Feature #1113 - SQL Schemas

builds/install/misc/replication.conf

@@ -11,15 +11,25 @@ database
 	#
 	# plugin =
 
+	# Pattern (regular expression) that defines what schemas must be included into
+	# replication. By default, tables from all schemas are replicated.
+	#
+	# include_schema_filter =
+
+	# Pattern (regular expression) that defines what schemas must be excluded from
+	# replication. By default, tables from all schemas are replicated.
+	#
+	# exclude_schema_filter =
+
 	# Pattern (regular expression) that defines what tables must be included into
 	# replication. By default, all tables are replicated.
 	#
-	# include_filter = 
+	# include_filter =
 
 	# Pattern (regular expression) that defines what tables must be excluded from
 	# replication. By default, all tables are replicated.
 	#
-	# exclude_filter = 
+	# exclude_filter =
 
 	# Boolean parameters describing how replication errors must be handled.
 	#
@@ -42,20 +52,20 @@ database
 
 	# Directory to store replication journal files.
 	#
-	# journal_directory = 
+	# journal_directory =
 
 	# Prefix for replication journal file names. It will be automatically suffixed
 	# with an ordinal sequential number. If not specified, database filename
 	# (without path) is used as a prefix.
 	#
-	# journal_file_prefix =  
+	# journal_file_prefix =
 
 	# Maximum allowed size for a single replication segment.
 	#
 	# journal_segment_size = 16777216 # 16MB
 
 	# Maximum allowed number of full replication segments. Once this limit is reached,
-	# the replication process is temporarily delayed to allow the archiving to catch up. 
+	# the replication process is temporarily delayed to allow the archiving to catch up.
  	# If any of the full segments is not archived during one minute,
  	# the replication fails with an error.
  	#
@@ -76,7 +86,7 @@ database
 	# Directory to store archived replication segments.
 	# It also defines the $(archpathname) substitution macro (see below).
 	#
-	# journal_archive_directory = 
+	# journal_archive_directory =
 
 	# Program (complete command line with arguments) that is executed when some
 	# replication segment gets full and needs archiving.
@@ -97,7 +107,7 @@ database
 	# or
 	# Windows: "copy $(pathname) $(archivepathname)"
 	#
-	# journal_archive_command = 
+	# journal_archive_command =
 
 	# Timeout, in seconds, to wait until incomplete segment is scheduled for archiving.
 	# It allows to minimize the replication gap if the database is modified rarely.
@@ -121,7 +131,7 @@ database
 	#
 	# Multiple entries are allowed (for different synchronous replicas).
 	#
-	# sync_replica = 
+	# sync_replica =
 	#
 	# It's also possible to configure replicas as separate sub-sections, e.g.:
 	#
@@ -175,13 +185,13 @@ database
 
 	# Directory to search for the journal files to be replicated.
 	#
-	# journal_source_directory = 
+	# journal_source_directory =
 
 	# Filter to limit replication to the particular source database (based on its GUID).
 	# Expected format: "{XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX}"
 	# Note that double quotes are mandatory, as well as curly braces.
 	#
-	# source_guid = 
+	# source_guid =
 
 	# If enabled, replication.log contains the detailed log of operations performed
 	# by the replication server. Otherwise (by default), only errors and warnings are logged.
@@ -202,6 +212,14 @@ database
 	# then reconnects back and tries to re-apply the latest segments from the point of failure.
 	#
 	# apply_error_timeout = 60
+
+	# Schema search path for compatibility with Firebird versions below 6.0
+	#
+	# Firebird master databases below v6 has no schemas, so use this search path in the replica to
+	# locate the objects.
+	# Used only with asynchronous replication.
+	#
+	# schema_search_path =
 }
 
