fix(http): add uri decoder for http_server (#357)

src/dist/http/http_server.cpp

@@ -11,6 +11,7 @@
 #include "root_http_service.h"
 #include "pprof_http_service.h"
 #include "perf_counter_http_service.h"
+#include "uri_decoder.h"
 
 namespace dsn {
 
@@ -98,6 +99,13 @@ void http_server::add_service(http_service *service)
         unresolved_path.resize(data_length + 1);
         strncpy(&unresolved_path[0], ret.full_url.data() + u.field_data[UF_PATH].off, data_length);
         unresolved_path[data_length] = '\0';
+
+        // decode resolved path
+        auto decoded_unresolved_path = uri::decode(unresolved_path);
+        if (!decoded_unresolved_path.is_ok()) {
+            return decoded_unresolved_path.get_error();
+        }
+        unresolved_path = decoded_unresolved_path.get_value();
     }
 
     std::string unresolved_query;
@@ -106,6 +114,13 @@ void http_server::add_service(http_service *service)
         unresolved_query.resize(data_length);
         strncpy(
             &unresolved_query[0], ret.full_url.data() + u.field_data[UF_QUERY].off, data_length);
+
+        // decode resolved query
+        auto decoded_unresolved_query = uri::decode(unresolved_query);
+        if (!decoded_unresolved_query.is_ok()) {
+            return decoded_unresolved_query.get_error();
+        }
+        unresolved_query = decoded_unresolved_query.get_value();
     }
 
     std::vector<std::string> args;

src/dist/http/test/uri_decoder_test.cpp

@@ -0,0 +1,78 @@
+// Copyright (c) 2018, Xiaomi, Inc.  All rights reserved.
+// This source code is licensed under the Apache License Version 2.0, which
+// can be found in the LICENSE file in the root directory of this source tree.
+
+#include <gtest/gtest.h>
+#include <dsn/utility/error_code.h>
+#include <dist/http/uri_decoder.h>
+
+namespace dsn {
+namespace uri {
+
+class uri_decoder_test : public testing::Test
+{
+};
+
+TEST_F(uri_decoder_test, decode)
+{
+    /// Extract from https://github.com/cpp-netlib/uri/blob/master/test/uri_encoding_test.cpp
+    struct test_case
+    {
+        std::string to_decode_uri;
+        error_code err;
+        std::string decoded_uri;
+        std::string description;
+    } tests[] = {
+        {"http%3A%2F%2F127.0.0.1%3A34101%2FperfCounter%3Fname%3Dcollector*app%23_all_",
+         ERR_OK,
+         "http://127.0.0.1:34101/perfCounter?name=collector*app#_all_",
+         "ERR_OK"},
+        {"%EB%B2%95%EC%A0%95%EB%8F%99", ERR_OK, "\xEB\xB2\x95\xEC\xA0\x95\xEB\x8F\x99", "ERR_OK"},
+        {"%21%23%24%26%27%28%29%2A%2B%2C%2F%3A%3B%3D%3F%40%5B%5D",
+         ERR_OK,
+         "!#$&'()*+,/:;=?@[]",
+         "ERR_OK"},
+        {"%",
+         ERR_INVALID_PARAMETERS,
+         "",
+         "ERR_INVALID_PARAMETERS: Encountered partial escape sequence at end of string"},
+        {"%2",
+         ERR_INVALID_PARAMETERS,
+         "",
+         "ERR_INVALID_PARAMETERS: Encountered partial escape sequence at end of string"},
+        {"%%%",
+         ERR_INVALID_PARAMETERS,
+         "",
+         "ERR_INVALID_PARAMETERS: The characters %% do not form a hex value. "
+         "Please escape it or pass a valid hex value"},
+        {"%2%",
+         ERR_INVALID_PARAMETERS,
+         "",
+         "ERR_INVALID_PARAMETERS: The characters 2% do not form a hex value. "
+         "Please escape it or pass a valid hex value"},
+        {"%G0",
+         ERR_INVALID_PARAMETERS,
+         "",
+         "ERR_INVALID_PARAMETERS: The characters G0 do not form a hex value. "
+         "Please escape it or pass a valid hex value"},
+        {"%0G",
+         ERR_INVALID_PARAMETERS,
+         "",
+         "ERR_INVALID_PARAMETERS: The characters 0G do not form a hex value. "
+         "Please escape it or pass a valid hex value"},
+        {"%20", ERR_OK, "\x20", "ERR_OK"},
+        {"%80", ERR_OK, "\x80", "ERR_OK"}};
+
+    for (auto test : tests) {
+        auto decode_res = decode(test.to_decode_uri);
+
+        ASSERT_EQ(decode_res.get_error().code(), test.err);
+        if (ERR_OK == test.err) {
+            ASSERT_EQ(decode_res.get_value(), test.decoded_uri);
+        }
+        ASSERT_EQ(decode_res.get_error().description(), test.description);
+    }
+}
+
+} // namespace dsn
+} // namespace uri

