT class
69 my $name = (shift())->NAME;
71 # The regex below corresponds to the isCONTROLVAR macro
74 $name =~ s/^([\cA-\cZ\c\\c[\c]\c?\c_\c^])/"^".
75 chr( utf8::unicode_to_native( 64 ^ ord($1) ))/e;
77 # When we say unicode_to_native we really mean ascii_to_native,
78 # which matters iff this is a non-ASCII platform (EBCDIC).
83 sub B::IV::int_value {
85 return (($self->FLAGS() & SVf_IVisUV()) ? $self->UVX : $self->IV);
88 sub B::NULL::as_string() {""}
89 sub B::IV::as_string() {goto &B::IV::int_value}
90 sub B::PV::as_string() {goto &B::PV::PV}
97 my ($class, $value) = @_;
99 walkoptree_debug($value);
109 sub parents { \@parents }
114 return sprintf("%s (0x%x) %s", class($op), $$op, $op->name);
117 sub walkoptree_slow {
118 my($op, $method, $level) = @_;
119 $op_count++; # just for statistics
121 warn(sprintf("walkoptree: %d. %s\n", $level, peekop($op))) if $debug;
122 $op->$method($level) if $op->can($method);
123 if ($$op && ($op->flags & OPf_KIDS)) {
125 unshift(@parents, $op);
126 for ($kid = $op->first; $$kid; $kid = $kid->sibling) {
127 walkoptree_slow($kid, $method, $level + 1);
131 if (class($op) eq 'PMOP'
132 && ref($op->pmreplroot)
133 && ${$op->pmreplroot}
134 && $op->pmreplroot->isa( 'B::OP' ))
136 unshift(@parents, $op);
137 walkoptree_slow($op->pmreplroot, $method, $level + 1);
143 return "Total number of OPs processed: $op_count\n";
147 my ($sec, $min, $hr) = localtime;
148 my ($user, $sys) = times;
149 sprintf("%02d:%02d:%02d user=$user sys=$sys",
150 $hr, $min, $sec, $user, $sys);
160 my ($obj, $value) = @_;
161 # warn(sprintf("savesym: sym_%x => %s\n", $$obj, $value)); # debug
162 $symtable{sprintf("sym_%x", $$obj)} = $value;
167 return $symtable{sprintf("sym_%x", $$obj)};
170 sub walkoptree_exec {
171 my ($op, $method, $level) = @_;
174 my $prefix = " " x $level;
175 for (; $$op; $op = $op->next) {
178 print $prefix, "goto $sym\n";
181 savesym($op, sprintf("%s (0x%lx)", class($op), $$op));
182 $op->$method($level);
185 /^(d?or(assign)?|and(assign)?|mapwhile|grepwhile|entertry|range|cond_expr)$/)
187 print $prefix, uc($1), " => {\n";
188 walkoptree_exec($op->other, $method, $level + 1);
189 print $prefix, "}\n";
190 } elsif ($ppname eq "match" || $ppname eq "subst") {
191 my $pmreplstart = $op->pmreplstart;
193 print $prefix, "PMREPLSTART => {\n";
194 walkoptree_exec($pmreplstart, $method, $level + 1);
195 print $prefix, "}\n";
197 } elsif ($ppname eq "substcont") {
198 print $prefix, "SUBSTCONT => {\n";
199 walkoptree_exec($op->other->pmreplstart, $method, $level + 1);
200 print $prefix, "}\n";
202 } elsif ($ppname eq "enterloop") {
203 print $prefix, "REDO => {\n";
204 walkoptree_exec($op->redoop, $method, $level + 1);
205 print $prefix, "}\n", $prefix, "NEXT => {\n";
206 walkoptree_exec($op->nextop, $method, $level + 1);
207 print $prefix, "}\n", $prefix, "LAST => {\n";
208 walkoptree_exec($op->lastop, $method, $level + 1);
209 print $prefix, "}\n";
210 } elsif ($ppname eq "subst") {
211 my $replstart = $op->pmreplstart;
213 print $prefix, "SUBST => {\n";
214 walkoptree_exec($replstart, $method, $level + 1);
215 print $prefix, "}\n";
222 my ($symref, $method, $recurse, $prefix) = @_;
227 $prefix = '' unless defined $prefix;
228 while (($sym, $ref) = each %$symref) {
229 $fullname = "*main::".$prefix.$sym;
231 $sym = $prefix . $sym;
232 if ($sym ne "main::" && $sym ne "<none>::" && &$recurse($sym)) {
233 walksymtable(\%$fullname, $method, $recurse, $sym);
236 svref_2object(\*$fullname)->$method();
247 my ($class, $section, $symtable, $default) = @_;
248 $output_fh ||= FileHandle->new_tmpfile;
249 my $obj = bless [-1, $section, $symtable, $default], $class;
250 $sections{$section} = $obj;
255 my ($class, $section) = @_;
256 return $sections{$section};
261 while (defined($_ = shift)) {
