This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Allow trie use for /iaa matching
[perl5.git] / regcomp.sym
CommitLineData
03363afd
YO
1# regcomp.sym
2#
3# File has two sections, divided by a line of dashes '-'.
4#
65aa4ca7
FC
5# Lines beginning with # are ignored, except for those that start with #*
6# which are included in pod/perldebguts.pod. # within a line may be part
7# of a description.
03363afd 8#
486ec47a 9# First section is for regops, second section is for regmatch-states
03363afd 10#
3dab1dad
YO
11# Note that the order in this file is important.
12#
03363afd 13# Format for first section:
519ebf3d
KW
14# NAME \s+ TYPE, arg-description [num-args] [flags] [longjump-len] ; DESCRIPTION
15# flag <S> means is REGNODE_SIMPLE; flag <V> means is REGNODE_VARIES
03363afd 16#
3dab1dad 17#
c476f425 18# run perl regen.pl after editing this file
3dab1dad 19
03363afd
YO
20
21
381b57db 22#* Exit points
1de06328 23
f8abb37e
NC
24END END, no ; End of program.
25SUCCEED END, no ; Return from a subroutine, basically.
d09b2d29 26
381b57db 27#* Anchors:
1de06328 28
f8abb37e
NC
29BOL BOL, no ; Match "" at beginning of line.
30MBOL BOL, no ; Same, assuming multiline.
31SBOL BOL, no ; Same, assuming singleline.
32EOS EOL, no ; Match "" at end of string.
33EOL EOL, no ; Match "" at end of line.
34MEOL EOL, no ; Same, assuming multiline.
35SEOL EOL, no ; Same, assuming singleline.
693fefec
KW
36# The regops that have varieties that vary depending on the character set regex
37# modifiers have to ordered thusly: /d, /l, /u, /a, /aa. This is because code
38# in regcomp.c uses the enum value of the modifier as an offset from the /d
39# version. The complements must come after the non-complements.
3018b823 40# BOUND, POSIX and their complements are affected, as well as EXACTF.
1e355c70 41BOUND BOUND, no ; Match "" at any word boundary using native charset semantics for non-utf8
f2284805 42BOUNDL BOUND, no ; Match "" at any locale word boundary
1e355c70 43BOUNDU BOUND, no ; Match "" at any word boundary using Unicode semantics
0c6e81eb 44BOUNDA BOUND, no ; Match "" at any word boundary using ASCII semantics
693fefec 45# All NBOUND nodes are required by code in regexec.c to be greater than all BOUND ones
1e355c70 46NBOUND NBOUND, no ; Match "" at any word non-boundary using native charset semantics for non-utf8
f2284805 47NBOUNDL NBOUND, no ; Match "" at any locale word non-boundary
1e355c70 48NBOUNDU NBOUND, no ; Match "" at any word non-boundary using Unicode semantics
0c6e81eb 49NBOUNDA NBOUND, no ; Match "" at any word non-boundary using ASCII semantics
f8abb37e 50GPOS GPOS, no ; Matches where last m//g left off.
d09b2d29 51
381b57db 52#* [Special] alternatives:
1de06328 53
f9ef50a7
NC
54REG_ANY REG_ANY, no 0 S ; Match any one character (except newline).
55SANY REG_ANY, no 0 S ; Match any one character.
56CANY REG_ANY, no 0 S ; Match any one byte.
0e019ad6 57ANYOF ANYOF, sv 0 S ; Match character in (or not in) this class, single char match only
954a2af6 58ANYOF_WARN_SUPER ANYOF, sv 0 S ; Match character in (or not in) this class, warn (if enabled) upon matching a char above Unicode max;
9aa1e39f 59ANYOF_SYNTHETIC ANYOF, sv 0 S ; Synthetic start class
693fefec 60
3018b823
KW
61# Order of the below is important. See ordering comment above.
