Added first version of conformance tests.

Change-Id: Ib75664194491643f8e4f1503a2ed942a2d1e1655

Makefile.am

@@ -9,7 +9,7 @@ AUTOMAKE_OPTIONS = foreign
 SUBDIRS = . src
 
 # Always include gtest in distributions.
-DIST_SUBDIRS = $(subdirs) src
+DIST_SUBDIRS = $(subdirs) src conformance
 
 # Build gtest before we build protobuf tests.  We don't add gtest to SUBDIRS
 # because then "make check" would also build and run all of gtest's own tests,
@@ -30,6 +30,10 @@ clean-local:
 	@if test -e gtest/Makefile; then \
 	  echo "Making clean in gtest"; \
 	  cd gtest && $(MAKE) $(AM_MAKEFLAGS) clean; \
+	fi; \
+	if test -e conformance/Makefile; then \
+	  echo "Making clean in conformance"; \
+	  cd conformance && $(MAKE) $(AM_MAKEFLAGS) clean; \
 	fi
 
 pkgconfigdir = $(libdir)/pkgconfig

configure.ac

@@ -164,5 +164,5 @@ export CFLAGS
 export CXXFLAGS
 AC_CONFIG_SUBDIRS([gtest])
 
-AC_CONFIG_FILES([Makefile src/Makefile protobuf.pc protobuf-lite.pc])
+AC_CONFIG_FILES([Makefile src/Makefile conformance/Makefile protobuf.pc protobuf-lite.pc])
 AC_OUTPUT

conformance/Makefile.am

@@ -0,0 +1,50 @@
+## Process this file with automake to produce Makefile.in
+
+protoc_inputs =                                                \
+  conformance.proto
+
+protoc_outputs =                                               \
+  conformance.pb.cc                                            \
+  conformance.pb.h
+
+bin_PROGRAMS = conformance-test conformance-cpp
+
+conformance_test_LDADD = $(top_srcdir)/src/libprotobuf.la
+conformance_test_SOURCES = conformance_test.cc
+nodist_conformance_test_SOURCES = conformance.pb.cc
+conformance_test_CPPFLAGS = -I$(top_srcdir)/src
+
+conformance_cpp_LDADD = $(top_srcdir)/src/libprotobuf.la
+conformance_cpp_SOURCES = conformance_cpp.cc
+nodist_conformance_cpp_SOURCES = conformance.pb.cc
+conformance_cpp_CPPFLAGS = -I$(top_srcdir)/src
+
+if USE_EXTERNAL_PROTOC
+
+unittest_proto_middleman: $(protoc_inputs)
+	$(PROTOC) -I$(srcdir) --cpp_out=. $^
+	touch unittest_proto_middleman
+
+else
+
+# We have to cd to $(srcdir) before executing protoc because $(protoc_inputs) is
+# relative to srcdir, which may not be the same as the current directory when
+# building out-of-tree.
+unittest_proto_middleman: $(top_srcdir)/src/protoc$(EXEEXT) $(protoc_inputs)
+	oldpwd=`pwd` && ( cd $(srcdir) && $$oldpwd/../src/protoc$(EXEEXT) -I. --cpp_out=$$oldpwd $(protoc_inputs) )
+	touch unittest_proto_middleman
+
+endif
+
+$(protoc_outputs): unittest_proto_middleman
+
+BUILT_SOURCES = $(protoc_outputs)
+
+CLEANFILES = $(protoc_outputs) unittest_proto_middleman
+
+MAINTAINERCLEANFILES =   \
+  Makefile.in
+
+# Targets for actually running tests.
+test_cpp: unittest_proto_middleman conformance-test conformance-cpp
+	./conformance-test ./conformance-cpp