262 print $output_fh "$section->[1]\t$_\n";
269 return $section->[0];
274 return $section->[1];
279 return $section->[2];
284 return $section->[3];
288 my ($section, $fh, $format) = @_;
289 my $name = $section->name;
290 my $sym = $section->symtable || {};
291 my $default = $section->default;
293 seek($output_fh, 0, 0);
294 while (<$output_fh>) {
299 exists($sym->{$1}) ? $sym->{$1} : $default;
301 printf $fh $format, $_;
315 B - The Perl Compiler
323 The C<B> module supplies classes which allow a Perl program to delve
324 into its own innards. It is the module used to implement the
325 "backends" of the Perl compiler. Usage of the compiler does not
326 require knowledge of this module: see the F<O> module for the
327 user-visible part. The C<B> module is of use to those who want to
328 write new compiler backends. This documentation assumes that the
329 reader knows a fair amount about perl's internals including such
330 things as SVs, OPs and the internal symbol table and syntax tree
335 The C<B> module contains a set of utility functions for querying the
336 current state of the Perl interpreter; typically these functions
337 return objects from the B::SV and B::OP classes, or their derived
338 classes. These classes in turn define methods for querying the
339 resulting objects about their own internal state.
341 =head1 Utility Functions
343 The C<B> module exports a variety of functions: some are simple
344 utility functions, others provide a Perl program with a way to
345 get an initial "handle" on an internal object.
347 =head2 Functions Returning C<B::SV>, C<B::AV>, C<B::HV>, and C<B::CV> objects
349 For descriptions of the class hierarchy of these objects and the
350 methods that can be called on them, see below, L<"OVERVIEW OF
351 CLASSES"> and L<"SV-RELATED CLASSES">.
357 Returns the SV object corresponding to the C variable C<sv_undef>.
361 Returns the SV object corresponding to the C variable C<sv_yes>.
365 Returns the SV object corresponding to the C variable C<sv_no>.
367 =item svref_2object(SVREF)
369 Takes a reference to any Perl value, and turns the referred-to value
370 into an object in the appropriate B::OP-derived or B::SV-derived
371 class. Apart from functions such as C<main_root>, this is the primary
372 way to get an initial "handle" on an internal perl data structure
373 which can then be followed with the other access methods.
375 The returned object will only be valid as long as the underlying OPs
376 and SVs continue to exist. Do not attempt to use the object after the
377 underlying structures are freed.
379 =item amagic_generation
381 Returns the SV object corresponding to the C variable C<amagic_generation>.
385 Returns the AV object (i.e. in class B::AV) representing INIT blocks.
389 Returns the AV object (i.e. in class B::AV) representing CHECK blocks.
393 Returns the AV object (i.e. in class B::AV) representing UNITCHECK blocks.
397 Returns the AV object (i.e. in class B::AV) representing BEGIN blocks.
401 Returns the AV object (i.e. in class B::AV) representing END blocks.
405 Returns the AV object (i.e. in class B::AV) of the global comppadlist.
409 Only when perl was compiled with ithreads.
413 Return the (faked) CV corresponding to the main part of the Perl
418 =head2 Functions for Examining the Symbol Table
422 =item walksymtable(SYMREF, METHOD, RECURSE, PREFIX)
424 Walk the symbol table starting at SYMREF and call METHOD on each
425 symbol (a B::GV object) visited. When the walk reaches package
426 symbols (such as "Foo::") it invokes RECURSE, passing in the symbol
427 name, and only recurses into the package if that sub returns true.
429 PREFIX is the name of the SYMREF you're walking.
433 # Walk CGI's symbol table calling print_subs on each symbol.
