1 # LR(1) Parser Generator #
3 This parser generator takes a grammar description combined with code
4 fragments, analyses it, and can report details about the analysis and
5 write out C code files which can be compiled to make a parser.
7 There are several distinct sections.
9 1. `grammar_read` will read a grammar from a literate-code file and
10 store the productions, symbols, and code fragments.
11 2. `grammar_analyse` will take that grammar and build LR parsing
12 states and look-ahead tables.
13 3. `grammar_report` will present the details of the analysis
14 in a readable format and will report any conflicts.
15 4. `parser_generate` will write out C code files with various
16 tables and with the code fragments arranged in useful places.
17 5. `parser_run` is a library function intended to be linked together
18 with the generated parser tables to complete the implementation of
20 6. `calc.cgm` is a test grammar for a simple calculator.
22 ###### File: parsergen.c
27 ## forward declarations
38 ###### File: libparser.c
47 ###### File: parsergen.mk
50 parsergen.c parsergen.mk calc.cgm libparser.c parser.h : parsergen.mdc
53 ## Reading the grammar
55 The grammar must be stored in a markdown literate code file as parsed
56 by "[mdcode][]". It must have three top level (i.e. unreferenced)
57 sections: `header`, `code`, and `grammar`. The first two will be
58 literally copied into the generated `.c` and `.h`. files. The last
59 contains the grammar. This is tokenised with "[scanner][]".
62 [scanner]: scanner.html
68 ###### parser includes
72 The grammar contains production sets, precedence/associativity
73 declarations, and data type declarations. These are all parsed with
74 _ad hoc_ parsing as we don't have a parser generator yet.
76 The precedence and associativity can be set for each production, but
77 can be inherited from symbols. The data type is potentially defined
78 for each non-terminal and describes what C structure is used to store
79 information about each symbol.
82 enum assoc {Left, Right, Non};
83 char *assoc_names[] = {"Left","Right","Non"};
86 struct text struct_name;
88 unsigned short precedence;
92 unsigned short precedence;
100 The strings reported by `mdcode` and `scanner` are `struct text` which have
101 length rather than being null terminated. To help with printing and
102 comparing we define `text_is` and `prtxt`, which should possibly go in
103 `mdcode`. `scanner` does provide `text_dump` which is useful for strings
104 which might contain control characters.
107 static int text_is(struct text t, char *s)
109 return (strlen(s) == t.len &&
110 strncmp(s, t.txt, t.len) == 0);
112 static void prtxt(struct text t)
114 printf("%.*s", t.len, t.txt);
119 Productions are comprised primarily of symbols - terminal and
120 non-terminal. We do not make a syntactic distinction between the two,
121 though non-terminals must be identifiers. Non-terminal symbols are
122 those which appear in the head of a production, terminal symbols are
123 those which don't. There are also "virtual" symbols used for precedence
124 marking discussed later, and sometimes we won't know what type a symbol
127 ###### forward declarations
128 enum symtype { Unknown, Virtual, Terminal, Nonterminal };
129 char *symtypes = "UVTN";
133 Symbols can be either `TK_ident` or `TK_mark`. They are saved in a
134 table of known symbols and the resulting parser will report them as
135 `TK_reserved + N`. A small set of identifiers are reserved for the
136 different token types that `scanner` can report.
139 static char *reserved_words[] = {
140 [TK_error] = "ERROR",
141 [TK_number] = "NUMBER",
142 [TK_ident] = "IDENTIFIER",
144 [TK_string] = "STRING",
145 [TK_multi_string] = "MULTI_STRING",
148 [TK_newline] = "NEWLINE",
154 Note that `TK_eof` and the two `TK_*_comment` tokens cannot be
155 recognised. The former is automatically expected at the end of the text
156 being parsed. The latter are always ignored by the parser.
158 All of these symbols are stored in a simple symbol table. We use the
159 `struct text` from `mdcode` to store the name and link them together
160 into a sorted list using an insertion sort.
162 We don't have separate `find` and `insert` functions as any symbol we
163 find needs to be remembered. We simply expect `find` to always return a
164 symbol, but its type might be `Unknown`.
173 ###### grammar fields
178 static int text_cmp(struct text a, struct text b)
183 int cmp = strncmp(a.txt, b.txt, len);
187 return a.len - b.len;
190 static struct symbol *sym_find(struct grammar *g, struct text s)
192 struct symbol **l = &g->syms;
197 (cmp = text_cmp((*l)->name, s)) < 0)
201 n = calloc(1, sizeof(*n));
210 static void symbols_init(struct grammar *g)
212 int entries = sizeof(reserved_words)/sizeof(reserved_words[0]);
214 for (i = 0; i < entries; i++) {
217 t.txt = reserved_words[i];
220 t.len = strlen(t.txt);
227 ### Data types and precedence.
229 Data type specification and precedence specification are both
230 introduced by a dollar sign at the start of the line. If the next
231 word is `LEFT`, `RIGHT` or `NON`, then the line specifies a
232 precedence, otherwise it specifies a data type.
234 The data type name is simply stored and applied to the head of all
235 subsequent productions. It must be the name of a structure, so `$expr`
236 maps to `struct expr`.
238 Any productions given before the first data type will have no data type
239 and can carry no information. In order to allow other non-terminals to
240 have no type, the data type `$void` can be given. This does *not* mean
241 that `struct void` will be used, but rather than no type will be
242 associated with future non-terminals.
244 The precedence line must contain a list of symbols - typically
245 terminal symbols, but not necessarily. It can only contain symbols
246 that have not been seen yet, so precedence declaration must precede
247 the productions that they affect.
249 A precedence line may also contain a "virtual" symbol which is an
250 identifier preceded by `$$`. Virtual terminals carry precedence
251 information but are not included in the grammar. A production can
252 declare that it inherits the precedence of a given virtual symbol.
254 This use for `$$` precludes it from being used as a symbol in the
255 described language. Two other symbols: `${` and `}$` are also
258 Each new precedence line introduces a new precedence level and
259 declares how it associates. This level is stored in each symbol
260 listed and may be inherited by any production which uses the symbol. A
261 production inherits from the last symbol which has a precedence.
263 ###### grammar fields
264 struct text current_type;
268 enum symbols { TK_virtual = TK_reserved, TK_open, TK_close };
269 static const char *known[] = { "$$", "${", "}$" };
272 static char *dollar_line(struct token_state *ts, struct grammar *g)
274 struct token t = token_next(ts);
279 if (t.num != TK_ident) {
280 err = "type or assoc expected after '$'";
283 if (text_is(t.txt, "LEFT"))
285 else if (text_is(t.txt, "RIGHT"))
287 else if (text_is(t.txt, "NON"))
290 g->current_type = t.txt;
291 if (text_is(t.txt, "void"))
292 g->current_type.txt = NULL;
294 if (t.num != TK_newline) {
295 err = "Extra tokens after type name";
301 // This is a precedence line, need some symbols.
305 while (t.num != TK_newline) {
306 enum symtype type = Terminal;
308 if (t.num == TK_virtual) {
311 if (t.num != TK_ident) {
312 err = "$$ must be followed by a word";
315 } else if (t.num != TK_ident &&
317 err = "Illegal token in precedence line";
320 s = sym_find(g, t.txt);
321 if (s->type != Unknown) {
322 err = "Symbols in precedence line must not already be known.";
326 s->precedence = g->prec_levels;
331 err = "No symbols given on precedence line";
335 while (t.num != TK_newline && t.num != TK_eof)
342 A production either starts with an identifier which is the head
343 non-terminal, or a vertical bar (`|`) in which case this production
344 uses the same head as the previous one. The identifier must be
345 followed by a `->` mark. All productions for a given non-terminal must
346 be grouped together with the `nonterminal ->` given only once.
348 After this a (possibly empty) sequence of identifiers and marks form
349 the body of the production. A virtual symbol may be given after the
350 body by preceding it with `$$`. If a virtual symbol is given, the
351 precedence of the production is that for the virtual symbol. If none
352 is given, the precedence is inherited from the last symbol in the
353 production which has a precedence specified.
355 After the optional precedence may come the `${` mark. This indicates
356 the start of a code fragment. If present, this must be on the same
357 line as the start of the production.
359 All of the text from the `${` through to the matching `}$` is
360 collected and forms the code-fragment for the production. It must all
361 be in one `code_node` of the literate code. The `}$` must be
362 at the end of a line.
364 Text in the code fragment will undergo substitutions where `$N` for
365 some numeric `N` will be replaced with a variable holding the parse
366 information for the particular symbol in the production. `$0` is the
367 head of the production, `$1` is the first symbol of the body, etc.