 #

builds/win32/msvc15/common.vcxproj

@@ -144,7 +144,7 @@
     <ClInclude Include="..\..\..\src\common\classes\NoThrowTimeStamp.h" />
     <ClInclude Include="..\..\..\src\common\classes\objects_array.h" />
     <ClInclude Include="..\..\..\src\common\classes\ParsedList.h" />
-    <ClInclude Include="..\..\..\src\common\classes\QualifiedName.h" />
+    <ClInclude Include="..\..\..\src\common\classes\QualifiedMetaString.h" />
     <ClInclude Include="..\..\..\src\common\classes\RefCounted.h" />
     <ClInclude Include="..\..\..\src\common\classes\RefMutex.h" />
     <ClInclude Include="..\..\..\src\common\classes\rwlock.h" />

builds/win32/msvc15/common.vcxproj.filters

@@ -449,9 +449,6 @@
     <ClInclude Include="..\..\..\src\common\classes\objects_array.h">
       <Filter>headers</Filter>
     </ClInclude>
-    <ClInclude Include="..\..\..\src\common\classes\QualifiedName.h">
-      <Filter>headers</Filter>
-    </ClInclude>
     <ClInclude Include="..\..\..\src\common\classes\RefCounted.h">
       <Filter>headers</Filter>
     </ClInclude>
@@ -611,6 +608,9 @@
     <ClInclude Include="..\..\..\src\common\classes\ParsedList.h">
       <Filter>headers</Filter>
     </ClInclude>
+    <ClInclude Include="..\..\..\src\common\classes\QualifiedMetaString.h">
+      <Filter>headers</Filter>
+    </ClInclude>
     <ClInclude Include="..\..\..\src\common\Int128.h">
       <Filter>headers</Filter>
     </ClInclude>

builds/win32/msvc15/common_test.vcxproj

@@ -181,7 +181,11 @@
     <ClCompile Include="..\..\..\src\common\tests\StringTest.cpp" />
     <ClCompile Include="..\..\..\src\common\classes\tests\AlignerTest.cpp" />
     <ClCompile Include="..\..\..\src\common\classes\tests\ArrayTest.cpp" />
+    <ClCompile Include="..\..\..\src\common\classes\tests\ClumpletTest.cpp" />
     <ClCompile Include="..\..\..\src\common\classes\tests\DoublyLinkedListTest.cpp" />
+    <ClCompile Include="..\..\..\src\common\classes\tests\MetaStringTest.cpp" />
+    <ClCompile Include="..\..\..\src\common\classes\tests\QualifiedMetaStringTest.cpp" />
+    <ClCompile Include="..\..\..\src\common\classes\tests\VectorTest.cpp" />
     <ClCompile Include="..\..\..\src\yvalve\gds.cpp" />
   </ItemGroup>
   <ItemGroup>
@@ -197,4 +201,4 @@
   <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
   <ImportGroup Label="ExtensionTargets">
   </ImportGroup>
-</Project>
+</Project>

builds/win32/msvc15/common_test.vcxproj.filters

@@ -30,11 +30,23 @@
     <ClCompile Include="..\..\..\src\common\classes\tests\ArrayTest.cpp">
       <Filter>source</Filter>
     </ClCompile>
+    <ClCompile Include="..\..\..\src\common\classes\tests\ClumpletTest.cpp">
+      <Filter>source</Filter>
+    </ClCompile>
     <ClCompile Include="..\..\..\src\common\classes\tests\DoublyLinkedListTest.cpp">
       <Filter>source</Filter>
     </ClCompile>
+    <ClCompile Include="..\..\..\src\common\classes\tests\MetaStringTest.cpp">
+      <Filter>source</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\src\common\classes\tests\QualifiedMetaStringTest.cpp">
+      <Filter>source</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\src\common\classes\tests\VectorTest.cpp">
+      <Filter>source</Filter>
+    </ClCompile>
     <ClCompile Include="..\..\..\src\yvalve\gds.cpp">
       <Filter>source</Filter>
     </ClCompile>
   </ItemGroup>
-</Project>
+</Project>

doc/README.replication.md

@@ -47,7 +47,8 @@ ALTER DATABASE EXCLUDE ALL FROM PUBLICATION
 -- to disable replication of specific tables: 
 ALTER DATABASE EXCLUDE TABLE T1, T2, T3 FROM PUBLICATION 
 
-Tables enabled for replicated can be additionally filtered using two settings in the configuration file: include\_filter and exclude\_filter. They are regular expressions that are applied to table names and define rules for inclusion table\(s\) into the replication set or excluding them from the replication set.
+Tables enabled for replicated can be additionally filtered using four settings in the configuration file: include\_schema\_filter, exclude\_schema\_filter, include\_filter and exclude\_filter.
+They are regular expressions that are applied to schema and table names and define rules for inclusion table\(s\) into the replication set or excluding them from the replication set.
 
