Validate decoder pipelines.

PDF decoders, AKA filters, can be chained together. There can be
an arbitrary number of decoding / decompressing filters in the pipeline,
but there should be at most 1 image decoder, and the image decoder
should only be at the end of the chain.

BUG=chromium:880675

Change-Id: Iffa27c70ec1ed7574e38e0de23413840ee900959
Reviewed-on: https://pdfium-review.googlesource.com/42711
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>

3 files changed, 103 insertions, 0 deletions

diff --git a/core/fpdfapi/parser/fpdf_parser_decode.cpp b/core/fpdfapi/parser/fpdf_parser_decode.cpp
index eafb8ee8a3..2f73b3a4cb 100644
--- a/core/fpdfapi/parser/fpdf_parser_decode.cpp
+++ b/core/fpdfapi/parser/fpdf_parser_decode.cpp
@@ -24,6 +24,7 @@
 #include "core/fxcodec/fx_codec.h"
 #include "core/fxcrt/fx_extension.h"
 #include "third_party/base/numerics/safe_math.h"
+#include "third_party/base/stl_util.h"
 
 namespace {
 
@@ -87,6 +88,22 @@ const uint16_t PDFDocEncoding[256] = {
     0x00f3, 0x00f4, 0x00f5, 0x00f6, 0x00f7, 0x00f8, 0x00f9, 0x00fa, 0x00fb,
     0x00fc, 0x00fd, 0x00fe, 0x00ff};
 
+bool ValidateDecoderPipeline(const CPDF_Array* pDecoders) {
+  size_t count = pDecoders->GetCount();
+  if (count <= 1)
+    return true;
+
+  // TODO(thestig): Consolidate all the places that use these filter names.
+  static const char kValidDecoders[][16] = {
+      "FlateDecode",    "Fl",  "LZWDecode",       "LZW", "ASCII85Decode", "A85",
+      "ASCIIHexDecode", "AHx", "RunLengthDecode", "RL"};
+  for (size_t i = 0; i < count - 1; ++i) {
+    if (!pdfium::ContainsValue(kValidDecoders, pDecoders->GetStringAt(i)))
+      return false;
+  }
+  return true;
+}
+
 uint32_t A85Decode(pdfium::span<const uint8_t> src_span,
                    std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
                    uint32_t* dest_size) {
@@ -359,6 +376,9 @@ bool PDF_DataDecode(pdfium::span<const uint8_t> src_span,
 
   std::vector<std::pair<ByteString, const CPDF_Object*>> DecoderArray;
   if (const CPDF_Array* pDecoders = pDecoder->AsArray()) {
+    if (!ValidateDecoderPipeline(pDecoders))
+      return false;
+
     const CPDF_Array* pParamsArray = ToArray(pParams);
     for (size_t i = 0; i < pDecoders->GetCount(); ++i) {
       DecoderArray.push_back(
diff --git a/core/fpdfapi/parser/fpdf_parser_decode.h b/core/fpdfapi/parser/fpdf_parser_decode.h
index e4fdce1037..57ee271656 100644
--- a/core/fpdfapi/parser/fpdf_parser_decode.h
+++ b/core/fpdfapi/parser/fpdf_parser_decode.h
@@ -15,11 +15,14 @@
 #include "third_party/base/span.h"
 
 class CCodec_ScanlineDecoder;
+class CPDF_Array;
 class CPDF_Dictionary;
 
 // Indexed by 8-bit char code, contains unicode code points.
 extern const uint16_t PDFDocEncoding[256];
 
+bool ValidateDecoderPipeline(const CPDF_Array* pDecoders);
+
 ByteString PDF_EncodeString(const ByteString& src, bool bHex);
 WideString PDF_DecodeText(const uint8_t* pData, uint32_t size);
 WideString PDF_DecodeText(const ByteString& bstr);
diff --git a/core/fpdfapi/parser/fpdf_parser_decode_unittest.cpp b/core/fpdfapi/parser/fpdf_parser_decode_unittest.cpp
index b6d7679ef5..a4572396e6 100644
--- a/core/fpdfapi/parser/fpdf_parser_decode_unittest.cpp
+++ b/core/fpdfapi/parser/fpdf_parser_decode_unittest.cpp
@@ -4,9 +4,89 @@
 
 #include "core/fpdfapi/parser/fpdf_parser_decode.h"
 
+#include "core/fpdfapi/parser/cpdf_array.h"
+#include "core/fpdfapi/parser/cpdf_name.h"
 #include "testing/gtest/include/gtest/gtest.h"
 #include "testing/test_support.h"
 
+TEST(fpdf_parser_decode, ValidateDecoderPipeline) {
+  {
+    // Empty decoder list is always valid.
+    CPDF_Array decoders;
+    EXPECT_TRUE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // 1 decoder is always valid.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("FlateEncode");
+    EXPECT_TRUE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // 1 decoder is always valid, even with an unknown decoder.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("FooBar");
+    EXPECT_TRUE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // Valid 2 decoder pipeline.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("AHx");
+    decoders.AddNew<CPDF_Name>("LZWDecode");
+    EXPECT_TRUE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // Valid 2 decoder pipeline.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("ASCII85Decode");
+    decoders.AddNew<CPDF_Name>("ASCII85Decode");
+    EXPECT_TRUE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // Valid 5 decoder pipeline.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("ASCII85Decode");
+    decoders.AddNew<CPDF_Name>("A85");
+    decoders.AddNew<CPDF_Name>("RunLengthDecode");
+    decoders.AddNew<CPDF_Name>("FlateDecode");
+    decoders.AddNew<CPDF_Name>("RL");
+    EXPECT_TRUE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // Valid 5 decoder pipeline, with an image decoder at the end.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("RunLengthDecode");
+    decoders.AddNew<CPDF_Name>("ASCII85Decode");
+    decoders.AddNew<CPDF_Name>("FlateDecode");
+    decoders.AddNew<CPDF_Name>("LZW");
+    decoders.AddNew<CPDF_Name>("DCTDecode");
+    EXPECT_TRUE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // Invalid 2 decoder pipeline, with 2 image decoders.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("DCTDecode");
+    decoders.AddNew<CPDF_Name>("CCITTFaxDecode");
+    EXPECT_FALSE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // Invalid 2 decoder pipeline, with 1 image decoder at the start.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("DCTDecode");
+    decoders.AddNew<CPDF_Name>("FlateDecode");
+    EXPECT_FALSE(ValidateDecoderPipeline(&decoders));
+  }
+  {
+    // Invalid 5 decoder pipeline.
+    CPDF_Array decoders;
+    decoders.AddNew<CPDF_Name>("FlateDecode");
+    decoders.AddNew<CPDF_Name>("FlateDecode");
+    decoders.AddNew<CPDF_Name>("DCTDecode");
+    decoders.AddNew<CPDF_Name>("FlateDecode");
+    decoders.AddNew<CPDF_Name>("FlateDecode");
+    EXPECT_FALSE(ValidateDecoderPipeline(&decoders));
+  }
+}
+
 TEST(fpdf_parser_decode, A85Decode) {
   const pdfium::DecodeTestData test_data[] = {
       // Empty src string.