-/* $Id: mandoc.c,v 1.64 2012/05/31 22:34:06 schwarze Exp $ */
+/* $Id: mandoc.c,v 1.80 2014/06/20 17:24:00 schwarze Exp $ */
/*
* Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
- * Copyright (c) 2011, 2012 Ingo Schwarze <schwarze@openbsd.org>
+ * Copyright (c) 2011, 2012, 2013, 2014 Ingo Schwarze <schwarze@openbsd.org>
*
* Permission to use, copy, modify, and distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
#include <time.h>
#include "mandoc.h"
+#include "mandoc_aux.h"
#include "libmandoc.h"
#define DATESIZE 32
enum mandoc_esc
mandoc_escape(const char **end, const char **start, int *sz)
{
- char c, term;
- int i, rlim;
- const char *cp, *rstart;
- enum mandoc_esc gly;
-
- cp = *end;
- rstart = cp;
- if (start)
- *start = rstart;
- i = rlim = 0;
+ const char *local_start;
+ int local_sz;
+ char term;
+ enum mandoc_esc gly;
+
+ /*
+ * When the caller doesn't provide return storage,
+ * use local storage.
+ */
+
+ if (NULL == start)
+ start = &local_start;
+ if (NULL == sz)
+ sz = &local_sz;
+
+ /*
+ * 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.
+ */
+
gly = ESCAPE_ERROR;
+ *start = ++*end;
+ *sz = 0;
term = '\0';
- switch ((c = cp[i++])) {
+ switch ((*start)[-1]) {
/*
* First the glyphs. There are several different forms of
* these, but each eventually returns a substring of the glyph
* name.
*/
- case ('('):
+ case '(':
gly = ESCAPE_SPECIAL;
- rlim = 2;
+ *sz = 2;
break;
- case ('['):
+ case '[':
gly = ESCAPE_SPECIAL;
/*
* Unicode escapes are defined in groff as \[uXXXX] to
* Unicode codepoint. Here, however, only check whether
* it's not a zero-width escape.
*/
- if ('u' == cp[i] && ']' != cp[i + 1])
+ if ('u' == (*start)[0] && ']' != (*start)[1])
gly = ESCAPE_UNICODE;
term = ']';
break;
- case ('C'):
- if ('\'' != cp[i])
+ case 'C':
+ if ('\'' != **start)
return(ESCAPE_ERROR);
- gly = ESCAPE_SPECIAL;
+ *start = ++*end;
+ if ('u' == (*start)[0] && '\'' != (*start)[1])
+ gly = ESCAPE_UNICODE;
+ else
+ gly = ESCAPE_SPECIAL;
term = '\'';
break;
+ /*
+ * Escapes taking no arguments at all.
+ */
+ case 'd':
+ /* FALLTHROUGH */
+ case 'u':
+ return(ESCAPE_IGNORE);
+
/*
* 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'):
- (*end)++;
+ 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'):
+ case 'F':
/* FALLTHROUGH */
- case ('g'):
+ case 'g':
/* FALLTHROUGH */
- case ('k'):
+ case 'k':
/* FALLTHROUGH */
- case ('M'):
+ case 'M':
/* FALLTHROUGH */
- case ('m'):
+ case 'm':
/* FALLTHROUGH */
- case ('n'):
+ case 'n':
/* FALLTHROUGH */
- case ('V'):
+ case 'V':
/* FALLTHROUGH */
- case ('Y'):
+ case 'Y':
gly = ESCAPE_IGNORE;
/* FALLTHROUGH */
- case ('f'):
+ case 'f':
if (ESCAPE_ERROR == gly)
gly = ESCAPE_FONT;
-
- rstart= &cp[i];
- if (start)
- *start = rstart;
-
- switch (cp[i++]) {
- case ('('):
- rlim = 2;
+ switch (**start) {
+ case '(':
+ *start = ++*end;
+ *sz = 2;
break;
- case ('['):
+ case '[':
+ *start = ++*end;
term = ']';
break;
default:
- rlim = 1;
- i--;
+ *sz = 1;
break;
}
break;
/*
* 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'):
+ case 'A':
/* FALLTHROUGH */
- case ('b'):
+ case 'b':
/* FALLTHROUGH */
- case ('D'):
+ case 'D':
/* FALLTHROUGH */
- case ('o'):
+ case 'o':
/* FALLTHROUGH */
- case ('R'):
+ case 'R':
/* FALLTHROUGH */
- case ('X'):
+ case 'X':
/* FALLTHROUGH */
- case ('Z'):
- if ('\'' != cp[i++])
+ case 'Z':
+ if ('\0' == **start)
return(ESCAPE_ERROR);
gly = ESCAPE_IGNORE;
- term = '\'';
+ term = **start;
+ *start = ++*end;
break;
/*
* These escapes are of the form \X'N', where 'X' is the trigger
* and 'N' resolves to a numerical expression.
*/
- case ('B'):
+ case 'h':
/* FALLTHROUGH */
- case ('h'):
+ case 'H':
/* FALLTHROUGH */
- case ('H'):
+ case 'L':
/* FALLTHROUGH */
- case ('L'):
+ case 'l':
/* FALLTHROUGH */
- case ('l'):
- gly = ESCAPE_NUMBERED;
+ case 'S':
/* FALLTHROUGH */
- case ('S'):
+ case 'v':
/* FALLTHROUGH */
- case ('v'):
- /* FALLTHROUGH */
- case ('w'):
- /* FALLTHROUGH */
- case ('x'):
- if (ESCAPE_ERROR == gly)
- gly = ESCAPE_IGNORE;
- if ('\'' != cp[i++])
+ case 'x':
+ if (strchr("\0 %&()*+-./0123456789:<=>", **start))
return(ESCAPE_ERROR);
- term = '\'';
+ gly = ESCAPE_IGNORE;
+ term = **start;
+ *start = ++*end;
break;
/*
* Special handling for the numbered character escape.