62POSIXD POSIXD, none 0 S ; Some [[:class:]] under /d; the FLAGS field gives which one
63POSIXL POSIXD, none 0 S ; Some [[:class:]] under /l; the FLAGS field gives which one
64POSIXU POSIXD, none 0 S ; Some [[:class:]] under /u; the FLAGS field gives which one
3615ea58 65POSIXA POSIXD, none 0 S ; Some [[:class:]] under /a; the FLAGS field gives which one
3018b823
KW
66NPOSIXD NPOSIXD, none 0 S ; complement of POSIXD, [[:^class:]]
67NPOSIXL NPOSIXD, none 0 S ; complement of POSIXL, [[:^class:]]
68NPOSIXU NPOSIXD, none 0 S ; complement of POSIXU, [[:^class:]]
9e84774b 69NPOSIXA NPOSIXD, none 0 S ; complement of POSIXA, [[:^class:]]
3018b823 70# End of order is important
693fefec 71
2448cf39 72CLUMP CLUMP, no 0 V ; Match any extended grapheme cluster sequence
d09b2d29 73
381b57db 74#* Alternation
1de06328 75
65aa4ca7
FC
76#* BRANCH The set of branches constituting a single choice are
77#* hooked together with their "next" pointers, since
78#* precedence prevents anything being concatenated to
79#* any individual branch. The "next" pointer of the last
80#* BRANCH in a choice points to the thing following the
81#* whole choice. This is also where the final "next"
82#* pointer of each individual branch points; each branch
83#* starts with the operand node of a BRANCH node.
84#*
f9ef50a7 85BRANCH BRANCH, node 0 V ; Match this alternative, or the next...
d09b2d29 86
381b57db 87#*Back pointer
1de06328 88
65aa4ca7
FC
89#* BACK Normal "next" pointers all implicitly point forward;
90#* BACK exists to make loop structures possible.
91#* not used
f9ef50a7 92BACK BACK, no 0 V ; Match "", "next" ptr points backward.
d09b2d29 93
65aa4ca7
FC
94#*Literals
95# NOTE: the relative ordering of these types is important do not change it
1de06328 96
f8abb37e 97EXACT EXACT, str ; Match this string (preceded by length).
a36afff1
KW
98EXACTF EXACT, str ; Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len).
99EXACTFL EXACT, str ; Match this string (not guaranteed to be folded) using /il rules (w/len).
3c760661 100EXACTFU EXACT, 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).
8c1182fd 101EXACTFA EXACT, str ; Match this string (not guaranteed to be folded) using /iaa rules (w/len).
3c760661 102EXACTFU_SS EXACT, 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).
098b07d5 103EXACTFA_NO_TRIE EXACT, str ; Match this string (which is not trie-able; not guaranteed to be folded) using /iaa rules (w/len).
d09b2d29 104
381b57db 105#*Do nothing types
1de06328 106
f8abb37e 107NOTHING NOTHING, no ; Match empty string.
65aa4ca7 108#*A variant of above which delimits a group, thus stops optimizations
f8abb37e 109TAIL NOTHING, no ; Match empty string. Can jump here from outside.
d09b2d29 110
381b57db 111#*Loops
1de06328 112
65aa4ca7
FC
113#* STAR,PLUS '?', and complex '*' and '+', are implemented as
114#* circular BRANCH structures using BACK. Simple cases
115#* (one character per match) are implemented with STAR
116#* and PLUS for speed and to minimize recursive plunges.
117#*
f9ef50a7
NC
118STAR STAR, node 0 V ; Match this (simple) thing 0 or more times.
119PLUS PLUS, node 0 V ; Match this (simple) thing 1 or more times.
d09b2d29 120
f9ef50a7
NC
121CURLY CURLY, sv 2 V ; Match this simple thing {n,m} times.
122CURLYN CURLY, no 2 V ; Capture next-after-this simple thing
123CURLYM CURLY, no 2 V ; Capture this medium-complex thing {n,m} times.
