=================================================================== RCS file: /cvs/mandoc/mandoc.c,v retrieving revision 1.98 retrieving revision 1.112 diff -u -p -r1.98 -r1.112 --- mandoc/mandoc.c 2015/11/12 22:44:27 1.98 +++ mandoc/mandoc.c 2018/12/16 00:17:02 1.112 @@ -1,7 +1,7 @@ -/* $Id: mandoc.c,v 1.98 2015/11/12 22:44:27 schwarze Exp $ */ +/* $Id: mandoc.c,v 1.112 2018/12/16 00:17:02 schwarze Exp $ */ /* * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons - * Copyright (c) 2011-2015 Ingo Schwarze + * Copyright (c) 2011-2015, 2017, 2018 Ingo Schwarze * * Permission to use, copy, modify, and distribute this software for any * purpose with or without fee is hereby granted, provided that the above @@ -28,8 +28,9 @@ #include #include -#include "mandoc.h" #include "mandoc_aux.h" +#include "mandoc.h" +#include "roff.h" #include "libmandoc.h" static int a2time(time_t *, const char *, const char *); @@ -37,10 +38,63 @@ static char *time2a(time_t); enum mandoc_esc +mandoc_font(const char *cp, int sz) +{ + switch (sz) { + case 0: + return ESCAPE_FONTPREV; + case 1: + switch (cp[0]) { + case 'B': + case '3': + return ESCAPE_FONTBOLD; + case 'I': + case '2': + return ESCAPE_FONTITALIC; + case 'P': + return ESCAPE_FONTPREV; + case 'R': + case '1': + return ESCAPE_FONTROMAN; + case '4': + return ESCAPE_FONTBI; + default: + return ESCAPE_ERROR; + } + case 2: + switch (cp[0]) { + case 'B': + switch (cp[1]) { + case 'I': + return ESCAPE_FONTBI; + default: + return ESCAPE_ERROR; + } + case 'C': + switch (cp[1]) { + case 'B': + return ESCAPE_FONTBOLD; + case 'I': + return ESCAPE_FONTITALIC; + case 'R': + case 'W': + return ESCAPE_FONTCW; + default: + return ESCAPE_ERROR; + } + default: + return ESCAPE_ERROR; + } + default: + return ESCAPE_ERROR; + } +} + +enum mandoc_esc mandoc_escape(const char **end, const char **start, int *sz) { const char *local_start; - int local_sz; + int local_sz, c, i; char term; enum mandoc_esc gly; @@ -55,6 +109,14 @@ mandoc_escape(const char **end, const char **start, in sz = &local_sz; /* + * Treat "\E" just like "\"; + * it only makes a difference in copy mode. + */ + + if (**end == 'E') + ++*end; + + /* * Beyond the backslash, at least one input character * is part of the escape sequence. With one exception * (see below), that character won't be returned. @@ -76,6 +138,10 @@ mandoc_escape(const char **end, const char **start, in *sz = 2; break; case '[': + if (**start == ' ') { + ++*end; + return ESCAPE_ERROR; + } gly = ESCAPE_SPECIAL; term = ']'; break; @@ -90,11 +156,28 @@ mandoc_escape(const char **end, const char **start, in /* * Escapes taking no arguments at all. */ - case 'd': - case 'u': + case '!': + case '?': + return ESCAPE_UNSUPP; + case '%': + case '&': + case ')': case ',': case '/': + case '^': + case 'a': + case 'd': + case 'r': + case 't': + case 'u': + case '{': + case '|': + case '}': return ESCAPE_IGNORE; + case 'c': + return ESCAPE_NOSPACE; + case 'p': + return ESCAPE_BREAK; /* * The \z escape is supposed to output the following @@ -110,32 +193,57 @@ mandoc_escape(const char **end, const char **start, in * 'X' is the trigger. These have opaque sub-strings. */ case 'F': + case 'f': case 'g': case 'k': case 'M': case 'm': case 'n': + case 'O': case 'V': case 'Y': - gly = ESCAPE_IGNORE; - /* FALLTHROUGH */ - case 'f': - if (ESCAPE_ERROR == gly) - gly = ESCAPE_FONT; + gly = (*start)[-1] == 'f' ? ESCAPE_FONT : ESCAPE_IGNORE; switch (**start) { case '(': + if ((*start)[-1] == 'O') + gly = ESCAPE_ERROR; *start = ++*end; *sz = 2; break; case '[': + if ((*start)[-1] == 'O') + gly = (*start)[1] == '5' ? + ESCAPE_UNSUPP : ESCAPE_ERROR; *start = ++*end; term = ']'; break; default: + if ((*start)[-1] == 'O') { + switch (**start) { + case '0': + gly = ESCAPE_UNSUPP; + break; + case '1': + case '2': + case '3': + case '4': + break; + default: + gly = ESCAPE_ERROR; + break; + } + } *sz = 1; break; } break; + case '*': + if (strncmp(*start, "(.T", 3) != 0) + abort(); + gly = ESCAPE_DEVICE; + *start = ++*end; + *sz = 2; + break; /* * These escapes are of the form \X'Y', where 'X' is the trigger @@ -175,7 +283,17 @@ mandoc_escape(const char **end, const char **start, in ++*end; return ESCAPE_ERROR; } - gly = ESCAPE_IGNORE; + switch ((*start)[-1]) { + case 'h': + gly = ESCAPE_HORIZ; + break; + case 'l': + gly = ESCAPE_HLINE; + break; + default: + gly = ESCAPE_IGNORE; + break; + } term = **start; *start = ++*end; break; @@ -237,18 +355,29 @@ mandoc_escape(const char **end, const char **start, in break; /* - * Anything else is assumed to be a glyph. - * In this case, pass back the