368 The type of `$N` for a terminal symbol is `struct token`. For
369 a non-terminal, it is whatever has been declared for that symbol.
371 While building productions we will need to add to an array which needs to
375 static void array_add(void *varray, int *cnt, void *new)
377 void ***array = varray;
380 current = ((*cnt-1) | (step-1))+1;
381 if (*cnt == current) {
384 *array = realloc(*array, current * sizeof(void*));
386 (*array)[*cnt] = new;
390 Collecting the code fragment simply involves reading tokens until we
391 hit the end token `}$`, and noting the character position of the start and
395 static struct text collect_code(struct token_state *state,
400 code.txt = start.txt.txt + start.txt.len;
402 t = token_next(state);
403 while (t.node == start.node &&
404 t.num != TK_close && t.num != TK_error &&
406 if (t.num == TK_close && t.node == start.node)
407 code.len = t.txt.txt - code.txt;
413 Now we have all the bit we need to parse a full production.
418 ###### grammar fields
419 struct production **productions;
420 int production_count;
422 ###### production fields
424 struct symbol **body;
429 int first_production;
432 static char *parse_production(struct grammar *g,
434 struct token_state *state)
436 /* Head has already been parsed. */
439 struct production p, *pp;
441 memset(&p, 0, sizeof(p));
443 tk = token_next(state);
444 while (tk.num == TK_ident || tk.num == TK_mark) {
445 struct symbol *bs = sym_find(g, tk.txt);
446 if (bs->type == Unknown)
448 if (bs->type == Virtual) {
449 err = "Virtual symbol not permitted in production";
452 if (bs->precedence) {
453 p.precedence = bs->precedence;
456 array_add(&p.body, &p.body_size, bs);
457 tk = token_next(state);
459 if (tk.num == TK_virtual) {
461 tk = token_next(state);
462 if (tk.num != TK_ident) {
463 err = "word required after $$";
466 vs = sym_find(g, tk.txt);
467 if (vs->type != Virtual) {
468 err = "symbol after $$ must be virtual";
471 p.precedence = vs->precedence;
473 tk = token_next(state);
475 if (tk.num == TK_open) {
476 p.code = collect_code(state, tk);
477 if (p.code.txt == NULL) {
478 err = "code fragment not closed properly";
481 tk = token_next(state);
483 if (tk.num != TK_newline && tk.num != TK_eof) {
484 err = "stray tokens at end of line";
487 pp = malloc(sizeof(*pp));
489 array_add(&g->productions, &g->production_count, pp);
492 while (tk.num != TK_newline && tk.num != TK_eof)
493 tk = token_next(state);
497 With the ability to parse production and dollar-lines, we have nearly all
498 that we need to parse a grammar from a `code_node`.
500 The head of the first production will effectively be the `start` symbol of
501 the grammar. However it won't _actually_ be so. Processing the grammar is
502 greatly simplified if the real start symbol only has a single production,
503 and expects `$eof` as the final terminal. So when we find the first
504 explicit production we insert an extra production as production zero which
507 ###### Example: production 0
510 where `START` is the first non-terminal given.
512 ###### create production zero
513 struct production *p = calloc(1,sizeof(*p));
514 struct text start = {"$start",6};
515 struct text eof = {"$eof",4};
516 p->head = sym_find(g, start);
517 p->head->type = Nonterminal;
518 array_add(&p->body, &p->body_size, head);
519 array_add(&p->body, &p->body_size, sym_find(g, eof));
520 g->start = p->head->num;
521 p->head->first_production = g->production_count;
522 array_add(&g->productions, &g->production_count, p);
524 Now we are ready to read in the grammar.
526 ###### grammar fields
527 int start; // the 'start' symbol.
530 static struct grammar *grammar_read(struct code_node *code)
532 struct token_config conf = {
535 .words_marks = known,
536 .known_count = sizeof(known)/sizeof(known[0]),
538 .ignored = (1 << TK_line_comment)
539 | (1 << TK_block_comment)
542 | (1 << TK_multi_string)
547 struct token_state *state = token_open(code, &conf);
548 struct token tk = token_next(state);
549 struct symbol *head = NULL;
553 g = calloc(1, sizeof(*g));
556 for (tk = token_next(state); tk.num != TK_eof;
557 tk = token_next(state)) {
558 if (tk.num == TK_newline)
560 if (tk.num == TK_ident) {
562 head = sym_find(g, tk.txt);
563 if (head->type == Nonterminal)
564 err = "This non-terminal has already be used.";
565 else if (head->type == Virtual)
566 err = "Virtual symbol not permitted in head of production";
568 head->type = Nonterminal;
569 head->struct_name = g->current_type;
571 ## create production zero
573 head->first_production = g->production_count;
574 tk = token_next(state);
575 if (tk.num == TK_mark &&
576 text_is(tk.txt, "->"))
577 err = parse_production(g, head, state);
579 err = "'->' missing in production";
581 } else if (tk.num == TK_mark
582 && text_is(tk.txt, "|")) {
583 // another production for same non-term
585 err = parse_production(g, head, state);
587 err = "First production must have a head";
588 } else if (tk.num == TK_mark
589 && text_is(tk.txt, "$")) {
590 err = dollar_line(state, g);
592 err = "Unrecognised token at start of line.";
600 fprintf(stderr, "Error at line %d: %s\n",
607 ## Analysing the grammar
609 The central task in analysing the grammar is to determine a set of
610 states to drive the parsing process. This will require finding
611 various sets of symbols and of "items". Some of these sets will need
612 to attach information to each element in the set, so they are more
615 Each "item" may have a 'look-ahead' or `LA` set associated with
616 it. Many of these will be the same as each other. To avoid memory
617 wastage, and to simplify some comparisons of sets, these sets will be
618 stored in a list of unique sets, each assigned a number.
620 Once we have the data structures in hand to manage these sets and
621 lists, we can start setting the 'nullable' flag, build the 'FIRST'
622 sets, and then create the item sets which define the various states.
626 Though we don't only store symbols in these sets, they are the main
627 things we store, so they are called symbol sets or "symsets".
629 A symset has a size, an array of shorts, and an optional array of data
630 values, which are also shorts. If the array of data is not present,
631 we store an impossible pointer, as `NULL` is used to indicate that no
632 memory has been allocated yet;
637 unsigned short *syms, *data;
639 #define NO_DATA ((unsigned short *)1)
640 const struct symset INIT_SYMSET = { 0, NULL, NO_DATA };
641 const struct symset INIT_DATASET = { 0, NULL, NULL };
643 The arrays of shorts are allocated in blocks of 8 and are kept sorted
644 by using an insertion sort. We don't explicitly record the amount of
645 allocated space as it can be derived directly from the current `cnt` using
646 `((cnt - 1) | 7) + 1`.
649 static void symset_add(struct symset *s, unsigned short key, unsigned short val)
652 int current = ((s->cnt-1) | 7) + 1;
653 if (current == s->cnt) {
655 s->syms = realloc(s->syms, sizeof(*s->syms) * current);
656 if (s->data != NO_DATA)
657 s->data = realloc(s->data, sizeof(*s->data) * current);
660 while (i > 0 && s->syms[i-1] > key) {
661 s->syms[i] = s->syms[i-1];
662 if (s->data != NO_DATA)
663 s->data[i] = s->data[i-1];
667 if (s->data != NO_DATA)
672 Finding a symbol (or item) in a `symset` uses a simple binary search.
673 We return the index where the value was found (so data can be accessed),
674 or `-1` to indicate failure.
676 static int symset_find(struct symset *ss, unsigned short key)
683 while (lo + 1 < hi) {
684 int mid = (lo + hi) / 2;
685 if (ss->syms[mid] <= key)
690 if (ss->syms[lo] == key)
695 We will often want to form the union of two symsets. When we do, we
696 will often want to know if anything changed (as they might mean there
697 is more work to do). So `symset_union` reports whether anything was
698 added to the first set. We use a slow quadratic approach as these
699 sets don't really get very big. If profiles shows this to be a problem is
700 can be optimised later.
702 static int symset_union(struct symset *a, struct symset *b)
706 for (i = 0; i < b->cnt; i++)
707 if (symset_find(a, b->syms[i]) < 0) {
708 unsigned short data = 0;
709 if (b->data != NO_DATA)
711 symset_add(a, b->syms[i], data);
717 And of course we must be able to free a symset.
719 static void symset_free(struct symset ss)
722 if (ss.data != NO_DATA)
728 Some symsets are simply stored somewhere appropriate in the `grammar`
729 data structure, others needs a bit of indirection. This applies
730 particularly to "LA" sets. These will be paired with "items" in an
731 "itemset". As itemsets will be stored in a symset, the "LA" set needs to be
732 stored in the `data` field, so we need a mapping from a small (short)
733 number to an LA `symset`.
