Webware for Python

Update of /cvsroot/webware/Webware/MiddleKit/Design
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv16037/Design

Modified Files:
	MSSQLSQLGenerator.py SQLGenerator.py 
Log Message:
- tweaks


Index: MSSQLSQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/MSSQLSQLGenerator.py,v
retrieving revision 1.17
retrieving revision 1.18
diff -C2 -d -r1.17 -r1.18
*** MSSQLSQLGenerator.py	1 Mar 2004 01:44:34 -0000	1.17
--- MSSQLSQLGenerator.py	1 Mar 2004 05:04:04 -0000	1.18
***************
*** 142,146 ****
  		Returns a one liner that becomes part of the CREATE statement for creating the primary key of the table. SQL generators often override this mix-in method to customize the creation of the primary key for their SQL variant. This method should use self.sqlIdName() and often ljust()s it by self.maxNameWidth().
  		'''
! 		z = '	%s int primary key not null IDENTITY (1, 1),\n' % self.sqlSerialColumnName().ljust(self.maxNameWidth())
  		return z
  
--- 142,146 ----
  		Returns a one liner that becomes part of the CREATE statement for creating the primary key of the table. SQL generators often override this mix-in method to customize the creation of the primary key for their SQL variant. This method should use self.sqlIdName() and often ljust()s it by self.maxNameWidth().
  		'''
! 		z = '	%s int primary key not null identity(1, 1),\n' % self.sqlSerialColumnName().ljust(self.maxNameWidth())
  		return z
  

Index: SQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/SQLGenerator.py,v
retrieving revision 1.44
retrieving revision 1.45
diff -C2 -d -r1.44 -r1.45
*** SQLGenerator.py	1 Mar 2004 01:44:34 -0000	1.44
--- SQLGenerator.py	1 Mar 2004 05:04:04 -0000	1.45
***************
*** 674,678 ****
  			else:
  				objIdRef = ''
! 			out.write('\t%s %s%s%s references _MKClassIds,\n' % (classIdName, self.sqlType(), notNull, classIdDefault))
  			out.write('\t%s %s%s%s' % (objIdName, self.sqlType(), notNull, objIdRef))
  
--- 674,678 ----
  			else:
  				objIdRef = ''
! 			out.write('\t%s %s%s%s references _MKClassIds, /* %s */ \n' % (classIdName, self.sqlType(), notNull, classIdDefault, self.targetClassName()))
  			out.write('\t%s %s%s%s' % (objIdName, self.sqlType(), notNull, objIdRef))

Update of /cvsroot/webware/Webware/MiddleKit/Docs
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv17967/Docs

Modified Files:
	RelNotes-X.Y.html UsersGuide.html 
Log Message:
- change the default primary key serial number column name to 'serialNum'
- added a new setting, SQLSerialColumnName, that controls that


Index: RelNotes-X.Y.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/RelNotes-X.Y.html,v
retrieving revision 1.8
retrieving revision 1.9
diff -C2 -d -r1.8 -r1.9
*** RelNotes-X.Y.html	28 Feb 2004 03:25:29 -0000	1.8
--- RelNotes-X.Y.html	1 Mar 2004 01:44:34 -0000	1.9
***************
*** 30,37 ****
  
  <a name=Incompatible><h2>Incompatible Changes</h2></a>
  <ul>
! 	<li> <font color=red>There are some MiddleKit improvements that break compatibility with previous versions of MiddleKit, although all are easy to address. Read below.</font>
  </ul>
  
  <h3>Obj Ref Columns</h3>
  
--- 30,48 ----
  
  <a name=Incompatible><h2>Incompatible Changes</h2></a>
+ 
+ <p> <font color=red>There are some MiddleKit improvements that break compatibility with previous versions of MiddleKit, although all are easy to address.</font> In a nutshell, if you are upgrading MiddleKit for an existing project, take the following steps:
+ 
  <ul>
! 	<li> Add these to Settings.config inside the model:
! <pre>{
! 	'SQLSerialColumnName': '%(className)sId',
! 	'UseBigIntObjRefColumns': True,
! }</pre>
! 	<li> If your model has a SQLGenerator.config file, move its contents to Settings.config.
! 	<li> If you use Microsoft SQL Server and your model uses the float type then read the section below detailing that.
  </ul>
  
+ <p> Below are more details about the incompatible changes. Below that are more general notes on this release.
+ 
  <h3>Obj Ref Columns</h3>
  
***************
*** 45,48 ****
--- 56,63 ----
  
  
+ <h3>Serial Number Column</h3>
+ 
+ <p> MiddleKit creates a SQL table for every concrete class, and each table has a column for the serial number of a record. This column is the primary key for the table. Its new default name is <span class=name>serialNum</span> which matches the MiddleObject method of the same name and fits MiddleKit naming conventions. You can control this with a <a href=UsersGuide.html#Configuration_SQLSerialColumnName>new setting</a>.
+ 
  <h3>SQLGenerator.config</h3>
  

Index: UsersGuide.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/UsersGuide.html,v
retrieving revision 1.26
retrieving revision 1.27
diff -C2 -d -r1.26 -r1.27
*** UsersGuide.html	28 Feb 2004 01:20:31 -0000	1.26
--- UsersGuide.html	1 Mar 2004 01:44:34 -0000	1.27
***************
*** 313,316 ****
--- 313,324 ----
  
