This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Avoid setting PL_cv_has_eval unnecessarily
[perl5.git] / regcomp.sym
CommitLineData
03363afd
YO
1# regcomp.sym
2#
3# File has two sections, divided by a line of dashes '-'.
4#
5# Empty rows after #-comment are removed from input are ignored
6#
7# First section is for regops, second sectionis for regmatch-states
8#
3dab1dad
YO
9# Note that the order in this file is important.
10#
03363afd 11# Format for first section:
f9ef50a7 12# NAME \s+ TYPE, arg-description [flags] [num-args] [longjump-len] ; DESCRIPTION
03363afd 13#
3dab1dad 14#
c476f425 15# run perl regen.pl after editing this file
3dab1dad 16
03363afd
YO
17
18
1de06328
YO
19#* Exit points (0,1)
20
f8abb37e
NC
21END END, no ; End of program.
22SUCCEED END, no ; Return from a subroutine, basically.
d09b2d29 23
1de06328
YO
24#* Anchors: (2..13)
25
f8abb37e
NC
26BOL BOL, no ; Match "" at beginning of line.
27MBOL BOL, no ; Same, assuming multiline.
28SBOL BOL, no ; Same, assuming singleline.
29EOS EOL, no ; Match "" at end of string.
30EOL EOL, no ; Match "" at end of line.
31MEOL EOL, no ; Same, assuming multiline.
32SEOL EOL, no ; Same, assuming singleline.
33BOUND BOUND, no ; Match "" at any word boundary
34BOUNDL BOUND, no ; Match "" at any word boundary
35NBOUND NBOUND, no ; Match "" at any word non-boundary
36NBOUNDL NBOUND, no ; Match "" at any word non-boundary
37GPOS GPOS, no ; Matches where last m//g left off.
d09b2d29 38
1de06328
YO
39#* [Special] alternatives: (14..30)
40
f9ef50a7
NC
41REG_ANY REG_ANY, no 0 S ; Match any one character (except newline).
42SANY REG_ANY, no 0 S ; Match any one character.
43CANY REG_ANY, no 0 S ; Match any one byte.
617b7ae7 44ANYOF ANYOF, sv 0 S ; Match character in (or not in) this class, folding is native charset for non-utf8.
f9ef50a7
NC
45ALNUM ALNUM, no 0 S ; Match any alphanumeric character
46ALNUML ALNUM, no 0 S ; Match any alphanumeric char in locale
47NALNUM NALNUM, no 0 S ; Match any non-alphanumeric character
48NALNUML NALNUM, no 0 S ; Match any non-alphanumeric char in locale
49SPACE SPACE, no 0 S ; Match any whitespace character
50SPACEL SPACE, no 0 S ; Match any whitespace char in locale
51NSPACE NSPACE, no 0 S ; Match any non-whitespace character
52NSPACEL NSPACE, no 0 S ; Match any non-whitespace char in locale
53DIGIT DIGIT, no 0 S ; Match any numeric character
f8abb37e 54DIGITL DIGIT, no ; Match any numeric character in locale
f9ef50a7 55NDIGIT NDIGIT, no 0 S ; Match any non-numeric character
f8abb37e 56NDIGITL NDIGIT, no ; Match any non-numeric character in locale
2448cf39 57CLUMP CLUMP, no 0 V ; Match any extended grapheme cluster sequence
d09b2d29 58
1de06328
YO
59#* Alternation (31)
60
f8abb37e
NC
61# BRANCH The set of branches constituting a single choice are hooked
62# together with their "next" pointers, since precedence prevents
63# anything being concatenated to any individual branch. The
64# "next" pointer of the last BRANCH in a choice points to the
65# thing following the whole choice. This is also where the
66# final "next" pointer of each individual branch points; each
67# branch starts with the operand node of a BRANCH node.
d09b2d29 68#
f9ef50a7 69BRANCH BRANCH, node 0 V ; Match this alternative, or the next...
d09b2d29 70
1de06328
YO
71#*Back pointer (32)
72
f8abb37e
NC
73# BACK Normal "next" pointers all implicitly point forward; BACK
74# exists to make loop structures possible.
d09b2d29 75# not used
f9ef50a7 76BACK BACK, no 0 V ; Match "", "next" ptr points backward.
d09b2d29 77
1de06328
YO
78#*Literals (33..35)
79
f8abb37e 80EXACT EXACT, str ; Match this string (preceded by length).