 Synchronous replication can be turned on using the sync\_replica setting \(multiple entries are allowed\). It must specify a connection string to the replica database, prefixed with username/password. In SuperServer and SuperClassic architectures, replica database is being internally attached when the first user gets connected to the master database and detached when the last user disconnects from the master database. In Classic Server architecture, every server process keeps an active connection to the replica database.
 

doc/sql.extensions/README.ddl.txt

@@ -569,7 +569,7 @@ With support for various system privileges in engine it's getting very convenien
 rights to users already having specific system privilege. Therefore appropriate grantee type is
 suppoprted now. Example:
 
-GRANT ALL ON PLG$SRP_VIEW TO SYSTEM PRIVILEGE USER_MANAGEMENT
+GRANT ALL ON PLG$SRP.PLG$SRP_VIEW TO SYSTEM PRIVILEGE USER_MANAGEMENT
 
 Grants all rights to view (used in SRP management plugin) to users having USER_MANAGEMENT privilege.
 
@@ -715,6 +715,6 @@ ALTER TABLE <table> ADD CONSTRAINT [IF NOT EXISTS] <constraint name> ...
 CREATE [UNIQUE] [ASC[ENDING] | DESC[ENDING]]
   INDEX indexname [{ACTIVE | INACTIVE}]
   ON tablename {(col [, col ...]) | COMPUTED BY (<expression>)}
-  [WHERE <search_condition>] 
+  [WHERE <search_condition>]
 
 'isql -x' generates script accordingly.

doc/sql.extensions/README.ddl_triggers.txt

@@ -119,6 +119,7 @@ DDL_TRIGGER context namespace:
 	- EVENT_TYPE: event type (CREATE, ALTER, DROP)
 	- OBJECT_TYPE: object type (TABLE, VIEW, etc)
 	- DDL_EVENT: event name (<ddl event item>), where <ddl_event_item> is EVENT_TYPE || ' ' || OBJECT_TYPE
+	- SCHEMA_NAME: object's schema name
 	- OBJECT_NAME: metadata object name
 	- OLD_OBJECT_NAME: metadata object name before a rename
 	- NEW_OBJECT_NAME: metadata object name after a rename
@@ -294,51 +295,51 @@ commit;
 
 select id, ddl_event, object_name, old_object_name, new_object_name, sql_text, ok from ddl_log order by id;
 
