From ca1d38914ab622f9b46c7d0417958ecc5d846d3a Mon Sep 17 00:00:00 2001
From: Alexey Milovidov <milovidov@yandex-team.ru>
Date: Fri, 30 Nov 2018 22:37:31 +0300
Subject: [PATCH] Don't log to stderr within RE2 library [#CLICKHOUSE-2]

---
 dbms/src/Common/ErrorCodes.cpp                |  1 +
 ...n.inl.h => OptimizedRegularExpression.cpp} | 34 ++++++++++++-------
 dbms/src/Common/OptimizedRegularExpression.h  |  8 ++---
 3 files changed, 25 insertions(+), 18 deletions(-)
 rename dbms/src/Common/{OptimizedRegularExpression.inl.h => OptimizedRegularExpression.cpp} (93%)

diff --git a/dbms/src/Common/ErrorCodes.cpp b/dbms/src/Common/ErrorCodes.cpp
index 719e593e3b1..669e7f9c36b 100644
--- a/dbms/src/Common/ErrorCodes.cpp
+++ b/dbms/src/Common/ErrorCodes.cpp
@@ -401,6 +401,7 @@ namespace ErrorCodes
     extern const int CANNOT_LINK = 424;
     extern const int SYSTEM_ERROR = 425;
     extern const int NULL_POINTER_DEREFERENCE = 426;
+    extern const int CANNOT_COMPILE_REGEXP = 427;
 
     extern const int KEEPER_EXCEPTION = 999;
     extern const int POCO_EXCEPTION = 1000;
diff --git a/dbms/src/Common/OptimizedRegularExpression.inl.h b/dbms/src/Common/OptimizedRegularExpression.cpp
similarity index 93%
rename from dbms/src/Common/OptimizedRegularExpression.inl.h
rename to dbms/src/Common/OptimizedRegularExpression.cpp
index 04860622fe7..16251ae1ee0 100644
--- a/dbms/src/Common/OptimizedRegularExpression.inl.h
+++ b/dbms/src/Common/OptimizedRegularExpression.cpp
@@ -1,7 +1,4 @@
-#include <iostream>
-
-#include <Poco/Exception.h>
-
+#include <Common/Exception.h>
 #include <Common/OptimizedRegularExpression.h>
 
 
@@ -9,6 +6,15 @@
 #define MAX_SUBPATTERNS 5
 
 
+namespace DB
+{
+    namespace ErrorCodes
+    {
+        extern const int CANNOT_COMPILE_REGEXP;
+    }
+}
+
+
 template <bool thread_safe>
 void OptimizedRegularExpressionImpl<thread_safe>::analyze(
     const std::string & regexp,
@@ -254,11 +260,11 @@ OptimizedRegularExpressionImpl<thread_safe>::OptimizedRegularExpressionImpl(cons
 
     /// Just three following options are supported
     if (options & (~(RE_CASELESS | RE_NO_CAPTURE | RE_DOT_NL)))
-        throw Poco::Exception("OptimizedRegularExpression: Unsupported option.");
+        throw DB::Exception("OptimizedRegularExpression: Unsupported option.", DB::ErrorCodes::CANNOT_COMPILE_REGEXP);
 
-    is_case_insensitive   = options & RE_CASELESS;
-    bool is_no_capture    = options & RE_NO_CAPTURE;
-    bool is_dot_nl        = options & RE_DOT_NL;
+    is_case_insensitive = options & RE_CASELESS;
+    bool is_no_capture = options & RE_NO_CAPTURE;
+    bool is_dot_nl = options & RE_DOT_NL;
 
     number_of_subpatterns = 0;
     if (!is_trivial)
@@ -266,6 +272,9 @@ OptimizedRegularExpressionImpl<thread_safe>::OptimizedRegularExpressionImpl(cons
         /// Compile the re2 regular expression.
         typename RegexType::Options regexp_options;
 
+        /// Never write error messages to stderr. It's ignorant to do it from library code.
+        regexp_options.set_log_errors(false);
+
         if (is_case_insensitive)
             regexp_options.set_case_sensitive(false);
 
@@ -274,13 +283,13 @@ OptimizedRegularExpressionImpl<thread_safe>::OptimizedRegularExpressionImpl(cons
 
         re2 = std::make_unique<RegexType>(regexp_, regexp_options);
         if (!re2->ok())
-            throw Poco::Exception("OptimizedRegularExpression: cannot compile re2: " + regexp_ + ", error: " + re2->error());
+            throw DB::Exception("OptimizedRegularExpression: cannot compile re2: " + regexp_ + ", error: " + re2->error() + ". Look at https://github.com/google/re2/wiki/Syntax for reference.", DB::ErrorCodes::CANNOT_COMPILE_REGEXP);
 
         if (!is_no_capture)
         {
             number_of_subpatterns = re2->NumberOfCapturingGroups();
             if (number_of_subpatterns > MAX_SUBPATTERNS)
-                throw Poco::Exception("OptimizedRegularExpression: too many subpatterns in regexp: " + regexp_);
+                throw DB::Exception("OptimizedRegularExpression: too many subpatterns in regexp: " + regexp_, DB::ErrorCodes::CANNOT_COMPILE_REGEXP);
         }
     }
 }
@@ -432,6 +441,5 @@ unsigned OptimizedRegularExpressionImpl<thread_safe>::match(const char * subject
     }
 }
 
-#undef MIN_LENGTH_FOR_STRSTR
-#undef MAX_SUBPATTERNS
-
+template class OptimizedRegularExpressionImpl<true>;
+template class OptimizedRegularExpressionImpl<false>;
diff --git a/dbms/src/Common/OptimizedRegularExpression.h b/dbms/src/Common/OptimizedRegularExpression.h
index 60566c9f029..729711277b1 100644
--- a/dbms/src/Common/OptimizedRegularExpression.h
+++ b/dbms/src/Common/OptimizedRegularExpression.h
@@ -45,9 +45,9 @@ class OptimizedRegularExpressionImpl
 public:
     enum Options
     {
-        RE_CASELESS        = 0x00000001,
-        RE_NO_CAPTURE    = 0x00000010,
-        RE_DOT_NL        = 0x00000100
+        RE_CASELESS   = 0x00000001,
+        RE_NO_CAPTURE = 0x00000010,
+        RE_DOT_NL     = 0x00000100
     };
 
     using Match = OptimizedRegularExpressionDetails::Match;
@@ -106,5 +106,3 @@ private:
 };
 
 using OptimizedRegularExpression = OptimizedRegularExpressionImpl<true>;
-
-#include "OptimizedRegularExpression.inl.h"