124CURLYX CURLY, sv 2 V ; Match this complex thing {n,m} times.
d09b2d29 125
65aa4ca7 126#*This terminator creates a loop structure for CURLYX
f9ef50a7 127WHILEM WHILEM, no 0 V ; Do curly processing and see if rest matches.
d09b2d29 128
381b57db 129#*Buffer related
1de06328 130
65aa4ca7 131#*OPEN,CLOSE,GROUPP ...are numbered at compile time.
f8abb37e
NC
132OPEN OPEN, num 1 ; Mark this point in input as start of #n.
133CLOSE CLOSE, num 1 ; Analogous to OPEN.
d09b2d29 134
f9ef50a7 135REF REF, num 1 V ; Match some already matched string
2448cf39 136REFF REF, num 1 V ; Match already matched string, folded using native charset semantics for non-utf8
f9ef50a7 137REFFL REF, num 1 V ; Match already matched string, folded in loc.
781aab5c 138# N?REFF[AU] could have been implemented using the FLAGS field of the
01f98ec2
KW
139# regnode, but by having a separate node type, we can use the existing switch
140# statement to avoid some tests
141REFFU REF, num 1 V ; Match already matched string, folded using unicode semantics for non-utf8
781aab5c 142REFFA REF, num 1 V ; Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII
d09b2d29 143
65aa4ca7
FC
144#*Named references. Code in regcomp.c assumes that these all are after
145#*the numbered references
01f98ec2
KW
146NREF REF, no-sv 1 V ; Match some already matched string
147NREFF REF, no-sv 1 V ; Match already matched string, folded using native charset semantics for non-utf8
148NREFFL REF, no-sv 1 V ; Match already matched string, folded in loc.
149NREFFU REF, num 1 V ; Match already matched string, folded using unicode semantics for non-utf8
781aab5c 150NREFFA REF, num 1 V ; Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII
1de06328 151
f9ef50a7
NC
152IFMATCH BRANCHJ, off 1 . 2 ; Succeeds if the following matches.
153UNLESSM BRANCHJ, off 1 . 2 ; Fails if the following matches.
154SUSPEND BRANCHJ, off 1 V 1 ; "Independent" sub-RE.
65aa4ca7 155IFTHEN BRANCHJ, off 1 V 1 ; Switch, should be preceded by switcher.
f8abb37e 156GROUPP GROUPP, num 1 ; Whether the group matched.
d09b2d29 157
381b57db 158#*Support for long RE
1de06328 159
f9ef50a7
NC
160LONGJMP LONGJMP, off 1 . 1 ; Jump far away.
161BRANCHJ BRANCHJ, off 1 V 1 ; BRANCH with long offset.
d09b2d29 162
381b57db 163#*The heavy worker
1de06328 164
f8abb37e 165EVAL EVAL, evl 1 ; Execute some Perl code.
d09b2d29 166
381b57db 167#*Modifiers
1de06328 168
f8abb37e
NC
169MINMOD MINMOD, no ; Next operator is not greedy.
170LOGICAL LOGICAL, no ; Next opcode should set the flag only.
d09b2d29 171
65aa4ca7 172#*This is not used yet
f9ef50a7 173RENUM BRANCHJ, off 1 . 1 ; Group with independently numbered parens.
d09b2d29 174
381b57db 175#*Trie Related
1de06328 176
65aa4ca7
FC
177#* Behave the same as A|LIST|OF|WORDS would. The '..C' variants
178#* have inline charclass data (ascii only), the 'C' store it in the
179#* structure.
