=================================================================== RCS file: /cvs/mandoc/mdoc_macro.c,v retrieving revision 1.43 retrieving revision 1.75 diff -u -p -r1.43 -r1.75 --- mandoc/mdoc_macro.c 2010/03/29 19:28:04 1.43 +++ mandoc/mdoc_macro.c 2010/05/31 10:28:04 1.75 @@ -1,4 +1,4 @@ -/* $Id: mdoc_macro.c,v 1.43 2010/03/29 19:28:04 kristaps Exp $ */ +/* $Id: mdoc_macro.c,v 1.75 2010/05/31 10:28:04 kristaps Exp $ */ /* * Copyright (c) 2008, 2009 Kristaps Dzonsons * @@ -25,40 +25,45 @@ #include #include +#include "mandoc.h" #include "libmdoc.h" +#include "libmandoc.h" -#define REWIND_REWIND (1 << 0) -#define REWIND_NOHALT (1 << 1) -#define REWIND_HALT (1 << 2) +enum rew { + REWIND_REWIND, + REWIND_NOHALT, + REWIND_HALT +}; -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); -static int in_line_argn(MACRO_PROT_ARGS); -static int in_line(MACRO_PROT_ARGS); -static int blk_full(MACRO_PROT_ARGS); -static int blk_exp_close(MACRO_PROT_ARGS); -static int blk_part_imp(MACRO_PROT_ARGS); +static int blk_full(MACRO_PROT_ARGS); +static int blk_exp_close(MACRO_PROT_ARGS); +static int blk_part_exp(MACRO_PROT_ARGS); +static int blk_part_imp(MACRO_PROT_ARGS); +static int ctx_synopsis(MACRO_PROT_ARGS); +static int in_line_eoln(MACRO_PROT_ARGS); +static int in_line_argn(MACRO_PROT_ARGS); +static int in_line(MACRO_PROT_ARGS); +static int obsolete(MACRO_PROT_ARGS); +static int phrase_ta(MACRO_PROT_ARGS); -static int phrase(struct mdoc *, int, int, char *); -static int rew_dohalt(int, enum mdoc_type, - const struct mdoc_node *); -static int rew_alt(int); -static int rew_dobreak(int, const struct mdoc_node *); -static int rew_elem(struct mdoc *, int); -static int rew_sub(enum mdoc_type, struct mdoc *, - int, int, int); -static int rew_last(struct mdoc *, - const struct mdoc_node *); -static int append_delims(struct mdoc *, int, int *, 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 *); +static int append_delims(struct mdoc *, + int, int *, char *); +static enum mdoct lookup(enum mdoct, const char *); +static enum mdoct lookup_raw(const char *); +static int phrase(struct mdoc *, int, int, char *); +static enum mdoct rew_alt(enum mdoct); +static int rew_dobreak(enum mdoct, + const struct mdoc_node *); +static enum rew rew_dohalt(enum mdoct, enum mdoc_type, + const struct mdoc_node *); +static int rew_elem(struct mdoc *, enum mdoct); +static int rew_last(struct mdoc *, + const struct mdoc_node *); +static int rew_sub(enum mdoc_type, struct mdoc *, + enum mdoct, int, int); +static int swarn(struct mdoc *, enum mdoc_type, int, + int, const struct mdoc_node *); -/* Central table of library: who gets parsed how. */ - const struct mdoc_macro __mdoc_macros[MDOC_MAX] = { { in_line_argn, MDOC_CALLABLE | MDOC_PARSED }, /* Ap */ { in_line_eoln, MDOC_PROLOGUE }, /* Dd */ @@ -137,7 +142,7 @@ const struct mdoc_macro __mdoc_macros[MDOC_MAX] = { { in_line_argn, MDOC_CALLABLE | MDOC_PARSED }, /* Nx */ { in_line_argn, MDOC_CALLABLE | MDOC_PARSED }, /* Ox */ { blk_exp_close, MDOC_EXPLICIT | MDOC_CALLABLE | MDOC_PARSED }, /* Pc */ - { in_line_argn, MDOC_PARSED | MDOC_IGNDELIM }, /* Pf */ + { in_line_argn, MDOC_CALLABLE | MDOC_PARSED | MDOC_IGNDELIM }, /* Pf */ { blk_part_exp, MDOC_CALLABLE | MDOC_PARSED | MDOC_EXPLICIT }, /* Po */ { blk_part_imp, MDOC_CALLABLE | MDOC_PARSED }, /* Pq */ { blk_exp_close, MDOC_EXPLICIT | MDOC_CALLABLE | MDOC_PARSED }, /* Qc */ @@ -166,7 +171,7 @@ const struct mdoc_macro __mdoc_macros[MDOC_MAX] = { { in_line_eoln, 0 }, /* Hf */ { obsolete, 0 }, /* Fr */ { in_line_eoln, 0 }, /* Ud */ - { in_line_eoln, 0 }, /* Lb */ + { in_line, 0 }, /* Lb */ { in_line_eoln, 0 }, /* Lp */ { in_line, MDOC_CALLABLE | MDOC_PARSED }, /* Lk */ { in_line, MDOC_CALLABLE | MDOC_PARSED }, /* Mt */ @@ -181,6 +186,7 @@ const struct mdoc_macro __mdoc_macros[MDOC_MAX] = { { in_line_eoln, 0 }, /* br */ { in_line_eoln, 0 }, /* sp */ { in_line_eoln, 0 }, /* %U */ + { phrase_ta, MDOC_CALLABLE | MDOC_PARSED }, /* Ta */ }; const struct mdoc_macro * const mdoc_macros = __mdoc_macros; @@ -191,6 +197,7 @@ swarn(struct mdoc *mdoc, enum mdoc_type type, int line, int pos, const struct mdoc_node *p) { const char *n, *t, *tt; + enum mandocerr ec; n = t = ""; tt = "block"; @@ -223,13 +230,12 @@ swarn(struct mdoc *mdoc, enum mdoc_type type, break; } - if ( ! (MDOC_IGN_SCOPE & mdoc->pflags)) - return(mdoc_verr(mdoc, line, pos, - "%s scope breaks %s scope of %s", + ec = (MDOC_IGN_SCOPE & mdoc->pflags) ? + MANDOCERR_SCOPE : MANDOCERR_SYNTSCOPE; + + return(mdoc_vmsg(mdoc, ec, line, pos, + "%s scope breaks %s of %s", tt, t, n)); - return(mdoc_vwarn(mdoc, line, pos, - "%s scope breaks %s scope of %s", - tt, t, n)); } @@ -252,7 +258,8 @@ mdoc_macroend(struct mdoc *m) continue; if ( ! (MDOC_EXPLICIT & mdoc_macros[n->tok].flags)) continue; - return(mdoc_nerr(m, n, EOPEN)); + mdoc_nmsg(m, n, MANDOCERR_SYNTSCOPE); + return(0); } /* Rewind to the first. */ @@ -264,8 +271,8 @@ mdoc_macroend(struct mdoc *m) /* * Look up a macro from within a subsequent context. */ -static int -lookup(int from, const char *p) +static enum mdoct +lookup(enum mdoct from, const char *p) { /* FIXME: make -diag lists be un-PARSED. */ @@ -278,10 +285,10 @@ lookup(int from, const char *p) /* * Lookup a macro following the initial line macro. */ -static int +static enum mdoct lookup_raw(const char *p) { - int res; + enum mdoct res; if (MDOC_MAX == (res = mdoc_hash_find(p))) return(MDOC_MAX); @@ -318,8 +325,8 @@ rew_last(struct mdoc *mdoc, const struct mdoc_node *to * Return the opening macro of a closing one, e.g., `Ec' has `Eo' as its * matching pair. */ -static int -rew_alt(int tok) +static enum mdoct +rew_alt(enum mdoct tok) { switch (tok) { case (MDOC_Ac): @@ -368,8 +375,9 @@ rew_alt(int tok) * close our current scope (REWIND_REWIND), or continue (REWIND_NOHALT). * The scope-closing and so on occurs in the various rew_* routines. */ -static int -rew_dohalt(int tok, enum mdoc_type type, const struct mdoc_node *p) +static enum rew +rew_dohalt(enum mdoct tok, enum mdoc_type type, + const struct mdoc_node *p) { if (MDOC_ROOT == p->type) @@ -508,7 +516,7 @@ rew_dohalt(int tok, enum mdoc_type type, const struct * REWIND_NOHALT). */ static int -rew_dobreak(int tok, const struct mdoc_node *p) +rew_dobreak(enum mdoct tok, const struct mdoc_node *p) { assert(MDOC_ROOT != p->type); @@ -537,7 +545,6 @@ rew_dobreak(int tok, const struct mdoc_node *p) return(1); break; case (MDOC_Oc): - /* XXX - experimental! */ if (MDOC_Op == p->tok) return(1); break; @@ -555,7 +562,7 @@ rew_dobreak(int tok, const struct mdoc_node *p) static int -rew_elem(struct mdoc *mdoc, int tok) +rew_elem(struct mdoc *mdoc, enum mdoct tok) { struct mdoc_node *n; @@ -571,10 +578,10 @@ rew_elem(struct mdoc *mdoc, int tok) static int rew_sub(enum mdoc_type t, struct mdoc *m, - int tok, int line, int ppos) + enum mdoct tok, int line, int ppos) { struct mdoc_node *n; - int c; + enum rew c; /* LINTED */ for (n = m->last; n; n = n->parent) { @@ -584,7 +591,9 @@ rew_sub(enum mdoc_type t, struct mdoc *m, return(1); if ( ! (MDOC_EXPLICIT & mdoc_macros[tok].flags)) return(1); - return(mdoc_perr(m, line, ppos, ENOCTX)); + /* FIXME: shouldn't raise an error */ + mdoc_pmsg(m, line, ppos, MANDOCERR_SYNTNOSCOPE); + return(0); } if (REWIND_REWIND == c) break; @@ -595,31 +604,64 @@ rew_sub(enum mdoc_type t, struct mdoc *m, } assert(n); - return(rew_last(m, n)); + if ( ! rew_last(m, n)) + return(0); + +#ifdef UGLY + /* + * The current block extends an enclosing block beyond a line + * break. Now that the current block ends, close the enclosing + * block, too. + */ + if (NULL != (n = n->pending)) { + assert(MDOC_HEAD == n->type); + if ( ! rew_last(m, n)) + return(0); + if ( ! mdoc_body_alloc(m, n->line, n->pos, n->tok)) + return(0); + } +#endif + + return(1); } static int -append_delims(struct mdoc *mdoc, int line, int *pos, char *buf) +append_delims(struct mdoc *m, int line, int *pos, char *buf) { - int c, lastarg; + int la; + enum margserr ac; char *p; - if (0 == buf[*pos]) + if ('\0' == buf[*pos]) return(1); for (;;) { - lastarg = *pos; - c = mdoc_zargs(mdoc, line, pos, buf, ARGS_NOWARN, &p); - assert(ARGS_PHRASE != c); + la = *pos; + ac = mdoc_zargs(m, line, pos, buf, ARGS_NOWARN, &p); - if (ARGS_ERROR == c) + if (ARGS_ERROR == ac) return(0); - else