Use struct zip_error to pass error around, conform to coding style.

--HG--
branch : HEAD
diff --git a/lib/zip_utf-8.c b/lib/zip_utf-8.c
index 83d35de..636fba4 100644
--- a/lib/zip_utf-8.c
+++ b/lib/zip_utf-8.c
@@ -130,17 +130,18 @@
 	    continue;
 
 	ret = ZIP_ENCODING_UTF8;
-	if ((name[i] & UTF_8_LEN_2_MASK) == UTF_8_LEN_2_MATCH) {
+	if ((name[i] & UTF_8_LEN_2_MASK) == UTF_8_LEN_2_MATCH)
 	    ulen = 1;
-	} else if ((name[i] & UTF_8_LEN_3_MASK) == UTF_8_LEN_3_MATCH) {
+	else if ((name[i] & UTF_8_LEN_3_MASK) == UTF_8_LEN_3_MATCH)
 	    ulen = 2;
-	} else if ((name[i] & UTF_8_LEN_4_MASK) == UTF_8_LEN_4_MATCH) {
+	else if ((name[i] & UTF_8_LEN_4_MASK) == UTF_8_LEN_4_MATCH)
 	    ulen = 3;
-	} else
+	else
 	    return ZIP_ENCODING_CP437;
 
 	if (i + ulen >= len)
 	    return ZIP_ENCODING_CP437;
+
 	for (j=1; j<=ulen; j++) {
 	    if ((name[i+j] & UTF_8_CONTINUE_MASK) != UTF_8_CONTINUE_MATCH)
 		return ZIP_ENCODING_CP437;
@@ -196,7 +197,7 @@
 
 zip_uint8_t *
 _zip_cp437_to_utf8(const zip_uint8_t * const cp437buf, zip_uint32_t len,
-		   int *zep)
+		   struct zip_error *error)
 {
     zip_uint8_t *utf8buf;
     zip_uint32_t buflen, i, offset;
@@ -205,9 +206,8 @@
     for (i=0; i<len; i++)
 	buflen += _zip_unicode_to_utf8_len(_cp437_to_unicode[cp437buf[i]]);
 
-    /* XXX: adapt error handling to caller usage */
     if ((utf8buf=(zip_uint8_t*)malloc(buflen)) == NULL) {
-	*zep = ZIP_ER_MEMORY;
+	_zip_error_set(error, ZIP_ER_MEMORY, 0);
 	return NULL;
     }