  
+ <p> <a name=Configuration_SQLSerialColumnName></a> The <span class=name>SQLSerialColumnName</span> controls the name that is used for the serial number of a given database record, which is also the primary key. The default is 'serialNum' which matches MiddleKit naming conventions. You can change this:
+ 
+ <pre>{
+     'SQLSerialColumnName': 'SerialNum',  # capitalized, or
+     'SQLSerialColumnName': '%(className)sId',  # the name used by older MiddleKits
+         # you can use className for lower, ClassName for upper, or _ClassName for as-is
+ }</pre>
+ 
  <p> <a name=Configuration_ObjRefSuffixes></a> The <span class=name>ObjRefSuffixes</span> controls the suffixes that are appended for the names of the two SQL column that are created for each obj ref attribute. The suffixes must be different from each other.

Update of /cvsroot/webware/Webware/MiddleKit/Design
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv17967/Design

Modified Files:
	MSSQLSQLGenerator.py MySQLSQLGenerator.py SQLGenerator.py 
Log Message:
- change the default primary key serial number column name to 'serialNum'
- added a new setting, SQLSerialColumnName, that controls that


Index: MSSQLSQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/MSSQLSQLGenerator.py,v
retrieving revision 1.16
retrieving revision 1.17
diff -C2 -d -r1.16 -r1.17
*** MSSQLSQLGenerator.py	24 Feb 2004 02:04:13 -0000	1.16
--- MSSQLSQLGenerator.py	1 Mar 2004 01:44:34 -0000	1.17
***************
*** 138,179 ****
  class Klass:
  
- 
- 
- # Create table.
- #			wr("print 'Creating table %s'\n" % name)
- #			wr('create table [%s] (\n' % name)
- #			wr('	%s int primary key not null IDENTITY (1, 1),\n' % ljust(sqlIdName, self.maxNameWidth()))
- 
- #			for attr in self.allAttrs():
- #				attr.writeSQL(generator, out)
- 
- #			wr('	unique (%s)\n' % sqlIdName)
- #			wr(')\ngo\n\n\n')
- 
- 	def sqlIdName(self):
- 		name = self._name # do not need or want protected name
- 		if name:
- 			name = lower(name[0]) + name[1:] + 'Id'
- 		return name
- 
- 	def maxNameWidth(self):
- 		return 30   # @@ 2000-09-15 ce: Ack! Duplicated from Attr class below
- 
  	def primaryKeySQLDef(self, generator):
  		'''
  		Returns a one liner that becomes part of the CREATE statement for creating the primary key of the table. SQL generators often override this mix-in method to customize the creation of the primary key for their SQL variant. This method should use self.sqlIdName() and often ljust()s it by self.maxNameWidth().
  		'''
! 
! #		print("print 'Creating table %s'\n" % name)
! #		print('create table [%s] (\n' % name)
! 		z = '	%s int primary key not null IDENTITY (1, 1),\n' % self.sqlIdName().ljust(self.maxNameWidth())
! #		print(z)
  		return z
  
- 
- #	def name(self):
- #		return '[' + self._name + ']'
- 
- 
  	def sqlTableName(self):
  		"""
--- 138,148 ----
  class Klass:
  
  	def primaryKeySQLDef(self, generator):
  		'''
  		Returns a one liner that becomes part of the CREATE statement for creating the primary key of the table. SQL generators often override this mix-in method to customize the creation of the primary key for their SQL variant. This method should use self.sqlIdName() and often ljust()s it by self.maxNameWidth().
  		'''
! 		z = '	%s int primary key not null IDENTITY (1, 1),\n' % self.sqlSerialColumnName().ljust(self.maxNameWidth())
  		return z
  
  	def sqlTableName(self):
  		"""

Index: MySQLSQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/MySQLSQLGenerator.py,v
retrieving revision 1.12
retrieving revision 1.13
diff -C2 -d -r1.12 -r1.13
*** MySQLSQLGenerator.py	24 Feb 2004 02:04:13 -0000	1.12
--- MySQLSQLGenerator.py	1 Mar 2004 01:44:34 -0000	1.13
***************
*** 39,43 ****
  
  	def primaryKeySQLDef(self, generator):
! 		return '    %s int not null primary key auto_increment,\n' % self.sqlIdName().ljust(self.maxNameWidth())
  
  
--- 39,43 ----
  
  	def primaryKeySQLDef(self, generator):
! 		return '    %s int not null primary key auto_increment,\n' % self.sqlSerialColumnName().ljust(self.maxNameWidth())
  
  

Index: SQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/SQLGenerator.py,v
retrieving revision 1.43
retrieving revision 1.44
diff -C2 -d -r1.43 -r1.44
*** SQLGenerator.py	28 Feb 2004 03:13:09 -0000	1.43
--- SQLGenerator.py	1 Mar 2004 01:44:34 -0000	1.44
***************
*** 141,145 ****
  					attrs = []
  					for name in names:
! 						if name == klass.sqlIdName():
  							attrs.append(PrimaryKey(name, klass))
  						else:
--- 141,145 ----
  					attrs = []
  					for name in names:
! 						if name==klass.sqlSerialColumnName():
  							attrs.append(PrimaryKey(name, klass))
  						else:
***************
*** 345,348 ****
--- 345,351 ----
  