486ec47a 180# NOTE: the relative order of the TRIE-like regops is significant
ce5e9471 181
7986cb47 182TRIE TRIE, trie 1 ; Match many EXACT(F[ALU]?)? at once. flags==type
f8abb37e 183TRIEC TRIE,trie charclass ; Same as TRIE, but with embedded charclass data
3dab1dad 184
1de06328 185# For start classes, contains an added fail table.
f8abb37e
NC
186AHOCORASICK TRIE, trie 1 ; Aho Corasick stclass. flags==type
187AHOCORASICKC TRIE,trie charclass ; Same as AHOCORASICK, but with embedded charclass data
1de06328 188
381b57db 189#*Regex Subroutines
f8abb37e
NC
190GOSUB GOSUB, num/ofs 2L ; recurse to paren arg1 at (signed) ofs arg2
191GOSTART GOSTART, no ; recurse to start of pattern
03363afd 192
381b57db 193#*Special conditionals
f8abb37e
NC
194NGROUPP NGROUPP, no-sv 1 ; Whether the group matched.
195INSUBP INSUBP, num 1 ; Whether we are in a specific recurse.
196DEFINEP DEFINEP, none 1 ; Never execute directly.
0a4db386 197
486ec47a 198#*Backtracking Verbs
f8abb37e
NC
199ENDLIKE ENDLIKE, none ; Used only for the type field of verbs
200OPFAIL ENDLIKE, none ; Same as (?!)
201ACCEPT ENDLIKE, parno 1 ; Accepts the current matched string.
5d458dd8
YO
202
203
204#*Verbs With Arguments
f8abb37e
NC
205VERB VERB, no-sv 1 ; Used only for the type field of verbs
206PRUNE VERB, no-sv 1 ; Pattern fails at this startpoint if no-backtracking through this
207MARKPOINT VERB, no-sv 1 ; Push the current location for rollback by cut.
208SKIP VERB, no-sv 1 ; On failure skip forward (to the mark) before retrying
209COMMIT VERB, no-sv 1 ; Pattern fails outright if backtracking through this
210CUTGROUP VERB, no-sv 1 ; On failure go to the next alternation in the group
e2e6a0f1 211
ee9b8eae 212#*Control what to keep in $&.
f8abb37e 213KEEPS KEEPS, no ; $& begins here.
ee9b8eae 214
e1d1eefb 215#*New charclass like patterns
f8abb37e 216LNBREAK LNBREAK, none ; generic newline pattern
3172e3fd 217
381b57db 218# NEW STUFF SOMEWHERE ABOVE THIS LINE
1de06328 219
03363afd
YO
220################################################################################
221
7f69552c 222#*SPECIAL REGOPS
1de06328 223
65aa4ca7
FC
224#* This is not really a node, but an optimized away piece of a "long"
225#* node. To simplify debugging output, we mark it as if it were a node
f8abb37e 226OPTIMIZED NOTHING, off ; Placeholder for dump.
1de06328 227
65aa4ca7
FC
228#* Special opcode with the property that no opcode in a compiled program
229#* will ever be of this type. Thus it can be used as a flag value that
230#* no other opcode has been seen. END is used similarly, in that an END
231#* node cant be optimized. So END implies "unoptimizable" and PSEUDO
232#* mean "not seen anything to optimize yet".
f8abb37e 233PSEUDO PSEUDO, off ; Pseudo opcode for internal use.
1de06328 234
03363afd
YO
235-------------------------------------------------------------------------------
236# Format for second section:
65aa4ca7 237# REGOP \t typelist [ \t typelist]
03363afd
YO
238# typelist= namelist
239# = namelist:FAIL
240# = name:count
241
242# Anything below is a state
243#
244#
f8abb37e
NC
245TRIE next:FAIL
246EVAL AB:FAIL
247CURLYX end:FAIL
248WHILEM A_pre,A_min,A_max,B_min,B_max:FAIL
249BRANCH next:FAIL
250CURLYM A,B:FAIL
251IFMATCH A:FAIL
252CURLY B_min_known,B_min,B_max:FAIL
253COMMIT next:FAIL
254MARKPOINT next:FAIL
255SKIP next:FAIL
256CUTGROUP next:FAIL
257KEEPS next:FAIL