2448cf39 81EXACTF EXACT, str ; Match this string, folded, native charset semantics for non-utf8 (prec. by length).
f8abb37e 82EXACTFL EXACT, str ; Match this string, folded in locale (w/len).
d09b2d29 83
1de06328
YO
84#*Do nothing types (36..37)
85
f8abb37e 86NOTHING NOTHING, no ; Match empty string.
d09b2d29 87# A variant of above which delimits a group, thus stops optimizations
f8abb37e 88TAIL NOTHING, no ; Match empty string. Can jump here from outside.
d09b2d29 89
1de06328
YO
90#*Loops (38..44)
91
f8abb37e
NC
92# STAR,PLUS '?', and complex '*' and '+', are implemented as circular
93# BRANCH structures using BACK. Simple cases (one character
94# per match) are implemented with STAR and PLUS for speed
95# and to minimize recursive plunges.
d09b2d29 96#
f9ef50a7
NC
97STAR STAR, node 0 V ; Match this (simple) thing 0 or more times.
98PLUS PLUS, node 0 V ; Match this (simple) thing 1 or more times.
d09b2d29 99
f9ef50a7
NC
100CURLY CURLY, sv 2 V ; Match this simple thing {n,m} times.
101CURLYN CURLY, no 2 V ; Capture next-after-this simple thing
102CURLYM CURLY, no 2 V ; Capture this medium-complex thing {n,m} times.
103CURLYX CURLY, sv 2 V ; Match this complex thing {n,m} times.
d09b2d29
IZ
104
105# This terminator creates a loop structure for CURLYX
f9ef50a7 106WHILEM WHILEM, no 0 V ; Do curly processing and see if rest matches.
d09b2d29 107
1de06328
YO
108#*Buffer related (45..49)
109
f8abb37e
NC
110# OPEN,CLOSE,GROUPP ...are numbered at compile time.
111OPEN OPEN, num 1 ; Mark this point in input as start of #n.
112CLOSE CLOSE, num 1 ; Analogous to OPEN.
d09b2d29 113
f9ef50a7 114REF REF, num 1 V ; Match some already matched string
2448cf39 115REFF REF, num 1 V ; Match already matched string, folded using native charset semantics for non-utf8
f9ef50a7 116REFFL REF, num 1 V ; Match already matched string, folded in loc.
d09b2d29 117
1de06328 118
f9ef50a7
NC
119IFMATCH BRANCHJ, off 1 . 2 ; Succeeds if the following matches.
120UNLESSM BRANCHJ, off 1 . 2 ; Fails if the following matches.
121SUSPEND BRANCHJ, off 1 V 1 ; "Independent" sub-RE.
122IFTHEN BRANCHJ, off 1 V 1 ; Switch, should be preceeded by switcher .
f8abb37e 123GROUPP GROUPP, num 1 ; Whether the group matched.
d09b2d29 124
1de06328
YO
125#*Support for long RE (55..56)
126
f9ef50a7
NC
127LONGJMP LONGJMP, off 1 . 1 ; Jump far away.
128BRANCHJ BRANCHJ, off 1 V 1 ; BRANCH with long offset.
d09b2d29 129
5014f7da 130#*The heavy worker (57)
1de06328 131
f8abb37e 132EVAL EVAL, evl 1 ; Execute some Perl code.
d09b2d29 133
5014f7da 134#*Modifiers (58..59)
1de06328 135
f8abb37e
NC
136MINMOD MINMOD, no ; Next operator is not greedy.
137LOGICAL LOGICAL, no ; Next opcode should set the flag only.
d09b2d29 138
5014f7da 139# This is not used yet (60)
f9ef50a7 140RENUM BRANCHJ, off 1 . 1 ; Group with independently numbered parens.
d09b2d29 141
5014f7da 142#*Trie Related (61..62)
1de06328
YO
143
144# Behave the same as A|LIST|OF|WORDS would. The '..C' variants have
145# inline charclass data (ascii only), the 'C' store it in the structure.
146# NOTE: the relative order of the TRIE-like regops is signifigant
ce5e9471 147
60c7e672 148TRIE TRIE, trie 1 ; Match many EXACT(F[LU]?)? at once. flags==type
f8abb37e 149TRIEC TRIE,trie charclass ; Same as TRIE, but with embedded charclass data
3dab1dad 150
1de06328 151# For start classes, contains an added fail table.
f8abb37e
NC
152AHOCORASICK TRIE, trie 1 ; Aho Corasick stclass. flags==type
153AHOCORASICKC TRIE,trie charclass ; Same as AHOCORASICK, but with embedded charclass data
1de06328 154
1a147d38 155#*Regex Subroutines (65..66)
f8abb37e
NC
156GOSUB GOSUB, num/ofs 2L ; recurse to paren arg1 at (signed) ofs arg2
157GOSTART GOSTART, no ; recurse to start of pattern
03363afd 158
81714fb9 159#*Named references (67..69)
f9ef50a7 160NREF REF, no-sv 1 V ; Match some already matched string
2448cf39 161NREFF REF, no-sv 1 V ; Match already matched string, folded using native charset semantics for non-utf8
f9ef50a7 162NREFFL REF, no-sv 1 V ; Match already matched string, folded in loc.
