=================================================================== RCS file: /cvs/mandoc/mandoc.c,v retrieving revision 1.65 retrieving revision 1.105 diff -u -p -r1.65 -r1.105 --- mandoc/mandoc.c 2012/05/31 22:38:16 1.65 +++ mandoc/mandoc.c 2018/08/10 22:12:44 1.105 @@ -1,7 +1,7 @@ -/* $Id: mandoc.c,v 1.65 2012/05/31 22:38:16 schwarze Exp $ */ +/* $Id: mandoc.c,v 1.105 2018/08/10 22:12:44 schwarze Exp $ */ /* - * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons - * Copyright (c) 2011, 2012 Ingo Schwarze + * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons + * 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 @@ -15,9 +15,7 @@ * 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 @@ -30,11 +28,11 @@ #include #include +#include "mandoc_aux.h" #include "mandoc.h" +#include "roff.h" #include "libmandoc.h" -#define DATESIZE 32 - static int a2time(time_t *, const char *, const char *); static char *time2a(time_t); @@ -43,9 +41,9 @@ 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; + enum mandoc_esc gly; /* * When the caller doesn't provide return storage, @@ -74,69 +72,65 @@ mandoc_escape(const char **end, const char **start, in * these, but each eventually returns a substring of the glyph * name. */ - case ('('): + case '(': gly = ESCAPE_SPECIAL; *sz = 2; break; - case ('['): + case '[': gly = ESCAPE_SPECIAL; - /* - * Unicode escapes are defined in groff as \[uXXXX] to - * \[u10FFFF], where the contained value must be a valid - * Unicode codepoint. Here, however, only check whether - * it's not a zero-width escape. - */ - if ('u' == (*start)[0] && ']' != (*start)[1]) - gly = ESCAPE_UNICODE; term = ']'; break; - case ('C'): + case 'C': if ('\'' != **start) - return(ESCAPE_ERROR); - gly = ESCAPE_SPECIAL; + return ESCAPE_ERROR; *start = ++*end; + gly = ESCAPE_SPECIAL; term = '\''; break; /* + * Escapes taking no arguments at all. + */ + case 'd': + case 'u': + case ',': + case '/': + return ESCAPE_IGNORE; + case 'p': + return ESCAPE_BREAK; + + /* * The \z escape is supposed to output the following - * character without advancing the cursor position. + * character without advancing the cursor position. * Since we are mostly dealing with terminal mode, * let us just skip the next character. */ - case ('z'): - return(ESCAPE_SKIPCHAR); + case 'z': + return ESCAPE_SKIPCHAR; /* * Handle all triggers matching \X(xy, \Xx, and \X[xxxx], where * 'X' is the trigger. These have opaque sub-strings. */ - case ('F'): - /* FALLTHROUGH */ - case ('g'): - /* FALLTHROUGH */ - case ('k'): - /* FALLTHROUGH */ - case ('M'): - /* FALLTHROUGH */ - case ('m'): - /* FALLTHROUGH */ - case ('n'): - /* FALLTHROUGH */ - case ('V'): - /* FALLTHROUGH */ - case ('Y'): + case 'F': + case 'g': + case 'k': + case 'M': + case 'm': + case 'n': + case 'V': + case 'Y': gly = ESCAPE_IGNORE; /* FALLTHROUGH */ - case ('f'): + case 'f': if (ESCAPE_ERROR == gly) gly = ESCAPE_FONT; switch (**start) { - case ('('): + case '(': *start = ++*end; *sz = 2; break; - case ('['): + case '[': *start = ++*end; term = ']'; break; @@ -149,68 +143,67 @@ mandoc_escape(const char **end, const char **start, in /* * These escapes are of the form \X'Y', where 'X' is the trigger * and 'Y' is any string. These have