compatibility: reduce the assumption the builtin libc locale is ASCII-compatible

This is not clearly defined by the C99 standard.
Define ctype-like macros to force it to be ASCII / UTF-8 (not extended ASCII or
something like noticed on OpenBSD 3.8).

(In practise modern libc libraries are all ASCII and UTF-8-compatible. Otherwise
this would break many programs)

7 files changed, 48 insertions, 46 deletions

diff --git a/sfeed.c b/sfeed.c
index fc1249a..b969874 100644
--- a/sfeed.c
+++ b/sfeed.c
@@ -246,7 +246,7 @@ gettag(enum FeedType feedtype, const char *name, size_t namelen)
 static char *
 ltrim(const char *s)
 {
-	for (; isspace((unsigned char)*s); s++)
+	for (; ISSPACE((unsigned char)*s); s++)
 		;
 	return (char *)s;
 }
@@ -256,7 +256,7 @@ rtrim(const char *s)
 {
 	const char *e;
 
-	for (e = s + strlen(s); e > s && isspace((unsigned char)*(e - 1)); e--)
+	for (e = s + strlen(s); e > s && ISSPACE((unsigned char)*(e - 1)); e--)
 		;
 	return (char *)e;
 }
@@ -341,7 +341,7 @@ printtrimmed(const char *s)
 	p = ltrim(s);
 	e = rtrim(p);
 	for (; *p && p != e; p++) {
-		if (isspace((unsigned char)*p))
+		if (ISSPACE((unsigned char)*p))
 			putchar(' '); /* any whitespace to space */
 		else if (!ISCNTRL((unsigned char)*p))
 			/* ignore other control chars */
@@ -514,20 +514,20 @@ gettzoffset(const char *s)
 	long tzhour = 0, tzmin = 0;
 	size_t i;
 
-	for (; isspace((unsigned char)*s); s++)
+	for (; ISSPACE((unsigned char)*s); s++)
 		;
 	switch (*s) {
 	case '-': /* offset */
 	case '+':
-		for (i = 0, p = s + 1; i < 2 && isdigit((unsigned char)*p); i++, p++)
+		for (i = 0, p = s + 1; i < 2 && ISDIGIT((unsigned char)*p); i++, p++)
 			tzhour = (tzhour * 10) + (*p - '0');
 		if (*p == ':')
 			p++;
-		for (i = 0; i < 2 && isdigit((unsigned char)*p); i++, p++)
+		for (i = 0; i < 2 && ISDIGIT((unsigned char)*p); i++, p++)
 			tzmin = (tzmin * 10) + (*p - '0');
 		return ((tzhour * 3600) + (tzmin * 60)) * (s[0] == '-' ? -1 : 1);
 	default: /* timezone name */
-		for (i = 0; isalpha((unsigned char)s[i]); i++)
+		for (i = 0; ISALPHA((unsigned char)s[i]); i++)
 			;
 		if (i != 3)
 			return 0;
@@ -565,35 +565,35 @@ parsetime(const char *s, long long *tp)
 	int va[6] = { 0 }, i, j, v, vi;
 	size_t m;
 
-	for (; isspace((unsigned char)*s); s++)
+	for (; ISSPACE((unsigned char)*s); s++)
 		;
-	if (!isdigit((unsigned char)*s) && !isalpha((unsigned char)*s))
+	if (!ISDIGIT((unsigned char)*s) && !ISALPHA((unsigned char)*s))
 		return -1;
 