  
+ import KlassSQLSerialColumnName
+ 
+ 
  class Klass:
  
***************
*** 356,360 ****
  		name = self.name()
  		wr = out.write
- 		sqlIdName = self.sqlIdName()
  		wr('create table %s (\n' % self.sqlTableName())
  		wr(self.primaryKeySQLDef(generator))
--- 359,362 ----
***************
*** 383,403 ****
  	def primaryKeySQLDef(self, generator):
  		"""
! 		Returns a one liner that becomes part of the CREATE statement for creating the primary key of the table. SQL generators often override this mix-in method to customize the creation of the primary key for their SQL variant. This method should use self.sqlIdName() and often ljust()s it by self.maxNameWidth().
  		"""
! 		return '    %s int not null primary key,\n' % self.sqlIdName().ljust(self.maxNameWidth())
  
  	def deletedSQLDef(self, generator):
  		"""
! 		Returns a one liner that becomes part of the CREATE statement for creating the deleted timestamp field of the table.
  		This is used if DeleteBehavior is set to "mark".
  		"""
  		return '    %s datetime null,\n' % ('deleted'.ljust(self.maxNameWidth()))
  
- 	def sqlIdName(self):
- 		name = self.name()
- 		if name:
- 			name = name[0].lower() + name[1:] + 'Id'
- 		return name
- 
  	def maxNameWidth(self):
  		return 30   # @@ 2000-09-15 ce: Ack! Duplicated from Attr class below
--- 385,405 ----
  	def primaryKeySQLDef(self, generator):
  		"""
! 		Returns a one liner that becomes part of the CREATE statement
! 		for creating the primary key of the table. SQL generators often
! 		override this mix-in method to customize the creation of the
! 		primary key for their SQL variant. This method should use
! 		self.sqlSerialColumnName() and often ljust()s it by
! 		self.maxNameWidth().
  		"""
! 		return '    %s int not null primary key,\n' % self.sqlSerialColumnName().ljust(self.maxNameWidth())
  
  	def deletedSQLDef(self, generator):
  		"""
! 		Returns a one liner that becomes part of the CREATE statement
! 		for creating the deleted timestamp field of the table.
  		This is used if DeleteBehavior is set to "mark".
  		"""
  		return '    %s datetime null,\n' % ('deleted'.ljust(self.maxNameWidth()))
  
  	def maxNameWidth(self):
  		return 30   # @@ 2000-09-15 ce: Ack! Duplicated from Attr class below
***************
*** 410,417 ****
  		wr('\n')
  
- 
  	def sqlTableName(self):
  		"""
! 		Can be overiddent to allow for table names that do not conflict with SQL
  		reserved words. dr 08-08-2002 - MSSQL uses [tablename]
  		"""
--- 412,418 ----
  		wr('\n')
  
  	def sqlTableName(self):
  		"""
! 		Can be overidden to allow for table names that do not conflict with SQL
  		reserved words. dr 08-08-2002 - MSSQL uses [tablename]
  		"""

Update of /cvsroot/webware/Webware/MiddleKit/Run
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv17967/Run

Modified Files:
	SQLObjectStore.py 
Log Message:
- change the default primary key serial number column name to 'serialNum'
- added a new setting, SQLSerialColumnName, that controls that


Index: SQLObjectStore.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/SQLObjectStore.py,v
retrieving revision 1.61
retrieving revision 1.62
diff -C2 -d -r1.61 -r1.62
*** SQLObjectStore.py	28 Feb 2004 03:13:09 -0000	1.61
--- SQLObjectStore.py	1 Mar 2004 01:44:34 -0000	1.62
***************
*** 55,61 ****
  		assert sourceKlass
  		sourceTableName = sourceKlass.sqlTableName()
! 		sourceSqlIdName = sourceKlass.sqlIdName()
  		return 'update %s set %s where %s=%s;' % (
! 			sourceTableName, self.sourceAttr.sqlUpdateExpr(self.targetObject), sourceSqlIdName, self.sourceObject.serialNum())
  
  	def __repr__(self):
--- 55,61 ----
  		assert sourceKlass
  		sourceTableName = sourceKlass.sqlTableName()
! 		sourceSqlSerialName = sourceKlass.sqlSerialColumnName()
  		return 'update %s set %s where %s=%s;' % (
! 			sourceTableName, self.sourceAttr.sqlUpdateExpr(self.targetObject), sourceSqlSerialName, self.sourceObject.serialNum())
  
  	def __repr__(self):
***************
*** 293,297 ****
  			className = klass.name()
  			if serialNum is not None:
! 				clauses = 'where %s=%d' % (klass.sqlIdName(), serialNum)
  			if self._markDeletes:
  				clauses = self.addDeletedToClauses(clauses)
--- 293,297 ----
  			className = klass.name()
  			if serialNum is not None:
! 				clauses = 'where %s=%d' % (klass.sqlSerialColumnName(), serialNum)
  			if self._markDeletes:
  				clauses = self.addDeletedToClauses(clauses)
***************
*** 452,456 ****
  				return obj
  
