1XS::Parse::Keyword(3) User Contributed Perl DocumentationXS::Parse::Keyword(3)
2
6 "XS::Parse::Keyword" - XS functions to assist in parsing keyword syntax
7
9 This module provides some XS functions to assist in writing syntax
10 modules that provide new perl-visible syntax, primarily for authors of
11 keyword plugins using the "PL_keyword_plugin" hook mechanism. It is
12 unlikely to be of much use to anyone else; and highly unlikely to be
13 any use when writing perl code using these. Unless you are writing a
14 keyword plugin using XS, this module is not for you.
15 This module is also currently experimental, and the design is still
17 evolving and subject to change. Later versions may break ABI
18 compatibility, requiring changes or at least a rebuild of any module
19 that depends on it.
20
22 boot_xs_parse_keyword
23 void boot_xs_parse_keyword(double ver);
24 Call this function from your "BOOT" section in order to initialise the
26 module and parsing hooks.
27 ver should either be 0 or a decimal number for the module version
29 requirement; e.g.
30 boot_xs_parse_keyword(0.14);
32 register_xs_parse_keyword
34 void register_xs_parse_keyword(const char *keyword,
35 const struct XSParseKeywordHooks *hooks, void *hookdata);
36 This function installs a set of parsing hooks to be associated with the
38 given keyword. Such a keyword will then be handled automatically by a
39 keyword parser installed by "XS::Parse::Keyword" itself.
40
42 The "XSParseKeywordHooks" structure provides the following hook stages,
43 which are invoked in the given order.
44 flags
46 The following flags are defined:
47 "XPK_FLAG_EXPR"
49 The parse or build function is expected to return
50 "KEYWORD_PLUGIN_EXPR".
51 "XPK_FLAG_STMT"
53 The parse or build function is expected to return
54 "KEYWORD_PLUGIN_STMT".
55 These two flags are largely for the benefit of giving static
57 information at registration time to assist static parsing or other
58 related tasks to know what kind of grammatical element this keyword
59 will produce.
60 "XPK_FLAG_AUTOSEMI"
62 The syntax forms a complete statement, which should be followed by
63 a statement separator semicolon (";"). This semicolon is optional
64 at the end of a block.
65 The semicolon, if present, will be consumed automatically.
67 The "permit" Stage
69 const char *permit_hintkey;
70 bool (*permit) (pTHX_ void *hookdata);
71 Called by the installed keyword parser hook which is used to handle
73 keywords registered by "register_xs_parse_keyword".
74 As a shortcut for the common case, the "permit_hintkey" may point to a
76 string to look up from the hints hash. If the given key name is not
77 found in the hints hash then the keyword is not permitted. If the key
78 is present then the "permit" function is invoked as normal.
79 If not rejected by a hint key that was not found in the hints hash, the
81 function part of the stage is called next and should inspect whether
82 the keyword is permitted at this time perhaps by inspecting other
83 lexical clues, and return true only if the keyword is permitted.
84 Both the string and the function are optional. Either or both may be
86 present. If neither is present then the keyword is always permitted -
87 which is likely not what you wanted to do.
88 The "check" Stage
90 void (*check)(pTHX_ void *hookdata);
91 Invoked once the keyword has been permitted. If present, this hook
93 function can check the surrounding lexical context, state, or other
94 information and throw an exception if it is unhappy that the keyword
95 should apply in this position.
96 The "parse" Stage
98 This stage is invoked once the keyword has been checked, and actually
99 parses the incoming text into an optree. It is implemented by calling
100 the first of the following function pointers which is not NULL. The
101 invoked function may optionally build an optree to represent the parsed
102 syntax, and place it into the variable addressed by "out". If it does
103 not, then a simple "OP_NULL" will be constructed in its place.
104 "lex_read_space()" is called both before and after this stage is
106 invoked, so in many simple cases the hook function itself does not need
107 to bother with it.
108 int (*parse)(pTHX_ OP **out, void *hookdata);
110 If present, this should consume text from the parser buffer by invoking
112 "lex_*" or "parse_*" functions and eventually return a
113 "KEYWORD_PLUGIN_*" result value.
114 This is the most generic and powerful of the options, but requires the
116 most amount of implementation work.