conformance/README.md

@@ -0,0 +1,45 @@
+Protocol Buffers - Google's data interchange format
+===================================================
+
+[![Build Status](https://travis-ci.org/google/protobuf.svg?branch=master)](https://travis-ci.org/google/protobuf)
+
+Copyright 2008 Google Inc.
+
+This directory contains conformance tests for testing completeness and
+correctness of Protocol Buffers implementations.  These tests are designed
+to be easy to run against any Protocol Buffers implementation.
+
+This directory contains the tester process `conformance-test`, which
+contains all of the tests themselves.  Then separate programs written
+in whatever language you want to test communicate with the tester
+program over a pipe.
+
+Before running any of these tests, make sure you run `make` in the base
+directory to build `protoc`, since all the tests depend on it.
+
+    $ make
+
+Then to run the tests against the C++ implementation, run:
+
+    $ cd conformance && make test_cpp
+
+More tests and languages will be added soon!
+
+Testing other Protocol Buffer implementations
+---------------------------------------------
+
+To run these tests against a new Protocol Buffers implementation, write a
+program in your language that uses the protobuf implementation you want
+to test.  This program should implement the testing protocol defined in
+[conformance.proto](https://github.com/google/protobuf/blob/master/conformance/conformance.proto).
+This is designed to be as easy as possible: the C++ version is only
+150 lines and is a good example for what this program should look like
+(see [conformance_cpp.cc](https://github.com/google/protobuf/blob/master/conformance/conformance_cpp.cc)).
+The program only needs to be able to read from stdin and write to stdout.
+
+Portability
+-----------
+
+Note that the test runner currently does not work on Windows.  Patches
+to fix this are welcome!  (But please get in touch first to settle on
+a general implementation strategy).

