[C++] read encrypted ORC file

c++/include/orc/Common.hh

@@ -224,6 +224,24 @@ namespace orc {
      * Get the writer timezone.
      */
     virtual const std::string& getWriterTimezone() const = 0;
+    /**
+     *Get the list of Localkeys for all columns in the Stripe, with each encrypted column
+     *corresponding to a Localkey.
+     * @return
+     */
+    virtual std::shared_ptr<std::vector<std::vector<unsigned char>>>
+    getEncryptedLocalKeys() const = 0;
+    /**
+     *Get the Localkey for a specific column in this Stripe.
+     */
+    virtual std::vector<unsigned char>& getEncryptedLocalKeyByVariantId(int32_t col) const = 0;
+    /**
+     * In general, only the first stripe in an ORC file will store the LocalKey.In this case, the
+     * stripeId and originalStripeId are equal. If an ORC file has multiple stripes storing the
+     * LocalKey, the values of stripeId and originalStripeId may not be equal.
+     * @return
+     */
+    virtual int64_t getOriginalStripeId() const = 0;
   };
 
   // Return true if val1 < val2; otherwise return false

c++/include/orc/Reader.hh

@@ -39,7 +39,7 @@ namespace orc {
   // classes that hold data members so we can maintain binary compatibility
   struct ReaderOptionsPrivate;
   struct RowReaderOptionsPrivate;
-
+  class KeyProvider;
   /**
    * Expose the reader metrics including the latency and
    * number of calls of the decompression/decoding/IO modules.
@@ -138,6 +138,10 @@ namespace orc {
      * Get the reader metrics.
      */
     ReaderMetrics* getReaderMetrics() const;
+
+    ReaderOptions& setKeyProvider(std::shared_ptr<KeyProvider> keyProvider);
+
+    std::shared_ptr<KeyProvider> getKeyProvider() const;
   };
 
   /**

c++/src/CMakeLists.txt

@@ -144,7 +144,6 @@ add_custom_command(OUTPUT orc_proto.pb.h orc_proto.pb.cc
         --cpp_out="${CMAKE_CURRENT_BINARY_DIR}"
         ../../orc-format_ep-prefix/src/orc-format_ep/src/main/proto/orc/proto/orc_proto.proto
 )
-
 set(SOURCE_FILES
   "${CMAKE_CURRENT_BINARY_DIR}/Adaptor.hh"
   orc_proto.pb.h
@@ -187,7 +186,10 @@ set(SOURCE_FILES
   Timezone.cc
   TypeImpl.cc
   Vector.cc
-  Writer.cc)
+  Writer.cc
+  security/InMemoryKeystore.cc
+  security/ReaderEncryption.cc
+)
 
 if(BUILD_LIBHDFSPP)
   set(SOURCE_FILES ${SOURCE_FILES} OrcHdfsFile.cc)
@@ -211,6 +213,8 @@ target_link_libraries (orc
     $<BUILD_INTERFACE:orc::lz4>
     $<BUILD_INTERFACE:orc::zstd>
     $<BUILD_INTERFACE:${LIBHDFSPP_LIBRARIES}>
+    ${OPENSSL_CRYPTO_LIBRARY}
+    ${OPENSSL_SSL_LIBRARY}
   )
 
 target_include_directories (orc

c++/src/Compression.cc

@@ -1272,7 +1272,13 @@ namespace orc {
         throw NotImplementedYet("compression codec");
     }
   }
-
+  std::unique_ptr<SeekableInputStream> createDecompressorAndDecryption(
+      CompressionKind kind, std::unique_ptr<SeekableInputStream> input, uint64_t blockSize,
+      MemoryPool& pool, ReaderMetrics* metrics,std::vector<unsigned char> key,
+      std::vector<unsigned char> iv,const EVP_CIPHER* cipher){
+    auto dec = std::make_unique<DecryptionInputStream>(std::move(input),key,iv,cipher,pool);
+    return createDecompressor(kind,std::move(dec),blockSize,pool,metrics);
+  }
   std::unique_ptr<SeekableInputStream> createDecompressor(
       CompressionKind kind, std::unique_ptr<SeekableInputStream> input, uint64_t blockSize,
       MemoryPool& pool, ReaderMetrics* metrics) {

c++/src/Compression.hh

@@ -36,6 +36,10 @@ namespace orc {
       CompressionKind kind, std::unique_ptr<SeekableInputStream> input, uint64_t bufferSize,
       MemoryPool& pool, ReaderMetrics* metrics);
 
+  std::unique_ptr<SeekableInputStream> createDecompressorAndDecryption(CompressionKind kind, std::unique_ptr<SeekableInputStream> input, uint64_t blockSize,
+                                                                       MemoryPool& pool, ReaderMetrics* metrics,std::vector<unsigned char> key,
+                                                                       std::vector<unsigned char> iv,const EVP_CIPHER* cipher);
+
   /**
    * Create a compressor for the given compression kind.
    * @param kind the compression type to implement

c++/src/Options.hh

@@ -43,6 +43,7 @@ namespace orc {
     MemoryPool* memoryPool;
     std::string serializedTail;
     ReaderMetrics* metrics;
+    std::shared_ptr<KeyProvider> keyProvider;
 
     ReaderOptionsPrivate() {
       tailLocation = std::numeric_limits<uint64_t>::max();
@@ -121,7 +122,14 @@ namespace orc {
   std::ostream* ReaderOptions::getErrorStream() const {
     return privateBits_->errorStream;
   }
+  ReaderOptions& ReaderOptions::setKeyProvider(std::shared_ptr<KeyProvider> keyProvider) {
+    privateBits_->keyProvider = keyProvider;
+    return *this;
+  }
 
+  std::shared_ptr<KeyProvider> ReaderOptions::getKeyProvider() const {
+    return privateBits_->keyProvider;
+  }
   /**
    * RowReaderOptions Implementation
    */