117 int (*build)(pTHX_ OP **out, XSParseKeywordPiece *args[], size_t nargs, void *hookdata);
119 If "parse" is not present, this is called instead after parsing a
121 sequence of arguments, of types given by the pieces field; which should
122 be a zero- terminated array of piece types.
123 This alternative is somewhat less generic and powerful than providing
125 "parse" yourself, but involves much less parsing work and is shorter
126 and easier to implement.
127 int (*build1)(pTHX_ OP **out, XSParseKeywordPiece *arg0, void *hookdata);
129 If neither "parse" nor "build" are present, this is called as a simpler
131 variant of "build" when only a single argument is required. It takes
132 its type from the "piece1" field instead.
133
135 When using the "build" or "build1" alternatives for the "parse" phase,
136 the actual syntax is parsed automatically by this module, according to
137 the specification given by the pieces or piece1 field. The result of
138 that parsing step is placed into the args or arg0 parameter to the
139 invoked function, using a "struct" type consisting of the following
140 fields:
141 typedef struct
143 union {
144 OP *op;
145 CV *cv;
146 SV *sv;
147 int i;
148 struct {
149 SV *name;
150 SV *value;
151 } attr;
152 PADOFFSET padix;
153 struct XSParseInfixInfo *infix;
154 };
155 int line;
156 } XSParseKeywordPiece;
157 Which field of the anonymous union is set depends on the type of the
159 piece. The line field contains the line number of the source file
160 where parsing of that piece began.
161 Some piece types are "atomic", whose definition is self-contained.
163 Others are structural, defined in terms of inner pieces. Together these
164 form an entire tree-shaped definition of the syntax that the keyword
165 expects to find.
166 Atomic types generally provide exactly one argument into the list of
168 args (with the exception of literal matches, which do not provide
169 anything). Structural types may provide an initial argument
170 themselves, followed by a list of the values of each sub-piece they
171 contained inside them. Thus, while the data structure defining the
172 syntax shape is a tree, the argument values it parses into is passed as
173 a flat array to the "build" function.
174 Some structural types need to be able to determine whether or not
176 syntax relating some optional part of them is present in the incoming
177 source text. In this case, the pieces relating to those optional parts
178 must support "probing". This ability is also noted below.
179 The type of each piece should be one of the following macro values.
181 XPK_BLOCK
183 atomic, can probe, emits op.
184 XPK_BLOCK
186 A brace-delimited block of code is expected, passed as an optree in the
188 op field. This will be parsed as a block within the current function
189 scope.
190 This can be probed by checking for the presence of an open-brace ("{")
192 character.
193 Be careful defining grammars with this because an open-brace is also a
195 valid character to start a term expression, for example. Given a choice
196 between "XPK_BLOCK" and "XPK_TERMEXPR", either of them could try to
197 consume such code as
198 { 123, 456 }
200 XPK_BLOCK_VOIDCTX, XPK_BLOCK_SCALARCTX, XPK_BLOCK_LISTCTX
202 Variants of "XPK_BLOCK" which wrap a void, scalar or list-context scope
203 around the block.
204 XPK_PREFIXED_BLOCK
206 structural, emits op.
207 XPK_PREFIXED_BLOCK(pieces ...)
209 Some pieces are expected, followed by a brace-delimited block of code,
211 which is passed as an optree in the op field. The prefix pieces are
212 parsed first, and their results are passed before the block itself.
213 The entire sequence, including the prefix items, is contained within a
215 pair of "block_start()" / "block_end()" calls. This permits the prefix
216 pieces to introduce new items into the lexical scope of the block - for
217 example by the use of "XPK_LEXVAR_MY".
218 A call to "intro_my()" is automatically made at the end of the prefix
220 pieces, before the block itself is parsed, ensuring any new lexical
221 variables are now visible.
222 In addition, the following extra piece types are recognised here:
224 XPK_SETUP
226 void setup(pTHX_ void *hookdata);
227 XPK_SETUP(&setup)
229 atomic, emits nothing.
231 This piece type runs a function given by pointer. Typically this
233 function may be used to introduce new lexical state into the
234 parser, or in some other way have some side-effect on the parsing
235 context of the block to be parsed.
236 XPK_PREFIXED_BLOCK_ENTERLEAVE
238 A variant of "XPK_PREFIXED_BLOCK" which additionally wraps the entire
239 parsing operation, including the "block_start()", "block_end()" and any
240 calls to "XPK_SETUP" functions, within a "ENTER"/"LEAVE" pair.