-                   ID DDL_EVENT                 OBJECT_NAME                     OLD_OBJECT_NAME                 NEW_OBJECT_NAME                          SQL_TEXT OK     
-===================== ========================= =============================== =============================== =============================== ================= ====== 
-                    2 CREATE TABLE              T1                              <null>                          <null>                                       80:0 Y      
+                   ID DDL_EVENT                 OBJECT_NAME                     OLD_OBJECT_NAME                 NEW_OBJECT_NAME                          SQL_TEXT OK
+===================== ========================= =============================== =============================== =============================== ================= ======
+                    2 CREATE TABLE              T1                              <null>                          <null>                                       80:0 Y
 ==============================================================================
-SQL_TEXT:  
+SQL_TEXT:
 recreate table t1 (
     n1 integer,
     n2 integer
 )
 ==============================================================================
-                    3 CREATE TABLE              T1                              <null>                          <null>                                       80:1 N      
+                    3 CREATE TABLE              T1                              <null>                          <null>                                       80:1 N
 ==============================================================================
-SQL_TEXT:  
+SQL_TEXT:
 create table t1 (
     n1 integer,
     n2 integer
 )
 ==============================================================================
-                    4 DROP TABLE                T1                              <null>                          <null>                                       80:2 Y      
+                    4 DROP TABLE                T1                              <null>                          <null>                                       80:2 Y
 ==============================================================================
-SQL_TEXT:  
+SQL_TEXT:
 recreate table t1 (
     n integer
 )
 ==============================================================================
-                    5 CREATE TABLE              T1                              <null>                          <null>                                       80:3 Y      
+                    5 CREATE TABLE              T1                              <null>                          <null>                                       80:3 Y
 ==============================================================================
-SQL_TEXT:  
+SQL_TEXT:
 recreate table t1 (
     n integer
 )
 ==============================================================================
-                    6 CREATE DOMAIN             DOM1                            <null>                          <null>                                       80:4 Y      
+                    6 CREATE DOMAIN             DOM1                            <null>                          <null>                                       80:4 Y
 ==============================================================================
-SQL_TEXT:  
+SQL_TEXT:
 create domain dom1 as integer
 ==============================================================================
-                    7 ALTER DOMAIN              DOM1                            <null>                          <null>                                       80:5 Y      
+                    7 ALTER DOMAIN              DOM1                            <null>                          <null>                                       80:5 Y
 ==============================================================================
-SQL_TEXT:  
+SQL_TEXT:
 alter domain dom1 type bigint
 ==============================================================================
-                    8 ALTER DOMAIN              DOM1                            DOM1                            DOM2                                         80:6 Y      
+                    8 ALTER DOMAIN              DOM1                            DOM1                            DOM2                                         80:6 Y
 ==============================================================================
-SQL_TEXT:  
+SQL_TEXT:
 alter domain dom1 to dom2
 ==============================================================================
 

doc/sql.extensions/README.name_resolution.md

@@ -0,0 +1,68 @@
+# Name resolution (FB 6.0)
+
+With the introduction of schemas in Firebird 6.0, the syntax `<name>.<name>` - used for tables, views, procedures,
+and functions (both standalone and packaged) - introduces ambiguity when resolving object names using the schema
+search path. The ambiguity arises between:
+
+- `<schema>.<object>` (a schema and its object)
+- `<package>.<object>` (a package and its object)
+
+This document focuses on name resolution rules for tables, views, procedures, and functions within queries and
+code blocks.
+
+## Scope specifier (`%`)
+
+To resolve these ambiguities, Firebird introduces a **scope specifier**, represented by the `%` symbol. This
+allows unambiguous referencing of objects.
+
+### Syntax
+
+```sql
+<name> % { SCHEMA | PACKAGE } . <name>
+```
+
+### Examples
+
+```sql
+select *
+    from plg$profiler%schema.plg$prof_sessions;
+
+execute procedure rdb$profiler%package.pause_session;
+
+call rdb$profiler%package.pause_session();
+
+select rdb$time_zone_util%package.database_version()
+    from system%schema.rdb$database;
+
+select *
+    from rdb$time_zone_util%package.transitions('America/Sao_Paulo', timestamp '2017-01-01', timestamp '2019-01-01');
+```
+
+## Detailed name resolution rules
+
+Firebird resolves object names following a structured sequence of rules. Once an object is located, the resolution
+process halts, ensuring no ambiguity errors occur.
+
+- **`name1.name2.name3`**
+  1. Look for routine `name3` inside package `name2`, inside schema `name1`.
+
+- **`name1%schema.name2`**
+  1. Look for object `name2` inside schema `name1`.
+
+- **`name1%package.name2`**
+  1. Look for object `name2` inside a package `name1` using the schema search path.
+
+- **`name1.name2`**
+  1. If inside a package named `name1`, look for routine `name2` in the same package.
+  2. Look in schema `name1` for object `name2`.
+  3. Look for object `name2` inside a package `name1` using the schema search path.
+
+- **`name`**
+  1. Look for subroutine `name`.
+  2. If inside a package, look for routine `name` in the same package.
+  3. Look for object `name` using the schema search path.
+
+> **_Note:_** Object resolution also depends on the context in which they are used. For example, in
+`select * from name1.name2`, `name2` could be a table, view, or procedure. However, in
+`execute procedure name1.name2`, `name2` must be a procedure. This distinction means that an `execute procedure`
+command versus a `select` command can resolve to different objects.