conformance/conformance.proto

@@ -0,0 +1,219 @@
+// Protocol Buffers - Google's data interchange format
+// Copyright 2008 Google Inc.  All rights reserved.
+// https://developers.google.com/protocol-buffers/
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//     * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//     * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following disclaimer
+// in the documentation and/or other materials provided with the
+// distribution.
+//     * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+syntax = "proto3";
+package conformance;
+
+// This defines the conformance testing protocol.  This protocol exists between
+// the conformance tester process (the "tester") and the process whose protobuf
+// implemention is being tested (the "testee").  The tester forks the testee and
+// communicates with it over its stdin/stdout:
+//
+//     +--------+   pipe   +----------+
+//     | tester | <------> | testee   |
+//     |        |          |          |
+//     |  C++   |          | any lang |
+//     +--------+          +----------+
+//
+// The tester contains all of the test cases and their expected output.
+// The testee is a simple program written in the target language that reads
+// each test case and attempts to produce acceptable output for it.
+//
+// Every test consists of a ConformanceRequest/ConformanceResponse
+// request/reply pair.  The protocol on the pipe is simply:
+//
+//   1. tester sends 4-byte length N
+//   2. tester sends N bytes representing a ConformanceRequest proto
+//   3. testee sends 4-byte length M
+//   4. testee sends M bytes representing a ConformanceResponse proto
+
+// Represents a single test case's input.  The testee should:
+//
+//   1. parse this proto (which should always succeed)
+//   2. parse the protobuf or JSON payload in "payload" (which may fail)
+//   3. if the parse succeeded, serialize the message in the requested format.
+message ConformanceRequest {
+  // The payload (whether protobuf of JSON) is always for a TestAllTypes proto
+  // (see below).
+  oneof payload {
+    bytes protobuf_payload = 1;
+    string json_payload = 2;
+  }
+
+  enum RequestedOutput {
+    UNSPECIFIED = 0;
+    PROTOBUF = 1;
+    JSON = 2;
+  }
+
+  // Which format should the testee serialize its message to?
+  optional RequestedOutput requested_output = 3;
+}
+
+// Represents a single test case's output.
+message ConformanceResponse {
+  oneof result {
+    // This string should be set to indicate parsing failed.  The string can
+    // provide more information about the parse error if it is available.
+    //
+    // Setting this string does not necessarily mean the testee failed the
+    // test.  Some of the test cases are intentionally invalid input.
+    string parse_error = 1;
+
+    // This should be set if some other error occurred.  This will always
+    // indicate that the test failed.  The string can provide more information
+    // about the failure.
+    string runtime_error = 2;
+
+    // If the input was successfully parsed and the requested output was
+    // protobuf, serialize it to protobuf and set it in this field.
+    bytes protobuf_payload = 3;
+
+    // If the input was successfully parsed and the requested output was JSON,
+    // serialize to JSON and set it in this field.
+    string json_payload = 4;
+  }
+}
+
+// This proto includes every type of field in both singular and repeated
+// forms.
+message TestAllTypes {
+  message NestedMessage {
+    optional int32 a = 1;
+    optional TestAllTypes corecursive = 2;
+  }
+
+  enum NestedEnum {
+    FOO = 0;
+    BAR = 1;
+    BAZ = 2;
+    NEG = -1;  // Intentionally negative.
+  }
+
+  // Singular
+  optional    int32 optional_int32    =  1;