! 			clauses = 'where %s=%d' % (klass.sqlIdName(), serialNum)
  			objs = self.fetchObjectsOfClass(klass, clauses, isDeep=0)
  			if len(objs)==1:
--- 452,456 ----
  				return obj
  
! 			clauses = 'where %s=%d' % (klass.sqlSerialColumnName(), serialNum)
  			objs = self.fetchObjectsOfClass(klass, clauses, isDeep=0)
  			if len(objs)==1:
***************
*** 515,519 ****
  			attrs = [ attr for attr in klass.allAttrs() if attr.hasSQLColumn() ]
  			colNames = [ attr.name() for attr in attrs ]
! 			colNames.insert(0, klass.sqlIdName())
  			out.write(CSVJoiner.joinCSVFields(colNames) + "\n")
  
--- 515,519 ----
  			attrs = [ attr for attr in klass.allAttrs() if attr.hasSQLColumn() ]
  			colNames = [ attr.name() for attr in attrs ]
! 			colNames.insert(0, klass.sqlSerialColumnName())
  			out.write(CSVJoiner.joinCSVFields(colNames) + "\n")
  
***************
*** 609,613 ****
  			res.append(attr.sqlUpdateExpr(self.valueForAttr(attr)))
  		res = ','.join(res)
! 		res = ('update ', klass.sqlTableName(), ' set ', res, ' where ', klass.sqlIdName(), '=', str(self.serialNum()))
  		return ''.join(res)
  
--- 609,613 ----
  			res.append(attr.sqlUpdateExpr(self.valueForAttr(attr)))
  		res = ','.join(res)
! 		res = ('update ', klass.sqlTableName(), ' set ', res, ' where ', klass.sqlSerialColumnName(), '=', str(self.serialNum()))
  		return ''.join(res)
  
***************
*** 623,629 ****
  		assert klass is not None
  		if self.store().model().setting('DeleteBehavior', 'delete')=='mark':
! 			return 'update %s set deleted=%s where %s=%d;' % (klass.sqlTableName(), self.store().sqlNowCall(), klass.sqlIdName(), self.serialNum())
  		else:
! 			return 'delete from %s where %s=%d;' % (klass.sqlTableName(), klass.sqlIdName(), self.serialNum())
  
  	def referencingObjectsAndAttrsFetchKeywordArgs(self, backObjRefAttr):
--- 623,629 ----
  		assert klass is not None
  		if self.store().model().setting('DeleteBehavior', 'delete')=='mark':
! 			return 'update %s set deleted=%s where %s=%d;' % (klass.sqlTableName(), self.store().sqlNowCall(), klass.sqlSerialColumnName(), self.serialNum())
  		else:
! 			return 'delete from %s where %s=%d;' % (klass.sqlTableName(), klass.sqlSerialColumnName(), self.serialNum())
  
  	def referencingObjectsAndAttrsFetchKeywordArgs(self, backObjRefAttr):
***************
*** 646,649 ****
--- 646,652 ----
  
  
+ import MiddleKit.Design.KlassSQLSerialColumnName
+ 
+ 
  class Klass:
  
***************
*** 660,672 ****
  		return self.name()
  
- 	def sqlIdName(self):
- 		name = self.name()
- 		return name[0].lower() + name[1:] + 'Id'
- 
  	def fetchSQLStart(self):
  		if self._fetchSQLStart is None:
  			attrs = self.allDataAttrs()
  			attrs = [attr for attr in attrs if attr.hasSQLColumn()]
! 			colNames = [self.sqlIdName()]
  			colNames.extend([attr.sqlColumnName() for attr in attrs])
  			self._fetchSQLStart = 'select %s from %s ' % (','.join(colNames), self.sqlTableName())
--- 663,671 ----
  		return self.name()
  
  	def fetchSQLStart(self):
  		if self._fetchSQLStart is None:
  			attrs = self.allDataAttrs()
  			attrs = [attr for attr in attrs if attr.hasSQLColumn()]
! 			colNames = [self.sqlSerialColumnName()]
  			colNames.extend([attr.sqlColumnName() for attr in attrs])
  			self._fetchSQLStart = 'select %s from %s ' % (','.join(colNames), self.sqlTableName())
***************
*** 683,687 ****
  			fieldNames = [attr.sqlColumnName() for attr in attrs]
  			if len(fieldNames)==0:
! 				fieldNames = [self.sqlIdName()]
  			res.append(','.join(fieldNames))
  			res.append(') values (')
--- 682,686 ----
  			fieldNames = [attr.sqlColumnName() for attr in attrs]
  			if len(fieldNames)==0:
! 				fieldNames = [self.sqlSerialColumnName()]
  			res.append(','.join(fieldNames))
  			res.append(') values (')

Update of /cvsroot/webware/Webware/MiddleKit/Core
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv17967/Core

Modified Files:
	Klasses.py Model.py 
Log Message:
- change the default primary key serial number column name to 'serialNum'
- added a new setting, SQLSerialColumnName, that controls that


Index: Klasses.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/Klasses.py,v
retrieving revision 1.13
retrieving revision 1.14
diff -C2 -d -r1.13 -r1.14
*** Klasses.py	19 Dec 2003 05:20:19 -0000	1.13
--- Klasses.py	1 Mar 2004 01:44:34 -0000	1.14
***************
*** 64,67 ****
--- 64,70 ----
  	## Accessing ##
  
