=================================================================== RCS file: /cvs/mandoc/roff.c,v retrieving revision 1.81 retrieving revision 1.89 diff -u -p -r1.81 -r1.89 --- mandoc/roff.c 2010/05/17 00:46:35 1.81 +++ mandoc/roff.c 2010/06/26 15:22:19 1.89 @@ -1,4 +1,4 @@ -/* $Id: roff.c,v 1.81 2010/05/17 00:46:35 kristaps Exp $ */ +/* $Id: roff.c,v 1.89 2010/06/26 15:22:19 kristaps Exp $ */ /* * Copyright (c) 2010 Kristaps Dzonsons * @@ -19,16 +19,30 @@ #endif #include +#include +#include +#include #include #include #include #include "mandoc.h" +#include "regs.h" #include "roff.h" +#define RSTACK_MAX 128 + #define ROFF_CTL(c) \ ('.' == (c) || '\'' == (c)) +#if 1 +#define ROFF_DEBUG(fmt, args...) \ + do { /* Nothing. */ } while (/*CONSTCOND*/ 0) +#else +#define ROFF_DEBUG(fmt, args...) \ + do { fprintf(stderr, fmt , ##args); } while (/*CONSTCOND*/ 0) +#endif + enum rofft { ROFF_am, ROFF_ami, @@ -36,28 +50,32 @@ enum rofft { ROFF_de, ROFF_dei, ROFF_de1, + ROFF_ds, + ROFF_el, + ROFF_ie, ROFF_if, ROFF_ig, + ROFF_rm, + ROFF_tr, ROFF_cblock, ROFF_ccond, -#if 0 - ROFF_ie, - ROFF_el, -#endif + ROFF_nr, ROFF_MAX }; +enum roffrule { + ROFFRULE_ALLOW, + ROFFRULE_DENY +}; + struct roff { struct roffnode *last; /* leaf of stack */ mandocmsg msg; /* err/warn/fatal messages */ void *data; /* privdata for messages */ + enum roffrule rstack[RSTACK_MAX]; /* stack of !`ie' rules */ + int rstackpos; /* position in rstack */ }; -enum roffrule { - ROFFRULE_ALLOW, - ROFFRULE_DENY -}; - struct roffnode { enum rofft tok; /* type of node */ struct roffnode *parent; /* up one in stack */ @@ -65,10 +83,11 @@ struct roffnode { int col; /* parse col */ char *end; /* end-rules: custom token */ int endspan; /* end-rules: next-line or infty */ - enum roffrule rule; + enum roffrule rule; /* current evaluation rule */ }; #define ROFF_ARGS struct roff *r, /* parse ctx */ \ + struct regset *regs, /* registers */ \ enum rofft tok, /* tok of macro */ \ char **bufp, /* input buffer */ \ size_t *szp, /* size of input buffer */ \ @@ -86,6 +105,7 @@ struct roffmac { roffproc sub; /* process as child of macro */ int flags; #define ROFFMAC_STRUCT (1 << 0) /* always interpret */ + struct roffmac *next; }; static enum rofferr roff_block(ROFF_ARGS); @@ -93,32 +113,75 @@ static enum rofferr roff_block_text(ROFF_ARGS); static enum rofferr roff_block_sub(ROFF_ARGS); static enum rofferr roff_cblock(ROFF_ARGS); static enum rofferr roff_ccond(ROFF_ARGS); -static enum rofferr roff_if(ROFF_ARGS); -static enum rofferr roff_if_text(ROFF_ARGS); -static enum rofferr roff_if_sub(ROFF_ARGS); +static enum rofferr roff_cond(ROFF_ARGS); +static enum rofferr roff_cond_text(ROFF_ARGS); +static enum rofferr roff_cond_sub(ROFF_ARGS); +static enum rofferr roff_line(ROFF_ARGS); +static enum rofferr roff_nr(ROFF_ARGS); +static enum roffrule roff_evalcond(const char *, int *); -const struct roffmac roffs[ROFF_MAX] = { - { "am", roff_block, roff_block_text, roff_block_sub, 0 }, - { "ami", roff_block, roff_block_text, roff_block_sub, 0 }, - { "am1", roff_block, roff_block_text, roff_block_sub, 0 }, - { "de", roff_block, roff_block_text, roff_block_sub, 