+  optional    int64 optional_int64    =  2;
+  optional   uint32 optional_uint32   =  3;
+  optional   uint64 optional_uint64   =  4;
+  optional   sint32 optional_sint32   =  5;
+  optional   sint64 optional_sint64   =  6;
+  optional  fixed32 optional_fixed32  =  7;
+  optional  fixed64 optional_fixed64  =  8;
+  optional sfixed32 optional_sfixed32 =  9;
+  optional sfixed64 optional_sfixed64 = 10;
+  optional    float optional_float    = 11;
+  optional   double optional_double   = 12;
+  optional     bool optional_bool     = 13;
+  optional   string optional_string   = 14;
+  optional    bytes optional_bytes    = 15;
+
+  optional group OptionalGroup = 16 {
+    optional int32 a = 17;
+  }
+
+  optional NestedMessage                        optional_nested_message  = 18;
+  optional ForeignMessage                       optional_foreign_message = 19;
+
+  optional NestedEnum                           optional_nested_enum     = 21;
+  optional ForeignEnum                          optional_foreign_enum    = 22;
+
+  optional string optional_string_piece = 24 [ctype=STRING_PIECE];
+  optional string optional_cord = 25 [ctype=CORD];
+
+  optional TestAllTypes recursive_message = 27;
+
+  // Repeated
+  repeated    int32 repeated_int32    = 31;
+  repeated    int64 repeated_int64    = 32;
+  repeated   uint32 repeated_uint32   = 33;
+  repeated   uint64 repeated_uint64   = 34;
+  repeated   sint32 repeated_sint32   = 35;
+  repeated   sint64 repeated_sint64   = 36;
+  repeated  fixed32 repeated_fixed32  = 37;
+  repeated  fixed64 repeated_fixed64  = 38;
+  repeated sfixed32 repeated_sfixed32 = 39;
+  repeated sfixed64 repeated_sfixed64 = 40;
+  repeated    float repeated_float    = 41;
+  repeated   double repeated_double   = 42;
+  repeated     bool repeated_bool     = 43;
+  repeated   string repeated_string   = 44;
+  repeated    bytes repeated_bytes    = 45;
+
+  repeated group RepeatedGroup = 46 {
+    optional int32 a = 47;
+  }
+
+  repeated NestedMessage                        repeated_nested_message  = 48;
+  repeated ForeignMessage                       repeated_foreign_message = 49;
+
+  repeated NestedEnum                           repeated_nested_enum     = 51;
+  repeated ForeignEnum                          repeated_foreign_enum    = 52;
+
+  repeated string repeated_string_piece = 54 [ctype=STRING_PIECE];
+  repeated string repeated_cord = 55 [ctype=CORD];
+
+  // Map
+  map <   int32, int32>    map_int32_int32 = 56;
+  map <   int64, int64>    map_int64_int64 = 57;
+  map <  uint32, uint32>   map_uint32_uint32 = 58;
+  map <  uint64, uint64>   map_uint64_uint64 = 59;
+  map <  sint32, sint32>   map_sint32_sint32 = 60;
+  map <  sint64, sint64>   map_sint64_sint64 = 61;
+  map < fixed32, fixed32>  map_fixed32_fixed32 = 62;
+  map < fixed64, fixed64>  map_fixed64_fixed64 = 63;
+  map <sfixed32, sfixed32> map_sfixed32_sfixed32 = 64;
+  map <sfixed64, sfixed64> map_sfixed64_sfixed64 = 65;
+  map <   int32, float>    map_int32_float = 66;
+  map <   int32, double>   map_int32_double = 67;
+  map <    bool, bool>     map_bool_bool = 68;
+  map <  string, string>   map_string_string = 69;
+  map <  string, bytes>    map_string_bytes = 70;
+  map <  string, NestedMessage>  map_string_nested_message = 71;
+  map <  string, ForeignMessage> map_string_foreign_message = 72;
+  map <  string, NestedEnum>     map_string_nested_enum = 73;
+  map <  string, ForeignEnum>    map_string_foreign_enum = 74;
+
+  oneof oneof_field {
+    uint32 oneof_uint32 = 111;
+    NestedMessage oneof_nested_message = 112;
+    string oneof_string = 113;
+    bytes oneof_bytes = 114;
+  }
+}
+
+message ForeignMessage {
+  optional int32 c = 1;
+}
+
+enum ForeignEnum {
+  FOREIGN_FOO = 0;
+  FOREIGN_BAR = 1;
+  FOREIGN_BAZ = 2;
+}