-	if (isdigit((unsigned char)s[0]) &&
-	    isdigit((unsigned char)s[1]) &&
-	    isdigit((unsigned char)s[2]) &&
-	    isdigit((unsigned char)s[3])) {
+	if (ISDIGIT((unsigned char)s[0]) &&
+	    ISDIGIT((unsigned char)s[1]) &&
+	    ISDIGIT((unsigned char)s[2]) &&
+	    ISDIGIT((unsigned char)s[3])) {
 		/* formats "%Y-%m-%d %H:%M:%S", "%Y-%m-%dT%H:%M:%S" or "%Y%m%d%H%M%S" */
 		vi = 0;
 	} else {
 		/* format: "[%a, ]%d %b %Y %H:%M:%S" */
 		/* parse "[%a, ]%d %b %Y " part, then use time parsing as above */
-		for (; isalpha((unsigned char)*s); s++)
+		for (; ISALPHA((unsigned char)*s); s++)
 			;
-		for (; isspace((unsigned char)*s); s++)
+		for (; ISSPACE((unsigned char)*s); s++)
 			;
 		if (*s == ',')
 			s++;
-		for (; isspace((unsigned char)*s); s++)
+		for (; ISSPACE((unsigned char)*s); s++)
 			;
-		for (v = 0, i = 0; i < 2 && isdigit((unsigned char)*s); s++, i++)
+		for (v = 0, i = 0; i < 2 && ISDIGIT((unsigned char)*s); s++, i++)
 			v = (v * 10) + (*s - '0');
 		va[2] = v; /* day */
-		for (; isspace((unsigned char)*s); s++)
+		for (; ISSPACE((unsigned char)*s); s++)
 			;
 		/* end of word month */
-		for (j = 0; isalpha((unsigned char)s[j]); j++)
+		for (j = 0; ISALPHA((unsigned char)s[j]); j++)
 			;
 		/* check month name */
 		if (j < 3 || j > 9)
@@ -609,15 +609,15 @@ parsetime(const char *s, long long *tp)
 		}
 		if (m >= 12)
 			return -1; /* no month found */
-		for (; isspace((unsigned char)*s); s++)
+		for (; ISSPACE((unsigned char)*s); s++)
 			;
-		for (v = 0, i = 0; i < 4 && isdigit((unsigned char)*s); s++, i++)
+		for (v = 0, i = 0; i < 4 && ISDIGIT((unsigned char)*s); s++, i++)
 			v = (v * 10) + (*s - '0');
 		/* obsolete short year: RFC2822 4.3 */
 		if (i <= 3)
 			v += (v >= 0 && v <= 49) ? 2000 : 1900;
 		va[0] = v; /* year */
-		for (; isspace((unsigned char)*s); s++)
+		for (; ISSPACE((unsigned char)*s); s++)
 			;
 		/* parse only regular time part, see below */
 		vi = 3;
@@ -626,20 +626,20 @@ parsetime(const char *s, long long *tp)
 	/* parse time parts (and possibly remaining date parts) */
 	for (; *s && vi < 6; vi++) {
 		for (i = 0, v = 0; i < ((vi == 0) ? 4 : 2) &&
-		                   isdigit((unsigned char)*s); s++, i++) {
+		                   ISDIGIT((unsigned char)*s); s++, i++) {
 			v = (v * 10) + (*s - '0');
 		}
 		va[vi] = v;
 
 		if ((vi < 2 && *s == '-') ||
-		    (vi == 2 && (*s == 'T' || isspace((unsigned char)*s))) ||
+		    (vi == 2 && (*s == 'T' || ISSPACE((unsigned char)*s))) ||
 		    (vi > 2 && *s == ':'))
 			s++;
 	}
 
 	/* skip milliseconds in for example: "%Y-%m-%dT%H:%M:%S.000Z" */
 	if (*s == '.') {
-		for (s++; isdigit((unsigned char)*s); s++)
+		for (s++; ISDIGIT((unsigned char)*s); s++)
 			;
 	}
 
diff --git a/sfeed_opml_import.c b/sfeed_opml_import.c
index 9a13e36..14b5444 100644
--- a/sfeed_opml_import.c
+++ b/sfeed_opml_import.c
@@ -1,4 +1,3 @@
-#include <ctype.h>
 #include <stdio.h>
 #include <strings.h>
 
