sfeed: Address more ctype(3) UBs

[Leonardo Taccari <leot%NetBSD.org@localhost>]

Module Name:	pkgsrc-wip
Committed By:	Leonardo Taccari <leot%NetBSD.org@localhost>
Pushed By:	leot
Date:		Fri Sep 7 18:18:05 2018 +0200
Changeset:	89f53cb8da228d9fe03c6de7789965ad1ce28528

Modified Files:
	sfeed/distinfo
	sfeed/patches/patch-sfeed.c

Log Message:
sfeed: Address more ctype(3) UBs

Now sfeed should no longer "eat" UTF-8 characters.

(This was accidentally previously not committed because the patches
was applied manually in the WRKSRC... but of course applying more
than a patch overwrite the `.orig'!, Whoops!)

To see a diff of this commit:
https://wip.pkgsrc.org/cgi-bin/gitweb.cgi?p=pkgsrc-wip.git;a=commitdiff;h=89f53cb8da228d9fe03c6de7789965ad1ce28528

Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.

diffstat:
 sfeed/distinfo              |   2 +-
 sfeed/patches/patch-sfeed.c | 137 +++++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 136 insertions(+), 3 deletions(-)

diffs:
diff --git a/sfeed/distinfo b/sfeed/distinfo
index be653000f9..662f47ffc2 100644
--- a/sfeed/distinfo
+++ b/sfeed/distinfo
@@ -4,7 +4,7 @@ SHA1 (sfeed-0.9.5.tar.gz) = 6de7b76ec357c7f6d5289e88a842924cb95e7846
 RMD160 (sfeed-0.9.5.tar.gz) = 349e363785877c0a0a9879038746085de797dbf7
 SHA512 (sfeed-0.9.5.tar.gz) = f9c667632fb86395588a7917bb92c48e82768021c54595db8cd5e8bbaf043def2678f54f77cbbb7c8d673eff04d5f99d00d1c244b1089d7b77f4e23e8fa3e079
 Size (sfeed-0.9.5.tar.gz) = 37374 bytes
-SHA1 (patch-sfeed.c) = 6ff9e8e7d8ab061fce5aa603008a4ec338bfdbb9
+SHA1 (patch-sfeed.c) = 7239a85c8c8a1fd207d2a8386d691c09b3b8cc4f
 SHA1 (patch-sfeed__opml__import.c) = f4870e075cf2675cfa6a104ecd594e7b373537e2
 SHA1 (patch-sfeed__web.c) = 361c9b985e8610bedf35176147c3b8c9511233ce
 SHA1 (patch-sfeed__xmlenc.c) = 2476f0f4ff215a60da57cfd7eec5c1b6af1e7c55
diff --git a/sfeed/patches/patch-sfeed.c b/sfeed/patches/patch-sfeed.c
index 8c8c63d9ae..2fa975094f 100644
--- a/sfeed/patches/patch-sfeed.c
+++ b/sfeed/patches/patch-sfeed.c
@@ -2,9 +2,58 @@ $NetBSD$
 
 Avoid possible UBs in ctype(3) functions.
 