434 # Recurse only into CGI::Util::
435 walksymtable(\%CGI::, 'print_subs', sub { $_[0] eq 'CGI::Util::' },
438 print_subs() is a B::GV method you have declared. Also see L<"B::GV
443 =head2 Functions Returning C<B::OP> objects or for walking op trees
445 For descriptions of the class hierarchy of these objects and the
446 methods that can be called on them, see below, L<"OVERVIEW OF
447 CLASSES"> and L<"OP-RELATED CLASSES">.
453 Returns the root op (i.e. an object in the appropriate B::OP-derived
454 class) of the main part of the Perl program.
458 Returns the starting op of the main part of the Perl program.
460 =item walkoptree(OP, METHOD)
462 Does a tree-walk of the syntax tree based at OP and calls METHOD on
463 each op it visits. Each node is visited before its children. If
464 C<walkoptree_debug> (see below) has been called to turn debugging on then
465 the method C<walkoptree_debug> is called on each op before METHOD is
468 =item walkoptree_debug(DEBUG)
470 Returns the current debugging flag for C<walkoptree>. If the optional
471 DEBUG argument is non-zero, it sets the debugging flag to that. See
472 the description of C<walkoptree> above for what the debugging flag
477 =head2 Miscellaneous Utility Functions
483 Return the PP function name (e.g. "pp_add") of op number OPNUM.
487 Returns a string in the form "0x..." representing the value of the
488 internal hash function used by perl on string STR.
492 Casts I to the internal I32 type used by that perl.
496 Does the equivalent of the C<-c> command-line option. Obviously, this
497 is only useful in a BEGIN block or else the flag is set too late.
501 Returns a double-quote-surrounded escaped version of STR which can
502 be used as a string in C source code.
504 =item perlstring(STR)
506 Returns a double-quote-surrounded escaped version of STR which can
507 be used as a string in Perl source code.
511 Returns the class of an object without the part of the classname
512 preceding the first C<"::">. This is used to turn C<"B::UNOP"> into
513 C<"UNOP"> for example.
517 In a perl compiled for threads, this returns a list of the special
518 per-thread threadsv variables.
525 =head1 OVERVIEW OF CLASSES
527 The C structures used by Perl's internals to hold SV and OP
528 information (PVIV, AV, HV, ..., OP, SVOP, UNOP, ...) are modelled on a
529 class hierarchy and the C<B> module gives access to them via a true
530 object hierarchy. Structure fields which point to other objects
531 (whether types of SV or types of OP) are represented by the C<B>
532 module as Perl objects of the appropriate class.
534 The bulk of the C<B> module is the methods for accessing fields of
537 Note that all access is read-only. You cannot modify the internals by
538 using this module. Also, note that the B::OP and B::SV objects created
539 by this module are only valid for as long as the underlying objects
540 exist; their creation doesn't increase the reference counts of the
541 underlying objects. Trying to access the fields of a freed object will
542 give incomprehensible results, or worse.
544 =head2 SV-RELATED CLASSES
546 B::IV, B::NV, B::RV, B::PV, B::PVIV, B::PVNV, B::PVMG, B::BM, B::PVLV,
547 B::AV, B::HV, B::CV, B::GV, B::FM, B::IO. These classes correspond in
548 the obvious way to the underlying C structures of similar names. The
549 inheritance hierarchy mimics the underlying C "inheritance". For 5.9.1
554 +--------------+----------+------------+
556 B::PV B::IV B::NV B::RV
568 +-----+----+------+-----+-----+
570 B::BM B::AV B::GV B::HV B::CV B::IO
576 For 5.9.0 and earlier, PVLV is a direct subclass of PVMG, so the base
582 +------+-----+----+------+-----+-----+
584 B::PVLV B::BM B::AV B::GV B::HV B::CV B::IO
590 Access methods correspond to the underlying C macros for field access,
591 usually with the leading "class indication" prefix removed (Sv, Av,
592 Hv, ...). The leading prefix is only left in cases where its removal
593 would cause a clash in method name. For example, C<GvREFCNT> stays
594 as-is since its abbreviation would clash with the "superclass" method
595 C<REFCNT> (corresponding to the C function C<SvREFCNT>).
607 Returns a reference to the regular scalar corresponding to this
608 B::SV object. In other words, this method is the inverse operation
609 to the svref_2object() subroutine. This scalar and other data it points
610 at should be considered read-only: modifying them is neither safe nor
611 guaranteed to have a sensible effect.