735 As mentioned earlier this involves creating a list of unique symsets.
737 For now, we just use a linear list sorted by insertion. A skiplist
738 would be more efficient and may be added later.
745 struct setlist *next;
748 ###### grammar fields
749 struct setlist *sets;
754 static int ss_cmp(struct symset a, struct symset b)
757 int diff = a.cnt - b.cnt;
760 for (i = 0; i < a.cnt; i++) {
761 diff = (int)a.syms[i] - (int)b.syms[i];
768 static int save_set(struct grammar *g, struct symset ss)
770 struct setlist **sl = &g->sets;
774 while (*sl && (cmp = ss_cmp((*sl)->ss, ss)) < 0)
781 s = malloc(sizeof(*s));
790 Finding a set by number is currently performed by a simple linear search.
791 If this turns out to hurt performance, we can store the sets in a dynamic
792 array like the productions.
794 static struct symset set_find(struct grammar *g, int num)
796 struct setlist *sl = g->sets;
797 while (sl && sl->num != num)
803 ### Setting `nullable`
805 We set `nullable` on the head symbol for any production for which all
806 the body symbols (if any) are nullable. As this is a recursive
807 definition, any change in the `nullable` setting means that we need to
808 re-evaluate where it needs to be set.
810 We simply loop around performing the same calculations until no more
817 static void set_nullable(struct grammar *g)
820 while (check_again) {
823 for (p = 0; p < g->production_count; p++) {
824 struct production *pr = g->productions[p];
827 if (pr->head->nullable)
829 for (s = 0; s < pr->body_size; s++)
830 if (! pr->body[s]->nullable)
832 if (s == pr->body_size) {
833 pr->head->nullable = 1;
840 ### Building the `first` sets
842 When calculating what can follow a particular non-terminal, we will need to
843 know what the "first" terminal in an subsequent non-terminal might be. So
844 we calculate the `first` set for every non-terminal and store them in an
845 array. We don't bother recording the "first" set for terminals as they are
848 As the "first" for one symbol might depend on the "first" of another,
849 we repeat the calculations until no changes happen, just like with
850 `set_nullable`. This makes use of the fact that `symset_union`
851 reports if any change happens.
853 The core of this which finds the "first" of part of a production body
854 will be reused for computing the "follow" sets, so we split it out
855 into a separate function.
857 ###### grammar fields
858 struct symset *first;
862 static int add_first(struct production *pr, int start,
863 struct symset *target, struct grammar *g,
868 for (s = start; s < pr->body_size; s++) {
869 struct symbol *bs = pr->body[s];
870 if (bs->type == Terminal) {
871 if (symset_find(target, bs->num) < 0) {
872 symset_add(target, bs->num, 0);
876 } else if (symset_union(target, &g->first[bs->num]))
882 *to_end = (s == pr->body_size);
886 static void build_first(struct grammar *g)
890 g->first = calloc(g->num_syms, sizeof(g->first[0]));
891 for (s = 0; s < g->num_syms; s++)
892 g->first[s] = INIT_SYMSET;
894 while (check_again) {
897 for (p = 0; p < g->production_count; p++) {
898 struct production *pr = g->productions[p];
899 struct symset *head = &g->first[pr->head->num];
901 if (add_first(pr, 0, head, g, NULL))
907 ### Building the `follow` sets.
909 There are two different situations when we will want to generate "follow"
910 sets. If we are doing an SLR analysis, we want to generate a single
911 "follow" set for each non-terminal in the grammar. That is what is
912 happening here. If we are doing an LALR or LR analysis we will want
913 to generate a separate "LA" set for each item. We do that later
916 There are two parts to generating a "follow" set. Firstly we look at
917 every place that any non-terminal appears in the body of any
918 production, and we find the set of possible "first" symbols after
919 there. This is done using `add_first` above and only needs to be done
920 once as the "first" sets are now stable and will not change.
924 for (p = 0; p < g->production_count; p++) {
925 struct production *pr = g->productions[p];
928 for (b = 0; b < pr->body_size - 1; b++) {
929 struct symbol *bs = pr->body[b];
930 if (bs->type == Terminal)
932 add_first(pr, b+1, &g->follow[bs->num], g, NULL);
936 The second part is to add the "follow" set of the head of a production
937 to the "follow" sets of the final symbol in the production, and any
938 other symbol which is followed only by `nullable` symbols. As this
939 depends on "follow" itself we need to repeatedly perform the process
940 until no further changes happen.
944 for (again = 0, p = 0;
945 p < g->production_count;
946 p < g->production_count-1
947 ? p++ : again ? (again = 0, p = 0)
949 struct production *pr = g->productions[p];
952 for (b = pr->body_size - 1; b >= 0; b--) {
953 struct symbol *bs = pr->body[b];
954 if (bs->type == Terminal)
956 if (symset_union(&g->follow[bs->num],
957 &g->follow[pr->head->num]))
964 We now just need to create and initialise the `follow` list to get a
967 ###### grammar fields
968 struct symset *follow;
971 static void build_follow(struct grammar *g)
974 g->follow = calloc(g->num_syms, sizeof(g->follow[0]));
975 for (s = 0; s < g->num_syms; s++)
976 g->follow[s] = INIT_SYMSET;
980 ### Building itemsets and states
982 There are three different levels of detail that can be chosen for
983 building the itemsets and states for the LR grammar. They are:
985 1. LR(0) or SLR(1), where no look-ahead is considered.
986 2. LALR(1) where we build look-ahead sets with each item and merge
987 the LA sets when we find two paths to the same "kernel" set of items.
988 3. LR(1) where different look-ahead for any item in the set means
989 a different state must be created.
991 ###### forward declarations
992 enum grammar_type { LR0, LR05, SLR, LALR, LR1 };
994 We need to be able to look through existing states to see if a newly
995 generated state already exists. For now we use a simple sorted linked
998 An item is a pair of numbers: the production number and the position of
999 "DOT", which is an index into the body of the production.
1000 As the numbers are not enormous we can combine them into a single "short"
1001 and store them in a `symset` - 4 bits for "DOT" and 12 bits for the
1002 production number (so 4000 productions with maximum of 15 symbols in the
1005 Comparing the itemsets will be a little different to comparing symsets
1006 as we want to do the lookup after generating the "kernel" of an
1007 itemset, so we need to ignore the offset=zero items which are added during
1010 To facilitate this, we modify the "DOT" number so that "0" sorts to
1011 the end of the list in the symset, and then only compare items before
1015 static inline unsigned short item_num(int production, int index)
1017 return production | ((31-index) << 11);
1019 static inline int item_prod(unsigned short item)
1021 return item & 0x7ff;
1023 static inline int item_index(unsigned short item)
1025 return (31-(item >> 11)) & 0x1f;
1028 For LR(1) analysis we need to compare not just the itemset in a state
1029 but also the LA sets. As we assign each unique LA set a number, we
1030 can just compare the symset and the data values together.
1033 static int itemset_cmp(struct symset a, struct symset b,
1034 enum grammar_type type)
1040 i < a.cnt && i < b.cnt &&
1041 item_index(a.syms[i]) > 0 &&
1042 item_index(b.syms[i]) > 0;
1044 int diff = a.syms[i] - b.syms[i];
1048 diff = a.data[i] - b.data[i];
1053 if (i == a.cnt || item_index(a.syms[i]) == 0)
1057 if (i == b.cnt || item_index(b.syms[i]) == 0)
1063 if (type < LR1 || av == -1)
1066 a.data[i] - b.data[i];
1069 And now we can build the list of itemsets. The lookup routine returns
1070 both a success flag and a pointer to where in the list an insert
1071 should happen, so we don't need to search a second time.
1075 struct itemset *next;
1077 struct symset items;
1078 struct symset go_to;
1082 ###### grammar fields
1083 struct itemset *items;
1087 static int itemset_find(struct grammar *g, struct itemset ***where,
1088 struct symset kernel, enum grammar_type type)
1090 struct itemset **ip;
1092 for (ip = &g->items; *ip ; ip = & (*ip)->next) {
1093 struct itemset *i = *ip;
1095 diff = itemset_cmp(i->items, kernel, type);
1108 Adding an itemset may require merging the LA sets if LALR analysis is
1109 happening. If any new LA set add symbol that weren't in the old LA set, we
1110 clear the `completed` flag so that the dependants of this itemset will be
1111 recalculated and their LA sets updated.
