[vm] RFC6874 support zoneID in uri parser
An implementation of https://tools.ietf.org/html/rfc6874.
IP-literal = "[" ( IPv6address / IPvFuture ) "]"
will be updated to:
IP-literal = "[" ( IPv6address / IPv6addrz / IPvFuture ) "]"
ZoneID = 1*( unreserved / pct-encoded )
IPv6addrz = IPv6address "%25" ZoneID
Bug: https://github.com/dart-lang/sdk/issues/29456
Change-Id: Ieac7b00e97d3ceff794f3b56ed4b6e4d9d6bbb47
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103544
Commit-Queue: Zichang Guo <zichangguo@google.com>
Reviewed-by: Lasse R.H. Nielsen <lrn@google.com>
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 299563f..acde42a 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -77,6 +77,16 @@
[1]: https://github.com/dart-lang/sdk/blob/master/CHANGELOG.md#200---2018-08-07
+#### `dart:core`
+
+* Update `Uri` class to support RFC6874: https://tools.ietf.org/html/rfc6874
+ "%25" or "%" can be appended to the end of a valid IPv6 representing a Zone
+ Identifier. A valid zone ID consists of unreversed character or Percent
+ encoded octet, which was defined in RFC3986.
+ IPv6addrz = IPv6address "%25" ZoneID
+
+ [29456]: https://github.com/dart-lang/sdk/issues/29456
+
### Dart VM
### Tools
diff --git a/sdk/lib/core/uri.dart b/sdk/lib/core/uri.dart
index 2c7d918..044c863 100644
--- a/sdk/lib/core/uri.dart
+++ b/sdk/lib/core/uri.dart
@@ -761,6 +761,9 @@
// authority = [ userinfo "@" ] host [ ":" port ]
// userinfo = *( unreserved / pct-encoded / sub-delims / ":" )
// host = IP-literal / IPv4address / reg-name
+ // IP-literal = "[" ( IPv6address / IPv6addrz / IPvFuture ) "]"
+ // IPv6addrz = IPv6address "%25" ZoneID
+ // ZoneID = 1*( unreserved / pct-encoded )
// port = *DIGIT
// reg-name = *( unreserved / pct-encoded / sub-delims )
//
@@ -1643,14 +1646,24 @@
if (hostStart < authority.length &&
authority.codeUnitAt(hostStart) == _LEFT_BRACKET) {
// IPv6 host.
+ int escapeForZoneID = -1;
for (; hostEnd < authority.length; hostEnd++) {
- if (authority.codeUnitAt(hostEnd) == _RIGHT_BRACKET) break;
+ int char = authority.codeUnitAt(hostEnd);
+ if (char == _PERCENT && escapeForZoneID < 0) {
+ escapeForZoneID = hostEnd;
+ if (authority.startsWith("25", hostEnd + 1)) {
+ hostEnd += 2; // Might as well skip the already checked escape.
+ }
+ } else if (char == _RIGHT_BRACKET) {
+ break;
+ }
}
if (hostEnd == authority.length) {
throw FormatException(
"Invalid IPv6 host entry.", authority, hostStart);
}
- Uri.parseIPv6Address(authority, hostStart + 1, hostEnd);
+ Uri.parseIPv6Address(authority, hostStart + 1,
+ (escapeForZoneID < 0) ? hostEnd : escapeForZoneID);
hostEnd++; // Skip the closing bracket.
if (hostEnd != authority.length &&
authority.codeUnitAt(hostEnd) != _COLON) {
@@ -1959,22 +1972,124 @@
if (host.codeUnitAt(end - 1) != _RIGHT_BRACKET) {
_fail(host, start, 'Missing end `]` to match `[` in host');
}
- Uri.parseIPv6Address(host, start + 1, end - 1);
+ String zoneID = "";
+ int index = _checkZoneID(host, start + 1, end - 1);
+ if (index < end - 1) {
+ int zoneIDstart =
+ (host.startsWith("25", index + 1)) ? index + 3 : index + 1;
+ zoneID = _normalizeZoneID(host, zoneIDstart, end - 1, "%25");
+ }
+ Uri.parseIPv6Address(host, start + 1, index);
// RFC 5952 requires hex digits to be lower case.