---- sfeed.c.orig	2018-09-07 13:27:10.299338895 +0000
+--- sfeed.c.orig	2018-09-07 16:16:51.071550475 +0000
 +++ sfeed.c
-@@ -381,11 +381,11 @@ gettzoffset(const char *s)
+@@ -247,10 +247,10 @@ string_print_encoded(String *s)
+ 		return;
+ 
+ 	/* skip leading whitespace */
+-	for (p = s->data; *p && isspace((int)*p); p++)
++	for (p = s->data; *p && isspace((unsigned char)*p); p++)
+ 		;
+ 	/* seek location of trailing whitespace */
+-	for (e = s->data + s->len; e > p && isspace((int)*(e - 1)); e--)
++	for (e = s->data + s->len; e > p && isspace((unsigned char)*(e - 1)); e--)
+ 		;
+ 
+ 	for (; *p && p != e; p++) {
+@@ -260,7 +260,7 @@ string_print_encoded(String *s)
+ 		case '\t': fputs("\\t",  stdout); break;
+ 		default:
+ 			/* ignore control chars */
+-			if (!iscntrl((int)*p))
++			if (!iscntrl((unsigned char)*p))
+ 				putchar(*p);
+ 			break;
+ 		}
+@@ -278,16 +278,16 @@ string_print_trimmed(String *s)
+ 		return;
+ 
+ 	/* skip leading whitespace */
+-	for (p = s->data; *p && isspace((int)*p); p++)
++	for (p = s->data; *p && isspace((unsigned char)*p); p++)
+ 		;
+ 	/* seek location of trailing whitespace */
+-	for (e = s->data + s->len; e > p && isspace((int)*(e - 1)); e--)
++	for (e = s->data + s->len; e > p && isspace((unsigned char)*(e - 1)); e--)
+ 		;
+ 
+ 	for (; *p && p != e; p++) {
+-		if (isspace((int)*p))
++		if (isspace((unsigned char)*p))
+ 			putchar(' '); /* any whitespace to space */
+-		else if (!iscntrl((int)*p))
++		else if (!iscntrl((unsigned char)*p))
+ 			/* ignore other control chars */
+ 			putchar((int)*p);
+ 	}
+@@ -376,20 +376,20 @@ gettzoffset(const char *s)
+ 	int tzhour = 0, tzmin = 0;
+ 	size_t i, namelen;
+ 
+-	for (; *s && isspace((int)*s); s++)
++	for (; *s && isspace((unsigned char)*s); s++)
+ 		;
  	switch (s[0]) {
  	case '-': /* offset */
  	case '+':
@@ -19,3 +68,87 @@ Avoid possible UBs in ctype(3) functions.
  			tzmin = (tzmin * 10) + (*p - '0');
  		return ((tzhour * 3600) + (tzmin * 60)) * (s[0] == '-' ? -1 : 1);
  	default: /* timezone name */
+-		for (i = 0; *s && isalpha((int)s[i]); i++)
++		for (i = 0; *s && isalpha((unsigned char)s[i]); i++)
+ 			;
+ 		namelen = i; /* end of name */
+ 		/* optimization: these are always non-matching */
+@@ -429,49 +429,49 @@ parsetime(const char *s, time_t *tp)
+ 	int va[6] = { 0 }, i, j, v, vi;
+ 	size_t m;
+ 
+-	for (; *s && isspace((int)*s); s++)
++	for (; *s && isspace((unsigned char)*s); s++)
+ 		;
+-	if (!isdigit((int)*s) && !isalpha((int)*s))
++	if (!isdigit((unsigned char)*s) && !isalpha((unsigned char)*s))
+ 		return -1;
+ 
+-	if (isdigit((int)*s)) {
++	if (isdigit((unsigned char)*s)) {
+ 		/* format "%Y-%m-%d %H:%M:%S" or "%Y-%m-%dT%H:%M:%S" */
+ 		vi = 0;
+ time:
+ 		for (; *s && vi < 6; vi++) {
+-			for (i = 0, v = 0; *s && i < 4 && isdigit((int)*s); s++, i++)
++			for (i = 0, v = 0; *s && i < 4 && isdigit((unsigned char)*s); s++, i++)
+ 				v = (v * 10) + (*s - '0');
+ 			va[vi] = v;
+ 			if ((vi < 2 && *s == '-') ||
+-			    (vi == 2 && (*s == 'T' || isspace((int)*s))) ||
++			    (vi == 2 && (*s == 'T' || isspace((unsigned char)*s))) ||
+ 			    (vi > 2 && *s == ':'))
+ 				s++;
+ 		}
+ 		/* TODO: only if seconds are parsed (vi == 5)? */
+ 		/* skip milliseconds for: %Y-%m-%dT%H:%M:%S.000Z */
+ 		if (*s == '.') {
+-			for (s++; *s && isdigit((int)*s); s++)
++			for (s++; *s && isdigit((unsigned char)*s); s++)
+ 				;
+ 		}
+ 		end = s;
+-	} else if (isalpha((int)*s)) {
++	} else if (isalpha((unsigned char)*s)) {
+ 		/* format: "%a, %d %b %Y %H:%M:%S" */
+ 		/* parse "%a, %d %b %Y " part, then use time parsing as above */
+-		for (; *s && isalpha((int)*s); s++)
++		for (; *s && isalpha((unsigned char)*s); s++)
+ 			;
+-		for (; *s && isspace((int)*s); s++)
++		for (; *s && isspace((unsigned char)*s); s++)
+ 			;
+ 		if (*s != ',')
+ 			return -1;
+-		for (s++; *s && isspace((int)*s); s++)
++		for (s++; *s && isspace((unsigned char)*s); s++)
+ 			;
+-		for (v = 0, i = 0; *s && i < 4 && isdigit((int)*s); s++, i++)
++		for (v = 0, i = 0; *s && i < 4 && isdigit((unsigned char)*s); s++, i++)
+ 			v = (v * 10) + (*s - '0');
+ 		va[2] = v; /* day */
+-		for (; *s && isspace((int)*s); s++)
++		for (; *s && isspace((unsigned char)*s); s++)
+ 			;
+ 		/* end of word month */
+-		for (j = 0; *s && isalpha((int)s[j]); j++)
++		for (j = 0; *s && isalpha((unsigned char)s[j]); j++)
+ 			;
+ 		/* check month name */
+ 		if (j < 3 || j > 9)
+@@ -487,12 +487,12 @@ time:
+ 		}
+ 		if (m >= 12)
+ 			return -1; /* no month found */
+-		for (; *s && isspace((int)*s); s++)
++		for (; *s && isspace((unsigned char)*s); s++)
+ 			;
+-		for (v = 0, i = 0; *s && i < 4 && isdigit((int)*s); s++, i++)
++		for (v = 0, i = 0; *s && i < 4 && isdigit((unsigned char)*s); s++, i++)
+ 			v = (v * 10) + (*s - '0');
+ 		va[0] = v; /* year */
+-		for (; *s && isspace((int)*s); s++)
++		for (; *s && isspace((unsigned char)*s); s++)
+ 			;
+ 		/* parse regular time, see above */
+ 		vi = 3;