+ 	def model(self):
+ 		return self._model
+ 
  	def filename(self):
  		return self._filename

Index: Model.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/Model.py,v
retrieving revision 1.24
retrieving revision 1.25
diff -C2 -d -r1.24 -r1.25
*** Model.py	28 Feb 2004 00:08:07 -0000	1.24
--- Model.py	1 Mar 2004 01:44:34 -0000	1.25
***************
*** 328,331 ****
--- 328,332 ----
  			'PostSQL': '',
  			'DropStatements': 'database',  # database, tables
+ 			'SQLSerialColumnName': 'serialNum',  # can use [cC]lassName, _ClassName
  			'ExternalEnumsSQLNames': {
  				'Enable': False,

Update of /cvsroot/webware/Webware/MiddleKit/Tests/MKList.mkmodel
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv17967/Tests/MKList.mkmodel

Modified Files:
	Settings2.config 
Log Message:
- change the default primary key serial number column name to 'serialNum'
- added a new setting, SQLSerialColumnName, that controls that


Index: Settings2.config
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Tests/MKList.mkmodel/Settings2.config,v
retrieving revision 1.1
retrieving revision 1.2
diff -C2 -d -r1.1 -r1.2
*** Settings2.config	24 Feb 2004 02:04:14 -0000	1.1
--- Settings2.config	1 Mar 2004 01:44:34 -0000	1.2
***************
*** 1,3 ****
--- 1,4 ----
  {
  	'UseBigIntObjRefColumns': True,
+ 	'SQLSerialColumnName': '%(className)sId',
  }

Update of /cvsroot/webware/Webware/MiddleKit/Tests/MKObjRef.mkmodel
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv17967/Tests/MKObjRef.mkmodel

Added Files:
	Settings4.config Settings5.config 
Log Message:
- change the default primary key serial number column name to 'serialNum'
- added a new setting, SQLSerialColumnName, that controls that


--- NEW FILE: Settings4.config ---
{
	'UseBigIntObjRefColumns': False,
	'SQLSerialColumnName': '%(className)sId',
}

--- NEW FILE: Settings5.config ---
{
	'UseBigIntObjRefColumns': True,
	'SQLSerialColumnName': '%(className)sId',  # the old approach
}

Update of /cvsroot/webware/Webware/MiddleKit/Docs
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv16724

Modified Files:
	RelNotes-X.Y.html 
Log Message:
- more updates


Index: RelNotes-X.Y.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/RelNotes-X.Y.html,v
retrieving revision 1.7
retrieving revision 1.8
diff -C2 -d -r1.7 -r1.8
*** RelNotes-X.Y.html	28 Feb 2004 01:37:11 -0000	1.7
--- RelNotes-X.Y.html	28 Feb 2004 03:25:29 -0000	1.8
***************
*** 40,43 ****
--- 40,45 ----
  <p> The new approach is to break these values into two SQL columns. If the attribute name is "foo" then the SQL columns will be named "fooClassId" and "fooObjId". If, for some reason, you don't like those suffixes, you can customize them through the setting <a href=UsersGuide.html#Configuration_ObjRefSuffixes >ObjRefSuffixes</a>. As always, the class ids are stored in the _MKClassIds table that MiddleKit automatically generates.
  
+ <p> Another useful benefit is that the fooClassId columns are now given default values to match the class id of the target class (in other words, the integer id of class Foo). When the target class has no subclasses then the fooClassId column can be safely ignored in INSERTs, UPDATEs and SELECTs. The point of this is to make the database even easier to work with from SQL.
+ 
  <p> <font color=red>If you are upgrading Webware then your database schema will no longer be compatible with MiddleKit.</font> The easy way to fix this is to set UseBigIntObjRefColumns to True in Settings.config (see <a href=UsersGuide.html#Configuration_UseBigIntObjRefColumns>User's Guide - Configuration</a> for more information). This will give you the old behavior. The hard way is to fix up your schema and migrate the data from the old columns to the new.

Update of /cvsroot/webware/Webware/MiddleKit/Run
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv14756/Run

Modified Files:
	SQLObjectStore.py 
Log Message:
- make fooClassId SQL columns default to the target class of the
  obj ref attr
- add error checking when loading a model that the target class
  of an obj ref attr actually exists


Index: SQLObjectStore.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/SQLObjectStore.py,v
retrieving revision 1.60
retrieving revision 1.61
diff -C2 -d -r1.60 -r1.61
*** SQLObjectStore.py	24 Feb 2004 02:04:13 -0000	1.60
--- SQLObjectStore.py	28 Feb 2004 03:13:09 -0000	1.61
***************
*** 829,833 ****
  		# this does *not* get called under the old approach of single obj ref columns. see MiddleObject.readStoreData
  		classId, objId = row[i], row[i+1]
! 		if classId is None:
  			value = None
  		else:
--- 829,833 ----
  		# this does *not* get called under the old approach of single obj ref columns. see MiddleObject.readStoreData
  		classId, objId = row[i], row[i+1]
! 		if objId is None:
  			value = None
  		else:

Update of /cvsroot/webware/Webware/MiddleKit/Design
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv14756/Design

