3 # File has two sections, divided by a line of dashes '-'.
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
9 # First section is for regops, second section is for regmatch-states
11 # Note that the order in this file is important.
13 # Format for first section:
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
18 # run perl regen.pl after editing this file
24 END END, no ; End of program.
25 SUCCEED END, no ; Return from a subroutine, basically.
27 #* Line Start Anchors:
28 #Note flags field for SBOL indicates if it is a /^/ or a /\A/
29 SBOL BOL, no ; Match "" at beginning of line: /^/, /\A/
30 MBOL BOL, no ; Same, assuming multiline: /^/m
33 SEOL EOL, no ; Match "" at end of line: /$/
34 MEOL EOL, no ; Same, assuming multiline: /$/m
35 EOS EOL, no ; Match "" at end of string: /\z/
37 #* Match Start Anchors:
38 GPOS GPOS, no ; Matches where last m//g left off.
40 #* Word Boundary Opcodes:
41 # The regops that have varieties that vary depending on the character set regex
42 # modifiers have to ordered thusly: /d, /l, /u, /a, /aa. This is because code
43 # in regcomp.c uses the enum value of the modifier as an offset from the /d
44 # version. The complements must come after the non-complements.
45 # BOUND, POSIX and their complements are affected, as well as EXACTF.
46 BOUND BOUND, no ; Like BOUNDA for non-utf8, otherwise match "" between any Unicode \w\W or \W\w
47 BOUNDL BOUND, no ; Like BOUND/BOUNDU, but \w and \W are defined by current locale
48 BOUNDU BOUND, no ; Match "" at any boundary of a given type using Unicode rules
49 BOUNDA BOUND, no ; Match "" at any boundary between \w\W or \W\w, where \w is [_a-zA-Z0-9]
50 # All NBOUND nodes are required by code in regexec.c to be greater than all BOUND ones
51 NBOUND NBOUND, no ; Like NBOUNDA for non-utf8, otherwise match "" between any Unicode \w\w or \W\W
52 NBOUNDL NBOUND, no ; Like NBOUND/NBOUNDU, but \w and \W are defined by current locale
53 NBOUNDU NBOUND, no ; Match "" at any non-boundary of a given type using using Unicode rules
54 NBOUNDA NBOUND, no ; Match "" betweeen any \w\w or \W\W, where \w is [_a-zA-Z0-9]
56 #* [Special] alternatives:
57 REG_ANY REG_ANY, no 0 S ; Match any one character (except newline).
58 SANY REG_ANY, no 0 S ; Match any one character.
59 ANYOF ANYOF, sv 1 S ; Match character in (or not in) this class, single char match only
60 ANYOFD ANYOF, sv 1 S ; Like ANYOF, but /d is in effect
61 ANYOFL ANYOF, sv 1 S ; Like ANYOF, but /l is in effect
63 #* POSIX Character Classes:
64 # Order of the below is important. See ordering comment above.
65 POSIXD POSIXD, none 0 S ; Some [[:class:]] under /d; the FLAGS field gives which one
66 POSIXL POSIXD, none 0 S ; Some [[:class:]] under /l; the FLAGS field gives which one
67 POSIXU POSIXD, none 0 S ; Some [[:class:]] under /u; the FLAGS field gives which one
68 POSIXA POSIXD, none 0 S ; Some [[:class:]] under /a; the FLAGS field gives which one
69 NPOSIXD NPOSIXD, none 0 S ; complement of POSIXD, [[:^class:]]
70 NPOSIXL NPOSIXD, none 0 S ; complement of POSIXL, [[:^class:]]
71 NPOSIXU NPOSIXD, none 0 S ; complement of POSIXU, [[:^class:]]
72 NPOSIXA NPOSIXD, none 0 S ; complement of POSIXA, [[:^class:]]
73 # End of order is important
75 CLUMP CLUMP, no 0 V ; Match any extended grapheme cluster sequence
79 #* BRANCH The set of branches constituting a single choice are
80 #* hooked together with their "next" pointers, since
81 #* precedence prevents anything being concatenated to
82 #* any individual branch. The "next" pointer of the last
83 #* BRANCH in a choice points to the thing following the
84 #* whole choice. This is also where the final "next"
85 #* pointer of each individual branch points; each branch
86 #* starts with the operand node of a BRANCH node.
88 BRANCH BRANCH, node 0 V ; Match this alternative, or the next...
91 # NOTE: the relative ordering of these types is important do not change it
93 EXACT EXACT, str ; Match this string (preceded by length).
94 EXACTL EXACT, str ; Like EXACT, but /l is in effect (used so locale-related warnings can be checked for).
95 EXACTF EXACT, str ; Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len).
96 EXACTFL EXACT, str ; Match this string (not guaranteed to be folded) using /il rules (w/len).
97 EXACTFU 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).
98 EXACTFA EXACT, str ; Match this string (not guaranteed to be folded) using /iaa rules (w/len).
100 # End of important relative ordering.
102 EXACTFU_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).
103 EXACTFLU8 EXACT, str ; Rare cirucmstances: like EXACTFU, but is under /l, UTF-8, folded, and everything in it is above 255.
104 EXACTFA_NO_TRIE EXACT, str ; Match this string (which is not trie-able; not guaranteed to be folded) using /iaa rules (w/len).
108 NOTHING NOTHING, no ; Match empty string.
109 #*A variant of above which delimits a group, thus stops optimizations
110 TAIL NOTHING, no ; Match empty string. Can jump here from outside.
114 #* STAR,PLUS '?', and complex '*' and '+', are implemented as
115 #* circular BRANCH structures. Simple cases
116 #* (one character per match) are implemented with STAR
117 #* and PLUS for speed and to minimize recursive plunges.