src/dist/http/uri_decoder.cpp

@@ -0,0 +1,69 @@
+// Copyright (c) 2017-present, Xiaomi, Inc.  All rights reserved.
+// This source code is licensed under the Apache License Version 2.0, which
+// can be found in the LICENSE file in the root directory of this source tree.
+
+#include <fmt/format.h>
+#include "uri_decoder.h"
+
+namespace dsn {
+namespace uri {
+
+error_with<char> from_hex(const char c)
+{
+    switch (c) {
+    case '0' ... '9':
+        return c - '0';
+    case 'a' ... 'f':
+        return c - 'a' + 10;
+    case 'A' ... 'F':
+        return c - 'A' + 10;
+    default:
+        return error_s::make(ERR_INVALID_PARAMETERS);
+    }
+}
+
+error_with<char> decode_char(const string_view &hex)
+{
+    assert(2 == hex.size());
+
+    auto high = from_hex(hex[0]);
+    auto low = from_hex(hex[1]);
+    if (high.is_ok() && low.is_ok()) {
+        return (high.get_value() << 4) | low.get_value();
+    }
+
+    return error_s::make(ERR_INVALID_PARAMETERS);
+}
+
+error_with<std::string> decode(const string_view &encoded_uri)
+{
+    std::string decoded_uri;
+    for (size_t i = 0; i < encoded_uri.size(); ++i) {
+        // '%' is followed by 2 hex chars
+        if ('%' == encoded_uri[i]) {
+            if (i + 2 >= encoded_uri.size()) {
+                return error_s::make(ERR_INVALID_PARAMETERS,
+                                     "Encountered partial escape sequence at end of string");
+            }
+
+            const string_view encoded_char(encoded_uri.data() + i + 1, 2);
+            auto decoded_char = decode_char(encoded_char);
+            if (!decoded_char.is_ok()) {
+                return error_s::make(
+                    ERR_INVALID_PARAMETERS,
+                    fmt::format("The characters {} do not "
+                                "form a hex value. Please escape it or pass a valid hex value",
+                                encoded_char.data()));
+            }
+            decoded_uri += decoded_char.get_value();
+            i += 2;
+        } else {
+            decoded_uri += encoded_uri[i];
+        }
+    }
+
+    return decoded_uri;
+}
+
+} // namespace uri
+} // namespace dsn

src/dist/http/uri_decoder.h

@@ -0,0 +1,18 @@
+// Copyright (c) 2017-present, Xiaomi, Inc.  All rights reserved.
+// This source code is licensed under the Apache License Version 2.0, which
+// can be found in the LICENSE file in the root directory of this source tree.
+
+#pragma once
+
+#include <string>
+#include <dsn/utility/errors.h>
+
+namespace dsn {
+namespace uri {
+
+/// \brief Decodes a sequence according to the percent decoding rules.
+/// \returns the decoded uri path
+error_with<std::string> decode(const string_view &encoded_uri);
+
+} // namespace uri
+} // namespace dsn