241 This should not make a difference to the standard parser pieces
243 provided here, but may be useful behaviour for the code in the setup
244 function, especially if it wishes to modify parser state and use the
245 savestack to ensure it is restored again when parsing has finished.
246 XPK_ANONSUB
248 atomic, emits op.
249 A brace-delimited block of code is expected, and assembled into the
251 body of a new anonymous subroutine. This will be passed as a protosub
252 CV in the cv field.
253 XPK_ARITHEXPR
255 atomic, emits op.
256 XPK_ARITHEXPR
258 An arithmetic expression is expected, parsed using "parse_arithexpr()",
260 and passed as an optree in the op field.
261 XPK_ARITHEXPR_VOIDCTX, XPK_ARITHEXPR_SCALARCTX
263 Variants of "XPK_ARITHEXPR" which puts the expression in void or scalar
264 context.
265 XPK_TERMEXPR
267 atomic, emits op.
268 XPK_TERMEXPR
270 A term expression is expected, parsed using "parse_termexpr()", and
272 passed as an optree in the op field.
273 XPK_TERMEXPR_VOIDCTX, XPK_TERMEXPR_SCALARCTX
275 Variants of "XPK_TERMEXPR" which puts the expression in void or scalar
276 context.
277 XPK_LISTEXPR
279 atomic, emits op.
280 XPK_LISTEXPR
282 A list expression is expected, parsed using "parse_listexpr()", and
284 passed as an optree in the op field.
285 XPK_LISTEXPR_LISTCTX
287 Variant of "XPK_LISTEXPR" which puts the expression in list context.
288 XPK_IDENT, XPK_IDENT_OPT
290 atomic, can probe, emits sv.
291 A bareword identifier name is expected, and passed as an SV containing
293 a PV in the sv field. An identifier is not permitted to contain a
294 double colon ("::").
295 The "_OPT"-suffixed version is optional; if no identifier is found then
297 sv is set to "NULL".
298 XPK_PACKAGENAME, XPK_PACKAGENAME_OPT
300 atomic, can probe, emits sv.
301 A bareword package name is expected, and passed as an SV containing a
303 PV in the sv field. A package name is similar to an identifier, except
304 it permits double colons in the middle.
305 The "_OPT"-suffixed version is optional; if no package name is found
307 then sv is set to "NULL".
308 XPK_LEXVARNAME
310 atomic, emits sv.
311 XPK_LEXVARNAME(kind)
313 A lexical variable name is expected, and passed as an SV containing a
315 PV in the sv field. The "kind" argument specifies what kinds of
316 variable are permitted, and should be a bitmask of one or more bits
317 from "XPK_LEXVAR_SCALAR", "XPK_LEXVAR_ARRAY" and "XPK_LEXVAR_HASH". A
318 convenient shortcut "XPK_LEXVAR_ANY" permits all three.
319 XPK_ATTRIBUTES
321 atomic, emits i followed by more args.
322 A list of ":"-prefixed attributes is expected, in the same format as
324 sub or variable attributes. An optional leading ":" indicates the
325 presence of attributes, then one or more of them are parsed. Attributes
326 may be optionally separated by additional ":"s, but this is not
327 required.
328 Each attribute is expected to be an identifier name, followed by an
330 optional value wrapped in parentheses. Whitespace is NOT permitted
331 between the name and value, as per standard Perl parsing rules.
332 :attrname
334 :attrname(value)
335 The i field indicates how many attributes were found. That number of
337 additional arguments are then passed, each containing two SVs in the
338 attr.name and attr.value fields. This number may be zero.
339 It is not an error for there to be no attributes present, or for the
341 optional colon to be missing. In this case i will be set to zero.
342 XPK_VSTRING, XPK_VSTRING_OPT
344 atomic, can probe, emits sv.
345 A version string is expected, of the form "v1.234" including the
347 leading "v" character. It is passed as a version SV object in the sv
348 field.
349 The "_OPT"-suffixed version is optional; if no version string is found
351 then sv is set to "NULL".
352 XPK_LEXVAR_MY
354 atomic, emits padix.
355 XPK_LEXVAR_MY(kind)
357 A lexical variable name is expected, added to the current pad as if
359 specified in a "my" expression, and passed as the pad index in the
360 padix field.