- return host.substring(start, end).toLowerCase();
+ return host.substring(start, index).toLowerCase() + zoneID + ']';
}
if (!strictIPv6) {
// TODO(lrn): skip if too short to be a valid IPv6 address?
for (int i = start; i < end; i++) {
if (host.codeUnitAt(i) == _COLON) {
- Uri.parseIPv6Address(host, start, end);
- return '[$host]';
+ String zoneID = "";
+ int index = _checkZoneID(host, start, end);
+ if (index < end) {
+ int zoneIDstart =
+ (host.startsWith("25", index + 1)) ? index + 3 : index + 1;
+ zoneID = _normalizeZoneID(host, zoneIDstart, end, "%25");
+ }
+ Uri.parseIPv6Address(host, start, index);
+ return '[${host.substring(start, index)}' + zoneID + ']';
}
}
}
return _normalizeRegName(host, start, end);
}
+ // RFC 6874 check for ZoneID
+ // Return the index of first appeared `%`.
+ static int _checkZoneID(String host, int start, int end) {
+ int index = host.indexOf('%', start);
+ index = (index >= start && index < end) ? index : end;
+ return index;
+ }
+
+ static bool _isZoneIDChar(int char) {
+ return char < 127 && (_zoneIDTable[char >> 4] & (1 << (char & 0xf))) != 0;
+ }
+
+ /**
+ * Validates and does case- and percent-encoding normalization.
+ *
+ * The same as [_normalizeOrSubstring]
+ * except this function does not convert characters to lower case.
+ * The [host] must be an RFC6874 "ZoneID".
+ * ZoneID = 1*(unreserved / pct-encoded)
+ */
+ static String _normalizeZoneID(String host, int start, int end,
+ [String prefix = '']) {
+ StringBuffer buffer;
+ if (prefix != '') {
+ buffer = StringBuffer(prefix);
+ }
+ int sectionStart = start;
+ int index = start;
+ // Whether all characters between sectionStart and index are normalized,
+ bool isNormalized = true;
+
+ while (index < end) {
+ int char = host.codeUnitAt(index);
+ if (char == _PERCENT) {
+ String replacement = _normalizeEscape(host, index, true);
+ if (replacement == null && isNormalized) {
+ index += 3;
+ continue;
+ }
+ buffer ??= StringBuffer();
+ String slice = host.substring(sectionStart, index);
+ buffer.write(slice);
+ int sourceLength = 3;
+ if (replacement == null) {
+ replacement = host.substring(index, index + 3);
+ } else if (replacement == "%") {
+ _fail(host, index, "ZoneID should not contain % anymore");
+ }
+ buffer.write(replacement);
+ index += sourceLength;
+ sectionStart = index;
+ isNormalized = true;
+ } else if (_isZoneIDChar(char)) {
+ if (isNormalized && _UPPER_CASE_A <= char && _UPPER_CASE_Z >= char) {
+ // Put initial slice in buffer and continue in non-normalized mode
+ buffer ??= StringBuffer();
+ if (sectionStart < index) {
+ buffer.write(host.substring(sectionStart, index));
+ sectionStart = index;
+ }
+ isNormalized = false;
+ }
+ index++;
+ } else {
+ int sourceLength = 1;
+ if ((char & 0xFC00) == 0xD800 && (index + 1) < end) {
+ int tail = host.codeUnitAt(index + 1);
+ if ((tail & 0xFC00) == 0xDC00) {
+ char = 0x10000 | ((char & 0x3ff) << 10) | (tail & 0x3ff);
+ sourceLength = 2;
+ }
+ }
+ buffer ??= StringBuffer();
+ String slice = host.substring(sectionStart, index);
+ buffer.write(slice);
+ buffer.write(_escapeChar(char));
+ index += sourceLength;
+ sectionStart = index;
+ }
+ }
+ if (buffer == null) return host.substring(start, end);
+ if (sectionStart < end) {
+ String slice = host.substring(sectionStart, end);
+ buffer.write(slice);
+ }
+ return buffer.toString();
+ }
+
static bool _isRegNameChar(int char) {
return char < 127 && (_regNameTable[char >> 4] & (1 << (char & 0xf))) != 0;
}
@@ -3120,6 +3235,27 @@
// pqrstuvwxyz ~
0x47ff, // 0x70 - 0x7f 1111111111100010
];
+
+ // Characters allowed in the ZoneID as of RFC 6874.
+ // ZoneID = 1*( unreserved / pct-encoded )
+ static const _zoneIDTable = <int>[
+ // LSB MSB
+ // | |
+ 0x0000, // 0x00 - 0x0f 0000000000000000
+ 0x0000, // 0x10 - 0x1f 0000000000000000
+ // ! $%&'()*+,-.
