===================================================================
RCS file: /cvs/mandoc/mandoc.c,v
retrieving revision 1.48
retrieving revision 1.57
diff -u -p -r1.48 -r1.57
--- mandoc/mandoc.c	2011/04/19 16:38:48	1.48
+++ mandoc/mandoc.c	2011/07/27 07:06:29	1.57
@@ -1,4 +1,4 @@
-/*	$Id: mandoc.c,v 1.48 2011/04/19 16:38:48 kristaps Exp $ */
+/*	$Id: mandoc.c,v 1.57 2011/07/27 07:06:29 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2011 Ingo Schwarze <schwarze@openbsd.org>
@@ -23,6 +23,8 @@
 
 #include <assert.h>
 #include <ctype.h>
+#include <errno.h>
+#include <limits.h>
 #include <stdlib.h>
 #include <stdio.h>
 #include <string.h>
@@ -123,6 +125,14 @@ mandoc_escape(const char **end, const char **start, in
 		break;
 	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' == cp[i] && ']' != cp[i + 1])
+			gly = ESCAPE_UNICODE;
 		term = ']';
 		break;
 	case ('C'):
@@ -154,10 +164,6 @@ mandoc_escape(const char **end, const char **start, in
 		if (ESCAPE_ERROR == gly)
 			gly = ESCAPE_IGNORE;
 		/* FALLTHROUGH */
-	case ('*'):
-		if (ESCAPE_ERROR == gly)
-			gly = ESCAPE_PREDEF;
-		/* FALLTHROUGH */
 	case ('f'):
 		if (ESCAPE_ERROR == gly)
 			gly = ESCAPE_FONT;
@@ -426,7 +432,17 @@ mandoc_realloc(void *ptr, size_t size)
 	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)
 {
@@ -526,7 +542,10 @@ a2time(time_t *t, const char *fmt, const char *p)
 
 	memset(&tm, 0, sizeof(struct tm));
 
+	pp = NULL;
+#ifdef	HAVE_STRPTIME
 	pp = strptime(p, fmt, &tm);
+#endif
 	if (NULL != pp && '\0' == *pp) {
 		*t = mktime(&tm);
 		return(1);
@@ -538,12 +557,12 @@ a2time(time_t *t, const char *fmt, const char *p)
 static char *
 time2a(time_t t)
 {
-	struct tm	 tm;
+	struct tm	*tm;
 	char		*buf, *p;
 	size_t		 ssz;
 	int		 isz;
 
-	localtime_r(&t, &tm);
+	tm = localtime(&t);
 
 	/*
 	 * Reserve space:
@@ -553,15 +572,15 @@ time2a(time_t t)
 	 */
 	p = buf = mandoc_malloc(10 + 4 + 4 + 1);
 
-	if (0 == (ssz = strftime(p, 10 + 1, "%B ", &tm)))
+	if (0 == (ssz = strftime(p, 10 + 1, "%B ", tm)))
 		goto fail;
 	p += (int)ssz;
 
-	if (-1 == (isz = snprintf(p, 4 + 1, "%d, ", tm.tm_mday)))
+	if (-1 == (isz = snprintf(p, 4 + 1, "%d, ", tm->tm_mday)))
 		goto fail;
 	p += isz;
 
-	if (0 == strftime(p, 4 + 1, "%Y", &tm))
+	if (0 == strftime(p, 4 + 1, "%Y", tm))
 		goto fail;
 	return(buf);
 
@@ -603,7 +622,7 @@ mandoc_eos(const char *p, size_t sz, int enclosed)
 	/*
 	 * End-of-sentence recognition must include situations where
 	 * some symbols, such as `)', allow prior EOS punctuation to
-	 * propogate outward.
+	 * propagate outward.
 	 */
 
 	found = 0;
@@ -634,28 +653,39 @@ mandoc_eos(const char *p, size_t sz, int enclosed)
 	return(found && !enclosed);
 }
 
+/*
+ * Choose whether to break at a hyphenated character (identified by the
+ * ASCII_HYPH value in the input string).
+ */
 int
 mandoc_hyph(const char *start, const char *c)
 {
+	char		l, r;
 
-	/*
-	 * Choose whether to break at a hyphenated character.  We only
-	 * do this if it's free-standing within a word.
-	 */
+	l = *(c - 1);
+	r = *(c + 1);
 
 	/* Skip first/last character of buffer. */
-	if (c == start || '\0' == *(c + 1))
+	if (c == start || '\0' == r)
 		return(0);
+
+	/* Skip a number on either side of the hyphen. */
+	if (isdigit((unsigned char)r) || isdigit((unsigned char)l))
+		return(0);
+
 	/* Skip first/last character of word. */
-	if ('\t' == *(c + 1) || '\t' == *(c - 1))
+	if ('\t' == r || '\t' == l)
 		return(0);
-	if (' ' == *(c + 1) || ' ' == *(c - 1))
+
+	if (' ' == r || ' ' == l)
 		return(0);
+
 	/* Skip double invocations. */
-	if ('-' == *(c + 1) || '-' == *(c - 1))
+	if ('-' == r || '-' == l)
 		return(0);
+
 	/* Skip escapes. */
-	if ('\\' == *(c - 1))
+	if ('\\' == l)
 		return(0);
 
 	return(1);
@@ -685,4 +715,35 @@ mandoc_getcontrol(const char *cp, int *ppos)
 
 	*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.
+ */
+int
+mandoc_strntoi(const char *p, size_t sz, int base)
+{
+	char		 buf[32];
+	char		*ep;
+	long		 v;
+
+	if (sz > 31)
+		return(-1);
+
+	memcpy(buf, p, sz);
+	buf[(int)sz] = '\0';
+
+	errno = 0;
+	v = strtol(buf, &ep, base);
+
+	if (buf[0] == '\0' || *ep != '\0')
+		return(-1);
+
+	if (v > INT_MAX)
+		v = INT_MAX;
+	if (v < INT_MIN)
+		v = INT_MIN;
+
+	return((int)v);
 }