=================================================================== RCS file: /cvs/mandoc/mdoc_macro.c,v retrieving revision 1.55 retrieving revision 1.67 diff -u -p -r1.55 -r1.67 --- mandoc/mdoc_macro.c 2010/05/07 05:48:29 1.55 +++ mandoc/mdoc_macro.c 2010/05/15 08:54:04 1.67 @@ -1,4 +1,4 @@ -/* $Id: mdoc_macro.c,v 1.55 2010/05/07 05:48:29 kristaps Exp $ */ +/* $Id: mdoc_macro.c,v 1.67 2010/05/15 08:54:04 schwarze Exp $ */ /* * Copyright (c) 2008, 2009 Kristaps Dzonsons * @@ -26,6 +26,7 @@ #include #include "libmdoc.h" +#include "libmandoc.h" enum rew { REWIND_REWIND, @@ -45,9 +46,10 @@ static int obsolete(MACRO_PROT_ARGS); static int append_delims(struct mdoc *, int, int *, char *); -static enum mdoct lookup(int, const 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 int phrase(struct mdoc *, int, int, + char *, enum margserr); static enum mdoct rew_alt(enum mdoct); static int rew_dobreak(enum mdoct, const struct mdoc_node *); @@ -267,7 +269,7 @@ mdoc_macroend(struct mdoc *m) * Look up a macro from within a subsequent context. */ static enum mdoct -lookup(int from, const char *p) +lookup(enum mdoct from, const char *p) { /* FIXME: make -diag lists be un-PARSED. */ @@ -283,7 +285,7 @@ lookup(int from, const char *p) static enum mdoct lookup_raw(const char *p) { - int res; + enum mdoct res; if (MDOC_MAX == (res = mdoc_hash_find(p))) return(MDOC_MAX); @@ -620,27 +622,41 @@ rew_sub(enum mdoc_type t, struct mdoc *m, 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 lastarg; + int la; enum margserr ac; char *p; - if (0 == buf[*pos]) + if ('\0' == buf[*pos]) return(1); for (;;) { - lastarg = *pos; - ac = mdoc_zargs(mdoc, line, pos, buf, ARGS_NOWARN, &p); - assert(ARGS_PHRASE != ac); + la = *pos; + ac = mdoc_zargs(m, line, pos, buf, ARGS_NOWARN, &p); if (ARGS_ERROR == ac) return(0); 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); @@ -653,11 +669,13 @@ append_delims(struct mdoc *mdoc, int line, int *pos, c static int blk_exp_close(MACRO_PROT_ARGS) { - int j, 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; @@ -723,7 +741,7 @@ blk_exp_close(MACRO_PROT_ARGS) 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)); } @@ -732,12 +750,16 @@ blk_exp_close(MACRO_PROT_ARGS) static int in_line(MACRO_PROT_ARGS) { - int la, lastpunct, c, cnt, d, nc; + int la, lastpunct, 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. @@ -764,15 +786,15 @@ 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); @@ -814,7 +836,7 @@ in_line(MACRO_PROT_ARGS) } 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)); } @@ -825,9 +847,9 @@ in_line(MACRO_PROT_ARGS) * the word. */ - d = ARGS_QWORD == ac ? 