1113 `add_itemset` must consume the symsets it is passed, either by adding
1114 them to a data structure, of freeing them.
1116 static int add_itemset(struct grammar *g, struct symset ss,
1117 enum grammar_type type)
1119 struct itemset **where, *is;
1121 int found = itemset_find(g, &where, ss, type);
1123 is = calloc(1, sizeof(*is));
1124 is->state = g->states;
1128 is->go_to = INIT_DATASET;
1137 for (i = 0; i < ss.cnt; i++) {
1138 struct symset temp = INIT_SYMSET, s;
1139 if (ss.data[i] == is->items.data[i])
1141 s = set_find(g, is->items.data[i]);
1142 symset_union(&temp, &s);
1143 s = set_find(g, ss.data[i]);
1144 if (symset_union(&temp, &s)) {
1145 is->items.data[i] = save_set(g, temp);
1156 To build all the itemsets, we first insert the initial itemset made from the
1157 start symbol, complete each itemset, and then generate new itemsets from old
1158 until no new ones can be made.
1160 Completing an itemset means finding all the items where "DOT" is followed by
1161 a nonterminal and adding "DOT=0" items for every production from that
1162 non-terminal - providing each item hasn't already been added.
1164 If LA sets are needed, the LA set for each new item is found using
1165 `add_first` which was developed earlier for `FIRST` and `FOLLOW`. This may
1166 be supplemented by the LA set for the item which produce the new item.
1168 We also collect a set of all symbols which follow "DOT" (in `done`) as this
1169 is used in the next stage.
1171 NOTE: precedence handling should happen here - I haven't written this yet
1174 ###### complete itemset
1175 for (i = 0; i < is->items.cnt; i++) {
1176 int p = item_prod(is->items.syms[i]);
1177 int bs = item_index(is->items.syms[i]);
1178 struct production *pr = g->productions[p];
1181 struct symset LA = INIT_SYMSET;
1182 unsigned short sn = 0;
1184 if (bs == pr->body_size)
1187 if (symset_find(&done, s->num) < 0)
1188 symset_add(&done, s->num, 0);
1189 if (s->type != Nonterminal)
1195 add_first(pr, bs+1, &LA, g, &to_end);
1197 struct symset ss = set_find(g, is->items.data[i]);
1198 symset_union(&LA, &ss);
1200 sn = save_set(g, LA);
1201 LA = set_find(g, sn);
1203 /* Add productions for this symbol */
1204 for (p2 = s->first_production;
1205 p2 < g->production_count &&
1206 g->productions[p2]->head == s;
1208 int itm = item_num(p2, 0);
1209 int pos = symset_find(&is->items, itm);
1211 symset_add(&is->items, itm, sn);
1212 /* Will have re-ordered, so start
1213 * from beginning again */
1215 } else if (type >= LALR) {
1216 struct symset ss = set_find(g, is->items.data[pos]);
1217 struct symset tmp = INIT_SYMSET;
1219 symset_union(&tmp, &ss);
1220 if (symset_union(&tmp, &LA)) {
1221 is->items.data[pos] = save_set(g, tmp);
1229 For each symbol we found (and placed in `done`) we collect all the items for
1230 which this symbol is next, and create a new itemset with "DOT" advanced over
1231 the symbol. This is then added to the collection of itemsets (or merged
1232 with a pre-existing itemset).
1234 ###### derive itemsets
1235 // Now we have a completed itemset, so we need to
1236 // create all the 'next' itemset and create them
1237 // if they don't exist.
1238 for (i = 0; i < done.cnt; i++) {
1240 unsigned short state;
1241 struct symset newitemset = INIT_SYMSET;
1243 newitemset = INIT_DATASET;
1245 for (j = 0; j < is->items.cnt; j++) {
1246 int itm = is->items.syms[j];
1247 int p = item_prod(itm);
1248 int bp = item_index(itm);
1249 struct production *pr = g->productions[p];
1250 unsigned short la = 0;
1253 if (bp == pr->body_size)
1255 if (pr->body[bp]->num != done.syms[i])
1258 la = is->items.data[j];
1259 pos = symset_find(&newitemset, pr->head->num);
1261 symset_add(&newitemset, item_num(p, bp+1), la);
1262 else if (type >= LALR) {
1263 // Need to merge la set.
1264 int la2 = newitemset.data[pos];
1266 struct symset ss = set_find(g, la2);
1267 struct symset LA = INIT_SYMSET;
1268 symset_union(&LA, &ss);
1269 ss = set_find(g, la);
1270 if (symset_union(&LA, &ss))
1271 newitemset.data[pos] = save_set(g, LA);
1277 state = add_itemset(g, newitemset, type);
1278 if (symset_find(&is->go_to, done.syms[i]) < 0)
1279 symset_add(&is->go_to, done.syms[i], state);
1282 All that is left is to crate the initial itemset from production zero, and
1283 with `TK_eof` as the LA set.
1286 static void build_itemsets(struct grammar *g, enum grammar_type type)
1288 struct symset first = INIT_SYMSET;
1291 unsigned short la = 0;
1293 // LA set just has eof
1294 struct symset eof = INIT_SYMSET;
1295 symset_add(&eof, TK_eof, 0);
1296 la = save_set(g, eof);
1297 first = INIT_DATASET;
1299 // production 0, offset 0 (with no data)
1300 symset_add(&first, item_num(0, 0), la);
1301 add_itemset(g, first, type);
1302 for (again = 0, is = g->items;
1304 is = is->next ?: again ? (again = 0, g->items) : NULL) {
1306 struct symset done = INIT_SYMSET;
1317 ### Completing the analysis.
1319 The exact process of analysis depends on which level was requested. For
1320 `LR0` and `LR05` we don't need first and follow sets at all. For `LALR` and
1321 `LR1` we need first, but not follow. For `SLR` we need both.
1323 We don't build the "action" tables that you might expect as the parser
1324 doesn't use them. They will be calculated to some extent if needed for
1327 Once we have built everything we allocate arrays for the two lists:
1328 symbols and itemsets. This allows more efficient access during reporting.
1329 The symbols are grouped as terminals and non-terminals and we record the
1330 changeover point in `first_nonterm`.
1332 ###### grammar fields
1333 struct symbol **symtab;
1334 struct itemset **statetab;
1337 ###### grammar_analyse
1339 static void grammar_analyse(struct grammar *g, enum grammar_type type)
1343 int snum = TK_reserved;
1344 for (s = g->syms; s; s = s->next)
1345 if (s->num < 0 && s->type == Terminal) {
1349 g->first_nonterm = snum;
1350 for (s = g->syms; s; s = s->next)
1356 g->symtab = calloc(g->num_syms, sizeof(g->symtab[0]));
1357 for (s = g->syms; s; s = s->next)
1358 g->symtab[s->num] = s;
1367 build_itemsets(g, type);
1369 g->statetab = calloc(g->states, sizeof(g->statetab[0]));
1370 for (is = g->items; is ; is = is->next)
1371 g->statetab[is->state] = is;
1374 ## Reporting on the Grammar
1376 The purpose of the report is to give the grammar developer insight into
1377 how the grammar parser will work. It is basically a structured dump of
1378 all the tables that have been generated, plus an description of any conflicts.
1380 ###### grammar_report
1381 static int grammar_report(struct grammar *g, enum grammar_type type)
1387 return report_conflicts(g, type);
1390 Firstly we have the complete list of symbols, together with the "FIRST"
1391 set if that was generated.
1395 static void report_symbols(struct grammar *g)
1399 printf("SYMBOLS + FIRST:\n");
1401 printf("SYMBOLS:\n");
1403 for (n = 0; n < g->num_syms; n++) {
1404 struct symbol *s = g->symtab[n];
1408 printf(" %c%3d%c: ",
1409 s->nullable ? '*':' ',
1410 s->num, symtypes[s->type]);
1413 printf(" (%d%s)", s->precedence,
1414 assoc_names[s->assoc]);
1416 if (g->first && s->type == Nonterminal) {
1419 for (j = 0; j < g->first[n].cnt; j++) {
1422 prtxt(g->symtab[g->first[n].syms[j]]->name);
1429 Then we have to follow sets if they were computed.
1431 static void report_follow(struct grammar *g)
1434 printf("FOLLOW:\n");
1435 for (n = 0; n < g->num_syms; n++)
1436 if (g->follow[n].cnt) {
1440 prtxt(g->symtab[n]->name);
1441 for (j = 0; j < g->follow[n].cnt; j++) {
1444 prtxt(g->symtab[g->follow[n].syms[j]]->name);
1450 And finally the item sets. These include the GO TO tables and, for
1451 LALR and LR1, the LA set for each item. Lots of stuff, so we break
1452 it up a bit. First the items, with production number and associativity.
