Turn off encoding only when required

Author: ianton-ru

src/IO/S3/URI.cpp

@@ -16,12 +16,12 @@ namespace DB
 
 struct URIConverter
 {
-    static void modifyURI(Poco::URI & uri, std::unordered_map<std::string, std::string> mapper)
+    static void modifyURI(Poco::URI & uri, std::unordered_map<std::string, std::string> mapper, bool enable_url_encoding)
     {
         Macros macros({{"bucket", uri.getHost()}});
         uri = macros.expand(mapper[uri.getScheme()]).empty()
             ? uri
-            : Poco::URI(macros.expand(mapper[uri.getScheme()]) + uri.getPathAndQuery(), /*enable_url_encoding*/ false);
+            : Poco::URI(macros.expand(mapper[uri.getScheme()]) + uri.getPathAndQuery(), enable_url_encoding);
     }
 };
 
@@ -33,7 +33,7 @@ namespace ErrorCodes
 namespace S3
 {
 
-URI::URI(const std::string & uri_, bool allow_archive_path_syntax)
+URI::URI(const std::string & uri_, bool allow_archive_path_syntax, bool enable_url_encoding)
 {
     /// Case when bucket name represented in domain name of S3 URL.
     /// E.g. (https://bucket-name.s3.region.amazonaws.com/key)
@@ -55,7 +55,7 @@ URI::URI(const std::string & uri_, bool allow_archive_path_syntax)
     else
         uri_str = uri_;
 
-    uri = Poco::URI(uri_str, /*enable_url_encoding*/ false);
+    uri = Poco::URI(uri_str, enable_url_encoding);
 
     std::unordered_map<std::string, std::string> mapper;
     auto context = Context::getGlobalContextInstance();
@@ -77,7 +77,7 @@ URI::URI(const std::string & uri_, bool allow_archive_path_syntax)
         }
 
         if (!mapper.empty())
-            URIConverter::modifyURI(uri, mapper);
+            URIConverter::modifyURI(uri, mapper, enable_url_encoding);
     }
 
     storage_name = "S3";

src/IO/S3/URI.h

@@ -36,7 +36,7 @@ struct URI
     bool is_virtual_hosted_style;
 
     URI() = default;
-    explicit URI(const std::string & uri_, bool allow_archive_path_syntax = false);
+    explicit URI(const std::string & uri_, bool allow_archive_path_syntax = false, bool enable_url_encoding = true);
     void addRegionToURI(const std::string & region);
 
     static void validateBucket(const std::string & bucket, const Poco::URI & uri);

src/Storages/ObjectStorage/Utils.cpp

@@ -341,7 +341,8 @@ std::pair<DB::ObjectStoragePtr, std::string> resolveObjectStorageForPath(
         {
             normalized_path = "s3://" + target_decomposed.authority + "/" + target_decomposed.key;
         }
-        S3::URI s3_uri(normalized_path);
+        // enable_url_encoding=false, path from metadata must have correct encoding already
+        S3::URI s3_uri(normalized_path, /*allow_archive_path_syntax*/ false, /*enable_url_encoding*/ false);
 
         std::string key_to_use = s3_uri.key;
 
@@ -365,7 +366,7 @@ std::pair<DB::ObjectStoragePtr, std::string> resolveObjectStorageForPath(
             {
                 normalized_table_location = "s3://" + table_location_decomposed.authority + "/" + table_location_decomposed.key;
             }
-            S3::URI base_s3_uri(normalized_table_location);
+            S3::URI base_s3_uri(normalized_table_location, /*allow_archive_path_syntax*/ false, /*enable_url_encoding*/ false);
 
             if (s3URIMatches(s3_uri, base_s3_uri.bucket, base_s3_uri.endpoint, target_scheme_normalized))
                 use_base_storage = true;