0 }, - { "dei", roff_block, roff_block_text, roff_block_sub, 0 }, - { "de1", roff_block, roff_block_text, roff_block_sub, 0 }, - { "if", roff_if, roff_if_text, roff_if_sub, ROFFMAC_STRUCT }, - { "ig", roff_block, roff_block_text, roff_block_sub, 0 }, - { ".", roff_cblock, NULL, NULL, 0 }, - { "\\}", roff_ccond, NULL, NULL, 0 }, +/* See roff_hash_find() */ + +#define ASCII_HI 126 +#define ASCII_LO 33 +#define HASHWIDTH (ASCII_HI - ASCII_LO + 1) + +static struct roffmac *hash[HASHWIDTH]; + +static struct roffmac roffs[ROFF_MAX] = { + { "am", roff_block, roff_block_text, roff_block_sub, 0, NULL }, + { "ami", roff_block, roff_block_text, roff_block_sub, 0, NULL }, + { "am1", roff_block, roff_block_text, roff_block_sub, 0, NULL }, + { "de", roff_block, roff_block_text, roff_block_sub, 0, NULL }, + { "dei", roff_block, roff_block_text, roff_block_sub, 0, NULL }, + { "de1", roff_block, roff_block_text, roff_block_sub, 0, NULL }, + { "ds", roff_line, NULL, NULL, 0, NULL }, + { "el", roff_cond, roff_cond_text, roff_cond_sub, ROFFMAC_STRUCT, NULL }, + { "ie", roff_cond, roff_cond_text, roff_cond_sub, ROFFMAC_STRUCT, NULL }, + { "if", roff_cond, roff_cond_text, roff_cond_sub, ROFFMAC_STRUCT, NULL }, + { "ig", roff_block, roff_block_text, roff_block_sub, 0, NULL }, + { "rm", roff_line, NULL, NULL, 0, NULL }, + { "tr", roff_line, NULL, NULL, 0, NULL }, + { ".", roff_cblock, NULL, NULL, 0, NULL }, + { "\\}", roff_ccond, NULL, NULL, 0, NULL }, + { "nr", roff_nr, NULL, NULL, 0, NULL }, }; static void roff_free1(struct roff *); static enum rofft roff_hash_find(const char *); +static void roff_hash_init(void); static void roffnode_cleanscope(struct roff *); static int roffnode_push(struct roff *, enum rofft, int, int); static void roffnode_pop(struct roff *); static enum rofft roff_parse(const char *, int *); +static int roff_parse_nat(const char *, int *); +/* See roff_hash_find() */ +#define ROFF_HASH(p) (p[0] - ASCII_LO) +static void +roff_hash_init(void) +{ + struct roffmac *n; + int buc, i; + + for (i = 0; i < (int)ROFF_MAX; i++) { + assert(roffs[i].name[0] >= ASCII_LO); + assert(roffs[i].name[0] <= ASCII_HI); + + buc = ROFF_HASH(roffs[i].name); + + if (NULL != (n = hash[buc])) { + for ( ; n->next; n = n->next) + /* Do nothing. */ ; + n->next = &roffs[i]; + } else + hash[buc] = &roffs[i]; + } +} + + /* * Look up a roff token by its name. Returns ROFF_MAX if no macro by * the nil-terminated string name could be found. @@ -126,14 +189,27 @@ static enum rofft roff_parse(const char *, int *); static enum rofft roff_hash_find(const char *p) { - int i; + int buc; + struct roffmac *n; - /* FIXME: make this be fast and efficient. */ + /* + * libroff has an extremely simple hashtable, for the time + * being, which simply keys on the first character, which must + * be printable, then walks a chain. It works well enough until + * optimised. + */ - for (i = 0; i < (int)ROFF_MAX; i++) - if (0 == strcmp(roffs[i].name, p)) - return((enum rofft)i); + if (p[0] < ASCII_LO || p[0] > ASCII_HI) + return(ROFF_MAX); + buc = ROFF_HASH(p); + + if (NULL == (n = hash[buc])) + return(ROFF_MAX); + for ( ; n; n = n->next) + if (0 == strcmp(n->name, p)) + return((enum rofft)(n - roffs)); + return(ROFF_MAX); } @@ -149,6 +225,11 @@ roffnode_pop(struct roff *r) assert(r->last); p = r->last; + + if (ROFF_el == p->tok) + if (r->rstackpos > -1) + r->rstackpos--; + r->last = r->last->parent; if (p->end) free(p->end); @@ -219,12 +300,15 @@ roff_alloc(const mandocmsg msg, void *data) r->msg = msg; r->data = data; + r->rstackpos = -1; + + roff_hash_init(); return(r); } enum rofferr -roff_parseln(struct roff *r, int ln, +roff_parseln(struct roff *r, struct regset *regs, int ln, char **bufp, size_t *szp, int pos, int *offs) { enum rofft t; @@ -239,10 +323,16 @@ roff_parseln(struct roff *r, int ln, if (r->last && ! ROFF_CTL((*bufp)[pos])) { t = r->last->tok; assert(roffs[t].text); + ROFF_DEBUG("roff: intercept scoped text: %s, [%s]\n", + roffs[t].name, &(*bufp)[pos]); return((*roffs[t].text) - (r, t, bufp, szp, ln, pos, pos, offs)); - } else if ( ! ROFF_CTL((*bufp)[pos])) + (r, regs, t, bufp, + szp, ln, pos, pos, offs)); + } else if ( ! ROFF_CTL((*bufp)[pos])) { + ROFF_DEBUG("roff: pass non-scoped text: [%s]\n", + &(*bufp)[pos]); return(ROFF_CONT); + } /* * If a scope is open, go to the child handler for that macro, @@ -252,8 +342,11 @@ roff_parseln(struct roff *r, int ln, if (r->last) { t = r->last->tok; assert(roffs[t].sub); + ROFF_DEBUG("roff: intercept scoped context: %s\n", + roffs[t].name); return((*roffs[t].sub) - (r, t, bufp, szp, ln, pos, pos, offs)); + (r, regs, t, bufp, + szp, ln, pos, pos, offs)); } /* @@ -263,12 +356,18 @@ roff_parseln(struct roff *r, int ln, */ ppos = pos; - if (ROFF_MAX == (t = roff_parse(*bufp, &pos))) + if (ROFF_MAX == (t = roff_parse(*bufp, &pos))) { + ROFF_DEBUG("roff: pass non-scoped non-macro: [%s]\n", + &(*bufp)[pos]); return(ROFF_CONT); + } + ROFF_DEBUG("roff: intercept new-scope: %s, [%s]\n", + roffs[t].name, &(*bufp)[pos]); assert(roffs[t].proc); return((*roffs[t].proc) - (r, t, bufp, szp, ln, ppos, pos, offs)); + (r, regs, t, bufp, + szp, ln, ppos, pos, offs)); } @@ -306,7 +405,7 @@ roff_parse(const char *buf, int *pos) for (j = 0; j < 4; j++, (*pos)++) if ('\0' == (mac[j] = buf[*pos])) break; - else if (' ' == buf[*pos]) + else if (' ' == buf[*pos] || (j && '\\' == buf[*pos])) break; if (j == 4 || j < 1) @@ -324,6 +423,26 @@ roff_parse(const char *buf, int *pos) } +static int +roff_parse_nat(const char *buf, int *res) +{ + char *ep; + long lval; + + errno = 0; + lval = strtol(buf, &ep, 10); + if (buf[0] == '\0' || *ep != '\0') + return(0); + if ((errno == ERANGE && + (lval == LONG_MAX || lval == LONG_MIN)) || + (lval > INT_MAX || lval <= 0)) + return(0); + + *res = (int)lval; + return(1); +} + + /* ARGSUSED */ static enum rofferr roff_cblock(ROFF_ARGS) @@ -395,7 +514,14 @@ roff_ccond(ROFF_ARGS) return(ROFF_IGN); } - if (ROFF_if != r->last->tok) { + switch (r->last->tok) { + case (ROFF_el): + /* FALLTHROUGH */ + case (ROFF_ie): + /* FALLTHROUGH */ + case (ROFF_if): + break; + default: if ( ! (*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL)) return(ROFF_ERR); return(ROFF_IGN); @@ -478,36 +604,6 @@ roff_block(ROFF_ARGS) /* ARGSUSED */ static enum rofferr -roff_if_sub(ROFF_ARGS) -{ - enum rofft t; - enum roffrule rr; - - ppos = pos; - rr = r->last->rule; - roffnode_cleanscope(r); - - if (ROFF_MAX == (t = roff_parse(*bufp, &pos))) - return(ROFFRULE_DENY == rr ? ROFF_IGN : ROFF_CONT); - - /* - * A denied conditional must evaluate its children if and only - * if they're either structurally required (such as loops and - * conditionals) or a closing macro. - */ - if (ROFFRULE_DENY == rr) - if ( ! (ROFFMAC_STRUCT & roffs[t].flags)) - if (ROFF_ccond != t) - return(ROFF_IGN); - - assert(roffs[t].proc); - return((*roffs[t].proc) - (r, t, bufp, szp, ln, ppos, pos, offs)); -} - - -/* ARGSUSED */ -static enum rofferr roff_block_sub(ROFF_ARGS) { enum rofft t; @@ -557,7 +653,7 @@ roff_block_sub(ROFF_ARGS) return(ROFF_IGN); assert(roffs[t].proc); - return((*roffs[t].proc)(r, t, bufp, + return((*roffs[t].proc)(r, regs, t, bufp, szp, ln, ppos, pos, offs)); } @@ -573,41 +669,137 @@ roff_block_text(ROFF_ARGS) /* ARGSUSED */ static enum rofferr -roff_if_text(ROFF_ARGS) +roff_cond_sub(ROFF_ARGS) { + enum rofft t; + enum roffrule rr; + struct roffnode *l; + + ppos = pos; + rr = r->last->rule; + + /* + * Clean out scope. If we've closed ourselves, then don't + * continue. + */ + + l = r->last; + roffnode_cleanscope(r); + + if (l != r->last) + return(ROFFRULE_DENY == rr ? ROFF_IGN : ROFF_CONT); + + if (ROFF_MAX == (t = roff_parse(*bufp, &pos))) + return(ROFFRULE_DENY == rr ? ROFF_IGN : ROFF_CONT); + + /* + * A denied conditional must evaluate its children if and only + * if they're either structurally required (such as loops and + * conditionals) or a closing macro. + */ + if (ROFFRULE_DENY == rr) + if ( ! (ROFFMAC_STRUCT & roffs[t].flags)) + if (ROFF_ccond != t) + return(ROFF_IGN); + + assert(roffs[t].proc); + return((*roffs[t].proc) + (r, regs, t, bufp, szp, ln, ppos, pos, offs)); +} + + +/* ARGSUSED */ +static enum rofferr +roff_cond_text(ROFF_ARGS) +{ char *ep, *st; + enum roffrule rr; + rr = r->last->rule; + + /* + * We display the value of the text if out current evaluation + * scope permits us to do so. + */ + st = &(*bufp)[pos]; if (NULL == (ep = strstr(st, "\\}"))) { roffnode_cleanscope(r); - return(ROFF_IGN); + return(ROFFRULE_DENY == rr ? ROFF_IGN : ROFF_CONT); } - if (ep > st && '\\' != *(ep - 1)) + if (ep == st || (ep > st && '\\' != *(ep - 1))) roffnode_pop(r); roffnode_cleanscope(r); + return(ROFFRULE_DENY == rr ? ROFF_IGN : ROFF_CONT); +} + + +static enum roffrule +roff_evalcond(const char *v, int *pos) +{ + + switch (v[*pos]) { + case ('n'): + (*pos)++; + return(ROFFRULE_ALLOW); + case ('e'): + /* FALLTHROUGH */ + case ('o'): + /* FALLTHROUGH */ + case ('t'): + (*pos)++; + return(ROFFRULE_DENY); + default: + break; + } + + while (v[*pos] && ' ' != v[*pos]) + (*pos)++; + return(ROFFRULE_DENY); +} + + +/* ARGSUSED */ +static enum rofferr +roff_line(ROFF_ARGS) +{ + return(ROFF_IGN); } /* ARGSUSED */ static enum rofferr -roff_if(ROFF_ARGS) +roff_cond(ROFF_ARGS) { int sv; + enum roffrule rule; - /* - * Read ahead past the conditional. - * FIXME: this does not work, as conditionals