1454 static void report_item(struct grammar *g, int itm)
1456 int p = item_prod(itm);
1457 int dot = item_index(itm);
1458 struct production *pr = g->productions[p];
1462 prtxt(pr->head->name);
1464 for (i = 0; i < pr->body_size; i++) {
1465 printf(" %s", (dot == i ? ". ": ""));
1466 prtxt(pr->body[i]->name);
1468 if (dot == pr->body_size)
1472 printf(" (%d%s)", pr->precedence,
1473 assoc_names[pr->assoc]);
1477 The LA sets which are (possibly) reported with each item:
1479 static void report_la(struct grammar *g, int lanum)
1481 struct symset la = set_find(g, lanum);
1485 printf(" LOOK AHEAD(%d)", lanum);
1486 for (i = 0; i < la.cnt; i++) {
1489 prtxt(g->symtab[la.syms[i]]->name);
1494 Then the go to sets:
1497 static void report_goto(struct grammar *g, struct symset gt)
1502 for (i = 0; i < gt.cnt; i++) {
1504 prtxt(g->symtab[gt.syms[i]]->name);
1505 printf(" -> %d\n", gt.data[i]);
1509 Now we can report all the item sets complete with items, LA sets, and GO TO.
1511 static void report_itemsets(struct grammar *g)
1514 printf("ITEM SETS(%d)\n", g->states);
1515 for (s = 0; s < g->states; s++) {
1517 struct itemset *is = g->statetab[s];
1518 printf(" Itemset %d:\n", s);
1519 for (j = 0; j < is->items.cnt; j++) {
1520 report_item(g, is->items.syms[j]);
1521 if (is->items.data != NO_DATA)
1522 report_la(g, is->items.data[j]);
1524 report_goto(g, is->go_to);
1528 ### Reporting conflicts
1530 Conflict detection varies a lot among different analysis levels. However
1531 LR0 and LR0.5 are quite similar - having no follow sets, and SLR, LALR and
1532 LR1 are also similar as they have FOLLOW or LA sets.
1536 ## conflict functions
1538 static int report_conflicts(struct grammar *g, enum grammar_type type)
1541 printf("Conflicts:\n");
1543 cnt = conflicts_lr0(g, type);
1545 cnt = conflicts_slr(g, type);
1547 printf(" - no conflicts\n");
1551 LR0 conflicts are any state which have both a reducible item and
1554 LR05 conflicts only occurs if two possibly reductions exist,
1555 as shifts always over-ride reductions.
1557 ###### conflict functions
1558 static int conflicts_lr0(struct grammar *g, enum grammar_type type)
1562 for (i = 0; i < g->states; i++) {
1563 struct itemset *is = g->statetab[i];
1564 int last_reduce = -1;
1565 int prev_reduce = -1;
1566 int last_shift = -1;
1570 for (j = 0; j < is->items.cnt; j++) {
1571 int itm = is->items.syms[j];
1572 int p = item_prod(itm);
1573 int bp = item_index(itm);
1574 struct production *pr = g->productions[p];
1576 if (bp == pr->body_size) {
1577 prev_reduce = last_reduce;
1581 if (pr->body[bp]->type == Terminal)
1584 if (type == LR0 && last_reduce >= 0 && last_shift >= 0) {
1585 printf(" State %d has both SHIFT and REDUCE:\n", i);
1586 report_item(g, is->items.syms[last_shift]);
1587 report_item(g, is->items.syms[last_reduce]);
1590 if (prev_reduce >= 0) {
1591 printf(" State %d has 2 (or more) reducible items\n", i);
1592 report_item(g, is->items.syms[prev_reduce]);
1593 report_item(g, is->items.syms[last_reduce]);
1600 SLR, LALR, and LR1 conflicts happen if two reducible items have over-lapping
1601 look ahead, or if a symbol in a look-ahead can be shifted. The differ only
1602 in the source of the look ahead set.
1604 We build a dataset mapping terminal to item for possible SHIFTs and then
1605 another for possible REDUCE operations. We report when we get conflicts
1608 static int conflicts_slr(struct grammar *g, enum grammar_type type)
1613 for (i = 0; i < g->states; i++) {
1614 struct itemset *is = g->statetab[i];
1615 struct symset shifts = INIT_DATASET;
1616 struct symset reduce = INIT_DATASET;
1620 /* First collect the shifts */
1621 for (j = 0; j < is->items.cnt; j++) {
1622 unsigned short itm = is->items.syms[j];
1623 int p = item_prod(itm);
1624 int bp = item_index(itm);
1625 struct production *pr = g->productions[p];
1627 if (bp < pr->body_size &&
1628 pr->body[bp]->type == Terminal) {
1630 int sym = pr->body[bp]->num;
1631 if (symset_find(&shifts, sym) < 0)
1632 symset_add(&shifts, sym, itm);
1635 /* Now look for reduction and conflicts */
1636 for (j = 0; j < is->items.cnt; j++) {
1637 unsigned short itm = is->items.syms[j];
1638 int p = item_prod(itm);
1639 int bp = item_index(itm);
1640 struct production *pr = g->productions[p];
1642 if (bp < pr->body_size)
1647 la = g->follow[pr->head->num];
1649 la = set_find(g, is->items.data[j]);
1651 for (k = 0; k < la.cnt; k++) {
1652 int pos = symset_find(&shifts, la.syms[k]);
1654 printf(" State %d has SHIFT/REDUCE conflict on ", i);
1655 prtxt(g->symtab[la.syms[k]]->name);
1657 report_item(g, shifts.data[pos]);
1658 report_item(g, itm);
1661 pos = symset_find(&reduce, la.syms[k]);
1663 symset_add(&reduce, la.syms[k], itm);
1666 printf(" State %d has REDUCE/REDUCE conflict on ", i);
1667 prtxt(g->symtab[la.syms[k]]->name);
1669 report_item(g, itm);
1670 report_item(g, reduce.data[pos]);
1674 symset_free(shifts);
1675 symset_free(reduce);
1681 ## Generating the parser
1683 The export part of the parser is the `parse_XX` function, where the name
1684 `XX` is based on the name of the parser files.
1686 This takes a `code_node`, a partially initialized `token_config`, and an
1687 optional `FILE` to send tracing to. The `token_config` gets the list of
1688 known words added and then is used with the `code_node` to initialize the
1691 `parse_XX` then call the library function `parser_run` to actually complete
1692 the parse. This needs the `states` table and function to call the various
1693 pieces of code provided in the grammar file, so they are generated first.
1695 ###### parser_generate
1697 static void gen_parser(FILE *f, struct grammar *g, char *file, char *name)
1703 gen_reduce(f, g, file);
1706 fprintf(f, "#line 0 \"gen_parser\"\n");
1707 fprintf(f, "void *parse_%s(struct code_node *code, struct token_config *config, FILE *trace)\n",
1710 fprintf(f, "\tstruct token_state *tokens;\n");
1711 fprintf(f, "\tconfig->words_marks = known;\n");
1712 fprintf(f, "\tconfig->known_count = sizeof(known)/sizeof(known[0]);\n");
1713 fprintf(f, "\tconfig->ignored |= (1 << TK_line_comment) | (1 << TK_block_comment);\n");
1714 fprintf(f, "\ttokens = token_open(code, config);\n");
1715 fprintf(f, "\tvoid *rv = parser_run(tokens, states, do_reduce, do_free, trace, non_term, config->known_count);\n");
1716 fprintf(f, "\ttoken_close(tokens);\n");
1717 fprintf(f, "\treturn rv;\n");
1718 fprintf(f, "}\n\n");
1721 ### Table words table
1723 The know words is simply an array of terminal symbols.
1724 The table of nonterminals used for tracing is a similar array.
1728 static void gen_known(FILE *f, struct grammar *g)
1731 fprintf(f, "#line 0 \"gen_known\"\n");
1732 fprintf(f, "static const char *known[] = {\n");
1733 for (i = TK_reserved;
1734 i < g->num_syms && g->symtab[i]->type == Terminal;
1736 fprintf(f, "\t\"%.*s\",\n", g->symtab[i]->name.len,
1737 g->symtab[i]->name.txt);
1738 fprintf(f, "};\n\n");
1741 static void gen_non_term(FILE *f, struct grammar *g)
1744 fprintf(f, "#line 0 \"gen_non_term\"\n");
1745 fprintf(f, "static const char *non_term[] = {\n");
1746 for (i = TK_reserved;
1749 if (g->symtab[i]->type == Nonterminal)
1750 fprintf(f, "\t\"%.*s\",\n", g->symtab[i]->name.len,
1751 g->symtab[i]->name.txt);
1752 fprintf(f, "};\n\n");
1755 ### States and the goto tables.