Modified Files:
	PythonGenerator.py SQLGenerator.py 
Log Message:
- make fooClassId SQL columns default to the target class of the
  obj ref attr
- add error checking when loading a model that the target class
  of an obj ref attr actually exists


Index: PythonGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/PythonGenerator.py,v
retrieving revision 1.31
retrieving revision 1.32
diff -C2 -d -r1.31 -r1.32
*** PythonGenerator.py	28 Feb 2004 00:08:07 -0000	1.31
--- PythonGenerator.py	28 Feb 2004 03:13:09 -0000	1.32
***************
*** 472,476 ****
  			objSerialNum = parts[1]
  		else:
! 			className = self.className()
  			objSerialNum = string
  		klass = self.klass().klasses()._model.klass(className)
--- 472,476 ----
  			objSerialNum = parts[1]
  		else:
! 			className = self.targetClassName()
  			objSerialNum = string
  		klass = self.klass().klasses()._model.klass(className)
***************
*** 482,486 ****
  		name = self.name()
  		pySetName = self.pySetName()
! 		targetClassName = self.className()
  		package = self.setting('Package', '')
  		if package:
--- 482,486 ----
  		name = self.name()
  		pySetName = self.pySetName()
! 		targetClassName = self.targetClassName()
  		package = self.setting('Package', '')
  		if package:

Index: SQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/SQLGenerator.py,v
retrieving revision 1.42
retrieving revision 1.43
diff -C2 -d -r1.42 -r1.43
*** SQLGenerator.py	28 Feb 2004 01:20:31 -0000	1.42
--- SQLGenerator.py	28 Feb 2004 03:13:09 -0000	1.43
***************
*** 450,461 ****
  				out.write('\t/* %(Name)s %(Type)s - not a SQL column */' % self)
  		except:
  			print
  			print 'exception for attribute:'
  			print '%s.%s' % (self.klass().name(), self.name())
  			print
  			try:
! 				print self
  			except:
  				pass
  			print
  			raise
--- 450,465 ----
  				out.write('\t/* %(Name)s %(Type)s - not a SQL column */' % self)
  		except:
+ 			bar = '*'*78
  			print
+ 			print bar
  			print 'exception for attribute:'
  			print '%s.%s' % (self.klass().name(), self.name())
  			print
  			try:
! 				from pprint import pprint
! 				pprint(self.data)
  			except:
  				pass
+ 			print bar
  			print
  			raise
***************
*** 663,671 ****
  			else:
  				notNull = self.sqlNullSpec()
  			if self.get('Ref', None):
  				objIdRef = ' references %(Type)s(%(Type)sId) ' % self
  			else:
  				objIdRef = ''
! 			out.write('\t%s %s%s references _MKClassIds,\n' % (classIdName, self.sqlType(), notNull))
  			out.write('\t%s %s%s%s' % (objIdName, self.sqlType(), notNull, objIdRef))
  
--- 667,677 ----
  			else:
  				notNull = self.sqlNullSpec()
+ 			classIdDefault = ' default %s' % self.targetKlass().id()
+ 			#   ^ this makes the table a little to easier to work with in some cases (you can often just insert the obj id)
  			if self.get('Ref', None):
  				objIdRef = ' references %(Type)s(%(Type)sId) ' % self
  			else:
  				objIdRef = ''
! 			out.write('\t%s %s%s%s references _MKClassIds,\n' % (classIdName, self.sqlType(), notNull, classIdDefault))
  			out.write('\t%s %s%s%s' % (objIdName, self.sqlType(), notNull, objIdRef))
  
***************
*** 687,691 ****
  			objSerialNum = parts[1]
  		else:
! 			className = self.className()
  			objSerialNum = input
  		klass = self.klass().klasses()._model.klass(className)
--- 693,697 ----
  			objSerialNum = parts[1]
  		else:
! 			className = self.targetClassName()
  			objSerialNum = input
  		klass = self.klass().klasses()._model.klass(className)

Update of /cvsroot/webware/Webware/MiddleKit/Core
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv14756/Core

Modified Files:
	Attr.py Klass.py ObjRefAttr.py 
Log Message:
- make fooClassId SQL columns default to the target class of the
  obj ref attr
- add error checking when loading a model that the target class
  of an obj ref attr actually exists


Index: Attr.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/Attr.py,v
retrieving revision 1.11
retrieving revision 1.12
diff -C2 -d -r1.11 -r1.12
*** Attr.py	16 Feb 2004 22:14:52 -0000	1.11
--- Attr.py	28 Feb 2004 03:13:08 -0000	1.12
***************
*** 78,79 ****
--- 78,82 ----
  	def model(self):
  		return self._klass.klasses()._model
+ 
+ 	def awakeFromRead(self):
+ 		pass