conformance/conformance_cpp.cc

@@ -0,0 +1,156 @@
+// Protocol Buffers - Google's data interchange format
+// Copyright 2008 Google Inc.  All rights reserved.
+// https://developers.google.com/protocol-buffers/
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//     * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//     * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following disclaimer
+// in the documentation and/or other materials provided with the
+// distribution.
+//     * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+#include <errno.h>
+#include <stdarg.h>
+#include <unistd.h>
+
+#include "conformance.pb.h"
+
+using std::string;
+using conformance::ConformanceRequest;
+using conformance::ConformanceResponse;
+using conformance::TestAllTypes;
+
+int test_count = 0;
+bool verbose = false;
+
+bool CheckedRead(int fd, void *buf, size_t len) {
+  size_t ofs = 0;
+  while (len > 0) {
+    ssize_t bytes_read = read(fd, (char*)buf + ofs, len);
+
+    if (bytes_read == 0) return false;
+
+    if (bytes_read < 0) {
+      GOOGLE_LOG(FATAL) << "Error reading from test runner: " <<  strerror(errno);
+    }
+
+    len -= bytes_read;
+    ofs += bytes_read;
+  }
+
+  return true;
+}
+
+void CheckedWrite(int fd, const void *buf, size_t len) {
+  if (write(fd, buf, len) != len) {
+    GOOGLE_LOG(FATAL) << "Error writing to test runner: " << strerror(errno);
+  }
+}
+
+void DoTest(const ConformanceRequest& request, ConformanceResponse* response) {
+  TestAllTypes test_message;
+
+  switch (request.payload_case()) {
+    case ConformanceRequest::kProtobufPayload:
+      if (!test_message.ParseFromString(request.protobuf_payload())) {
+        // Getting parse details would involve something like:
+        //   http://stackoverflow.com/questions/22121922/how-can-i-get-more-details-about-errors-generated-during-protobuf-parsing-c
+        response->set_parse_error("Parse error (no more details available).");
+        return;
+      }
+      break;
+
+    case ConformanceRequest::kJsonPayload:
+      response->set_runtime_error("JSON input is not yet supported.");
+      break;
+
+    case ConformanceRequest::PAYLOAD_NOT_SET:
+      GOOGLE_LOG(FATAL) << "Request didn't have payload.";
+      break;
+  }
+
+  switch (request.requested_output()) {
+    case ConformanceRequest::UNSPECIFIED:
+      GOOGLE_LOG(FATAL) << "Unspecified output format";
+      break;
+
+    case ConformanceRequest::PROTOBUF:
+      test_message.SerializeToString(response->mutable_protobuf_payload());
+      break;
+
+    case ConformanceRequest::JSON:
+      response->set_runtime_error("JSON output is not yet supported.");
+      break;
+  }
+}
+
+bool DoTestIo() {
+  string serialized_input;
+  string serialized_output;
+  ConformanceRequest request;
+  ConformanceResponse response;
+  uint32_t bytes;
+
+  if (!CheckedRead(STDIN_FILENO, &bytes, sizeof(uint32_t))) {
+    // EOF.
+    return false;
+  }
+
+  serialized_input.resize(bytes);
+
+  if (!CheckedRead(STDIN_FILENO, (char*)serialized_input.c_str(), bytes)) {
+    GOOGLE_LOG(ERROR) << "Unexpected EOF on stdin. " << strerror(errno);
+  }
+
+  if (!request.ParseFromString(serialized_input)) {
+    GOOGLE_LOG(FATAL) << "Parse of ConformanceRequest proto failed.";
+    return false;
+  }
+
+  DoTest(request, &response);
+
+  response.SerializeToString(&serialized_output);
+
+  bytes = serialized_output.size();
+  CheckedWrite(STDOUT_FILENO, &bytes, sizeof(uint32_t));
+  CheckedWrite(STDOUT_FILENO, serialized_output.c_str(), bytes);
+
+  if (verbose) {
+    fprintf(stderr, "conformance-cpp: request=%s, response=%s\n",
+            request.ShortDebugString().c_str(),
+            response.ShortDebugString().c_str());
+  }
+
+  test_count++;
+
+  return true;
+}
+
+int main() {
+  while (1) {
+    if (!DoTestIo()) {
+      fprintf(stderr, "conformance-cpp: received EOF from test runner "
+                      "after %d tests, exiting\n", test_count);
+      return 0;
+    }
+  }
+}