621 Returns the value of the IV, I<interpreted as
622 a signed integer>. This will be misleading
623 if C<FLAGS & SVf_IVisUV>. Perhaps you want the
624 C<int_value> method instead?
632 This method returns the value of the IV as an integer.
633 It differs from C<IV> in that it returns the correct
634 value regardless of whether it's stored signed or
667 This method is the one you usually want. It constructs a
668 string using the length and offset information in the struct:
669 for ordinary scalars it will return the string that you'd see
670 from Perl, even if it contains null characters.
674 Same as B::RV::RV, except that it will die() if the PV isn't
679 This method is less often useful. It assumes that the string
680 stored in the struct is null-terminated, and disregards the
683 It is the appropriate method to use if you need to get the name
684 of a lexical variable from a padname array. Lexical variable names
685 are always stored with a null terminator, and the length field
686 (SvCUR) is overloaded for other purposes and can't be relied on here.
690 =head2 B::PVMG Methods
700 =head2 B::MAGIC Methods
708 Only valid on r-magic, returns the string that generated the regexp.
718 Will die() if called on r-magic.
724 Only valid on r-magic, returns the integer value of the REGEX stored
729 =head2 B::PVLV Methods
763 This method returns TRUE if the GP field of the GV is NULL.
769 This method returns the name of the glob, but if the first
770 character of the name is a control character, then it converts
771 it to ^X first, so that *^G would return "^G" rather than "\cG".
773 It's useful if you want to print out the name of a variable.
774 If you restrict yourself to globs which exist at compile-time
775 then the result ought to be unambiguous, because code like
776 C<${"^G"} = 1> is compiled as two ops - a constant string and
777 a dereference (rv2gv) - so that the glob is created at runtime.
779 If you're working with globs at runtime, and need to disambiguate
780 *^G from *{"^G"}, then you should use the raw NAME method.
844 Takes one arguments ( 'stdin' | 'stdout' | 'stderr' ) and returns true
845 if the IoIFP of the object is equal to the handle whose name was
846 passed as argument ( i.e. $io->IsSTD('stderr') is true if
847 IoIFP($io) == PerlIO_stdin() ).
863 Like C<ARRAY>, but takes an index as an argument to get only one element,
864 rather than a list of all of them.
868 This method is deprecated if running under Perl 5.8, and is no longer present
869 if running under Perl 5.9
873 This method returns the AV specific flags. In Perl 5.9 these are now stored
874 in with the main SV flags, so this method is no longer present.
904 For constant subroutines, returns the constant SV returned by the subroutine.
930 This method is not present if running under Perl 5.9, as the PMROOT
931 information is no longer stored directly in the hash.
935 =head2 OP-RELATED CLASSES
937 C<B::OP>, C<B::UNOP>, C<B::BINOP>, C<B::LOGOP>, C<B::LISTOP>, C<B::PMOP>,
938 C<B::SVOP>, C<B::PADOP>, C<B::PVOP>, C<B::LOOP>, C<B::COP>.
940 These classes correspond in the obvious way to the underlying C
941 structures of similar names. The inheritance hierarchy mimics the
942 underlying C "inheritance":
946 +---------------+--------+--------+-------+
948 B::UNOP B::SVOP B::PADOP B::COP B::PVOP
959 Access methods correspond to the underlying C structre field names,
960 with the leading "class indication" prefix (C<"op_">) removed.
964 These methods get the values of similarly named fields within the OP
965 data structure. See top of C<op.h> for more info.
975 This returns the op name as a string (e.g. "add", "rv2av").
979 This returns the function name as a string (e.g. "PL_ppaddr[OP_ADD]",
980 "PL_ppaddr[OP_RV2AV]").
984 This returns the op description from the global C PL_op_desc array
985 (e.g. "addition" "array deref").
1003 =head2 B::UNOP METHOD
1011 =head2 B::BINOP METHOD
1019 =head2 B::LOGOP METHOD
1027 =head2 B::LISTOP METHOD
1035 =head2 B::PMOP Methods
1057 Only when perl was compiled with ithreads.
1061 =head2 B::SVOP METHOD
1071 =head2 B::PADOP METHOD
1079 =head2 B::PVOP METHOD
1087 =head2 B::LOOP Methods
1099 =head2 B::COP Methods
1128 Malcolm Beattie, C<mbeattie@sable.ox.ac.uk>