119 STAR STAR, node 0 V ; Match this (simple) thing 0 or more times.
120 PLUS PLUS, node 0 V ; Match this (simple) thing 1 or more times.
122 CURLY CURLY, sv 2 V ; Match this simple thing {n,m} times.
123 CURLYN CURLY, no 2 V ; Capture next-after-this simple thing
124 CURLYM CURLY, no 2 V ; Capture this medium-complex thing {n,m} times.
125 CURLYX CURLY, sv 2 V ; Match this complex thing {n,m} times.
127 #*This terminator creates a loop structure for CURLYX
128 WHILEM WHILEM, no 0 V ; Do curly processing and see if rest matches.
132 #*OPEN,CLOSE,GROUPP ...are numbered at compile time.
133 OPEN OPEN, num 1 ; Mark this point in input as start of #n.
134 CLOSE CLOSE, num 1 ; Analogous to OPEN.
136 REF REF, num 1 V ; Match some already matched string
137 REFF REF, num 1 V ; Match already matched string, folded using native charset rules for non-utf8
138 REFFL REF, num 1 V ; Match already matched string, folded in loc.
139 # N?REFF[AU] could have been implemented using the FLAGS field of the
140 # regnode, but by having a separate node type, we can use the existing switch
141 # statement to avoid some tests
142 REFFU REF, num 1 V ; Match already matched string, folded using unicode rules for non-utf8
143 REFFA REF, num 1 V ; Match already matched string, folded using unicode rules for non-utf8, no mixing ASCII, non-ASCII
145 #*Named references. Code in regcomp.c assumes that these all are after
146 #*the numbered references
147 NREF REF, no-sv 1 V ; Match some already matched string
148 NREFF REF, no-sv 1 V ; Match already matched string, folded using native charset rules for non-utf8
149 NREFFL REF, no-sv 1 V ; Match already matched string, folded in loc.
150 NREFFU REF, num 1 V ; Match already matched string, folded using unicode rules for non-utf8
151 NREFFA REF, num 1 V ; Match already matched string, folded using unicode rules for non-utf8, no mixing ASCII, non-ASCII
153 #*Support for long RE
154 LONGJMP LONGJMP, off 1 . 1 ; Jump far away.
155 BRANCHJ BRANCHJ, off 1 V 1 ; BRANCH with long offset.
157 #*Special Case Regops
158 IFMATCH BRANCHJ, off 1 . 2 ; Succeeds if the following matches.
159 UNLESSM BRANCHJ, off 1 . 2 ; Fails if the following matches.
160 SUSPEND BRANCHJ, off 1 V 1 ; "Independent" sub-RE.
161 IFTHEN BRANCHJ, off 1 V 1 ; Switch, should be preceded by switcher.
162 GROUPP GROUPP, num 1 ; Whether the group matched.
167 EVAL EVAL, evl/flags 2L ; Execute some Perl code.
171 MINMOD MINMOD, no ; Next operator is not greedy.
172 LOGICAL LOGICAL, no ; Next opcode should set the flag only.
174 #*This is not used yet
175 RENUM BRANCHJ, off 1 . 1 ; Group with independently numbered parens.
179 #* Behave the same as A|LIST|OF|WORDS would. The '..C' variants
180 #* have inline charclass data (ascii only), the 'C' store it in the
182 # NOTE: the relative order of the TRIE-like regops is significant
184 TRIE TRIE, trie 1 ; Match many EXACT(F[ALU]?)? at once. flags==type
185 TRIEC TRIE,trie charclass ; Same as TRIE, but with embedded charclass data
187 # For start classes, contains an added fail table.
188 AHOCORASICK TRIE, trie 1 ; Aho Corasick stclass. flags==type
189 AHOCORASICKC TRIE,trie charclass ; Same as AHOCORASICK, but with embedded charclass data
192 GOSUB GOSUB, num/ofs 2L ; recurse to paren arg1 at (signed) ofs arg2
194 #*Special conditionals
195 NGROUPP NGROUPP, no-sv 1 ; Whether the group matched.
196 INSUBP INSUBP, num 1 ; Whether we are in a specific recurse.
197 DEFINEP DEFINEP, none 1 ; Never execute directly.
200 ENDLIKE ENDLIKE, none ; Used only for the type field of verbs
201 OPFAIL ENDLIKE, no-sv 1 ; Same as (?!), but with verb arg
202 ACCEPT ENDLIKE, no-sv/num 2L ; Accepts the current matched string, with verbar
204 #*Verbs With Arguments
205 VERB VERB, no-sv 1 ; Used only for the type field of verbs
206 PRUNE VERB, no-sv 1 ; Pattern fails at this startpoint if no-backtracking through this
207 MARKPOINT VERB, no-sv 1 ; Push the current location for rollback by cut.
208 SKIP VERB, no-sv 1 ; On failure skip forward (to the mark) before retrying
209 COMMIT VERB, no-sv 1 ; Pattern fails outright if backtracking through this
210 CUTGROUP VERB, no-sv 1 ; On failure go to the next alternation in the group
212 #*Control what to keep in $&.
213 KEEPS KEEPS, no ; $& begins here.
215 #*New charclass like patterns
216 LNBREAK LNBREAK, none ; generic newline pattern
218 # NEW STUFF SOMEWHERE ABOVE THIS LINE
220 ################################################################################
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
226 OPTIMIZED NOTHING, off ; Placeholder for dump.
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".
233 PSEUDO PSEUDO, off ; Pseudo opcode for internal use.
235 -------------------------------------------------------------------------------
236 # Format for second section:
237 # REGOP \t typelist [ \t typelist]
242 # Anything below is a state
248 WHILEM A_pre,A_min,A_max,B_min,B_max:FAIL
252 CURLY B_min_known,B_min,B_max:FAIL