conformance/conformance_test.cc

@@ -0,0 +1,395 @@
+// Protocol Buffers - Google's data interchange format
+// Copyright 2008 Google Inc.  All rights reserved.
+// https://developers.google.com/protocol-buffers/
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//     * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//     * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following disclaimer
+// in the documentation and/or other materials provided with the
+// distribution.
+//     * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+#include <errno.h>
+#include <stdarg.h>
+#include <unistd.h>
+#include <string>
+
+#include "conformance.pb.h"
+#include <google/protobuf/stubs/common.h>
+#include <google/protobuf/wire_format_lite.h>
+
+using conformance::ConformanceRequest;
+using conformance::ConformanceResponse;
+using conformance::TestAllTypes;
+using google::protobuf::Descriptor;
+using google::protobuf::FieldDescriptor;
+using google::protobuf::internal::WireFormatLite;
+using std::string;
+
+int write_fd;
+int read_fd;
+int successes;
+int failures;
+bool verbose = false;
+
+string Escape(const string& str) {
+  // TODO.
+  return str;
+}
+
+#define STRINGIFY(x) #x
+#define TOSTRING(x) STRINGIFY(x)
+#define CHECK_SYSCALL(call) \
+  if (call < 0) { \
+    perror(#call " " __FILE__ ":" TOSTRING(__LINE__)); \
+    exit(1); \
+  }
+
+// TODO(haberman): make this work on Windows, instead of using these
+// UNIX-specific APIs.
+//
+// There is a platform-agnostic API in
+//    src/google/protobuf/compiler/subprocess.h
+//
+// However that API only supports sending a single message to the subprocess.
+// We really want to be able to send messages and receive responses one at a
+// time:
+//
+// 1. Spawning a new process for each test would take way too long for thousands
+//    of tests and subprocesses like java that can take 100ms or more to start
+//    up.
+//
+// 2. Sending all the tests in one big message and receiving all results in one
+//    big message would take away our visibility about which test(s) caused a
+//    crash or other fatal error.  It would also give us only a single failure
+//    instead of all of them.
+void SpawnTestProgram(char *executable) {
+  int toproc_pipe_fd[2];
+  int fromproc_pipe_fd[2];
+  if (pipe(toproc_pipe_fd) < 0 || pipe(fromproc_pipe_fd) < 0) {
+    perror("pipe");
+    exit(1);
+  }
+
+  pid_t pid = fork();
+  if (pid < 0) {
+    perror("fork");
+    exit(1);
+  }
+
+  if (pid) {
+    // Parent.
+    CHECK_SYSCALL(close(toproc_pipe_fd[0]));
+    CHECK_SYSCALL(close(fromproc_pipe_fd[1]));
+    write_fd = toproc_pipe_fd[1];
+    read_fd = fromproc_pipe_fd[0];
+  } else {
+    // Child.
+    CHECK_SYSCALL(close(STDIN_FILENO));
+    CHECK_SYSCALL(close(STDOUT_FILENO));
+    CHECK_SYSCALL(dup2(toproc_pipe_fd[0], STDIN_FILENO));
+    CHECK_SYSCALL(dup2(fromproc_pipe_fd[1], STDOUT_FILENO));
+
+    CHECK_SYSCALL(close(toproc_pipe_fd[0]));
+    CHECK_SYSCALL(close(fromproc_pipe_fd[1]));
+    CHECK_SYSCALL(close(toproc_pipe_fd[1]));
+    CHECK_SYSCALL(close(fromproc_pipe_fd[0]));
+
+    char *const argv[] = {executable, NULL};
+    CHECK_SYSCALL(execv(executable, argv));  // Never returns.
+  }
+}
+
+/* Invoking of tests **********************************************************/
+
+void ReportSuccess() {
+  successes++;
+}
+
+void ReportFailure(const char *fmt, ...) {
+  va_list args;
+  va_start(args, fmt);
+  vfprintf(stderr, fmt, args);
+  va_end(args);
+  failures++;
+}
+
+void CheckedWrite(int fd, const void *buf, size_t len) {
+  if (write(fd, buf, len) != len) {
+    GOOGLE_LOG(FATAL) << "Error writing to test program: " << strerror(errno);
+  }
+}
+
+void CheckedRead(int fd, void *buf, size_t len) {
+  size_t ofs = 0;
+  while (len > 0) {
+    ssize_t bytes_read = read(fd, (char*)buf + ofs, len);
+
+    if (bytes_read == 0) {
+      GOOGLE_LOG(FATAL) << "Unexpected EOF from test program";
+    } else if (bytes_read < 0) {
+      GOOGLE_LOG(FATAL) << "Error reading from test program: " << strerror(errno);
+    }
+
+    len -= bytes_read;
+    ofs += bytes_read;
+  }
+}
+
+void RunTest(const ConformanceRequest& request, ConformanceResponse* response) {
+  string serialized;