don't end on - * whitespace, but are parsed according to a formal grammar. - * It's good enough for now, however. - */ + /* Stack overflow! */ - while ((*bufp)[pos] && ' ' != (*bufp)[pos]) - pos++; + if (ROFF_ie == tok && r->rstackpos == RSTACK_MAX - 1) { + (*r->msg)(MANDOCERR_MEM, r->data, ln, ppos, NULL); + return(ROFF_ERR); + } + /* First, evaluate the conditional. */ + + if (ROFF_el == tok) { + /* + * An `.el' will get the value of the current rstack + * entry set in prior `ie' calls or defaults to DENY. + */ + if (r->rstackpos < 0) + rule = ROFFRULE_DENY; + else + rule = r->rstack[r->rstackpos]; + } else + rule = roff_evalcond(*bufp, &pos); + sv = pos; + while (' ' == (*bufp)[pos]) pos++; @@ -619,22 +811,55 @@ roff_if(ROFF_ARGS) */ if ('\0' == (*bufp)[pos] && sv != pos) { - if ( ! (*r->msg)(MANDOCERR_NOARGS, r->data, ln, ppos, NULL)) - return(ROFF_ERR); - return(ROFF_IGN); + if ((*r->msg)(MANDOCERR_NOARGS, r->data, ln, ppos, NULL)) + return(ROFF_IGN); + return(ROFF_ERR); } if ( ! roffnode_push(r, tok, ln, ppos)) return(ROFF_ERR); - /* Don't evaluate: just assume NO. */ + r->last->rule = rule; + ROFF_DEBUG("roff: cond: %s -> %s\n", roffs[tok].name, + ROFFRULE_ALLOW == rule ? "allow" : "deny"); + + if (ROFF_ie == tok) { + /* + * An if-else will put the NEGATION of the current + * evaluated conditional into the stack. + */ + r->rstackpos++; + if (ROFFRULE_DENY == r->last->rule) + r->rstack[r->rstackpos] = ROFFRULE_ALLOW; + else + r->rstack[r->rstackpos] = ROFFRULE_DENY; + } + + /* If the parent has false as its rule, then so do we. */ + + if (r->last->parent && ROFFRULE_DENY == r->last->parent->rule) { + r->last->rule = ROFFRULE_DENY; + ROFF_DEBUG("roff: cond override: %s -> deny\n", + roffs[tok].name); + } + + /* + * Determine scope. If we're invoked with "\{" trailing the + * conditional, then we're in a multiline scope. Else our scope + * expires on the next line. + */ + r->last->endspan = 1; if ('\\' == (*bufp)[pos] && '{' == (*bufp)[pos + 1]) { r->last->endspan = -1; pos += 2; - } + ROFF_DEBUG("roff: cond-scope: %s, multi-line\n", + roffs[tok].name); + } else + ROFF_DEBUG("roff: cond-scope: %s, one-line\n", + roffs[tok].name); /* * If there are no arguments on the line, the next-line scope is @@ -648,4 +873,44 @@ roff_if(ROFF_ARGS) *offs = pos; return(ROFF_RERUN); +} + + +/* ARGSUSED */ +static enum rofferr +roff_nr(ROFF_ARGS) +{ + const char *key, *val; + + key = &(*bufp)[pos]; + + /* Parse register request. */ + while ((*bufp)[pos] && ' ' != (*bufp)[pos]) + pos++; + + /* + * Set our nil terminator. Because this line is going to be + * ignored anyway, we can munge it as we please. + */ + if ((*bufp)[pos]) + (*bufp)[pos++] = '\0'; + + /* Skip whitespace to register token. */ + while ((*bufp)[pos] && ' ' == (*bufp)[pos]) + pos++; + + val = &(*bufp)[pos]; + + /* Process register token. */ + + if (0 == strcmp(key, "nS")) { + if ( ! roff_parse_nat(val, ®s->regs[(int)REG_nS].i)) + regs->regs[(int)REG_nS].i = 0; + + ROFF_DEBUG("roff: register nS: %d\n", + regs->regs[(int)REG_nS].i); + } else + ROFF_DEBUG("roff: ignoring register: %s\n", key); + + return(ROFF_IGN); }