=================================================================== RCS file: /cvs/mandoc/mandoc.c,v retrieving revision 1.105 retrieving revision 1.114 diff -u -p -r1.105 -r1.114 --- mandoc/mandoc.c 2018/08/10 22:12:44 1.105 +++ mandoc/mandoc.c 2018/12/30 00:49:55 1.114 @@ -1,4 +1,4 @@ -/* $Id: mandoc.c,v 1.105 2018/08/10 22:12:44 schwarze Exp $ */ +/* $Id: mandoc.c,v 1.114 2018/12/30 00:49:55 schwarze Exp $ */ /* * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons * Copyright (c) 2011-2015, 2017, 2018 Ingo Schwarze @@ -32,12 +32,66 @@ #include "mandoc.h" #include "roff.h" #include "libmandoc.h" +#include "roff_int.h" static int a2time(time_t *, const char *, const char *); 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; @@ -56,6 +110,14 @@ 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. @@ -77,6 +139,10 @@ mandoc_escape(const char **end, const char **start, in *sz = 2; break; case '[': + if (**start == ' ') { + ++*end; + return ESCAPE_ERROR; + } gly = ESCAPE_SPECIAL; term = ']'; break; @@ -91,11 +157,26 @@ 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; @@ -113,32 +194,57 @@ 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 @@ -250,18 +356,29 @@ 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. @@ -284,6 +401,15 @@ 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)) @@ -295,46 +421,10 @@ 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; - break; - } - } else if (1 != *sz) - 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; - } + gly = mandoc_font(*start, *sz); break; 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) @@ -376,101 +466,6 @@ mandoc_escape(const char **end, const char **start, in return gly; } -/* - * 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. - * NUL-terminate the argument in place. - * Collapse pairs of quotes inside quoted arguments. - * Advance the argument pointer to the next argument, - * or to the NUL byte terminating the argument line. - */ -char * -mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos) -{ - char *start, *cp; - int quoted, pairs, white; - - /* Quoting can only start with a new word. */ - start = *cpp; - quoted = 0; - if ('"' == *start) { - quoted = 1; - start++; - } - - pairs = 0; - white = 0; - for (cp = start; '\0' != *cp; cp++) { - - /* - * Move the following text left - * after quoted quotes and after "\\" and "\t". - */ - if (pairs) - cp[-pairs] = cp[0]; - - if ('\\' == cp[0]) { - /* - * 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++; - break; - case ' ': - /* Skip escaped blanks. */ - if (0 == quoted) - cp++; - break; - default: - break; - } - } else if (0 == quoted) { - if (' ' == cp[0]) { - /* Unescaped blanks end unquoted args. */ - white = 1; - break; - } - } else if ('"' == cp[0]) { - if ('"' == cp[1]) { - /* Quoted quotes collapse. */ - pairs++; - cp++; - } else { - /* Unquoted quotes end quoted args. */ - quoted = 2; - break; - } - } - } - - /* Quoted argument without a closing quote. */ - if (1 == quoted) - mandoc_msg(MANDOCERR_ARG_QUOTE, parse, ln, *pos, NULL); - - /* NUL-terminate this argument and move to the next one. */ - if (pairs) - cp[-pairs] = '\0'; - if ('\0' != *cp) { - *cp++ = '\0'; - while (' ' == *cp) - cp++; - } - *pos += (int)(cp - start) + (quoted ? 1 : 0); - *cpp = cp; - - if ('\0' == *cp && (white || ' ' == cp[-1])) - mandoc_msg(MANDOCERR_SPACE_EOL, parse, ln, *pos, NULL); - - return start; -} - static int a2time(time_t *t, const char *fmt, const char *p) { @@ -547,7 +542,7 @@ mandoc_normdate(struct roff_man *man, char *in, int ln /* 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); + mandoc_msg(MANDOCERR_DATE_MISSING, ln, pos, NULL); return time2a(time(NULL)); } @@ -557,23 +552,20 @@ mandoc_normdate(struct roff_man *man, char *in, int ln a2time(&t, "%b %d, %Y", in)) { cp = time2a(t); if (t > time(NULL) + 86400) - mandoc_msg(MANDOCERR_DATE_FUTURE, man->parse, - ln, pos, cp); + mandoc_msg(MANDOCERR_DATE_FUTURE, ln, pos, "%s", cp); else if (*in != '$' && strcmp(in, cp) != 0) - mandoc_msg(MANDOCERR_DATE_NORM, man->parse, - ln, pos, cp); + mandoc_msg(MANDOCERR_DATE_NORM, ln, pos, "%s", cp); return cp; } /* 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); + mandoc_msg(MANDOCERR_DATE_BAD, ln, pos, "%s", 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); + mandoc_msg(MANDOCERR_DATE_FUTURE, ln, pos, "%s", in); + else if (man->meta.macroset == MACROSET_MDOC) + mandoc_msg(MANDOCERR_DATE_LEGACY, ln, pos, "Dd %s", in); /* Use any non-mdoc(7) date verbatim. */