Index: Klass.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/Klass.py,v
retrieving revision 1.18
retrieving revision 1.19
diff -C2 -d -r1.18 -r1.19
*** Klass.py	8 Dec 2003 06:08:40 -0000	1.18
--- Klass.py	28 Feb 2004 03:13:08 -0000	1.19
***************
*** 70,73 ****
--- 70,75 ----
  		# Invoking pyClass() makes that happen.
  		self.pyClass()
+ 		for attr in self.attrs():
+ 			attr.awakeFromRead()
  
  	def _makeAllAttrs(self):
***************
*** 283,287 ****
  				# find all ObjRefAttrs of klass that refer to one of our targetKlasses
  				for attr in klass.attrs():
! 					if isinstance(attr, ObjRefAttr) and targetKlasses.has_key(attr.className()):
  						backObjRefAttrs.append(attr)
  			self._backObjRefAttrs = backObjRefAttrs
--- 285,289 ----
  				# find all ObjRefAttrs of klass that refer to one of our targetKlasses
  				for attr in klass.attrs():
! 					if isinstance(attr, ObjRefAttr) and targetKlasses.has_key(attr.targetClassName()):
  						backObjRefAttrs.append(attr)
  			self._backObjRefAttrs = backObjRefAttrs

Index: ObjRefAttr.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/ObjRefAttr.py,v
retrieving revision 1.2
retrieving revision 1.3
diff -C2 -d -r1.2 -r1.3
*** ObjRefAttr.py	20 Jun 2002 17:47:44 -0000	1.2
--- ObjRefAttr.py	28 Feb 2004 03:13:08 -0000	1.3
***************
*** 21,25 ****
  		self._className = dict['Type']
  
! 	def className(self):
  		""" Returns the name of the base class that this obj ref attribute points to. """
  		return self._className
--- 21,43 ----
  		self._className = dict['Type']
  
! 	def targetClassName(self):
  		""" Returns the name of the base class that this obj ref attribute points to. """
  		return self._className
+ 
+ 	def targetKlass(self):
+ 		assert self._targetKlass, 'not yet fully initialized'
+ 		return self._targetKlass
+ 
+ 	def awakeFromRead(self):
+ 		"""
+ 		Check that the target class actually exists.
+ 		"""
+ 		self._targetKlass = self.model().klass(self.targetClassName(), None)
+ 		if not self._targetKlass:
+ 			from Model import ModelError
+ 			raise ModelError, 'class %s: attr %s: cannot locate target class %s for this obj ref.' % (
+ 				self.klass().name(), self.name(), self.targetClassName())
+ 
+ 	def className(self):
+ 		print 'DEPRECATED: ObjRefAttr.className() on 2004-02-27. Use targetClassName() instead.'
+ 		return self.targetClassName()

Update of /cvsroot/webware/Webware/MiddleKit/Docs
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv32164

Modified Files:
	RelNotes-X.Y.html 
Log Message:
- another edit


Index: RelNotes-X.Y.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/RelNotes-X.Y.html,v
retrieving revision 1.6
retrieving revision 1.7
diff -C2 -d -r1.6 -r1.7
*** RelNotes-X.Y.html	28 Feb 2004 01:31:45 -0000	1.6
--- RelNotes-X.Y.html	28 Feb 2004 01:37:11 -0000	1.7
***************
*** 63,67 ****
  <h2>Improvements and Refinements</h2>
  <ul>
! 	<li><b>Generate.py</b> does a better job of detecting errors in the class definitions and sample values than before. You are more likely to see a useful error message than a cryptic traceback. However, in those cases where an exception does occur, you can see that just before the exception, Generate.py prints the last attribute that it was processing before the error. You can then search for that attribute in you model and troubleshoot it.
  
  </ul>
--- 63,67 ----
  <h2>Improvements and Refinements</h2>
  <ul>
! 	<li><b>Generate.py</b> does a better job of detecting errors in the class definitions and sample values than before. You are more likely to see a useful error message than a cryptic traceback. However, in those cases where an exception does occur, you can see that just before the exception, Generate.py prints the last attribute that it was processing before the error. You can then search for that attribute in your model and troubleshoot it.
  
  </ul>

Update of /cvsroot/webware/Webware/MiddleKit/Docs
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv30459

Modified Files:
	RelNotes-X.Y.html 
Log Message:
- edits


Index: RelNotes-X.Y.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/RelNotes-X.Y.html,v
retrieving revision 1.5
retrieving revision 1.6
diff -C2 -d -r1.5 -r1.6
*** RelNotes-X.Y.html	28 Feb 2004 01:20:31 -0000	1.5
--- RelNotes-X.Y.html	28 Feb 2004 01:31:45 -0000	1.6
***************
*** 36,40 ****
  <h3>Obj Ref Columns</h3>
  
! <p> A long standing flaw in MiddleKit was the use of 64-bit integer SQL columns to store references/pointers to objects. The first 32-bits were for the class id and the second 32-bits were for the object id. These are difficult to work with from SQL. For example, an object reference to (classId=1, objId=1) showed as 4294967297 in SQL and couldn't not be readily joined with other tables.
  
  <p> The new approach is to break these values into two SQL columns. If the attribute name is "foo" then the SQL columns will be named "fooClassId" and "fooObjId". If, for some reason, you don't like those suffixes, you can customize them through the setting <a href=UsersGuide.html#Configuration_ObjRefSuffixes >ObjRefSuffixes</a>. As always, the class ids are stored in the _MKClassIds table that MiddleKit automatically generates.
--- 36,40 ----
  <h3>Obj Ref Columns</h3>
  
! <p> A long standing flaw in MiddleKit was the use of 64-bit integer SQL columns to store references/pointers to objects. The first 32-bits were for the class id and the second 32-bits were for the object id, information that is necessary because of inheritance and the creation of one SQL table per Python class. These 64-bit columns, packed with two numbers, are difficult to work with from SQL. For example, an object reference to (classId=1, objId=1) showed as 4294967297 in SQL and could not be readily joined with other tables.
  