0 : mdoc_isdelim(p); + d = ARGS_QWORD == ac ? DELIM_NONE : mdoc_isdelim(p); - if (ARGS_QWORD != ac && d) { + if (ARGS_QWORD != ac && DELIM_NONE != d) { if (0 == lastpunct && ! rew_elem(m, tok)) return(0); lastpunct = 1; @@ -837,7 +859,7 @@ in_line(MACRO_PROT_ARGS) lastpunct = 0; } - if ( ! d) + if (DELIM_NONE == d) cnt++; if ( ! mdoc_word_alloc(m, line, la, p)) return(0); @@ -874,7 +896,7 @@ in_line(MACRO_PROT_ARGS) return(0); } - if (ppos > 1) + if ( ! nl) return(1); return(append_delims(m, line, pos, buf)); } @@ -883,7 +905,7 @@ in_line(MACRO_PROT_ARGS) static int blk_full(MACRO_PROT_ARGS) { - int c, la; + int la, nl; struct mdoc_arg *arg; struct mdoc_node *head; /* save of head macro */ struct mdoc_node *body; /* save of body macro */ @@ -891,9 +913,12 @@ blk_full(MACRO_PROT_ARGS) struct mdoc_node *n; #endif enum mdoct ntok; - enum margserr ac; + enum margserr ac, lac; + enum margverr av; char *p; + nl = MDOC_NEWLINE & m->flags; + /* Close out prior implicit scope. */ if ( ! (MDOC_EXPLICIT & mdoc_macros[tok].flags)) { @@ -914,16 +939,16 @@ blk_full(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); @@ -951,8 +976,11 @@ blk_full(MACRO_PROT_ARGS) body = m->last; } - for (;;) { + ac = ARGS_ERROR; + + for ( ; ; ) { la = *pos; + lac = ac; ac = mdoc_args(m, line, pos, buf, tok, &p); if (ARGS_ERROR == ac) @@ -960,11 +988,20 @@ blk_full(MACRO_PROT_ARGS) if (ARGS_EOLN == ac) break; + if (ARGS_PEND == ac) { + if (ARGS_PPHRASE == lac) + ac = ARGS_PPHRASE; + else + ac = ARGS_PHRASE; + } + /* Don't emit leading punct. for phrases. */ - if (NULL == head && ARGS_PHRASE != ac && + if (NULL == head && + ARGS_PHRASE != ac && + ARGS_PPHRASE != ac && ARGS_QWORD != ac && - 1 == mdoc_isdelim(p)) { + DELIM_OPEN == mdoc_isdelim(p)) { if ( ! mdoc_word_alloc(m, line, la, p)) return(0); continue; @@ -972,15 +1009,20 @@ blk_full(MACRO_PROT_ARGS) /* Always re-open head for phrases. */ - if (NULL == head || ARGS_PHRASE == ac) { + if (NULL == head || + ARGS_PHRASE == ac || + ARGS_PPHRASE == ac) { if ( ! mdoc_head_alloc(m, line, ppos, tok)) return(0); head = m->last; } - if (ARGS_PHRASE == ac) { - if ( ! phrase(m, line, la, buf)) + if (ARGS_PHRASE == ac || ARGS_PPHRASE == ac) { + if (ARGS_PPHRASE == ac) + m->flags |= MDOC_PPHRASE; + if ( ! phrase(m, line, la, buf, ac)) return(0); + m->flags &= ~MDOC_PPHRASE; if ( ! rew_sub(MDOC_HEAD, m, tok, line, ppos)) return(0); continue; @@ -1005,7 +1047,7 @@ blk_full(MACRO_PROT_ARGS) head = m->last; } - if (1 == ppos && ! append_delims(m, line, pos, buf)) + if (nl && ! append_delims(m, line, pos, buf)) return(0); /* If we've already opened our body, exit now. */ @@ -1045,7 +1087,7 @@ blk_full(MACRO_PROT_ARGS) static int blk_part_imp(MACRO_PROT_ARGS) { - int la; + int la, nl; enum mdoct ntok; enum margserr ac; char *p; @@ -1053,6 +1095,8 @@ blk_part_imp(MACRO_PROT_ARGS) 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 @@ -1082,8 +1126,6 @@ blk_part_imp(MACRO_PROT_ARGS) la = *pos; ac = mdoc_args(m, line, pos, buf, tok, &p); - assert(ARGS_PHRASE != ac); - if (ARGS_ERROR == ac) return(0); if (ARGS_EOLN == ac) @@ -1092,7 +1134,7 @@ blk_part_imp(MACRO_PROT_ARGS) break; if (NULL == body && ARGS_QWORD != ac && - 1 == mdoc_isdelim(p)) { + DELIM_OPEN == mdoc_isdelim(p)) { if ( ! mdoc_word_alloc(m, line, la, p)) return(0); continue; @@ -1125,11 +1167,32 @@ blk_part_imp(MACRO_PROT_ARGS) 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. + * + * FIXME - this should be ifdef'd OpenBSD? */ for (n = m->last; n; n = n->parent) if (body == n) @@ -1143,7 +1206,7 @@ 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. */ @@ -1158,13 +1221,15 @@ blk_part_imp(MACRO_PROT_ARGS) static int blk_part_exp(MACRO_PROT_ARGS) { - int la; + 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 @@ -1185,12 +1250,10 @@ blk_part_exp(MACRO_PROT_ARGS) if (ARGS_EOLN == ac) break; - assert(ARGS_PHRASE != ac); - /* Flush out leading punctuation. */ if (NULL == head && ARGS_QWORD != ac && - 1 == mdoc_isdelim(p)) { + DELIM_OPEN == mdoc_isdelim(p)) { assert(NULL == body); if ( ! mdoc_word_alloc(m, line, la, p)) return(0); @@ -1259,22 +1322,25 @@ 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; + int la, flushed, j, maxargs, nl; enum margserr ac; + enum margverr av; struct mdoc_arg *arg; char *p; enum mdoct ntok; + 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 @@ -1303,16 +1369,16 @@ in_line_argn(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); @@ -1332,7 +1398,7 @@ in_line_argn(MACRO_PROT_ARGS) if ( ! (MDOC_IGNDELIM & mdoc_macros[tok].flags) && ARGS_QWORD != ac && - 0 == j && 1 == mdoc_isdelim(p)) { + 0 == j && DELIM_OPEN == mdoc_isdelim(p)) { if ( ! mdoc_word_alloc(m, line, la, p)) return(0); continue; @@ -1360,7 +1426,8 @@ in_line_argn(MACRO_PROT_ARGS) if ( ! (MDOC_IGNDELIM & mdoc_macros[tok].flags) && ARGS_QWORD != ac && - ! flushed && mdoc_isdelim(p)) { + ! flushed && + DELIM_NONE != mdoc_isdelim(p)) { if ( ! rew_elem(m, tok)) return(0); flushed = 1; @@ -1394,8 +1461,7 @@ in_line_argn(MACRO_PROT_ARGS) if ( ! flushed && ! rew_elem(m, tok)) return(0); - - if (ppos > 1) + if ( ! nl) return(1); return(append_delims(m, line, pos, buf)); } @@ -1404,8 +1470,9 @@ in_line_argn(MACRO_PROT_ARGS) static int in_line_eoln(MACRO_PROT_ARGS) { - int c, la; + int la; enum margserr ac; + enum margverr av; struct mdoc_arg *arg; char *p; enum mdoct ntok; @@ -1416,15 +1483,15 @@ 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); @@ -1470,13 +1537,16 @@ 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)); /* @@ -1504,25 +1574,26 @@ obsolete(MACRO_PROT_ARGS) * macro is encountered. */ static int -phrase(struct mdoc *m, int line, int ppos, char *buf) +phrase(struct mdoc *m, int line, int ppos, char *buf, enum margserr ac) { int la, pos; - enum margserr ac; + enum margserr aac; enum mdoct ntok; char *p; + assert(ARGS_PHRASE == ac || ARGS_PPHRASE == ac); + for (pos = ppos; ; ) { la = pos; - /* Note: no calling context! */ - ac = mdoc_zargs(m, line, &pos, buf, 0, &p); + aac = mdoc_zargs(m, line, &pos, buf, 0, &p); - if (ARGS_ERROR == ac) + if (ARGS_ERROR == aac) return(0); - if (ARGS_EOLN == ac) + if (ARGS_EOLN == aac) break; - ntok = ARGS_QWORD == ac ? MDOC_MAX : lookup_raw(p); + ntok = ARGS_QWORD == aac ? MDOC_MAX : lookup_raw(p); if (MDOC_MAX == ntok) { if ( ! mdoc_word_alloc(m, line, la, p))