From 2cde63a25ced0c252ff1929be7c242c6b40e3d9f Mon Sep 17 00:00:00 2001
From: avogar <avogar@clickhouse.com>
Date: Mon, 3 Apr 2023 16:25:22 +0000
Subject: [PATCH] Avoid abort in protobuf library in debug build

---
 src/Formats/ProtobufSchemas.cpp               | 31 ++++++++++++++++---
 .../02705_protobuf_debug_abort.reference      |  1 +
 .../0_stateless/02705_protobuf_debug_abort.sh | 18 +++++++++++
 3 files changed, 45 insertions(+), 5 deletions(-)
 create mode 100644 tests/queries/0_stateless/02705_protobuf_debug_abort.reference
 create mode 100755 tests/queries/0_stateless/02705_protobuf_debug_abort.sh

diff --git a/src/Formats/ProtobufSchemas.cpp b/src/Formats/ProtobufSchemas.cpp
index efc0a4e694f..048d88ff2d1 100644
--- a/src/Formats/ProtobufSchemas.cpp
+++ b/src/Formats/ProtobufSchemas.cpp
@@ -41,8 +41,19 @@ public:
             return descriptor;
 
         const auto * file_descriptor = importer.Import(schema_path);
-        // If there are parsing errors, AddError() throws an exception and in this case the following line
-        // isn't executed.
+        if (error)
+        {
+            auto info = error.value();
+            error.reset();
+            throw Exception(
+                ErrorCodes::CANNOT_PARSE_PROTOBUF_SCHEMA,
+                "Cannot parse '{}' file, found an error at line {}, column {}, {}",
+                info.filename,
+                std::to_string(info.line),
+                std::to_string(info.column),
+                info.message);
+        }
+
         assert(file_descriptor);
 
         if (with_envelope == WithEnvelope::No)
@@ -74,14 +85,24 @@ private:
     // Overrides google::protobuf::compiler::MultiFileErrorCollector:
     void AddError(const String & filename, int line, int column, const String & message) override
     {
-        throw Exception(ErrorCodes::CANNOT_PARSE_PROTOBUF_SCHEMA,
-                        "Cannot parse '{}' file, found an error at line {}, column {}, {}",
-                        filename, std::to_string(line), std::to_string(column), message);
+        /// Protobuf library code is not exception safe, we should
+        /// remember error and throw in later from our side.
+        error = ErrorInfo{filename, line, column, message};
     }
 
     google::protobuf::compiler::DiskSourceTree disk_source_tree;
     google::protobuf::compiler::Importer importer;
     const WithEnvelope with_envelope;
+
+    struct ErrorInfo
+    {
+        String filename;
+        int line;
+        int column;
+        String message;
+    };
+
+    std::optional<ErrorInfo> error;
 };
 
 
diff --git a/tests/queries/0_stateless/02705_protobuf_debug_abort.reference b/tests/queries/0_stateless/02705_protobuf_debug_abort.reference
new file mode 100644
index 00000000000..d00491fd7e5
--- /dev/null
+++ b/tests/queries/0_stateless/02705_protobuf_debug_abort.reference
@@ -0,0 +1 @@
+1
diff --git a/tests/queries/0_stateless/02705_protobuf_debug_abort.sh b/tests/queries/0_stateless/02705_protobuf_debug_abort.sh
new file mode 100755
index 00000000000..4a66cfca352
--- /dev/null
+++ b/tests/queries/0_stateless/02705_protobuf_debug_abort.sh
@@ -0,0 +1,18 @@
+#!/usr/bin/env bash
+# Tags: no-fasttest
+
+CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
+# shellcheck source=../shell_config.sh
+. "$CURDIR"/../shell_config.sh
+
+echo 'syntax = "proto3";
+
+message Message {
+    NotExisted x = 1;
+}' > 02705_schema.proto
+
+
+$CLICKHOUSE_LOCAL -q "select * from file(data.bin, Protobuf) settings format_schema='schema:Message'" 2>&1 | grep -c "CANNOT_PARSE_PROTOBUF_SCHEMA"
+
+rm 02705_schema.proto
+