opaque sub-strings. + * The \B and \w escapes are handled in roff.c, roff_res(). */ - case ('A'): - /* FALLTHROUGH */ - case ('b'): - /* FALLTHROUGH */ - case ('D'): - /* FALLTHROUGH */ - case ('o'): - /* FALLTHROUGH */ - case ('R'): - /* FALLTHROUGH */ - case ('X'): - /* FALLTHROUGH */ - case ('Z'): - if ('\'' != **start) - return(ESCAPE_ERROR); + case 'A': + case 'b': + case 'D': + case 'R': + case 'X': + case 'Z': gly = ESCAPE_IGNORE; + /* FALLTHROUGH */ + case 'o': + if (**start == '\0') + return ESCAPE_ERROR; + if (gly == ESCAPE_ERROR) + gly = ESCAPE_OVERSTRIKE; + term = **start; *start = ++*end; - term = '\''; break; /* * These escapes are of the form \X'N', where 'X' is the trigger * and 'N' resolves to a numerical expression. */ - case ('B'): - /* FALLTHROUGH */ - case ('h'): - /* FALLTHROUGH */ - case ('H'): - /* FALLTHROUGH */ - case ('L'): - /* FALLTHROUGH */ - case ('l'): - gly = ESCAPE_NUMBERED; - /* FALLTHROUGH */ - case ('S'): - /* FALLTHROUGH */ - case ('v'): - /* FALLTHROUGH */ - case ('w'): - /* FALLTHROUGH */ - case ('x'): - if ('\'' != **start) - return(ESCAPE_ERROR); - if (ESCAPE_ERROR == gly) + case 'h': + case 'H': + case 'L': + case 'l': + case 'S': + case 'v': + case 'x': + if (strchr(" %&()*+-./0123456789:<=>", **start)) { + if ('\0' != **start) + ++*end; + return ESCAPE_ERROR; + } + switch ((*start)[-1]) { + case 'h': + gly = ESCAPE_HORIZ; + break; + case 'l': + gly = ESCAPE_HLINE; + break; + default: gly = ESCAPE_IGNORE; + break; + } + term = **start; *start = ++*end; - term = '\''; break; /* * Special handling for the numbered character escape. * XXX Do any other escapes need similar handling? */ - case ('N'): + case 'N': if ('\0' == **start) - return(ESCAPE_ERROR); + return ESCAPE_ERROR; (*end)++; if (isdigit((unsigned char)**start)) { *sz = 1; - return(ESCAPE_IGNORE); + return ESCAPE_IGNORE; } (*start)++; while (isdigit((unsigned char)**end)) @@ -218,31 +211,37 @@ mandoc_escape(const char **end, const char **start, in *sz = *end - *start; if ('\0' != **end) (*end)++; - return(ESCAPE_NUMBERED); + return ESCAPE_NUMBERED; - /* + /* * Sizes get a special category of their own. */ - case ('s'): + case 's': gly = ESCAPE_IGNORE; /* See +/- counts as a sign. */ if ('+' == **end || '-' == **end || ASCII_HYPH == **end) - (*end)++; + *start = ++*end; switch (**end) { - case ('('): + case '(': *start = ++*end; *sz = 2; break; - case ('['): + case '[': *start = ++*end; term = ']'; break; - case ('\''): + case '\'': *start = ++*end; term = '\''; break; + case '3': + case '2': + case '1': + *sz = (*end)[-1] == 's' && + isdigit((unsigned char)(*end)[1]) ? 2 : 1; + break; default: *sz = 1; break; @@ -271,13 +270,13 @@ mandoc_escape(const char **end, const char **start, in if ('\0' != term) { while (**end != term) { switch (**end) { - case ('\0'): - return(ESCAPE_ERROR); - case ('\\'): + case '\0': + return ESCAPE_ERROR; + case '\\': (*end)++; if (ESCAPE_ERROR == mandoc_escape(end, NULL, NULL)) - return(ESCAPE_ERROR); + return ESCAPE_ERROR; break; default: (*end)++; @@ -288,132 +287,103 @@ mandoc_escape(const char **end, const char **start, in } else { assert(*sz > 0); if ((size_t)*sz > strlen(*start)) - return(ESCAPE_ERROR); + return ESCAPE_ERROR; *end += *sz; } /* Run post-processors. */ switch (gly) { - case (ESCAPE_FONT): - /* - * Pretend that the constant-width font modes are the - * same as the regular font modes. - */ - if (2 == *sz && 'C' == **start) { - (*start)++; - (*sz)--; + 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; + break; + } } else if (1 != *sz) break; switch (**start) { - case ('3'): - /* FALLTHROUGH */ - case ('B'): + case '3': + case 'B': gly = ESCAPE_FONTBOLD; break; - case ('2'): - /* FALLTHROUGH */ - case ('I'): + case '2': + case 'I': gly = ESCAPE_FONTITALIC; break; - case ('P'): + case 'P': gly = ESCAPE_FONTPREV; break; - case ('1'): - /* FALLTHROUGH */ - case ('R'): + case '1': + case 'R': gly = ESCAPE_FONTROMAN; break; } break; - case (ESCAPE_SPECIAL): - if (1 == *sz && 'c' == **start) - gly = ESCAPE_NOSPACE; + case ESCAPE_SPECIAL: + if (**start == 'c') { + if (*sz == 1) { + gly = ESCAPE_NOSPACE; + break; + } + if (*sz < 6 || *sz > 7 || + strncmp(*start, "char", 4) != 0 || + (int)strspn(*start + 4, "0123456789") + 4 < *sz) + break; + 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; + } + + /* + * Unicode escapes are defined in groff as \[u0000] + * to \[u10FFFF], where the contained value must be + * a valid Unicode codepoint. Here, however, only + * check the length and range. + */ + if (**start != 'u' || *sz < 5 || *sz > 7) + break; + if (*sz == 7 && ((*start)[1] != '1' || (*start)[2] != '0')) + break; + if (*sz == 6 && (*start)[1] == '0') + break; + if (*sz == 5 && (*start)[1] == 'D' && + strchr("89ABCDEF", (*start)[2]) != NULL) + break; + if ((int)strspn(*start + 1, "0123456789ABCDEFabcdef") + + 1 == *sz) + gly = ESCAPE_UNICODE; break; default: break; } - return(gly); + return gly; } -void * -mandoc_calloc(size_t num, size_t size) -{ - void *ptr; - - ptr = calloc(num, size); - if (NULL == ptr) { - perror(NULL); - exit((int)MANDOCLEVEL_SYSERR); - } - - return(ptr); -} - - -void * -mandoc_malloc(size_t size) -{ - void *ptr; - - ptr = malloc(size); - if (NULL == ptr) { - perror(NULL); - exit((int)MANDOCLEVEL_SYSERR); - } - - return(ptr); -} - - -void * -mandoc_realloc(void *ptr, size_t size) -{ - - ptr = realloc(ptr, size); - if (NULL == ptr) { - perror(NULL); - exit((int)MANDOCLEVEL_SYSERR); - } - - return(ptr); -} - -char * -mandoc_strndup(const char *ptr, size_t sz) -{ - char *p; - - p = mandoc_malloc(sz + 1); - memcpy(p, ptr, sz); - p[(int)sz] = '\0'; - return(p); -} - -char * -mandoc_strdup(const char *ptr) -{ - char *p; - - p = strdup(ptr); - if (NULL == p) { - perror(NULL); - exit((int)MANDOCLEVEL_SYSERR); - } - - return(p); -} - /* * Parse a quoted or unquoted roff-style request or macro argument. * Return a pointer to the parsed argument, which is either the original * pointer or advanced by one byte in case the argument is quoted. - * Null-terminate the argument in place. + * NUL-terminate the argument in place. * Collapse pairs of quotes inside quoted arguments. * Advance the argument pointer to the next argument, - * or to the null byte terminating the argument line. + * or to the NUL byte terminating the argument line. */ char * mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos) @@ -427,22 +397,40 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln if ('"' == *start) { quoted = 1; start++; - } + } pairs = 0; white = 0; for (cp = start; '\0' != *cp; cp++) { - /* Move left after quoted quotes and escaped backslashes. */ + + /* + * Move the following text left + * after quoted quotes and after "\\" and "\t". + */ if (pairs) cp[-pairs] = cp[0]; + if ('\\' == cp[0]) { - if ('\\' == cp[1]) { - /* Poor man's copy mode. */ + /* + * In copy mode, translate double to single + * backslashes and backslash-t to literal tabs. + */ + switch (cp[1]) { + case 't': + cp[0] = '\t'; + /* FALLTHROUGH */ + case '\\': pairs++; cp++; - } else if (0 == quoted && ' ' == cp[1]) + break; + case ' ': /* Skip escaped blanks. */ - cp++; + if (0 == quoted) + cp++; + break; + default: + break; + } } else if (0 == quoted) { if (' ' == cp[0]) { /* Unescaped blanks end unquoted args. */ @@ -464,9 +452,9 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln /* Quoted argument without a closing quote. */ if (1 == quoted) - mandoc_msg(MANDOCERR_BADQUOTE, parse, ln, *pos, NULL); + mandoc_msg(MANDOCERR_ARG_QUOTE, parse, ln, *pos, NULL); - /* Null-terminate this argument and move to the next one. */ + /* NUL-terminate this argument and move to the next one. */ if (pairs) cp[-pairs] = '\0'; if ('\0' != *cp) { @@ -478,9 +466,9 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln *cpp = cp; if ('\0' == *cp && (white || ' ' == cp[-1])) - mandoc_msg(MANDOCERR_EOLNSPACE, parse, ln, *pos, NULL); + mandoc_msg(MANDOCERR_SPACE_EOL, parse, ln, *pos, NULL); - return(start); + return start; } static int @@ -492,15 +480,15 @@ a2time(time_t *t, const char *fmt, const char *p) memset(&tm, 0, sizeof(struct tm)); pp = NULL; -#ifdef HAVE_STRPTIME +#if HAVE_STRPTIME pp = strptime(p, fmt, &tm); #endif if (NULL != pp && '\0' == *pp) { *t = mktime(&tm); - return(1); + return 1; } - return(0); + return 0; } static char * @@ -512,6 +500,8 @@ time2a(time_t t) int isz; tm = localtime(&t); + if (tm == NULL) + return NULL; /* * Reserve space: @@ -519,55 +509,85 @@ time2a(time_t t) * up to 2 characters for the day + comma + blank * 4 characters for the year and a terminating '\0' */ + p = buf = mandoc_malloc(10 + 4 + 4 + 1); - if (0 == (ssz = strftime(p, 10 + 1, "%B ", tm))) + if ((ssz = strftime(p, 10 + 1, "%B ", tm)) == 0) goto fail; p += (int)ssz; - if (-1 == (isz = snprintf(p, 4 + 1, "%d, ", tm->tm_mday))) + /* + * The output format is just "%d" here, not "%2d" or "%02d". + * That's also the reason why we can't just format the + * date as a whole with "%B %e, %Y" or "%B %d, %Y". + * Besides, the present approach is less prone to buffer + * overflows, in case anybody should ever introduce the bug + * of looking at LC_TIME. + */ + + if ((isz = snprintf(p, 4 + 1, "%d, ", tm->tm_mday)) == -1) goto fail; p += isz; - if (0 == strftime(p, 4 + 1, "%Y", tm)) + if (strftime(p, 4 + 1, "%Y", tm) == 0) goto fail; - return(buf); + return buf; fail: free(buf); - return(NULL); + return NULL; } 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 *out; + char *cp; time_t t; - if (NULL == in || '\0' == *in || - 0 == strcmp(in, "$" "Mdocdate$")) { - mandoc_msg(MANDOCERR_NODATE, parse, ln, pos, NULL); - time(&t); + /* No date specified: use today's date. */ + + if (in == NULL || *in == '\0' || strcmp(in, "$" "Mdocdate$") == 0) { + mandoc_msg(MANDOCERR_DATE_MISSING, man->parse, ln, pos, NULL); + return time2a(time(NULL)); } - else if (a2time(&t, "%Y-%m-%d", in)) - t = 0; - else if (!a2time(&t, "$" "Mdocdate: %b %d %Y $", in) && - !a2time(&t, "%b %d, %Y", in)) { - mandoc_msg(MANDOCERR_BADDATE, parse, ln, pos, NULL); - t = 0; + + /* Valid mdoc(7) date format. */ + + if (a2time(&t, "$" "Mdocdate: %b %d %Y $", in) || + a2time(&t, "%b %d, %Y", in)) { + cp = time2a(t); + if (t > time(NULL) + 86400) + mandoc_msg(MANDOCERR_DATE_FUTURE, man->parse, + ln, pos, cp); + else if (*in != '$' && strcmp(in, cp) != 0) + mandoc_msg(MANDOCERR_DATE_NORM, man->parse, + ln, pos, cp); + return cp; } - out = t ? time2a(t) : NULL; - return(out ? out : mandoc_strdup(in)); + + /* In man(7), do not warn about the legacy format. */ + + if (a2time(&t, "%Y-%m-%d", in) == 0) + mandoc_msg(MANDOCERR_DATE_BAD, man->parse, ln, pos, in); + else if (t > time(NULL) + 86400) + mandoc_msg(MANDOCERR_DATE_FUTURE, man->parse, ln, pos, in); + else if (man->macroset == MACROSET_MDOC) + mandoc_vmsg(MANDOCERR_DATE_LEGACY, man->parse, + ln, pos, "Dd %s", in); + + /* Use any non-mdoc(7) date verbatim. */ + + return mandoc_strdup(in); } int -mandoc_eos(const char *p, size_t sz, int enclosed) +mandoc_eos(const char *p, size_t sz) { - const char *q; - int found; + const char *q; + int enclosed, found; if (0 == sz) - return(0); + return 0; /* * End-of-sentence recognition must include situations where @@ -575,61 +595,31 @@ mandoc_eos(const char *p, size_t sz, int enclosed) * propagate outward. */ - found = 0; + enclosed = found = 0; for (q = p + (int)sz - 1; q >= p; q--) { switch (*q) { - case ('\"'): - /* FALLTHROUGH */ - case ('\''): - /* FALLTHROUGH */ - case (']'): - /* FALLTHROUGH */ - case (')'): + case '\"': + case '\'': + case ']': + case ')': if (0 == found) enclosed = 1; break; - case ('.'): - /* FALLTHROUGH */ - case ('!'): - /* FALLTHROUGH */ - case ('?'): + case '.': + case '!': + case '?': found = 1; break; default: - return(found && (!enclosed || isalnum((unsigned char)*q))); + return found && + (!enclosed || isalnum((unsigned char)*q)); } } - return(found && !enclosed); + return found && !enclosed; } /* - * Find out whether a line is a macro line or not. If it is, adjust the - * current position and return one; if it isn't, return zero and don't - * change the current position. - */ -int -mandoc_getcontrol(const char *cp, int *ppos) -{ - int pos; - - pos = *ppos; - - if ('\\' == cp[pos] && '.' == cp[pos + 1]) - pos += 2; - else if ('.' == cp[pos] || '\'' == cp[pos]) - pos++; - else - return(0); - - while (' ' == cp[pos] || '\t' == cp[pos]) - pos++; - - *ppos = pos; - return(1); -} - -/* * Convert a string to a long that may not be <0. * If the string is invalid, or is less than 0, return -1. */ @@ -641,7 +631,7 @@ mandoc_strntoi(const char *p, size_t sz, int base) long v; if (sz > 31) - return(-1); + return -1; memcpy(buf, p, sz); buf[(int)sz] = '\0'; @@ -650,12 +640,12 @@ mandoc_strntoi(const char *p, size_t sz, int base) v = strtol(buf, &ep, base); if (buf[0] == '\0' || *ep != '\0') - return(-1); + return -1; if (v > INT_MAX) v = INT_MAX; if (v < INT_MIN) v = INT_MIN; - return((int)v); + return (int)v; }