81714fb9
YO
163
164
7f69552c 165#*Special conditionals (70..72)
f8abb37e
NC
166NGROUPP NGROUPP, no-sv 1 ; Whether the group matched.
167INSUBP INSUBP, num 1 ; Whether we are in a specific recurse.
168DEFINEP DEFINEP, none 1 ; Never execute directly.
0a4db386 169
5d458dd8 170#*Bactracking Verbs
f8abb37e
NC
171ENDLIKE ENDLIKE, none ; Used only for the type field of verbs
172OPFAIL ENDLIKE, none ; Same as (?!)
173ACCEPT ENDLIKE, parno 1 ; Accepts the current matched string.
5d458dd8
YO
174
175
176#*Verbs With Arguments
f8abb37e
NC
177VERB VERB, no-sv 1 ; Used only for the type field of verbs
178PRUNE VERB, no-sv 1 ; Pattern fails at this startpoint if no-backtracking through this
179MARKPOINT VERB, no-sv 1 ; Push the current location for rollback by cut.
180SKIP VERB, no-sv 1 ; On failure skip forward (to the mark) before retrying
181COMMIT VERB, no-sv 1 ; Pattern fails outright if backtracking through this
182CUTGROUP VERB, no-sv 1 ; On failure go to the next alternation in the group
e2e6a0f1 183
ee9b8eae 184#*Control what to keep in $&.
f8abb37e 185KEEPS KEEPS, no ; $& begins here.
ee9b8eae 186
e1d1eefb 187#*New charclass like patterns
f8abb37e 188LNBREAK LNBREAK, none ; generic newline pattern
f9ef50a7
NC
189VERTWS VERTWS, none 0 S ; vertical whitespace (Perl 6)
190NVERTWS NVERTWS, none 0 S ; not vertical whitespace (Perl 6)
191HORIZWS HORIZWS, none 0 S ; horizontal whitespace (Perl 6)
192NHORIZWS NHORIZWS, none 0 S ; not horizontal whitespace (Perl 6)
e1d1eefb 193
f8abb37e 194FOLDCHAR FOLDCHAR, codepoint 1 ; codepoint with tricky case folding properties.
35f7ba49 195EXACTFU EXACT, str ; Match this string, folded, Unicode semantics for non-utf8 (prec. by length).
32e6a07c 196
7fcd3a28
KW
197# These could have been implemented using the FLAGS field of the regnode, but
198# by having a separate node type, we can use the existing switch statement to
199# avoid some tests
200REFFU REF, num 1 V ; Match already matched string, folded using unicode semantics for non-utf8
201NREFFU REF, num 1 V ; Match already matched string, folded using unicode semantics for non-utf8
202
203
e1d1eefb 204# NEW STUFF ABOVE THIS LINE
1de06328 205
03363afd
YO
206################################################################################
207
7f69552c 208#*SPECIAL REGOPS
1de06328
YO
209
210# This is not really a node, but an optimized away piece of a "long" node.
211# To simplify debugging output, we mark it as if it were a node
f8abb37e 212OPTIMIZED NOTHING, off ; Placeholder for dump.
1de06328 213
3dab1dad
YO
214# Special opcode with the property that no opcode in a compiled program
215# will ever be of this type. Thus it can be used as a flag value that
216# no other opcode has been seen. END is used similarly, in that an END
217# node cant be optimized. So END implies "unoptimizable" and PSEUDO mean
218# "not seen anything to optimize yet".
f8abb37e 219PSEUDO PSEUDO, off ; Pseudo opcode for internal use.
1de06328 220
03363afd
YO
221-------------------------------------------------------------------------------
222# Format for second section:
223# REGOP \t typelist [ \t typelist] [# Comment]
224# typelist= namelist
225# = namelist:FAIL
226# = name:count
227
228# Anything below is a state
229#
230#
f8abb37e
NC
231TRIE next:FAIL
232EVAL AB:FAIL
233CURLYX end:FAIL
234WHILEM A_pre,A_min,A_max,B_min,B_max:FAIL
235BRANCH next:FAIL
236CURLYM A,B:FAIL
237IFMATCH A:FAIL
238CURLY B_min_known,B_min,B_max:FAIL
239COMMIT next:FAIL
240MARKPOINT next:FAIL
241SKIP next:FAIL
242CUTGROUP next:FAIL
243KEEPS next:FAIL