1757 For each state we record the goto table, the reducible production if
1758 there is one, or a symbol to shift for error recovery.
1759 Some of the details of the reducible production are stored in the
1760 `do_reduce` function to come later. Here we store the production number,
1761 the body size (useful for stack management) and the resulting symbol (useful
1762 for knowing how to free data later).
1764 The go to table is stored in a simple array of `sym` and corresponding
1767 ###### exported types
1775 const struct lookup * go_to;
1785 static void gen_goto(FILE *f, struct grammar *g)
1788 fprintf(f, "#line 0 \"gen_goto\"\n");
1789 for (i = 0; i < g->states; i++) {
1791 fprintf(f, "static const struct lookup goto_%d[] = {\n",
1793 struct symset gt = g->statetab[i]->go_to;
1794 for (j = 0; j < gt.cnt; j++)
1795 fprintf(f, "\t{ %d, %d },\n",
1796 gt.syms[j], gt.data[j]);
1803 static void gen_states(FILE *f, struct grammar *g)
1806 fprintf(f, "#line 0 \"gen_states\"\n");
1807 fprintf(f, "static const struct state states[] = {\n");
1808 for (i = 0; i < g->states; i++) {
1809 struct itemset *is = g->statetab[i];
1810 int j, prod = -1, prod_len;
1812 int shift_len = 0, shift_remain = 0;
1813 for (j = 0; j < is->items.cnt; j++) {
1814 int itm = is->items.syms[j];
1815 int p = item_prod(itm);
1816 int bp = item_index(itm);
1817 struct production *pr = g->productions[p];
1819 if (bp < pr->body_size) {
1820 if (shift_sym < 0 ||
1821 (shift_len == bp && shift_remain > pr->body_size - bp)) {
1822 shift_sym = pr->body[bp]->num;
1824 shift_remain = pr->body_size - bp;
1828 /* This is what we reduce */
1829 if (prod < 0 || prod_len < pr->body_size) {
1831 prod_len = pr->body_size;
1836 fprintf(f, "\t[%d] = { %d, goto_%d, %d, %d, %d, 0 },\n",
1837 i, is->go_to.cnt, i, prod,
1838 g->productions[prod]->body_size,
1839 g->productions[prod]->head->num);
1841 fprintf(f, "\t[%d] = { %d, goto_%d, -1, -1, -1, %d },\n",
1842 i, is->go_to.cnt, i, shift_sym);
1844 fprintf(f, "};\n\n");
1847 ### The `do_reduce` function and the code
1849 When the parser engine decides to reduce a production, it calls `do_reduce`.
1850 This has two functions.
1852 Firstly, if a non-NULL `trace` file is passed, it prints out details of the
1853 production being reduced. Secondly it runs the code that was included with
1854 the production if any.
1856 This code needs to be able to store data somewhere. Rather than requiring
1857 `do_reduce` to `malloc` that "somewhere", we pass in a large buffer and have
1858 `do_reduce` return the size to be saved.
1860 The code fragment requires translation when written out. Any `$N` needs to
1861 be converted to a reference either to that buffer (if `$0`) or to the
1862 structure returned by a previous reduction. These pointer need to be cast
1863 to the appropriate type for each access. All this is handling in
1869 static void gen_code(struct production *p, FILE *f, struct grammar *g)
1872 fprintf(f, "\t\t\t");
1873 for (c = p->code.txt; c < p->code.txt + p->code.len; c++) {
1882 if (*c < '0' || *c > '9') {
1887 while (c[1] >= '0' && c[1] <= '9') {
1889 n = n * 10 + *c - '0';
1892 fprintf(f, "(*(struct %.*s*)ret)",
1893 p->head->struct_name.len,
1894 p->head->struct_name.txt);
1895 else if (n > p->body_size)
1896 fprintf(f, "$%d", n);
1897 else if (p->body[n-1]->type == Terminal)
1898 fprintf(f, "(*(struct token *)body[%d])",
1900 else if (p->body[n-1]->struct_name.txt == NULL)
1901 fprintf(f, "$%d", n);
1903 fprintf(f, "(*(struct %.*s*)body[%d])",
1904 p->body[n-1]->struct_name.len,
1905 p->body[n-1]->struct_name.txt, n-1);
1912 static void gen_reduce(FILE *f, struct grammar *g, char *file)
1915 fprintf(f, "#line 0 \"gen_reduce\"\n");
1916 fprintf(f, "static int do_reduce(int prod, void **body, void *ret)\n");
1918 fprintf(f, "\tint ret_size = 0;\n");
1920 fprintf(f, "\tswitch(prod) {\n");
1921 for (i = 0; i < g->production_count; i++) {
1922 struct production *p = g->productions[i];
1923 fprintf(f, "\tcase %d:\n", i);
1928 if (p->head->struct_name.txt)
1929 fprintf(f, "\t\tret_size = sizeof(struct %.*s);\n",
1930 p->head->struct_name.len,
1931 p->head->struct_name.txt);
1933 fprintf(f, "\t\tbreak;\n");
1935 fprintf(f, "\t}\n\treturn ret_size;\n}\n\n");
1940 As each non-terminal can potentially cause a different type of data
1941 structure to be allocated and filled in, we need to be able to free it when
1944 It is particularly important to have fine control over freeing during error
1945 recovery where individual stack frames might need to be freed.
1947 For this, the grammar author required to defined a `free_XX` function for
1948 each structure that is used by a non-terminal. `do_free` all call whichever
1949 is appropriate given a symbol number, and will call `free` (as is
1950 appropriate for tokens` on any terminal symbol.
1954 static void gen_free(FILE *f, struct grammar *g)
1958 fprintf(f, "#line 0 \"gen_free\"\n");
1959 fprintf(f, "static void do_free(short sym, void *asn)\n");
1961 fprintf(f, "\tif (!asn) return;\n");
1962 fprintf(f, "\tif (sym < %d) {\n", g->first_nonterm);
1963 fprintf(f, "\t\tfree(asn);\n\t\treturn;\n\t}\n");
1964 fprintf(f, "\tswitch(sym) {\n");
1966 for (i = 0; i < g->num_syms; i++) {
1967 struct symbol *s = g->symtab[i];
1969 s->type != Nonterminal ||
1970 s->struct_name.txt == NULL)
1973 fprintf(f, "\tcase %d:\n", s->num);
1974 fprintf(f, "\t\tfree_%.*s(asn);\n",
1976 s->struct_name.txt);
1977 fprintf(f, "\t\tbreak;\n");
1979 fprintf(f, "\t}\n}\n\n");
1982 ## The main routine.
1984 There are three key parts to the "main" function of parsergen: processing
1985 the arguments, loading the grammar file, and dealing with the grammar.
1987 ### Argument processing.
1989 Command line options allow the selection of analysis mode, name of output
1990 file, and whether or not a report should be generated. By default we create
1991 a report only if no code output was requested.
1993 The `parse_XX` function name uses the basename of the output file which
1994 should not have a suffix (`.c`). `.c` is added to the given name for the
1995 code, and `.h` is added for the header (if header text is specifed in the
2002 static const struct option long_options[] = {
2003 { "LR0", 0, NULL, '0' },
2004 { "LR05", 0, NULL, '5' },
2005 { "SLR", 0, NULL, 'S' },
2006 { "LALR", 0, NULL, 'L' },
2007 { "LR1", 0, NULL, '1' },
2008 { "report", 0, NULL, 'R' },
2009 { "output", 1, NULL, 'o' },
2010 { NULL, 0, NULL, 0 }
2012 const char *options = "05SL1Ro:";
2014 ###### process arguments
2016 char *outfile = NULL;
2020 enum grammar_type type = LR05;
2021 while ((opt = getopt_long(argc, argv, options,
2022 long_options, NULL)) != -1) {
2037 outfile = optarg; break;
2039 fprintf(stderr, "Usage: parsergen ...\n");
2044 infile = argv[optind++];
2046 fprintf(stderr, "No input file given\n");
2049 if (outfile && report == 1)
2052 if (name && strchr(name, '/'))
2053 name = strrchr(name, '/')+1;
2055 if (optind < argc) {
2056 fprintf(stderr, "Excess command line arguments\n");
2060 ### Loading the grammar file
2062 To be able to run `mdcode` and `scanner` on the grammar we need to memory
2065 One we have extracted the code (with `mdcode`) we expect to file three
2066 sections: header, code, and grammar. Anything else is an error.
