2 files changed, 28 insertions, 6 deletions
diff --git a/include/uriparser/Uri.h b/include/uriparser/Uri.h
index d2c8610..b80ac6c 100644
--- a/include/uriparser/Uri.h
+++ b/include/uriparser/Uri.h
@@ -1,4 +1,4 @@
-/* 4bf720e0ca97527a28e4c30f1c35b36a0b5f2697265c5ddc81080eaab4344ef2 (0.9.7+)
+/* e8e2c75d033ddfe256fe87c3fd5a330a6f2c9cbb376ebd83a1b3263e804c766a (0.9.8+)
  *
  * uriparser - RFC 3986 URI parsing library
  *
@@ -352,10 +352,19 @@ URI_PUBLIC int URI_FUNC(FreeUriMembersMm)(URI_TYPE(Uri) * uri,
 /**
  * Percent-encodes all unreserved characters from the input string and
  * writes the encoded version to the output string.
- * Be sure to allocate <b>3 times</b> the space of the input buffer for
+ *
+ * NOTE: Be sure to allocate <b>3 times</b> the space of the input buffer for
  * the output buffer for <c>normalizeBreaks == URI_FALSE</c> and <b>6 times</b>
  * the space for <c>normalizeBreaks == URI_TRUE</c>
- * (since e.g. "\x0d" becomes "%0D%0A" in that case)
+ * (since e.g. "\x0d" becomes "%0D%0A" in that case).
+ *
+ * NOTE: The implementation treats (both <c>char</c> and) <c>wchar_t</c> units
+ * as code point integers, which works well for code points <c>U+0001</c> to <c>U+00ff</c>
+ * in host-native endianness but nothing more;
+ * in particular, using <c>uriEscapeExW</c> with arbitrary Unicode input will
+ * not produce healthy results.
+ * Passing UTF-8 input to <c>uriEscapeExA</c> may be useful in some scenarios.
+ * Keep in mind that uriparser is about %URI (RFC 3986) not %IRI (RFC 3987).
  *
  * @param inFirst           <b>IN</b>: Pointer to first character of the input text
  * @param inAfterLast       <b>IN</b>: Pointer after the last character of the input text
@@ -377,10 +386,19 @@ URI_PUBLIC URI_CHAR * URI_FUNC(EscapeEx)(const URI_CHAR * inFirst,
 /**
  * Percent-encodes all unreserved characters from the input string and
  * writes the encoded version to the output string.
- * Be sure to allocate <b>3 times</b> the space of the input buffer for
+ *
+ * NOTE: Be sure to allocate <b>3 times</b> the space of the input buffer for
  * the output buffer for <c>normalizeBreaks == URI_FALSE</c> and <b>6 times</b>
  * the space for <c>normalizeBreaks == URI_TRUE</c>
- * (since e.g. "\x0d" becomes "%0D%0A" in that case)
+ * (since e.g. "\x0d" becomes "%0D%0A" in that case).
+ *
+ * NOTE: The implementation treats (both <c>char</c> and) <c>wchar_t</c> units
+ * as code point integers, which works well for code points <c>U+0001</c> to <c>U+00ff</c>
+ * in host-native endianness but nothing more;
+ * in particular, using <c>uriEscapeW</c> with arbitrary Unicode input will
+ * not produce healthy results.
+ * Passing UTF-8 input to <c>uriEscapeA</c> may be useful in some scenarios.
+ * Keep in mind that uriparser is about %URI (RFC 3986) not %IRI (RFC 3987).
  *
  * @param in                <b>IN</b>: Text source
  * @param out               <b>OUT</b>: Encoded text destination
@@ -608,6 +626,10 @@ URI_PUBLIC int URI_FUNC(ToStringCharsRequired)(const URI_TYPE(Uri) * uri,
  * Converts a %URI structure back to text as described in
  * <a href="http://tools.ietf.org/html/rfc3986#section-5.3">section 5.3 of RFC 3986</a>.
  *
+ * NOTE: Scheme-based normalization
+ * (<a href="http://tools.ietf.org/html/rfc3986#section-6.2.3">section 6.2.3 of RFC 3986</a>)
+ * is not applied and is considered a responsibility of the application using uriparser.
+ *
  * @param dest           <b>OUT</b>: Output destination
  * @param uri            <b>IN</b>: %URI to convert
  * @param maxChars       <b>IN</b>: Maximum number of characters to copy <b>including</b> terminator
diff --git a/include/uriparser/UriBase.h b/include/uriparser/UriBase.h
index 5216b1d..dc3883e 100644
--- a/include/uriparser/UriBase.h
+++ b/include/uriparser/UriBase.h
@@ -55,7 +55,7 @@
 /* Version */
 #define URI_VER_MAJOR           0
 #define URI_VER_MINOR           9
-#define URI_VER_RELEASE         7
+#define URI_VER_RELEASE         8
 #define URI_VER_SUFFIX_ANSI     ""
 #define URI_VER_SUFFIX_UNICODE  URI_ANSI_TO_UNICODE(URI_VER_SUFFIX_ANSI)