diff --git a/sfeed_web.c b/sfeed_web.c
index dfff0a3..630ab60 100644
--- a/sfeed_web.c
+++ b/sfeed_web.c
@@ -1,4 +1,3 @@
-#include <ctype.h>
 #include <stdio.h>
 #include <strings.h>
 
diff --git a/sfeed_xmlenc.c b/sfeed_xmlenc.c
index 67d7b0f..7fc93ae 100644
--- a/sfeed_xmlenc.c
+++ b/sfeed_xmlenc.c
@@ -1,4 +1,3 @@
-#include <ctype.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <strings.h>
@@ -26,10 +25,10 @@ xmlattr(XMLParser *p, const char *t, size_t tl, const char *n, size_t nl,
 		return;
 
 	for (; *v; v++) {
-		if (isalpha((unsigned char)*v) ||
-		    isdigit((unsigned char)*v) ||
+		if (ISALPHA((unsigned char)*v) ||
+		    ISDIGIT((unsigned char)*v) ||
 		    *v == '.' || *v == ':' || *v == '-' || *v == '_')
-			putchar(tolower((unsigned char)*v));
+			putchar(TOLOWER((unsigned char)*v));
 	}
 }
 
diff --git a/util.c b/util.c
index ed0b5c9..0b7da06 100644
--- a/util.c
+++ b/util.c
@@ -1,4 +1,3 @@
-#include <ctype.h>
 #include <errno.h>
 #include <stdarg.h>
 #include <stdio.h>
@@ -66,8 +65,8 @@ strcasestr(const char *h, const char *n)
 		return (char *)h;
 
 	for (; *h; ++h) {
-		for (i = 0; n[i] && tolower((unsigned char)n[i]) ==
-		            tolower((unsigned char)h[i]); ++i)
+		for (i = 0; n[i] && TOLOWER((unsigned char)n[i]) ==
+		            TOLOWER((unsigned char)h[i]); ++i)
 			;
 		if (n[i] == '\0')
 			return (char *)h;
@@ -82,7 +81,7 @@ uri_hasscheme(const char *s)
 {
 	const char *p = s;
 
-	for (; isalpha((unsigned char)*p) || isdigit((unsigned char)*p) ||
+	for (; ISALPHA((unsigned char)*p) || ISDIGIT((unsigned char)*p) ||
 		       *p == '+' || *p == '-' || *p == '.'; p++)
 		;
 	/* scheme, except if empty and starts with ":" then it is a path */
@@ -109,7 +108,7 @@ uri_parse(const char *s, struct uri *u)
 	}
 
 	/* scheme / protocol part */
-	for (; isalpha((unsigned char)*p) || isdigit((unsigned char)*p) ||
+	for (; ISALPHA((unsigned char)*p) || ISDIGIT((unsigned char)*p) ||
 		       *p == '+' || *p == '-' || *p == '.'; p++)
 		;
 	/* scheme, except if empty and starts with ":" then it is a path */
diff --git a/util.h b/util.h
index c68cef4..fac6424 100644
--- a/util.h
+++ b/util.h
@@ -8,8 +8,12 @@
 #define unveil(p1,p2) 0
 #endif
 
-/* control-character in the ASCII range 0-127: compatible with UTF-8 */
+/* ctype-like macros, but always compatible with ASCII / UTF-8 */
+#define ISALPHA(c) ((((unsigned)c) | 32) - 'a' < 26)
 #define ISCNTRL(c) ((c) < ' ' || (c) == 0x7f)
+#define ISDIGIT(c) (((unsigned)c) - '0' < 10)
+#define ISSPACE(c) ((c) == ' ' || ((((unsigned)c) - '\t') < 5))
+#define TOLOWER(c) ((((unsigned)c) - 'A' < 26) ? ((c) | 32) : (c))
 