2068 "header" and "code" are optional, though it is hard to build a working
2069 parser with neither. "grammar" must be provided.
2073 #include <sys/mman.h>
2078 static void pr_err(char *msg)
2081 fprintf(stderr, "%s\n", msg);
2085 struct section *table;
2089 fd = open(infile, O_RDONLY);
2091 fprintf(stderr, "parsergen: cannot open %s: %s\n",
2092 infile, strerror(errno));
2095 len = lseek(fd, 0, 2);
2096 file = mmap(NULL, len, PROT_READ, MAP_SHARED, fd, 0);
2097 table = code_extract(file, file+len, pr_err);
2099 struct code_node *hdr = NULL;
2100 struct code_node *code = NULL;
2101 struct code_node *gram = NULL;
2102 for (s = table; s; s = s->next) {
2103 if (text_is(s->section, "header"))
2105 else if (text_is(s->section, "code"))
2107 else if (text_is(s->section, "grammar"))
2110 fprintf(stderr, "Unknown content section: %.*s\n",
2111 s->section.len, s->section.txt);
2116 ### Processing the input
2118 As we need to append an extention to a filename and then open it for
2119 writing, and we need to do this twice, it helps to have a separate function.
2123 static FILE *open_ext(char *base, char *ext)
2125 char *fn = malloc(strlen(base) + strlen(ext) + 1);
2127 strcat(strcpy(fn, base), ext);
2133 If we can read the grammar, then we analyse and optionally report on it. If
2134 the report finds conflicts we will exit with an error status.
2136 ###### process input
2137 struct grammar *g = NULL;
2139 fprintf(stderr, "No grammar section provided\n");
2142 g = grammar_read(gram);
2144 fprintf(stderr, "Failure to parse grammar\n");
2149 grammar_analyse(g, type);
2151 if (grammar_report(g, type))
2155 If a headers section is defined, we write it out and include a declaration
2156 for the `parse_XX` function so it can be used from separate code.
2158 if (rv == 0 && hdr && outfile) {
2159 FILE *f = open_ext(outfile, ".h");
2161 code_node_print(f, hdr, infile);
2162 fprintf(f, "void *parse_%s(struct code_node *code, struct token_config *config, FILE *trace);\n",
2166 fprintf(stderr, "Cannot create %s.h\n",
2172 And if all goes well and an output file was provided, we create the `.c`
2173 file with the code section (if any) and the parser tables and function.
2175 if (rv == 0 && outfile) {
2176 FILE *f = open_ext(outfile, ".c");
2179 code_node_print(f, code, infile);
2180 gen_parser(f, g, infile, name);
2183 fprintf(stderr, "Cannot create %s.c\n",
2189 And that about wraps it up. We need to set the locale so that UTF-8 is
2190 recognised properly, and link with `libicuuc` is `libmdcode` requires that.
2192 ###### File: parsergen.mk
2193 parsergen : parsergen.o libscanner.o libmdcode.o
2194 $(CC) $(CFLAGS) -o parsergen parsergen.o libscanner.o libmdcode.o -licuuc
2201 int main(int argc, char *argv[])
2206 setlocale(LC_ALL,"");
2208 ## process arguments
2215 ## The SHIFT/REDUCE parser
2217 Having analysed the grammar and generated all the table, we only need the
2218 shift/reduce engine to bring it all together.
2220 ### Goto table lookup
2222 The parser generator has nicely provided us with goto tables sorted by
2223 symbol number. We need a binary search function to find a symbol in the
2226 ###### parser functions
2228 static int search(const struct state *l, int sym)
2231 int hi = l->go_to_cnt;
2235 while (lo + 1 < hi) {
2236 int mid = (lo + hi) / 2;
2237 if (l->go_to[mid].sym <= sym)
2242 if (l->go_to[lo].sym == sym)
2243 return l->go_to[lo].state;
2248 ### The state stack.
2250 The core data structure for the parser is the stack. This tracks all the
2251 symbols that have been recognised or partially recognised.
2253 The stack usually won't grow very large - maybe a few tens of entries. So
2254 we dynamically resize and array as required but never bother to shrink it
2257 We keep the stack as two separate allocations. One, `asn_stack` stores the
2258 "abstract syntax nodes" which are created by each reduction. When we call
2259 `do_reduce` we need to pass an array of the `asn`s of the body of the
2260 production, and by keeping a separate `asn` stack, we can just pass a
2261 pointer into this stack.
2263 The other allocation stores all other stack fields of which there are four.
2264 The `state` is the most important one and guides the parsing process. The
2265 `sym` is nearly unnecessary. However when we want to free entries from the
2266 `asn_stack`, it helps to know what type they are so we can call the right
2267 freeing function. The symbol leads us to the right free function through
2270 The `indents` count and the `starts_indented` flag track the line
2271 indents in the symbol. These are used to allow indent information to
2272 guide parsing and error recovery.
2274 As well as the stack of frames we have a `next` frame which is
2275 assembled from the incoming token and other information prior to
2276 pushing it onto the stack.
2278 ###### parser functions
2284 short starts_indented;
2294 Two operations are needed on the stack - shift (which is like push) and pop.
2296 Shift applies not only to terminals but also to non-terminals. When we
2297 reduce a production we will pop off entries corresponding to the body
2298 symbols, then push on an item for the head of the production. This last is
2299 exactly the same process as shifting in a terminal so we use the same
2302 To simplify other code we arrange for `shift` to fail if there is no `goto`
2303 state for the symbol. This is useful in basic parsing due to our design
2304 that we shift when we can, and reduce when we cannot. So the `shift`
2305 function reports if it could.
2307 So `shift` finds the next state. If that succeed it extends the allocations
2308 if needed and pushed all the information onto the stacks.
2310 ###### parser functions
2312 static int shift(struct parser *p,
2314 const struct state states[])
2316 // Push an entry onto the stack
2317 int newstate = search(&states[p->next.state], p->next.sym);
2320 if (p->tos >= p->stack_size) {
2321 p->stack_size += 10;
2322 p->stack = realloc(p->stack, p->stack_size
2323 * sizeof(p->stack[0]));
2324 p->asn_stack = realloc(p->asn_stack, p->stack_size
2325 * sizeof(p->asn_stack[0]));
2327 p->stack[p->tos] = p->next;
2328 p->asn_stack[p->tos] = asn;
2330 p->next.state = newstate;
2331 p->next.indents = 0;
2332 p->next.starts_indented = 0;
2336 `pop` simply moves the top of stack (`tos`) back down the required amount
2337 and frees any `asn` entries that need to be freed. It is called _after_ we
2338 reduce a production, just before we `shift` the nonterminal in.
2340 ###### parser functions
2342 static void pop(struct parser *p, int num,
2343 void(*do_free)(short sym, void *asn))
2347 for (i = 0; i < num; i++) {
2348 p->next.indents += p->stack[p->tos+i].indents;
2349 do_free(p->stack[p->tos+i].sym,
2350 p->asn_stack[p->tos+i]);
2354 p->next.state = p->stack[p->tos].state;
2355 p->next.starts_indented = p->stack[p->tos].starts_indented;
2359 ### Memory allocation
2361 The `scanner` returns tokens in a local variable - we want them in allocated
2362 memory so they can live in the `asn_stack`. Similarly the `asn` produced by
2363 a reduce is in a large buffer. Both of these require some allocation and
2364 copying, hence `memdup` and `tokcopy`.
2366 ###### parser includes
2369 ###### parser functions
2371 void *memdup(void *m, int len)
2377 memcpy(ret, m, len);
2381 static struct token *tok_copy(struct token tk)
2383 struct token *new = malloc(sizeof(*new));
2388 ### The heart of the parser.
2390 Now we have the parser. If we can shift, we do. If not and we can reduce
2391 we do. If the production we reduced was production zero, then we have
2392 accepted the input and can finish.
2394 We return whatever `asn` was returned by reducing production zero.
2396 If we can neither shift nor reduce we have an error to handle. We pop
2397 single entries off the stack until we can shift the `TK_error` symbol, then
2398 drop input tokens until we find one we can shift into the new error state.