361 The "kind" argument specifies what kinds of variable are permitted, as
363 per "XPK_LEXVARNAME".
364 XPK_COMMA, XPK_COLON, XPK_EQUALS
366 atomic, can probe, emits nothing.
367 A literal character (",", ":" or "=") is expected. No argument value is
369 passed.
370 XPK_AUTOSEMI
372 atomic, emits nothing.
373 A literal semicolon (";") as a statement terminator is optionally
375 expected. If the next token is a closing brace to indicate the end of
376 a block, then a semicolon is not required. If anything else is
377 encountered an error will be raised.
378 This piece type is the same as specifying the "XPK_FLAG_AUTOSEMI". It
380 is useful to put at the end of a sequence that forms part of a choice
381 of syntax, where some forms indicate a statement ending in a semicolon,
382 whereas others may end in a full block that does not need one.
383 XPK_INFIX_*
385 atomic, can probe, emits infix.
386 An infix operator as recognised by XS::Parse::Infix. The returned
388 pointer points to a structure allocated by "XS::Parse::Infix"
389 describing the operator.
390 Various versions of the macro are provided, each using a different
392 selection filter to choose certain available infix operators:
393 XPK_INFIX_RELATION # any relational operator
395 XPK_INFIX_EQUALITY # an equality operator like `==` or `eq`
396 XPK_INFIX_MATCH_NOSMART # any sort of "match"-like operator, except smartmatch
397 XPK_INFIX_MATCH_SMART # XPK_INFIX_MATCH_NOSMART plus smartmatch
398 XPK_LITERAL
400 atomic, can probe, emits nothing.
401 XPK_LITERAL("literal")
403 A literal string match is expected. No argument value is passed.
405 This form should generally be avoided if at all possible, because it is
407 very easy to abuse to make syntaxes which confuse humans and code tools
408 alike. Generally it is best reserved just for the first component of a
409 "XPK_OPTIONAL" or "XPK_REPEATED" sequence, to provide a "secondary
410 keyword" that such a repeated item can look out for.
411 XPK_KEYWORD
413 atomic, can probe, emits nothing.
414 XPK_KEYWORD("keyword")
416 A literal string match is expected. No argument value is passed.
418 This is similar to "XPK_LITERAL" except that it additionally checks
420 that the following character is not an identifier character. This
421 ensures that the expected keyword-like behaviour is preserved. For
422 example, given the input "keyword", the piece "XPK_LITERAL("key")"
423 would match it, whereas "XPK_KEYWORD("key")" would not because of the
424 subsequent "w" character.
425 XPK_SEQUENCE
427 structural, might support probe, emits nothing.
428 XPK_SEQUENCE(pieces ...)
430 A structural type which contains a number of pieces. This is normally
432 equivalent to simply placing the pieces in sequence inside their own
433 container, but it is useful inside "XPK_CHOICE" or "XPK_TAGGEDCHOICE".
434 An "XPK_SEQUENCE" supports probe if its first contained piece does;
436 i.e. is transparent to probing.
437 XPK_OPTIONAL
439 structural, emits i.
440 XPK_OPTIONAL(pieces ...)
442 A structural type which may expects to find its contained pieces, or is
444 happy not to. This will pass an argument whose i field contains either
445 1 or 0, depending whether the contents were found. The first piece type
446 within must support probe.
447 XPK_REPEATED
449 structural, emits i.
450 XPK_REPEATED(pieces ...)
452 A structural type which expects to find zero or more repeats of its
454 contained pieces. This will pass an argument whose i field contains the
455 count of the number of repeats it found. The first piece type within
456 must support probe.
457 XPK_CHOICE
459 structural, can probe, emits i.
460 XPK_CHOICE(options ...)
462 A structural type which expects to find one of a number of alternative
464 options. An ordered list of types is provided, all of which must
465 support probe. This will pass an argument whose i field gives the index
466 of the first choice that was accepted. The first option takes the value
467 0.
468 As each of the options is interpreted as an alternative, not a
470 sequence, you should use "XPK_SEQUENCE" if a sequence of multiple items
471 should be considered as a single alternative.
472 It is not an error if no choice matches. At that point, the i field
474 will be set to -1.