 #undef strcasestr
 char *strcasestr(const char *, const char *);
diff --git a/xml.c b/xml.c
index f16cbbb..3bd9557 100644
--- a/xml.c
+++ b/xml.c
@@ -1,4 +1,3 @@
-#include <ctype.h>
 #include <errno.h>
 #include <stdio.h>
 #include <stdlib.h>
@@ -6,6 +5,9 @@
 
 #include "xml.h"
 
+#define ISALPHA(c) ((((unsigned)c) | 32) - 'a' < 26)
+#define ISSPACE(c) ((c) == ' ' || ((((unsigned)c) - '\t') < 5))
+
 static void
 xml_parseattrs(XMLParser *x)
 {
@@ -13,7 +15,7 @@ xml_parseattrs(XMLParser *x)
 	int c, endsep, endname = 0, valuestart = 0;
 
 	while ((c = GETNEXT()) != EOF) {
-		if (isspace(c)) {
+		if (ISSPACE(c)) {
 			if (namelen)
 				endname = 1;
 			continue;
@@ -23,7 +25,7 @@ xml_parseattrs(XMLParser *x)
 			x->name[namelen] = '\0';
 			valuestart = 1;
 			endname = 1;
-		} else if (namelen && ((endname && !valuestart && isalpha(c)) || (c == '>' || c == '/'))) {
+		} else if (namelen && ((endname && !valuestart && ISALPHA(c)) || (c == '>' || c == '/'))) {
 			/* attribute without value */
 			x->name[namelen] = '\0';
 			if (x->xmlattrstart)
@@ -44,7 +46,7 @@ xml_parseattrs(XMLParser *x)
 			if (c == '\'' || c == '"') {
 				endsep = c;
 			} else {
-				endsep = ' '; /* isspace() */
+				endsep = ' '; /* ISSPACE() */
 				goto startvalue;
 			}
 
@@ -58,7 +60,7 @@ startvalue:
 					x->data[0] = c;
 					valuelen = 1;
 					while ((c = GETNEXT()) != EOF) {
-						if (c == endsep || (endsep == ' ' && (c == '>' || isspace(c))))
+						if (c == endsep || (endsep == ' ' && (c == '>' || ISSPACE(c))))
 							break;
 						if (valuelen < sizeof(x->data) - 1)
 							x->data[valuelen++] = c;
@@ -79,7 +81,7 @@ startvalue:
 							break;
 						}
 					}
-				} else if (c != endsep && !(endsep == ' ' && (c == '>' || isspace(c)))) {
+				} else if (c != endsep && !(endsep == ' ' && (c == '>' || ISSPACE(c)))) {
 					if (valuelen < sizeof(x->data) - 1) {
 						x->data[valuelen++] = c;
 					} else {
@@ -90,7 +92,7 @@ startvalue:
 						valuelen = 1;
 					}
 				}
-				if (c == endsep || (endsep == ' ' && (c == '>' || isspace(c)))) {
+				if (c == endsep || (endsep == ' ' && (c == '>' || ISSPACE(c)))) {
 					x->data[valuelen] = '\0';
 					if (x->xmlattr)
 						x->xmlattr(x, x->tag, x->taglen, x->name, namelen, x->data, valuelen);
@@ -328,7 +330,7 @@ xml_parse(XMLParser *x)
 				while ((c = GETNEXT()) != EOF) {
 					if (c == '/')
 						x->isshorttag = 1; /* short tag */
-					else if (c == '>' || isspace(c)) {
+					else if (c == '>' || ISSPACE(c)) {
 						x->tag[x->taglen] = '\0';
 						if (isend) { /* end tag, starts with </ */
 							if (x->xmltagend)
@@ -339,7 +341,7 @@ xml_parse(XMLParser *x)
 							/* start tag */
 							if (x->xmltagstart)
 								x->xmltagstart(x, x->tag, x->taglen);
-							if (isspace(c))
+							if (ISSPACE(c))
 								xml_parseattrs(x);
 							if (x->xmltagstartparsed)
 								x->xmltagstartparsed(x, x->tag, x->taglen, x->isshorttag);