+  request.SerializeToString(&serialized);
+  uint32_t len = serialized.size();
+  CheckedWrite(write_fd, &len, sizeof(uint32_t));
+  CheckedWrite(write_fd, serialized.c_str(), serialized.size());
+  CheckedRead(read_fd, &len, sizeof(uint32_t));
+  serialized.resize(len);
+  CheckedRead(read_fd, (void*)serialized.c_str(), len);
+  if (!response->ParseFromString(serialized)) {
+    GOOGLE_LOG(FATAL) << "Could not parse response proto from tested process.";
+  }
+
+  if (verbose) {
+    fprintf(stderr, "conformance_test: request=%s, response=%s\n",
+            request.ShortDebugString().c_str(),
+            response->ShortDebugString().c_str());
+  }
+}
+
+void DoExpectParseFailureForProto(const string& proto, int line) {
+  ConformanceRequest request;
+  ConformanceResponse response;
+  request.set_protobuf_payload(proto);
+
+  // We don't expect output, but if the program erroneously accepts the protobuf
+  // we let it send its response as this.  We must not leave it unspecified.
+  request.set_requested_output(ConformanceRequest::PROTOBUF);
+
+  RunTest(request, &response);
+  if (response.result_case() == ConformanceResponse::kParseError) {
+    ReportSuccess();
+  } else {
+    ReportFailure("Should have failed, but didn't. Line: %d, Request: %s, "
+                  "response: %s\n",
+                  line,
+                  request.ShortDebugString().c_str(),
+                  response.ShortDebugString().c_str());
+  }
+}
+
+// Expect that this precise protobuf will cause a parse error.
+#define ExpectParseFailureForProto(proto) DoExpectParseFailureForProto(proto, __LINE__)
+
+// Expect that this protobuf will cause a parse error, even if it is followed
+// by valid protobuf data.  We can try running this twice: once with this
+// data verbatim and once with this data followed by some valid data.
+//
+// TODO(haberman): implement the second of these.
+#define ExpectHardParseFailureForProto(proto) DoExpectParseFailureForProto(proto, __LINE__)
+
+
+/* Routines for building arbitrary protos *************************************/
+
+// We would use CodedOutputStream except that we want more freedom to build
+// arbitrary protos (even invalid ones).
+
+const string empty;
+
+string cat(const string& a, const string& b,
+           const string& c = empty,
+           const string& d = empty,
+           const string& e = empty,
+           const string& f = empty,
+           const string& g = empty,
+           const string& h = empty,
+           const string& i = empty,
+           const string& j = empty,
+           const string& k = empty,
+           const string& l = empty) {
+  string ret;
+  ret.reserve(a.size() + b.size() + c.size() + d.size() + e.size() + f.size() +
+              g.size() + h.size() + i.size() + j.size() + k.size() + l.size());
+  ret.append(a);
+  ret.append(b);
+  ret.append(c);
+  ret.append(d);
+  ret.append(e);
+  ret.append(f);
+  ret.append(g);
+  ret.append(h);
+  ret.append(i);
+  ret.append(j);
+  ret.append(k);
+  ret.append(l);
+  return ret;
+}
+
+// The maximum number of bytes that it takes to encode a 64-bit varint.
+#define VARINT_MAX_LEN 10
+
+size_t vencode64(uint64_t val, char *buf) {
+  if (val == 0) { buf[0] = 0; return 1; }
+  size_t i = 0;
+  while (val) {
+    uint8_t byte = val & 0x7fU;
+    val >>= 7;
+    if (val) byte |= 0x80U;
+    buf[i++] = byte;
+  }
+  return i;
+}
+
+string varint(uint64_t x) {
+  char buf[VARINT_MAX_LEN];
+  size_t len = vencode64(x, buf);
+  return string(buf, len);
+}
+
+// TODO: proper byte-swapping for big-endian machines.
+string fixed32(void *data) { return string(static_cast<char*>(data), 4); }
+string fixed64(void *data) { return string(static_cast<char*>(data), 8); }
+
+string delim(const string& buf) { return cat(varint(buf.size()), buf); }
+string uint32(uint32_t u32) { return fixed32(&u32); }
+string uint64(uint64_t u64) { return fixed64(&u64); }
+string flt(float f) { return fixed32(&f); }
+string dbl(double d) { return fixed64(&d); }
+string zz32(int32_t x) { return varint(WireFormatLite::ZigZagEncode32(x)); }
+string zz64(int64_t x) { return varint(WireFormatLite::ZigZagEncode64(x)); }
+
+string tag(uint32_t fieldnum, char wire_type) {