if (ARGS_EOLN == c) + else if (ARGS_EOLN == ac) break; - assert(mdoc_isdelim(p)); - if ( ! mdoc_word_alloc(mdoc, line, lastarg, p)) + + assert(DELIM_NONE != mdoc_isdelim(p)); + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); + + /* + * If we encounter end-of-sentence symbols, then trigger + * the double-space. + * + * XXX: it's easy to allow this to propogate outward to + * the last symbol, such that `. )' will cause the + * correct double-spacing. However, (1) groff isn't + * smart enough to do this and (2) it would require + * knowing which symbols break this behaviour, for + * example, `. ;' shouldn't propogate the double-space. + */ + if (mandoc_eos(p, strlen(p))) + m->last->flags |= MDOC_EOS; } return(1); @@ -632,9 +674,13 @@ append_delims(struct mdoc *mdoc, int line, int *pos, c static int blk_exp_close(MACRO_PROT_ARGS) { - int j, c, lastarg, maxargs, flushed; + int j, lastarg, maxargs, flushed, nl; + enum margserr ac; + enum mdoct ntok; char *p; + nl = MDOC_NEWLINE & m->flags; + switch (tok) { case (MDOC_Ec): maxargs = 1; @@ -645,8 +691,9 @@ blk_exp_close(MACRO_PROT_ARGS) } if ( ! (MDOC_CALLABLE & mdoc_macros[tok].flags)) { + /* FIXME: do this in validate */ if (buf[*pos]) - if ( ! mdoc_pwarn(m, line, ppos, ENOLINE)) + if ( ! mdoc_pmsg(m, line, ppos, MANDOCERR_ARGSLOST)) return(0); if ( ! rew_sub(MDOC_BODY, m, tok, line, ppos)) @@ -670,34 +717,37 @@ blk_exp_close(MACRO_PROT_ARGS) flushed = 1; } - c = mdoc_args(m, line, pos, buf, tok, &p); + ac = mdoc_args(m, line, pos, buf, tok, &p); - if (ARGS_ERROR == c) + if (ARGS_ERROR == ac) return(0); - if (ARGS_PUNCT == c) + if (ARGS_PUNCT == ac) break; - if (ARGS_EOLN == c) + if (ARGS_EOLN == ac) break; - if (MDOC_MAX != (c = lookup(tok, p))) { - if ( ! flushed) { - if ( ! rew_sub(MDOC_BLOCK, m, tok, line, ppos)) - return(0); - flushed = 1; - } - if ( ! mdoc_macro(m, c, line, lastarg, pos, buf)) + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup(tok, p); + + if (MDOC_MAX == ntok) { + if ( ! mdoc_word_alloc(m, line, lastarg, p)) return(0); - break; - } + continue; + } - if ( ! mdoc_word_alloc(m, line, lastarg, p)) + if ( ! flushed) { + if ( ! rew_sub(MDOC_BLOCK, m, tok, line, ppos)) + return(0); + flushed = 1; + } + if ( ! mdoc_macro(m, ntok, line, lastarg, pos, buf)) return(0); + break; } if ( ! flushed && ! rew_sub(MDOC_BLOCK, m, tok, line, ppos)) return(0); - if (ppos > 1) + if ( ! nl) return(1); return(append_delims(m, line, pos, buf)); } @@ -706,14 +756,21 @@ blk_exp_close(MACRO_PROT_ARGS) static int in_line(MACRO_PROT_ARGS) { - int la, lastpunct, c, w, cnt, d, nc; - struct mdoc_arg *arg; - char *p; + int la, scope, cnt, nc, nl; + enum margverr av; + enum mdoct ntok; + enum margserr ac; + enum mdelim d; + struct mdoc_arg *arg; + char *p; + nl = MDOC_NEWLINE & m->flags; + /* * Whether we allow ignored elements (those without content, * usually because of reserved words) to squeak by. */ + switch (tok) { case (MDOC_An): /* FALLTHROUGH */ @@ -735,36 +792,34 @@ in_line(MACRO_PROT_ARGS) for (arg = NULL;; ) { la = *pos; - c = mdoc_argv(m, line, tok, &arg, pos, buf); + av = mdoc_argv(m, line, tok, &arg, pos, buf); - if (ARGV_WORD == c) { + if (ARGV_WORD == av) { *pos = la; break; } - if (ARGV_EOLN == c) + if (ARGV_EOLN == av) break; - if (ARGV_ARG == c) + if (ARGV_ARG == av) continue; mdoc_argv_free(arg); return(0); } - for (cnt = 0, lastpunct = 1;; ) { + for (cnt = scope = 0;; ) { la = *pos; - w = mdoc_args(m, line, pos, buf, tok, &p); + ac = mdoc_args(m, line, pos, buf, tok, &p); - if (ARGS_ERROR == w) + if (ARGS_ERROR == ac) return(0); - if (ARGS_EOLN == w) + if (ARGS_EOLN == ac) break; - if (ARGS_PUNCT == w) + if (ARGS_PUNCT == ac) break; - /* Quoted words shouldn't be looked-up. */ + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup(tok, p); - c = ARGS_QWORD == w ? MDOC_MAX : lookup(tok, p); - /* * In this case, we've located a submacro and must * execute it. Close out scope, if open. If no @@ -772,8 +827,8 @@ in_line(MACRO_PROT_ARGS) * or raise a warning. */ - if (MDOC_MAX != c) { - if (0 == lastpunct && ! rew_elem(m, tok)) + if (MDOC_MAX != ntok) { + if (scope && ! rew_elem(m, tok)) return(0); if (nc && 0 == cnt) { if ( ! mdoc_elem_alloc(m, line, ppos, tok, arg)) @@ -782,13 +837,12 @@ in_line(MACRO_PROT_ARGS) return(0); } else if ( ! nc && 0 == cnt) { mdoc_argv_free(arg); - if ( ! mdoc_pwarn(m, line, ppos, EIGNE)) + if ( ! mdoc_pmsg(m, line, ppos, MANDOCERR_MACROEMPTY)) return(0); } - c = mdoc_macro(m, c, line, la, pos, buf); - if (0 == c) + if ( ! mdoc_macro(m, ntok, line, la, pos, buf)) return(0); - if (ppos > 1) + if ( ! nl) return(1); return(append_delims(m, line, pos, buf)); } @@ -799,19 +853,40 @@ in_line(MACRO_PROT_ARGS) * the word. */ - d = mdoc_isdelim(p); + d = ARGS_QWORD == ac ? DELIM_NONE : mdoc_isdelim(p); - if (ARGS_QWORD != w && d) { - if (0 == lastpunct && ! rew_elem(m, tok)) + if (DELIM_NONE != d) { + /* + * If we encounter closing punctuation, no word + * has been omitted, no scope is open, and we're + * allowed to have an empty element, then start + * a new scope. `Ar', `Fl', and `Li', only do + * this once per invocation. There may be more + * of these (all of them?). + */ + if (0 == cnt && (nc || MDOC_Li == tok) && + DELIM_CLOSE == d && ! scope) { + if ( ! mdoc_elem_alloc(m, line, ppos, tok, arg)) + return(0); + if (MDOC_Ar == tok || MDOC_Li == tok || + MDOC_Fl == tok) + cnt++; + scope = 1; + } + /* + * Close out our scope, if one is open, before + * any punctuation. + */ + if (scope && ! rew_elem(m, tok)) return(0); - lastpunct = 1; - } else if (lastpunct) { + scope = 0; + } else if ( ! scope) { if ( ! mdoc_elem_alloc(m, line, ppos, tok, arg)) return(0); - lastpunct = 0; + scope = 1; } - if ( ! d) + if (DELIM_NONE == d) cnt++; if ( ! mdoc_word_alloc(m, line, la, p)) return(0); @@ -821,22 +896,22 @@ in_line(MACRO_PROT_ARGS) * word so that the `-' can be added to each one without * having to parse out spaces. */ - if (0 == lastpunct && MDOC_Fl == tok) { + if (scope && MDOC_Fl == tok) { if ( ! rew_elem(m, tok)) return(0); - lastpunct = 1; + scope = 0; } } - if (0 == lastpunct && ! rew_elem(m, tok)) + if (scope && ! rew_elem(m, tok)) return(0); /* * If no elements have been collected and we're allowed to have * empties (nc), open a scope and close it out. Otherwise, * raise a warning. - * */ + if (nc && 0 == cnt) { if ( ! mdoc_elem_alloc(m, line, ppos, tok, arg)) return(0); @@ -844,11 +919,11 @@ in_line(MACRO_PROT_ARGS) return(0); } else if ( ! nc && 0 == cnt) { mdoc_argv_free(arg); - if ( ! mdoc_pwarn(m, line, ppos, EIGNE)) + if ( ! mdoc_pmsg(m, line, ppos, MANDOCERR_MACROEMPTY)) return(0); } - if (ppos > 1) + if ( ! nl) return(1); return(append_delims(m, line, pos, buf)); } @@ -857,25 +932,22 @@ in_line(MACRO_PROT_ARGS) static int blk_full(MACRO_PROT_ARGS) { - int c, lastarg, reopen, dohead; + int la, nl; struct mdoc_arg *arg; + struct mdoc_node *head; /* save of head macro */ + struct mdoc_node *body; /* save of body macro */ +#ifdef UGLY + struct mdoc_node *n; +#endif + enum mdoc_type mtt; + enum mdoct ntok; + enum margserr ac, lac; + enum margverr av; char *p; - /* - * Whether to process a block-head section. If this is - * non-zero, then a head will be opened for all line arguments. - * If not, then the head will always be empty and only a body - * will be opened, which will stay open at the eoln. - */ + nl = MDOC_NEWLINE & m->flags; - switch (tok) { - case (MDOC_Nd): - dohead = 0; - break; - default: - dohead = 1; - break; - } + /* Close out prior implicit scope. */ if ( ! (MDOC_EXPLICIT & mdoc_macros[tok].flags)) { if ( ! rew_sub(MDOC_BODY, m, tok, line, ppos)) @@ -884,18 +956,27 @@ blk_full(MACRO_PROT_ARGS) return(0); } + /* + * This routine accomodates implicitly- and explicitly-scoped + * macro openings. Implicit ones first close out prior scope + * (seen above). Delay opening the head until necessary to + * allow leading punctuation to print. Special consideration + * for `It -column', which has phrase-part syntax instead of + * regular child nodes. + */ + for (arg = NULL;; ) { - lastarg = *pos; - c = mdoc_argv(m, line, tok, &arg, pos, buf); + la = *pos; + av = mdoc_argv(m, line, tok, &arg, pos, buf); - if (ARGV_WORD == c) { - *pos = lastarg; + if (ARGV_WORD == av) { + *pos = la; break; } - if (ARGV_EOLN == c) + if (ARGV_EOLN == av) break; - if (ARGV_ARG == c) + if (ARGV_ARG == av) continue; mdoc_argv_free(arg); @@ -905,72 +986,160 @@ blk_full(MACRO_PROT_ARGS) if ( ! mdoc_block_alloc(m, line, ppos, tok, arg)) return(0); - if (0 == buf[*pos]) { + head = body = NULL; + + /* + * The `Nd' macro has all arguments in its body: it's a hybrid + * of block partial-explicit and full-implicit. Stupid. + */ + + if (MDOC_Nd == tok) { if ( ! mdoc_head_alloc(m, line, ppos, tok)) return(0); + head = m->last; if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) return(0); if ( ! mdoc_body_alloc(m, line, ppos, tok)) return(0); - return(1); - } + body = m->last; + } - if ( ! mdoc_head_alloc(m, line, ppos, tok)) - return(0); + ac = ARGS_ERROR; - /* Immediately close out head and enter body, if applicable. */ + for ( ; ; ) { + la = *pos; + /* Initialise last-phrase-type with ARGS_PEND. */ + lac = ARGS_ERROR == ac ? ARGS_PEND : ac; + ac = mdoc_args(m, line, pos, buf, tok, &p); - if (0 == dohead) { - if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) + if (ARGS_ERROR == ac) return(0); - if ( ! mdoc_body_alloc(m, line, ppos, tok)) - return(0); - } - for (reopen = 0;; ) { - lastarg = *pos; - c = mdoc_args(m, line, pos, buf, tok, &p); - - if (ARGS_ERROR == c) - return(0); - if (ARGS_EOLN == c) + if (ARGS_EOLN == ac) { + if (ARGS_PPHRASE != lac && ARGS_PHRASE != lac) + break; + /* + * This is necessary: if the last token on a + * line is a `Ta' or tab, then we'll get + * ARGS_EOLN, so we must be smart enough to + * reopen our scope if the last parse was a + * phrase or partial phrase. + */ + if ( ! rew_sub(MDOC_BODY, m, tok, line, ppos)) + return(0); + if ( ! mdoc_body_alloc(m, line, ppos, tok)) + return(0); + body = m->last; break; - if (ARGS_PHRASE == c) { - assert(dohead); - if (reopen && ! mdoc_head_alloc(m, line, ppos, tok)) + } + + /* + * Emit leading punctuation (i.e., punctuation before + * the MDOC_HEAD) for non-phrase types. + */ + + if (NULL == head && + ARGS_PEND != ac && + ARGS_PHRASE != ac && + ARGS_PPHRASE != ac && + ARGS_QWORD != ac && + DELIM_OPEN == mdoc_isdelim(p)) { + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); + continue; + } + + /* Open a head if one hasn't been opened. */ + + if (NULL == head) { + if ( ! mdoc_head_alloc(m, line, ppos, tok)) + return(0); + head = m->last; + } + + if (ARGS_PHRASE == ac || + ARGS_PEND == ac || + ARGS_PPHRASE == ac) { /* - * Phrases are self-contained macro phrases used - * in the columnar output of a macro. They need - * special handling. + * If we haven't opened a body yet, rewind the + * head; if we have, rewind that instead. */ - if ( ! phrase(m, line, lastarg, buf)) + + mtt = body ? MDOC_BODY : MDOC_HEAD; + if ( ! rew_sub(mtt, m, tok, line, ppos)) return(0); - if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) + + /* Then allocate our body context. */ + + if ( ! mdoc_body_alloc(m, line, ppos, tok)) return(0); + body = m->last; - reopen = 1; + /* + * Process phrases: set whether we're in a + * partial-phrase (this effects line handling) + * then call down into the phrase parser. + */ + + if (ARGS_PPHRASE == ac) + m->flags |= MDOC_PPHRASE; + if (ARGS_PEND == ac && ARGS_PPHRASE == lac) + m->flags |= MDOC_PPHRASE; + + if ( ! phrase(m, line, la, buf)) + return(0); + + m->flags &= ~MDOC_PPHRASE; continue; } - if (MDOC_MAX == (c = lookup(tok, p))) { - if ( ! mdoc_word_alloc(m, line, lastarg, p)) + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup(tok, p); + + if (MDOC_MAX == ntok) { + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); continue; - } + } - if ( ! mdoc_macro(m, c, line, lastarg, pos, buf)) + if ( ! mdoc_macro(m, ntok, line, la, pos, buf)) return(0); break; } + + if (NULL == head) { + if ( ! mdoc_head_alloc(m, line, ppos, tok)) + return(0); + head = m->last; + } - if (1 == ppos && ! append_delims(m, line, pos, buf)) + if (nl && ! append_delims(m, line, pos, buf)) return(0); - /* If the body's already open, then just return. */ - if (0 == dohead) + /* If we've already opened our body, exit now. */ + + if (NULL != body) return(1); +#ifdef UGLY + /* + * If there is an open (i.e., unvalidated) sub-block requiring + * explicit close-out, postpone switching the current block from + * head to body until the rew_sub() call closing out that + * sub-block. + */ + for (n = m->last; n && n != head; n = n->parent) { + if (MDOC_BLOCK == n->type && + MDOC_EXPLICIT & mdoc_macros[n->tok].flags && + ! (MDOC_VALID & n->flags)) { + assert( ! (MDOC_ACTED & n->flags)); + n->pending = head; + return(1); + } + } +#endif + + /* Close out scopes to remain in a consistent state. */ + if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) return(0); if ( ! mdoc_body_alloc(m, line, ppos, tok)) @@ -983,12 +1152,16 @@ blk_full(MACRO_PROT_ARGS) static int blk_part_imp(MACRO_PROT_ARGS) { - int la, c; + int la, nl; + enum mdoct ntok; + enum margserr ac; char *p; struct mdoc_node *blk; /* saved block context */ struct mdoc_node *body; /* saved body context */ struct mdoc_node *n; + nl = MDOC_NEWLINE & m->flags; + /* * A macro that spans to the end of the line. This is generally * (but not necessarily) called as the first macro. The block @@ -1016,18 +1189,17 @@ blk_part_imp(MACRO_PROT_ARGS) for (body = NULL; ; ) { la = *pos; - c = mdoc_args(m, line, pos, buf, tok, &p); + ac = mdoc_args(m, line, pos, buf, tok, &p); - assert(ARGS_PHRASE != c); - - if (ARGS_ERROR == c) + if (ARGS_ERROR == ac) return(0); - if (ARGS_EOLN == c) + if (ARGS_EOLN == ac) break; - if (ARGS_PUNCT == c) + if (ARGS_PUNCT == ac) break; - if (1 == mdoc_isdelim(p) && NULL == body) { + if (NULL == body && ARGS_QWORD != ac && + DELIM_OPEN == mdoc_isdelim(p)) { if ( ! mdoc_word_alloc(m, line, la, p)) return(0); continue; @@ -1039,35 +1211,59 @@ blk_part_imp(MACRO_PROT_ARGS) body = m->last; } - if (MDOC_MAX != (c = lookup(tok, p))) { - if ( ! mdoc_macro(m, c, line, la, pos, buf)) + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup(tok, p); + + if (MDOC_MAX == ntok) { + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); - break; - } + continue; + } - if ( ! mdoc_word_alloc(m, line, la, p)) + if ( ! mdoc_macro(m, ntok, line, la, pos, buf)) return(0); + break; } /* Clean-ups to leave in a consistent state. */ - if (NULL == body && ! mdoc_body_alloc(m, line, ppos, tok)) - return(0); + if (NULL == body) { + if ( ! mdoc_body_alloc(m, line, ppos, tok)) + return(0); + body = m->last; + } - body = m->last; + for (n = body->child; n && n->next; n = n->next) + /* Do nothing. */ ; + + /* + * End of sentence spacing: if the last node is a text node and + * has a trailing period, then mark it as being end-of-sentence. + */ + if (n && MDOC_TEXT == n->type && n->string) + if (mandoc_eos(n->string, strlen(n->string))) + n->flags |= MDOC_EOS; + + /* Up-propogate the end-of-space flag. */ + + if (n && (MDOC_EOS & n->flags)) { + body->flags |= MDOC_EOS; + body->parent->flags |= MDOC_EOS; + } + /* * 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. + * crufty use of `Op' breakage. + * + * FIXME - this should be ifdef'd OpenBSD? */ - for (n = m->last; n; n = n->parent) if (body == n) break; - if (NULL == n && ! mdoc_nwarn(m, body, EIMPBRK)) + if (NULL == n && ! mdoc_nmsg(m, body, MANDOCERR_SCOPE)) return(0); if (n && ! rew_last(m, body)) @@ -1075,7 +1271,7 @@ blk_part_imp(MACRO_PROT_ARGS) /* Standard appending of delimiters. */ - if (1 == ppos && ! append_delims(m, line, pos, buf)) + if (nl && ! append_delims(m, line, pos, buf)) return(0); /* Rewind scope, if applicable. */ @@ -1090,11 +1286,15 @@ blk_part_imp(MACRO_PROT_ARGS) static int blk_part_exp(MACRO_PROT_ARGS) { - int la, c; + int la, nl; + enum margserr ac; struct mdoc_node *head; /* keep track of head */ struct mdoc_node *body; /* keep track of body */ char *p; + enum mdoct ntok; + nl = MDOC_NEWLINE & m->flags; + /* * The opening of an explicit macro having zero or more leading * punctuation nodes; a head with optional single element (the @@ -1106,20 +1306,19 @@ blk_part_exp(MACRO_PROT_ARGS) for (head = body = NULL; ; ) { la = *pos; - c = mdoc_args(m, line, pos, buf, tok, &p); + ac = mdoc_args(m, line, pos, buf, tok, &p); - if (ARGS_ERROR == c) + if (ARGS_ERROR == ac) return(0); - if (ARGS_PUNCT == c) + if (ARGS_PUNCT == ac) break; - if (ARGS_EOLN == c) + if (ARGS_EOLN == ac) break; - assert(ARGS_PHRASE != c); - /* Flush out leading punctuation. */ - if (1 == mdoc_isdelim(p) && NULL == head) { + if (NULL == head && ARGS_QWORD != ac && + DELIM_OPEN == mdoc_isdelim(p)) { assert(NULL == body); if ( ! mdoc_word_alloc(m, line, la, p)) return(0); @@ -1157,14 +1356,17 @@ blk_part_exp(MACRO_PROT_ARGS) assert(NULL != head && NULL != body); - if (MDOC_MAX != (c = lookup(tok, p))) { - if ( ! mdoc_macro(m, c, line, la, pos, buf)) + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup(tok, p); + + if (MDOC_MAX == ntok) { + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); - break; + continue; } - if ( ! mdoc_word_alloc(m, line, la, p)) + if ( ! mdoc_macro(m, ntok, line, la, pos, buf)) return(0); + break; } /* Clean-up to leave in a consistent state. */ @@ -1185,22 +1387,33 @@ blk_part_exp(MACRO_PROT_ARGS) /* Standard appending of delimiters. */ - if (ppos > 1) + if ( ! nl) return(1); - return(append_delims(m, line, pos, buf)); } +/* ARGSUSED */ static int in_line_argn(MACRO_PROT_ARGS) { - int la, flushed, j, c, maxargs; - struct mdoc_arg *arg; - char *p; + int la, flushed, j, maxargs, nl; + enum margserr ac; + enum margverr av; + struct mdoc_arg *arg; + char *p; + enum mdoct ntok; - /* Fixed maximum arguments per macro, if applicable. */ + nl = MDOC_NEWLINE & m->flags; + /* + * A line macro that has a fixed number of arguments (maxargs). + * Only open the scope once the first non-leading-punctuation is + * found (unless MDOC_IGNDELIM is noted, like in `Pf'), then + * keep it open until the maximum number of arguments are + * exhausted. + */ + switch (tok) { case (MDOC_Ap): /* FALLTHROUGH */ @@ -1219,62 +1432,67 @@ in_line_argn(MACRO_PROT_ARGS) break; } - /* Macro argument processing. */ - - for (arg = NULL;; ) { + for (arg = NULL; ; ) { la = *pos; - c = mdoc_argv(m, line, tok, &arg, pos, buf); + av = mdoc_argv(m, line, tok, &arg, pos, buf); - if (ARGV_WORD == c) { + if (ARGV_WORD == av) { *pos = la; break; } - if (ARGV_EOLN == c) + if (ARGV_EOLN == av) break; - if (ARGV_ARG == c) + if (ARGV_ARG == av) continue; mdoc_argv_free(arg); return(0); } - /* Open the element scope. */ + for (flushed = j = 0; ; ) { + la = *pos; + ac = mdoc_args(m, line, pos, buf, tok, &p); - if ( ! mdoc_elem_alloc(m, line, ppos, tok, arg)) - return(0); + if (ARGS_ERROR == ac) + return(0); + if (ARGS_PUNCT == ac) + break; + if (ARGS_EOLN == ac) + break; - /* Process element arguments. */ + if ( ! (MDOC_IGNDELIM & mdoc_macros[tok].flags) && + ARGS_QWORD != ac && + 0 == j && DELIM_OPEN == mdoc_isdelim(p)) { + if ( ! mdoc_word_alloc(m, line, la, p)) + return(0); + continue; + } else if (0 == j) + if ( ! mdoc_elem_alloc(m, line, la, tok, arg)) + return(0); - for (flushed = j = 0; ; j++) { - la = *pos; - if (j == maxargs && ! flushed) { if ( ! rew_elem(m, tok)) return(0); flushed = 1; } - c = mdoc_args(m, line, pos, buf, tok, &p); + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup(tok, p); - if (ARGS_ERROR == c) - return(0); - if (ARGS_PUNCT == c) - break; - if (ARGS_EOLN == c) - break; - - if (MDOC_MAX != (c = lookup(tok, p))) { + if (MDOC_MAX != ntok) { if ( ! flushed && ! rew_elem(m, tok)) return(0); flushed = 1; - if ( ! mdoc_macro(m, c, line, la, pos, buf)) + if ( ! mdoc_macro(m, ntok, line, la, pos, buf)) return(0); + j++; break; } if ( ! (MDOC_IGNDELIM & mdoc_macros[tok].flags) && - ! flushed && mdoc_isdelim(p)) { + ARGS_QWORD != ac && + ! flushed && + DELIM_NONE != mdoc_isdelim(p)) { if ( ! rew_elem(m, tok)) return(0); flushed = 1; @@ -1287,23 +1505,28 @@ in_line_argn(MACRO_PROT_ARGS) * code is no here, it's unlikely to be removed. */ +#ifdef __OpenBSD__ if (MDOC_Xr == tok && j == maxargs) { - if ( ! mdoc_elem_alloc(m, line, ppos, MDOC_Ns, NULL)) + if ( ! mdoc_elem_alloc(m, line, la, MDOC_Ns, NULL)) return(0); if ( ! rew_elem(m, MDOC_Ns)) return(0); } +#endif if ( ! mdoc_word_alloc(m, line, la, p)) return(0); + j++; } - /* Close out and append delimiters. */ + if (0 == j && ! mdoc_elem_alloc(m, line, la, tok, arg)) + return(0); + /* Close out in a consistent state. */ + if ( ! flushed && ! rew_elem(m, tok)) return(0); - - if (ppos > 1) + if ( ! nl) return(1); return(append_delims(m, line, pos, buf)); } @@ -1312,9 +1535,12 @@ in_line_argn(MACRO_PROT_ARGS) static int in_line_eoln(MACRO_PROT_ARGS) { - int c, w, la; - struct mdoc_arg *arg; - char *p; + int la; + enum margserr ac; + enum margverr av; + struct mdoc_arg *arg; + char *p; + enum mdoct ntok; assert( ! (MDOC_PARSED & mdoc_macros[tok].flags)); @@ -1322,15 +1548,15 @@ in_line_eoln(MACRO_PROT_ARGS) for (arg = NULL; ; ) { la = *pos; - c = mdoc_argv(m, line, tok, &arg, pos, buf); + av = mdoc_argv(m, line, tok, &arg, pos, buf); - if (ARGV_WORD == c) { + if (ARGV_WORD == av) { *pos = la; break; } - if (ARGV_EOLN == c) + if (ARGV_EOLN == av) break; - if (ARGV_ARG == c) + if (ARGV_ARG == av) continue; mdoc_argv_free(arg); @@ -1346,23 +1572,24 @@ in_line_eoln(MACRO_PROT_ARGS) for (;;) { la = *pos; - w = mdoc_args(m, line, pos, buf, tok, &p); + ac = mdoc_args(m, line, pos, buf, tok, &p); - if (ARGS_ERROR == w) + if (ARGS_ERROR == ac) return(0); - if (ARGS_EOLN == w) + if (ARGS_EOLN == ac) break; - c = ARGS_QWORD == w ? MDOC_MAX : lookup(tok, p); + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup(tok, p); - if (MDOC_MAX != c) { - if ( ! rew_elem(m, tok)) + if (MDOC_MAX == ntok) { + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); - return(mdoc_macro(m, c, line, la, pos, buf)); - } + continue; + } - if ( ! mdoc_word_alloc(m, line, la, p)) + if ( ! rew_elem(m, tok)) return(0); + return(mdoc_macro(m, ntok, line, la, pos, buf)); } /* Close out (no delimiters). */ @@ -1375,13 +1602,16 @@ in_line_eoln(MACRO_PROT_ARGS) static int ctx_synopsis(MACRO_PROT_ARGS) { + int nl; + nl = MDOC_NEWLINE & m->flags; + /* 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) + if ( ! nl) return(in_line(m, tok, line, ppos, pos, buf)); /* @@ -1399,7 +1629,7 @@ static int obsolete(MACRO_PROT_ARGS) { - return(mdoc_pwarn(m, line, ppos, EOBS)); + return(mdoc_pmsg(m, line, ppos, MANDOCERR_MACROOBS)); } @@ -1411,33 +1641,96 @@ obsolete(MACRO_PROT_ARGS) static int phrase(struct mdoc *m, int line, int ppos, char *buf) { - int c, w, la, pos; - char *p; + int la, pos; + enum margserr ac; + enum mdoct ntok; + char *p; for (pos = ppos; ; ) { la = pos; - /* Note: no calling context! */ - w = mdoc_zargs(m, line, &pos, buf, 0, &p); + ac = mdoc_zargs(m, line, &pos, buf, 0, &p); - if (ARGS_ERROR == w) + if (ARGS_ERROR == ac) return(0); - if (ARGS_EOLN == w) + if (ARGS_EOLN == ac) break; - c = ARGS_QWORD == w ? MDOC_MAX : lookup_raw(p); + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup_raw(p); - if (MDOC_MAX != c) { - if ( ! mdoc_macro(m, c, line, la, &pos, buf)) + if (MDOC_MAX == ntok) { + if ( ! mdoc_word_alloc(m, line, la, p)) return(0); - return(append_delims(m, line, &pos, buf)); - } + continue; + } - if ( ! mdoc_word_alloc(m, line, la, p)) + if ( ! mdoc_macro(m, ntok, line, la, &pos, buf)) return(0); + return(append_delims(m, line, &pos, buf)); } return(1); } +static int +phrase_ta(MACRO_PROT_ARGS) +{ + int la; + enum mdoct ntok; + enum margserr ac; + struct mdoc_node *n; + char *p; + + /* + * FIXME: this is overly restrictive: if the `Ta' is unexpected, + * it should simply error out with ARGSLOST. + */ + + n = m->last; + if ( ! rew_sub(MDOC_BODY, m, MDOC_It, line, ppos)) + return(0); + + /* + * FIXME: this is necessary in bogus constructions like + * .Bl -column foo bar + * .Sy foo Ta bar + * It is, however, an ugly way to do it. + * + * XXX; remove this when the above construct builds an implied + * `It' marker. + */ + if (NULL == m->last || + MDOC_BODY != m->last->type || + MDOC_It != m->last->tok) { + swarn(m, tok, line, ppos, n); + return(0); + } + + if ( ! mdoc_body_alloc(m, line, ppos, MDOC_It)) + return(0); + + for (;;) { + la = *pos; + ac = mdoc_zargs(m, line, pos, buf, 0, &p); + + if (ARGS_ERROR == ac) + return(0); + if (ARGS_EOLN == ac) + break; + + ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup_raw(p); + + if (MDOC_MAX == ntok) { + if ( ! mdoc_word_alloc(m, line, la, p)) + return(0); + continue; + } + + if ( ! mdoc_macro(m, ntok, line, la, pos, buf)) + return(0); + return(append_delims(m, line, pos, buf)); + } + + return(1); +}