475 If you require a failure message in this case, set the final choice to
477 be of type "XPK_FAILURE". This will cause an error message to be
478 printed instead.
479 XPK_FAILURE("message string")
481 XPK_TAGGEDCHOICE
483 structural, can probe, emits i.
484 XPK_TAGGEDCHOICE(choice, tag, ...)
486 A structural type similar to "XPK_CHOICE", except that each choice type
488 is followed by an element of type "XPK_TAG" which gives an integer. It
489 is that integer value, rather than the positional index of the choice
490 within the list, which is passed in the i field.
491 XPK_TAG(value)
493 As each of the options is interpreted as an alternative, not a
495 sequence, you should use "XPK_SEQUENCE" if a sequence of multiple items
496 should be considered as a single alternative.
497 XPK_COMMALIST
499 structural, might support probe, emits i.
500 XPK_COMMALIST(pieces ...)
502 A structural type which expects to find one or more repeats of its
504 contained pieces, separated by literal comma (",") characters. This is
505 somewhat similar to "XPK_REPEATED", except that it needs at least one
506 copy, needs commas between its items, but does not require that the
507 first contained piece support probe (the comma itself is sufficient to
508 indicate a repeat).
509 An "XPK_COMMALIST" supports probe if its first contained piece does;
511 i.e. is transparent to probing.
512 XPK_PARENSCOPE
514 structural, can probe, emits nothing.
515 XPK_PARENSCOPE(pieces ...)
517 A structural type which expects to find a sequence of pieces, all
519 contained in parentheses as "( ... )". This will pass no extra
520 arguments.
521 XPK_ARGSCOPE
523 structural, emits nothing.
524 XPK_ARGSCOPE(pieces ...)
526 A structural type similar to "XPK_PARENSCOPE", except that the
528 parentheses themselves are optional; much like Perl's parsing of calls
529 to known functions.
530 If parentheses are encountered in the input, they will be consumed by
532 this piece and it will behave identically to "XPK_PARENSCOPE". If there
533 is no open parenthesis, this piece will behave like "XPK_SEQUENCE" and
534 consume all the pieces inside it, without expecting a closing
535 parenthesis.
536 XPK_BRACKETSCOPE
538 structural, can probe, emits nothing.
539 XPK_BRACKETSCOPE(pieces ...)
541 A structural type which expects to find a sequence of pieces, all
543 contained in square brackets as "[ ... ]". This will pass no extra
544 arguments.
545 XPK_BRACESCOPE
547 structural, can probe, emits nothing.
548 XPK_BRACESCOPE(pieces ...)
550 A structural type which expects to find a sequence of pieces, all
552 contained in braces as "{ ... }". This will pass no extra arguments.
553 Note that this is not necessary to use with "XPK_BLOCK" or
555 "XPK_ANONSUB"; those will already consume a set of braces. This is
556 intended for special constrained syntax that should not just accept an
557 arbitrary block.
558 XPK_CHEVRONSCOPE
560 structural, can probe, emits nothing.
561 XPK_CHEVRONSCOPE(pieces ...)
563 A structural type which expects to find a sequence of pieces, all
565 contained in angle brackets as "< ... >". This will pass no extra
566 arguments.
567 Remember that expressions like "a > b" are valid term expressions, so
569 the contents of this scope shouldn't allow arbitrary expressions or the
570 closing bracket will be ambiguous.
571 XPK_PARENSCOPE_OPT, XPK_BRACKETSCOPE_OPT, XPK_BRACESCOPE_OPT,
573 XPK_CHEVRONSCOPE_OPT
574 structural, can probe, emits i.
575 XPK_PARENSCOPE_OPT(pieces ...)
577 XPK_BRACKETSCOPE_OPT(pieces ...)
578 XPK_BRACESCOPE_OPT(pieces ...)
579 XPK_CHEVERONSCOPE_OPT(pieces ...)
580 Each of the four "XPK_...SCOPE" macros above has an optional variant,
582 whose name is suffixed by "_OPT". These pass an argument whose i field
583 is either true or false, indicating whether the scope was found,
584 followed by the values from the scope itself.
585 This is a convenient shortcut to nesting the scope within a
587 "XPK_OPTIONAL" macro.
588
590 Paul Evans <leonerd@leonerd.org.uk>
591perl v5.36.0 2022-07-26 XS::Parse::Keyword(3)