+  return varint((fieldnum << 3) | wire_type);
+}
+
+string submsg(uint32_t fn, const string& buf) {
+  return cat( tag(fn, WireFormatLite::WIRETYPE_LENGTH_DELIMITED), delim(buf) );
+}
+
+#define UNKNOWN_FIELD 666
+
+uint32_t GetFieldNumberForType(WireFormatLite::FieldType type, bool repeated) {
+  const Descriptor* d = TestAllTypes().GetDescriptor();
+  for (int i = 0; i < d->field_count(); i++) {
+    const FieldDescriptor* f = d->field(i);
+    if (static_cast<WireFormatLite::FieldType>(f->type()) == type &&
+        f->is_repeated() == repeated) {
+      return f->number();
+    }
+  }
+  GOOGLE_LOG(FATAL) << "Couldn't find field with type " << (int)type;
+  return 0;
+}
+
+void TestPrematureEOFForType(WireFormatLite::FieldType type) {
+  // Incomplete values for each wire type.
+  static const string incompletes[6] = {
+    string("\x80"),     // VARINT
+    string("abcdefg"),  // 64BIT
+    string("\x80"),     // DELIMITED (partial length)
+    string(),           // START_GROUP (no value required)
+    string(),           // END_GROUP (no value required)
+    string("abc")       // 32BIT
+  };
+
+  uint32_t fieldnum = GetFieldNumberForType(type, false);
+  uint32_t rep_fieldnum = GetFieldNumberForType(type, true);
+  WireFormatLite::WireType wire_type =
+      WireFormatLite::WireTypeForFieldType(type);
+  const string& incomplete = incompletes[wire_type];
+
+  // EOF before a known non-repeated value.
+  ExpectParseFailureForProto(tag(fieldnum, wire_type));
+
+  // EOF before a known repeated value.
+  ExpectParseFailureForProto(tag(rep_fieldnum, wire_type));
+
+  // EOF before an unknown value.
+  ExpectParseFailureForProto(tag(UNKNOWN_FIELD, wire_type));
+
+  // EOF inside a known non-repeated value.
+  ExpectParseFailureForProto(
+      cat( tag(fieldnum, wire_type), incomplete ));
+
+  // EOF inside a known repeated value.
+  ExpectParseFailureForProto(
+      cat( tag(rep_fieldnum, wire_type), incomplete ));
+
+  // EOF inside an unknown value.
+  ExpectParseFailureForProto(
+      cat( tag(UNKNOWN_FIELD, wire_type), incomplete ));
+
+  if (wire_type == WireFormatLite::WIRETYPE_LENGTH_DELIMITED) {
+    // EOF in the middle of delimited data for known non-repeated value.
+    ExpectParseFailureForProto(
+        cat( tag(fieldnum, wire_type), varint(1) ));
+
+    // EOF in the middle of delimited data for known repeated value.
+    ExpectParseFailureForProto(
+        cat( tag(rep_fieldnum, wire_type), varint(1) ));
+
+    // EOF in the middle of delimited data for unknown value.
+    ExpectParseFailureForProto(
+        cat( tag(UNKNOWN_FIELD, wire_type), varint(1) ));
+
+    if (type == WireFormatLite::TYPE_MESSAGE) {
+      // Submessage ends in the middle of a value.
+      string incomplete_submsg =
+          cat( tag(WireFormatLite::TYPE_INT32, WireFormatLite::WIRETYPE_VARINT),
+                incompletes[WireFormatLite::WIRETYPE_VARINT] );
+      ExpectHardParseFailureForProto(
+          cat( tag(fieldnum, WireFormatLite::WIRETYPE_LENGTH_DELIMITED),
+               varint(incomplete_submsg.size()),
+               incomplete_submsg ));
+    }
+  } else if (type != WireFormatLite::TYPE_GROUP) {
+    // Non-delimited, non-group: eligible for packing.
+
+    // Packed region ends in the middle of a value.
+    ExpectHardParseFailureForProto(
+        cat( tag(rep_fieldnum, WireFormatLite::WIRETYPE_LENGTH_DELIMITED),
+             varint(incomplete.size()),
+             incomplete ));
+
+    // EOF in the middle of packed region.
+    ExpectParseFailureForProto(
+        cat( tag(rep_fieldnum, WireFormatLite::WIRETYPE_LENGTH_DELIMITED),
+             varint(1) ));
+  }
+}
+
+
+int main(int argc, char *argv[]) {
+  if (argc < 2) {
+    fprintf(stderr, "Usage: conformance_test <test-program>\n");
+    exit(1);
+  }
+
+  SpawnTestProgram(argv[1]);
+
+  for (int i = 1; i <= FieldDescriptor::MAX_TYPE; i++) {
+    TestPrematureEOFForType(static_cast<WireFormatLite::FieldType>(i));
+  }
+
+  fprintf(stderr, "conformance_test: completed %d tests for %s, %d successes, "
+                  "%d failures.\n", successes + failures, argv[1], successes,
+                                     failures);
+}