+ 0x6000, // 0x20 - 0x2f 0000000000000110
+ // 0123456789 ; =
+ 0x03ff, // 0x30 - 0x3f 1111111111000000
+ // ABCDEFGHIJKLMNO
+ 0xfffe, // 0x40 - 0x4f 0111111111111111
+ // PQRSTUVWXYZ _
+ 0x87ff, // 0x50 - 0x5f 1111111111100001
+ // abcdefghijklmno
+ 0xfffe, // 0x60 - 0x6f 0111111111111111
+ // pqrstuvwxyz ~
+ 0x47ff, // 0x70 - 0x7f 1111111111100010
+ ];
}
// --------------------------------------------------------------------
diff --git a/tests/corelib_2/uri_http_test.dart b/tests/corelib_2/uri_http_test.dart
index 60d9732..399c403 100644
--- a/tests/corelib_2/uri_http_test.dart
+++ b/tests/corelib_2/uri_http_test.dart
@@ -33,6 +33,11 @@
new Uri.http("host", "/a/b", {"c=": "&d"}), "http://host/a/b?c%3D=%26d");
check(new Uri.http("[::]", "a"), "http://[::]/a");
check(new Uri.http("[::127.0.0.1]", "a"), "http://[::127.0.0.1]/a");
+ check(new Uri.http('[fe80::8eae:4c4d:fee9:8434%rename3]', ''),
+ 'http://[fe80::8eae:4c4d:fee9:8434%25rename3]');
+ check(new Uri.http('[ff02::1%1%41]', ''), 'http://[ff02::1%251a]');
+ check(new Uri.http('[ff02::1%321]', ''), 'http://[ff02::1%25321]');
+ check(new Uri.http('[ff02::1%%321]', ''), 'http://[ff02::1%2521]');
}
testHttpsUri() {
diff --git a/tests/corelib_2/uri_ipv6_test.dart b/tests/corelib_2/uri_ipv6_test.dart
index e6a7a4d..e5dbb9df 100644
--- a/tests/corelib_2/uri_ipv6_test.dart
+++ b/tests/corelib_2/uri_ipv6_test.dart
@@ -104,6 +104,76 @@
Expect.equals(80, uri.port);
Expect.equals('', uri.path);
Expect.equals(path.toLowerCase(), uri.toString());
+
+ // Checks for ZoneID in RFC 6874
+ path = 'https://[fe80::a%en1]:443/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('fe80::a%25en1', uri.host);
+ Expect.equals(443, uri.port);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[fe80::a%25en1]/index.html', uri.toString());
+
+ path = 'https://[fe80::a%25eE1]:443/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('fe80::a%25eE1', uri.host);
+ Expect.equals(443, uri.port);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[fe80::a%25eE1]/index.html', uri.toString());
+
+ // Recognize bare '%' and transform into '%25'
+ path = 'https://[fe80::a%1]:443/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('fe80::a%251', uri.host);
+ Expect.equals(443, uri.port);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[fe80::a%251]/index.html', uri.toString());
+
+ path = 'https://[ff02::5678%pvc1.3]/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('ff02::5678%25pvc1.3', uri.host);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[ff02::5678%25pvc1.3]/index.html', uri.toString());
+
+ // ZoneID contains percent encoded
+ path = 'https://[ff02::1%%321]/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('ff02::1%2521', uri.host);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[ff02::1%2521]/index.html', uri.toString());
+
+ path = 'https://[ff02::1%321]/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('ff02::1%25321', uri.host);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[ff02::1%25321]/index.html', uri.toString());
+
+ // Lower cases
+ path = 'https://[ff02::1%1%41]/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('ff02::1%251a', uri.host);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[ff02::1%251a]/index.html', uri.toString());
+
+ path = 'https://[fe80::8eae:4c4d:fee9:8434%rename3]/index.html';
+ uri = Uri.parse(path);
+ Expect.equals('https', uri.scheme);
+ Expect.equals('fe80::8eae:4c4d:fee9:8434%25rename3', uri.host);
+ Expect.equals('/index.html', uri.path);
+ Expect.equals('https://[fe80::8eae:4c4d:fee9:8434%25rename3]/index.html',
+ uri.toString());
+
+ // Test construtors with host name
+ uri = Uri(scheme: 'https', host: '[ff02::5678%pvc1.3]');
+ uri = Uri(scheme: 'https', host: '[fe80::a%1]');
+ uri = Uri(scheme: 'https', host: '[fe80::a%25eE1]');
+ uri = Uri(scheme: 'https', host: '[fe80::a%en1]');
}
void testParseIPv6Address() {