character after the backslash. + * Several special characters can be encoded as + * one-byte escape sequences without using \[]. */ - default: + case ' ': + case '\'': + case '-': + case '.': + case '0': + case ':': + case '_': + case '`': + case 'e': + case '~': gly = ESCAPE_SPECIAL; + /* FALLTHROUGH */ + default: + if (gly == ESCAPE_ERROR) + gly = ESCAPE_UNDEF; *start = --*end; *sz = 1; break; } - assert(ESCAPE_ERROR != gly); - /* * Read up to the terminating character, * paying attention to nested escapes. @@ -271,6 +400,15 @@ mandoc_escape(const char **end, const char **start, in } } *sz = (*end)++ - *start; + + /* + * The file chars.c only provides one common list + * of character names, but \[-] == \- is the only + * one of the characters with one-byte names that + * allows enclosing the name in brackets. + */ + if (gly == ESCAPE_SPECIAL && *sz == 1 && **start != '-') + return ESCAPE_ERROR; } else { assert(*sz > 0); if ((size_t)*sz > strlen(*start)) @@ -282,43 +420,25 @@ mandoc_escape(const char **end, const char **start, in switch (gly) { case ESCAPE_FONT: - if (2 == *sz) { - if ('C' == **start) { - /* - * Treat constant-width font modes - * just like regular font modes. - */ - (*start)++; - (*sz)--; - } else { - if ('B' == (*start)[0] && 'I' == (*start)[1]) - gly = ESCAPE_FONTBI; + gly = mandoc_font(*start, *sz); + break; + case ESCAPE_SPECIAL: + if (**start == 'c') { + if (*sz < 6 || *sz > 7 || + strncmp(*start, "char", 4) != 0 || + (int)strspn(*start + 4, "0123456789") + 4 < *sz) break; - } - } else if (1 != *sz) + c = 0; + for (i = 4; i < *sz; i++) + c = 10 * c + ((*start)[i] - '0'); + if (c < 0x21 || (c > 0x7e && c < 0xa0) || c > 0xff) + break; + *start += 4; + *sz -= 4; + gly = ESCAPE_NUMBERED; break; - - switch (**start) { - case '3': - case 'B': - gly = ESCAPE_FONTBOLD; - break; - case '2': - case 'I': - gly = ESCAPE_FONTITALIC; - break; - case 'P': - gly = ESCAPE_FONTPREV; - break; - case '1': - case 'R': - gly = ESCAPE_FONTROMAN; - break; } - break; - case ESCAPE_SPECIAL: - if (1 == *sz && 'c' == **start) - gly = ESCAPE_NOSPACE; + /* * Unicode escapes are defined in groff as \[u0000] * to \[u10FFFF], where the contained value must be @@ -355,7 +475,7 @@ mandoc_escape(const char **end, const char **start, in * or to the NUL byte terminating the argument line. */ char * -mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos) +mandoc_getarg(char **cpp, int ln, int *pos) { char *start, *cp; int quoted, pairs, white; @@ -385,6 +505,7 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln * backslashes and backslash-t to literal tabs. */ switch (cp[1]) { + case 'a': case 't': cp[0] = '\t'; /* FALLTHROUGH */ @@ -421,7 +542,7 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln /* Quoted argument without a closing quote. */ if (1 == quoted) - mandoc_msg(MANDOCERR_ARG_QUOTE, parse, ln, *pos, NULL); + mandoc_msg(MANDOCERR_ARG_QUOTE, ln, *pos, NULL); /* NUL-terminate this argument and move to the next one. */ if (pairs) @@ -435,7 +556,7 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln *cpp = cp; if ('\0' == *cp && (white || ' ' == cp[-1])) - mandoc_msg(MANDOCERR_SPACE_EOL, parse, ln, *pos, NULL); + mandoc_msg(MANDOCERR_SPACE_EOL, ln, *pos, NULL); return start; } @@ -508,27 +629,38 @@ fail: } char * -mandoc_normdate(struct mparse *parse, char *in, int ln, int pos) +mandoc_normdate(struct roff_man *man, char *in, int ln, int pos) { + char *cp; time_t t; /* No date specified: use today's date. */ if (in == NULL || *in == '\0' || strcmp(in, "$" "Mdocdate$") == 0) { - mandoc_msg(MANDOCERR_DATE_MISSING, parse, ln, pos, NULL); + mandoc_msg(MANDOCERR_DATE_MISSING, ln, pos, NULL); return time2a(time(NULL)); } /* Valid mdoc(7) date format. */ if (a2time(&t, "$" "Mdocdate: %b %d %Y $", in) || - a2time(&t, "%b %d, %Y", in)) - return time2a(t); + a2time(&t, "%b %d, %Y", in)) { + cp = time2a(t); + if (t > time(NULL) + 86400) + mandoc_msg(MANDOCERR_DATE_FUTURE, ln, pos, "%s", cp); + else if (*in != '$' && strcmp(in, cp) != 0) + mandoc_msg(MANDOCERR_DATE_NORM, ln, pos, "%s", cp); + return cp; + } - /* Do not warn about the legacy man(7) format. */ + /* In man(7), do not warn about the legacy format. */ - if ( ! a2time(&t, "%Y-%m-%d", in)) - mandoc_msg(MANDOCERR_DATE_BAD, parse, ln, pos, in); + if (a2time(&t, "%Y-%m-%d", in) == 0) + mandoc_msg(MANDOCERR_DATE_BAD, ln, pos, "%s", in); + else if (t > time(NULL) + 86400) + mandoc_msg(MANDOCERR_DATE_FUTURE, ln, pos, "%s", in); + else if (man->macroset == MACROSET_MDOC) + mandoc_msg(MANDOCERR_DATE_LEGACY, ln, pos, "Dd %s", in); /* Use any non-mdoc(7) date verbatim. */