2400 When we find `TK_in` and `TK_out` tokens which report indents we need
2401 to handle them directly as the grammar cannot express what we want to
2404 `TK_in` tokens are easy: we simply update the `next` stack frame to
2405 record how many indents there are and that the next token started with
2408 `TK_out` tokens must either be counted off against any pending indent,
2409 or must force reductions until there is a pending indent which isn't
2410 at the start of a production.
2412 ###### parser includes
2415 void *parser_run(struct token_state *tokens,
2416 const struct state states[],
2417 int (*do_reduce)(int, void**, void*),
2418 void (*do_free)(short, void*),
2419 FILE *trace, const char *non_term[], int knowns)
2421 struct parser p = { 0 };
2422 struct token *tk = NULL;
2427 struct token *err_tk;
2429 tk = tok_copy(token_next(tokens));
2430 p.next.sym = tk->num;
2432 parser_trace(trace, &p, tk, states, non_term, knowns);
2434 if (p.next.sym == TK_in) {
2435 p.next.starts_indented = 1;
2441 if (p.next.sym == TK_out) {
2442 if (p.stack[p.tos-1].indents > p.stack[p.tos-1].starts_indented ||
2443 (p.stack[p.tos-1].indents == 1 &&
2444 states[p.next.state].reduce_size > 1)) {
2445 p.stack[p.tos-1].indents -= 1;
2450 // fall through and force a REDUCE (as 'shift'
2453 if (shift(&p, tk, states)) {
2457 if (states[p.next.state].reduce_prod >= 0) {
2459 int prod = states[p.next.state].reduce_prod;
2460 int size = states[p.next.state].reduce_size;
2462 static char buf[16*1024];
2463 p.next.sym = states[p.next.state].reduce_sym;
2465 body = p.asn_stack +
2466 (p.tos - states[p.next.state].reduce_size);
2468 bufsize = do_reduce(prod, body, buf);
2470 pop(&p, size, do_free);
2471 shift(&p, memdup(buf, bufsize), states);
2476 if (tk->num == TK_out) {
2477 // Indent problem - synthesise tokens to get us
2479 fprintf(stderr, "Synthesize %d to handle indent problem\n", states[p.next.state].shift_sym);
2480 p.next.sym = states[p.next.state].shift_sym;
2481 shift(&p, tok_copy(*tk), states);
2482 // FIXME need to report this error somehow
2485 /* Error. We walk up the stack until we
2486 * find a state which will accept TK_error.
2487 * We then shift in TK_error and see what state
2488 * that takes us too.
2489 * Then we discard input tokens until
2490 * we find one that is acceptable.
2493 err_tk = tok_copy(*tk);
2494 p.next.sym = TK_error;
2495 while (shift(&p, err_tk, states) == 0
2497 // discard this state
2498 pop(&p, 1, do_free);
2501 // no state accepted TK_error
2504 while (search(&states[p.next.state], tk->num) < 0 &&
2505 tk->num != TK_eof) {
2507 tk = tok_copy(token_next(tokens));
2508 if (tk->num == TK_in)
2509 p.next.indents += 1;
2510 if (tk->num == TK_out) {
2511 if (p.next.indents == 0)
2513 p.next.indents -= 1;
2516 if (p.tos == 0 && tk->num == TK_eof)
2521 ret = p.asn_stack[0];
2523 pop(&p, p.tos, do_free);
2529 ###### exported functions
2530 void *parser_run(struct token_state *tokens,
2531 const struct state states[],
2532 int (*do_reduce)(int, void**, void*),
2533 void (*do_free)(short, void*),
2534 FILE *trace, const char *non_term[], int knowns);
2538 Being able to visualize the parser in action can be invaluable when
2539 debugging the parser code, or trying to understand how the parse of a
2540 particular grammar progresses. The stack contains all the important
2541 state, so just printing out the stack every time around the parse loop
2542 can make it possible to see exactly what is happening.
2544 This doesn't explicitly show each SHIFT and REDUCE action. However they
2545 are easily deduced from the change between consecutive lines, and the
2546 details of each state can be found by cross referencing the states list
2547 in the stack with the "report" that parsergen can generate.
2549 For terminal symbols, we just dump the token. For non-terminals we
2550 print the name of the symbol. The look ahead token is reported at the
2551 end inside square brackets.
2553 ###### parser functions
2555 static char *reserved_words[] = {
2556 [TK_error] = "ERROR",
2559 [TK_newline] = "NEWLINE",
2562 void parser_trace(FILE *trace, struct parser *p,
2563 struct token *tk, const struct state states[],
2564 const char *non_term[], int knowns)
2567 for (i = 0; i < p->tos; i++) {
2568 int sym = p->stack[i].sym;
2569 fprintf(trace, "(%d) ", p->stack[i].state);
2570 if (sym < TK_reserved &&
2571 reserved_words[sym] != NULL)
2572 fputs(reserved_words[sym], trace);
2573 else if (sym < TK_reserved + knowns) {
2574 struct token *t = p->asn_stack[i];
2575 text_dump(trace, t->txt, 20);
2577 fputs(non_term[sym - TK_reserved - knowns],
2581 fprintf(trace, "(%d) [", p->next.state);
2582 if (tk->num < TK_reserved &&
2583 reserved_words[tk->num] != NULL)
2584 fputs(reserved_words[tk->num], trace);
2586 text_dump(trace, tk->txt, 20);
2587 fputs("]\n", trace);
2592 The obvious example for a parser is a calculator.
2594 As `scanner` provides number parsing function using `libgmp` is it not much
2595 work to perform arbitrary rational number calculations.
2597 This calculator takes one expression, or an equality test per line. The
2598 results are printed and in any equality test fails, the program exits with
2601 Embedding mdcode inside mdcode is rather horrible. I'd like to find a
2602 better approach, but as the grammar file must have 3 components I need
2603 something like this.
2605 ###### File: parsergen.mk
2606 calc.c : parsergen calc.cgm
2607 ./parsergen -o calc calc.cgm
2608 calc : calc.o libparser.o libscanner.o libmdcode.o libnumber.o
2609 $(CC) $(CFLAGS) -o calc calc.o libparser.o libscanner.o libmdcode.o libnumber.o -licuuc -lgmp
2617 // what do we use for a demo-grammar? A calculator of course.
2631 #include <sys/mman.h>
2636 #include "scanner.h"
2642 static void free_number(struct number *n)
2648 int main(int argc, char *argv[])
2650 int fd = open(argv[1], O_RDONLY);
2651 int len = lseek(fd, 0, 2);
2652 char *file = mmap(NULL, len, PROT_READ, MAP_SHARED, fd, 0);
2653 struct section *s = code_extract(file, file+len, NULL);
2654 struct token_config config = {
2655 .ignored = (1 << TK_line_comment)
2656 | (1 << TK_block_comment)
2659 .number_chars = ".,_+-",
2663 parse_calc(s->code, &config, argc > 2 ? stderr : NULL);
2671 Session -> Session Line
2674 Line -> Expression NEWLINE ${ gmp_printf("Answer = %Qd\n", $1.val);
2675 { mpf_t fl; mpf_init2(fl, 20); mpf_set_q(fl, $1.val);
2676 gmp_printf(" or as a decimal: %Fg\n", fl);
2680 | Expression = Expression NEWLINE ${
2681 if (mpq_equal($1.val, $3.val))
2682 gmp_printf("Both equal %Qd\n", $1.val);
2684 gmp_printf("NOT EQUAL: %Qd\n != : %Qd\n",
2689 | NEWLINE ${ printf("Blank line\n"); }$
2690 | ERROR NEWLINE ${ printf("Skipped a bad line\n"); }$
2693 Expression -> Expression + Term ${ mpq_init($0.val); mpq_add($0.val, $1.val, $3.val); }$
2694 | Expression - Term ${ mpq_init($0.val); mpq_sub($0.val, $1.val, $3.val); }$
2695 | Term ${ mpq_init($0.val); mpq_set($0.val, $1.val); }$
2697 Term -> Term * Factor ${ mpq_init($0.val); mpq_mul($0.val, $1.val, $3.val); }$
2698 | Term / Factor ${ mpq_init($0.val); mpq_div($0.val, $1.val, $3.val); }$
2699 | Factor ${ mpq_init($0.val); mpq_set($0.val, $1.val); }$
2701 Factor -> NUMBER ${ if (number_parse($0.val, $0.tail, $1.txt) == 0) mpq_init($0.val); }$
2702 | ( Expression ) ${ mpq_init($0.val); mpq_set($0.val, $2.val); }$
code / ocean / blob