=================================================================== RCS file: /cvs/mandoc/mdoc_macro.c,v retrieving revision 1.34 retrieving revision 1.43 diff -u -p -r1.34 -r1.43 --- mandoc/mdoc_macro.c 2009/09/16 14:40:56 1.34 +++ mandoc/mdoc_macro.c 2010/03/29 19:28:04 1.43 @@ -1,4 +1,4 @@ -/* $Id: mdoc_macro.c,v 1.34 2009/09/16 14:40:56 kristaps Exp $ */ +/* $Id: mdoc_macro.c,v 1.43 2010/03/29 19:28:04 kristaps Exp $ */ /* * Copyright (c) 2008, 2009 Kristaps Dzonsons * @@ -14,11 +14,16 @@ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */ +#ifdef HAVE_CONFIG_H +#include "config.h" +#endif + #include #include #include #include #include +#include #include "libmdoc.h" @@ -26,6 +31,7 @@ #define REWIND_NOHALT (1 << 1) #define REWIND_HALT (1 << 2) +static int ctx_synopsis(MACRO_PROT_ARGS); static int obsolete(MACRO_PROT_ARGS); static int blk_part_exp(MACRO_PROT_ARGS); static int in_line_eoln(MACRO_PROT_ARGS); @@ -46,8 +52,8 @@ static int rew_sub(enum mdoc_type, struct mdoc *, static int rew_last(struct mdoc *, const struct mdoc_node *); static int append_delims(struct mdoc *, int, int *, char *); -static int lookup(struct mdoc *, int, const char *); -static int lookup_raw(struct mdoc *, const char *); +static int lookup(int, const char *); +static int lookup_raw(const char *); static int swarn(struct mdoc *, enum mdoc_type, int, int, const struct mdoc_node *); @@ -93,8 +99,8 @@ const struct mdoc_macro __mdoc_macros[MDOC_MAX] = { { in_line_eoln, 0 }, /* Rv */ { in_line_argn, MDOC_CALLABLE | MDOC_PARSED }, /* St */ { in_line, MDOC_CALLABLE | MDOC_PARSED }, /* Va */ - { in_line, MDOC_CALLABLE | MDOC_PARSED }, /* Vt */ - { in_line, MDOC_CALLABLE | MDOC_PARSED }, /* Xr */ + { ctx_synopsis, MDOC_CALLABLE | MDOC_PARSED }, /* Vt */ + { in_line_argn, MDOC_CALLABLE | MDOC_PARSED }, /* Xr */ { in_line_eoln, 0 }, /* %A */ { in_line_eoln, 0 }, /* %B */ { in_line_eoln, 0 }, /* %D */ @@ -174,6 +180,7 @@ const struct mdoc_macro __mdoc_macros[MDOC_MAX] = { { in_line_eoln, 0 }, /* %Q */ { in_line_eoln, 0 }, /* br */ { in_line_eoln, 0 }, /* sp */ + { in_line_eoln, 0 }, /* %U */ }; const struct mdoc_macro * const mdoc_macros = __mdoc_macros; @@ -258,12 +265,13 @@ mdoc_macroend(struct mdoc *m) * Look up a macro from within a subsequent context. */ static int -lookup(struct mdoc *mdoc, int from, const char *p) +lookup(int from, const char *p) { + /* FIXME: make -diag lists be un-PARSED. */ if ( ! (MDOC_PARSED & mdoc_macros[from].flags)) return(MDOC_MAX); - return(lookup_raw(mdoc, p)); + return(lookup_raw(p)); } @@ -271,7 +279,7 @@ lookup(struct mdoc *mdoc, int from, const char *p) * Lookup a macro following the initial line macro. */ static int -lookup_raw(struct mdoc *mdoc, const char *p) +lookup_raw(const char *p) { int res; @@ -391,6 +399,8 @@ rew_dohalt(int tok, enum mdoc_type type, const struct case (MDOC_Qq): /* FALLTHROUGH */ case (MDOC_Sq): + /* FALLTHROUGH */ + case (MDOC_Vt): assert(MDOC_TAIL != type); if (type == p->type && tok == p->tok) return(REWIND_REWIND); @@ -669,7 +679,7 @@ blk_exp_close(MACRO_PROT_ARGS) if (ARGS_EOLN == c) break; - if (MDOC_MAX != (c = lookup(m, tok, p))) { + if (MDOC_MAX != (c = lookup(tok, p))) { if ( ! flushed) { if ( ! rew_sub(MDOC_BLOCK, m, tok, line, ppos)) return(0); @@ -753,7 +763,7 @@ in_line(MACRO_PROT_ARGS) /* Quoted words shouldn't be looked-up. */ - c = ARGS_QWORD == w ? MDOC_MAX : lookup(m, tok, p); + c = ARGS_QWORD == w ? MDOC_MAX : lookup(tok, p); /* * In this case, we've located a submacro and must @@ -805,6 +815,17 @@ in_line(MACRO_PROT_ARGS) cnt++; if ( ! mdoc_word_alloc(m, line, la, p)) return(0); + + /* + * `Fl' macros have their scope re-opened with each new + * word so that the `-' can be added to each one without + * having to parse out spaces. + */ + if (0 == lastpunct && MDOC_Fl == tok) { + if ( ! rew_elem(m, tok)) + return(0); + lastpunct = 1; + } } if (0 == lastpunct && ! rew_elem(m, tok)) @@ -932,7 +953,7 @@ blk_full(MACRO_PROT_ARGS) continue; } - if (MDOC_MAX == (c = lookup(m, tok, p))) { + if (MDOC_MAX == (c = lookup(tok, p))) { if ( ! mdoc_word_alloc(m, line, lastarg, p)) return(0); continue; @@ -964,59 +985,91 @@ blk_part_imp(MACRO_PROT_ARGS) { int la, c; char *p; - struct mdoc_node *blk, *body, *n; + struct mdoc_node *blk; /* saved block context */ + struct mdoc_node *body; /* saved body context */ + struct mdoc_node *n; - /* If applicable, close out prior scopes. */ + /* + * A macro that spans to the end of the line. This is generally + * (but not necessarily) called as the first macro. The block + * has a head as the immediate child, which is always empty, + * followed by zero or more opening punctuation nodes, then the + * body (which may be empty, depending on the macro), then zero + * or more closing punctuation nodes. + */ if ( ! mdoc_block_alloc(m, line, ppos, tok, NULL)) return(0); - /* Saved for later close-out. */ + blk = m->last; + if ( ! mdoc_head_alloc(m, line, ppos, tok)) return(0); if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) return(0); - if ( ! mdoc_body_alloc(m, line, ppos, tok)) - return(0); - /* Saved for later close-out. */ - body = m->last; - /* Body argument processing. */ + /* + * Open the body scope "on-demand", that is, after we've + * processed all our the leading delimiters (open parenthesis, + * etc.). + */ - for (;;) { + for (body = NULL; ; ) { la = *pos; c = mdoc_args(m, line, pos, buf, tok, &p); + assert(ARGS_PHRASE != c); if (ARGS_ERROR == c) return(0); - if (ARGS_PUNCT == c) - break; if (ARGS_EOLN == c) break; + if (ARGS_PUNCT == c) + break; - if (MDOC_MAX == (c = lookup(m, tok, p))) { + if (1 == mdoc_isdelim(p) && NULL == body) { if ( ! mdoc_word_alloc(m, line, la, p)) return(0); continue; } - if ( ! mdoc_macro(m, c, line, la, pos, buf)) + if (NULL == body) { + if ( ! mdoc_body_alloc(m, line, ppos, tok)) + return(0); + body = m->last; + } + + if (MDOC_MAX != (c = lookup(tok, p))) { + if ( ! mdoc_macro(m, c, line, la, pos, buf)) + return(0); + break; + } + + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); - break; } + /* Clean-ups to leave in a consistent state. */ + + if (NULL == body && ! mdoc_body_alloc(m, line, ppos, tok)) + return(0); + + body = m->last; + /* * If we can't rewind to our body, then our scope has already * been closed by another macro (like `Oc' closing `Op'). This * is ugly behaviour nodding its head to OpenBSD's overwhelming - * crufty use of `Op' breakage--XXX, deprecate in time. + * crufty use of `Op' breakage. XXX: DEPRECATE. */ + for (n = m->last; n; n = n->parent) if (body == n) break; + if (NULL == n && ! mdoc_nwarn(m, body, EIMPBRK)) return(0); + if (n && ! rew_last(m, body)) return(0); @@ -1037,59 +1090,23 @@ blk_part_imp(MACRO_PROT_ARGS) static int blk_part_exp(MACRO_PROT_ARGS) { - int la, flushed, j, c, maxargs; + int la, c; + struct mdoc_node *head; /* keep track of head */ + struct mdoc_node *body; /* keep track of body */ char *p; - /* Number of head arguments. Only `Eo' has these, */ + /* + * The opening of an explicit macro having zero or more leading + * punctuation nodes; a head with optional single element (the + * case of `Eo'); and a body that may be empty. + */ - switch (tok) { - case (MDOC_Eo): - maxargs = 1; - break; - default: - maxargs = 0; - break; - } - - /* Begin the block scope. */ - if ( ! mdoc_block_alloc(m, line, ppos, tok, NULL)) return(0); - /* - * If no head arguments, open and then close out a head, noting - * that we've flushed our terms. `flushed' means that we've - * flushed out the head and the body is open. - */ - - if (0 == maxargs) { - if ( ! mdoc_head_alloc(m, line, ppos, tok)) - return(0); - if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) - return(0); - if ( ! mdoc_body_alloc(m, line, ppos, tok)) - return(0); - flushed = 1; - } else { - if ( ! mdoc_head_alloc(m, line, ppos, tok)) - return(0); - flushed = 0; - } - - /* Process the head/head+body line arguments. */ - - for (j = 0; ; j++) { + for (head = body = NULL; ; ) { la = *pos; - if (j == maxargs && ! flushed) { - if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) - return(0); - flushed = 1; - if ( ! mdoc_body_alloc(m, line, ppos, tok)) - return(0); - } - c = mdoc_args(m, line, pos, buf, tok, &p); - assert(ARGS_PHRASE != c); if (ARGS_ERROR == c) return(0); @@ -1098,44 +1115,79 @@ blk_part_exp(MACRO_PROT_ARGS) if (ARGS_EOLN == c) break; - if (MDOC_MAX != (c = lookup(m, tok, p))) { - if ( ! flushed) { - if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) - return(0); - flushed = 1; - if ( ! mdoc_body_alloc(m, line, ppos, tok)) - return(0); - } - if ( ! mdoc_macro(m, c, line, la, pos, buf)) + assert(ARGS_PHRASE != c); + + /* Flush out leading punctuation. */ + + if (1 == mdoc_isdelim(p) && NULL == head) { + assert(NULL == body); + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); - break; + continue; + } + + if (NULL == head) { + assert(NULL == body); + if ( ! mdoc_head_alloc(m, line, ppos, tok)) + return(0); + head = m->last; } - if ( ! flushed && mdoc_isdelim(p)) { + /* + * `Eo' gobbles any data into the head, but most other + * macros just immediately close out and begin the body. + */ + + if (NULL == body) { + assert(head); + /* No check whether it's a macro! */ + if (MDOC_Eo == tok) + if ( ! mdoc_word_alloc(m, line, la, p)) + return(0); + if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) return(0); - flushed = 1; if ( ! mdoc_body_alloc(m, line, ppos, tok)) return(0); + body = m->last; + + if (MDOC_Eo == tok) + continue; } - + + assert(NULL != head && NULL != body); + + if (MDOC_MAX != (c = lookup(tok, p))) { + if ( ! mdoc_macro(m, c, line, la, pos, buf)) + return(0); + break; + } + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); } - /* Close the head and open the body, if applicable. */ + /* Clean-up to leave in a consistent state. */ - if ( ! flushed) { + if (NULL == head) { + if ( ! mdoc_head_alloc(m, line, ppos, tok)) + return(0); + head = m->last; + } + + if (NULL == body) { if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) return(0); if ( ! mdoc_body_alloc(m, line, ppos, tok)) return(0); + body = m->last; } /* Standard appending of delimiters. */ if (ppos > 1) return(1); + return(append_delims(m, line, pos, buf)); } @@ -1159,6 +1211,9 @@ in_line_argn(MACRO_PROT_ARGS) case (MDOC_Ux): maxargs = 0; break; + case (MDOC_Xr): + maxargs = 2; + break; default: maxargs = 1; break; @@ -1209,7 +1264,7 @@ in_line_argn(MACRO_PROT_ARGS) if (ARGS_EOLN == c) break; - if (MDOC_MAX != (c = lookup(m, tok, p))) { + if (MDOC_MAX != (c = lookup(tok, p))) { if ( ! flushed && ! rew_elem(m, tok)) return(0); flushed = 1; @@ -1224,7 +1279,21 @@ in_line_argn(MACRO_PROT_ARGS) return(0); flushed = 1; } - + + /* + * XXX: this is a hack to work around groff's ugliness + * as regards `Xr' and extraneous arguments. It should + * ideally be deprecated behaviour, but because this is + * code is no here, it's unlikely to be removed. + */ + + if (MDOC_Xr == tok && j == maxargs) { + if ( ! mdoc_elem_alloc(m, line, ppos, MDOC_Ns, NULL)) + return(0); + if ( ! rew_elem(m, MDOC_Ns)) + return(0); + } + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); } @@ -1284,7 +1353,7 @@ in_line_eoln(MACRO_PROT_ARGS) if (ARGS_EOLN == w) break; - c = ARGS_QWORD == w ? MDOC_MAX : lookup(m, tok, p); + c = ARGS_QWORD == w ? MDOC_MAX : lookup(tok, p); if (MDOC_MAX != c) { if ( ! rew_elem(m, tok)) @@ -1304,6 +1373,29 @@ in_line_eoln(MACRO_PROT_ARGS) /* ARGSUSED */ static int +ctx_synopsis(MACRO_PROT_ARGS) +{ + + /* If we're not in the SYNOPSIS, go straight to in-line. */ + if (SEC_SYNOPSIS != m->lastsec) + return(in_line(m, tok, line, ppos, pos, buf)); + + /* If we're a nested call, same place. */ + if (ppos > 1) + return(in_line(m, tok, line, ppos, pos, buf)); + + /* + * XXX: this will open a block scope; however, if later we end + * up formatting the block scope, then child nodes will inherit + * the formatting. Be careful. + */ + + return(blk_part_imp(m, tok, line, ppos, pos, buf)); +} + + +/* ARGSUSED */ +static int obsolete(MACRO_PROT_ARGS) { @@ -1333,7 +1425,7 @@ phrase(struct mdoc *m, int line, int ppos, char *buf) if (ARGS_EOLN == w) break; - c = ARGS_QWORD == w ? MDOC_MAX : lookup_raw(m, p); + c = ARGS_QWORD == w ? MDOC_MAX : lookup_raw(p); if (MDOC_MAX != c) { if ( ! mdoc_macro(m, c, line, la, &pos, buf))