* XXX Do any other escapes need similar handling?
*/
- case ('N'):
- if ('\0' == cp[i])
+ case 'N':
+ if ('\0' == **start)
return(ESCAPE_ERROR);
- *end = &cp[++i];
- if (isdigit((unsigned char)cp[i-1]))
+ (*end)++;
+ if (isdigit((unsigned char)**start)) {
+ *sz = 1;
return(ESCAPE_IGNORE);
+ }
+ (*start)++;
while (isdigit((unsigned char)**end))
(*end)++;
- if (start)
- *start = &cp[i];
- if (sz)
- *sz = *end - &cp[i];
+ *sz = *end - *start;
if ('\0' != **end)
(*end)++;
return(ESCAPE_NUMBERED);
- /*
+ /*
* Sizes get a special category of their own.
*/
- case ('s'):
+ case 's':
gly = ESCAPE_IGNORE;
- rstart = &cp[i];
- if (start)
- *start = rstart;
-
/* See +/- counts as a sign. */
- c = cp[i];
- if ('+' == c || '-' == c || ASCII_HYPH == c)
- ++i;
+ if ('+' == **end || '-' == **end || ASCII_HYPH == **end)
+ (*end)++;
- switch (cp[i++]) {
- case ('('):
- rlim = 2;
+ switch (**end) {
+ case '(':
+ *start = ++*end;
+ *sz = 2;
break;
- case ('['):
+ case '[':
+ *start = ++*end;
term = ']';
break;
- case ('\''):
+ case '\'':
+ *start = ++*end;
term = '\'';
break;
default:
- rlim = 1;
- i--;
+ *sz = 1;
break;
}
- /* See +/- counts as a sign. */
- c = cp[i];
- if ('+' == c || '-' == c || ASCII_HYPH == c)
- ++i;
-
break;
/*
* Anything else is assumed to be a glyph.
+ * In this case, pass back the character after the backslash.
*/
default:
gly = ESCAPE_SPECIAL;
- rlim = 1;
- i--;
+ *start = --*end;
+ *sz = 1;
break;
}
assert(ESCAPE_ERROR != gly);
- *end = rstart = &cp[i];
- if (start)
- *start = rstart;
-
/*
* Read up to the terminating character,
* paying attention to nested escapes.
if ('\0' != term) {
while (**end != term) {
switch (**end) {
- case ('\0'):
+ case '\0':
return(ESCAPE_ERROR);
- case ('\\'):
+ case '\\':
(*end)++;
if (ESCAPE_ERROR ==
mandoc_escape(end, NULL, NULL))
break;
}
}
- rlim = (*end)++ - rstart;
+ *sz = (*end)++ - *start;
} else {
- assert(rlim > 0);
- if ((size_t)rlim > strlen(rstart))
+ assert(*sz > 0);
+ if ((size_t)*sz > strlen(*start))
return(ESCAPE_ERROR);
- *end += rlim;
+ *end += *sz;
}
- if (sz)
- *sz = rlim;
/* 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 == rlim && 'C' == *rstart)
- rstart++;
- else if (1 != rlim)
+ 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 (*rstart) {
- case ('3'):
+ switch (**start) {
+ case '3':
/* FALLTHROUGH */
- case ('B'):
+ case 'B':
gly = ESCAPE_FONTBOLD;
break;
- case ('2'):
+ case '2':
/* FALLTHROUGH */
- case ('I'):
+ case 'I':
gly = ESCAPE_FONTITALIC;
break;
- case ('P'):
+ case 'P':
gly = ESCAPE_FONTPREV;
break;
- case ('1'):
+ case '1':
/* FALLTHROUGH */
- case ('R'):
+ case 'R':
gly = ESCAPE_FONTROMAN;
break;
}
break;
- case (ESCAPE_SPECIAL):
- if (1 != rlim)
- break;
- if ('c' == *rstart)
+ case ESCAPE_SPECIAL:
+ if (1 == *sz && 'c' == **start)
gly = ESCAPE_NOSPACE;
break;
default:
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)
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. */
if (1 == quoted)
mandoc_msg(MANDOCERR_BADQUOTE, 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) {
if (NULL == in || '\0' == *in ||
0 == strcmp(in, "$" "Mdocdate$")) {
- mandoc_msg(MANDOCERR_NODATE, parse, ln, pos, NULL);
+ mandoc_msg(MANDOCERR_DATE_MISSING, parse, ln, pos, NULL);
time(&t);
}
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);
+ mandoc_msg(MANDOCERR_DATE_BAD, parse, ln, pos, NULL);
t = 0;
}
out = t ? time2a(t) : NULL;
}
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);
* propagate outward.
*/
- found = 0;
+ enclosed = found = 0;
for (q = p + (int)sz - 1; q >= p; q--) {
switch (*q) {
- case ('\"'):
+ case '\"':
/* FALLTHROUGH */
- case ('\''):
+ case '\'':
/* FALLTHROUGH */
- case (']'):
+ case ']':
/* FALLTHROUGH */
- case (')'):
+ case ')':
if (0 == found)
enclosed = 1;
break;
- case ('.'):
+ case '.':
/* FALLTHROUGH */
- case ('!'):
+ case '!':
/* FALLTHROUGH */
- case ('?'):
+ case '?':
found = 1;
break;
default:
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.
git.ckatri.com / mandoc.git / blobdiff