  <p> The new approach is to break these values into two SQL columns. If the attribute name is "foo" then the SQL columns will be named "fooClassId" and "fooObjId". If, for some reason, you don't like those suffixes, you can customize them through the setting <a href=UsersGuide.html#Configuration_ObjRefSuffixes >ObjRefSuffixes</a>. As always, the class ids are stored in the _MKClassIds table that MiddleKit automatically generates.
***************
*** 45,49 ****
  <h3>SQLGenerator.config</h3>
  
! <p> There used to be another config file, besides Settings.config, named SQLGenerator.config, with a single setting, DropStatements. That file is no longer used and the setting has been moved to Settings.config. If you have a SQLGenerator.config in your model directories, move the setting over.
  
  
--- 45,49 ----
  <h3>SQLGenerator.config</h3>
  
! <p> There used to be another config file, besides Settings.config, named SQLGenerator.config, with a single setting, DropStatements. That file is no longer used and the setting has been moved to Settings.config. If you have SQLGenerator.config in a model directory, move the setting over.
  
  
***************
*** 63,69 ****
  <h2>Improvements and Refinements</h2>
  <ul>
! 	<li>Improved error message when there are errors in the model (instead of
! 		cryptic tracebacks).  In many cases the filename of the model and the
! 	line number are printed, along with the error message.
  </ul>
  
--- 63,68 ----
  <h2>Improvements and Refinements</h2>
  <ul>
! 	<li><b>Generate.py</b> does a better job of detecting errors in the class definitions and sample values than before. You are more likely to see a useful error message than a cryptic traceback. However, in those cases where an exception does occur, you can see that just before the exception, Generate.py prints the last attribute that it was processing before the error. You can then search for that attribute in you model and troubleshoot it.
! 
  </ul>

Update of /cvsroot/webware/Webware/MiddleKit/Docs
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv28838/Docs

Modified Files:
	RelNotes-X.Y.html UsersGuide.html 
Log Message:
- added new SQLDefault key for attributes


Index: RelNotes-X.Y.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/RelNotes-X.Y.html,v
retrieving revision 1.4
retrieving revision 1.5
diff -C2 -d -r1.4 -r1.5
*** RelNotes-X.Y.html	24 Feb 2004 02:04:13 -0000	1.4
--- RelNotes-X.Y.html	28 Feb 2004 01:20:31 -0000	1.5
***************
*** 57,61 ****
  <h2>New Features</h2>
  <ul>
! 	<li> *
  </ul>
  
--- 57,61 ----
  <h2>New Features</h2>
  <ul>
! 	<li> You can specify a <b>SQLDefault</b> for attributes. <a href=UsersGuide.html#MT_SQLDefault>docs</a>
  </ul>
  

Index: UsersGuide.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/UsersGuide.html,v
retrieving revision 1.25
retrieving revision 1.26
diff -C2 -d -r1.25 -r1.26
*** UsersGuide.html	24 Feb 2004 02:04:13 -0000	1.25
--- UsersGuide.html	28 Feb 2004 01:20:31 -0000	1.26
***************
*** 38,41 ****
--- 38,42 ----
  <br> &nbsp; &nbsp; &nbsp; <a href="#MT_DeletingObjects">Deleting objects</a>
  <br> &nbsp; &nbsp; &nbsp; <a href="#MT_DerivedAttributes">Derived attributes</a>
+ <br> &nbsp; &nbsp; &nbsp; <a href="#MT_SQLDefault">SQLDefault</a>
  <br> <a href="#ModelInheritance"><font size=+1>Model Inheritance</font></a>
  <br> <a href="#RelatedLinks"><font size=+1>Related Links</font></a>
***************
*** 437,440 ****
--- 438,446 ----
  
  
+ <a name=MT_SQLDefault><h3>SQLDefault</h3></a>
+ 
+ <p> MiddleKit will use the <b>Default</b> of attributes to generate a <b>DEFAULT <i>sqlValue</i></b> in the attribute's portion of the SQL CREATE statement, taking care to quote strings properly. This default value is also used in the Python class. But on occasion you may have a need to specify an alternative SQL default (such as <b>GetDate()</b>). When that happens, specify a <b>SQLDefault</b> for the attribute. If you do this in the <b>Extras</b> column, quote the SQL; for example, SQLDefault='GetDate()'. MiddleKit will pass this Python string down to the CREATE statement.
+ 
+ 
  <a name=ModelInheritance><h2>Model Inheritance</h2></a>

Update of /cvsroot/webware/Webware/MiddleKit/Design
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv28838/Design

Modified Files:
	SQLGenerator.py 
Log Message:
- added new SQLDefault key for attributes


Index: SQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/SQLGenerator.py,v
retrieving revision 1.41
retrieving revision 1.42
diff -C2 -d -r1.41 -r1.42
*** SQLGenerator.py	28 Feb 2004 00:08:07 -0000	1.41
--- SQLGenerator.py	28 Feb 2004 01:20:31 -0000	1.42
***************
*** 483,492 ****
  
  	def createDefaultSQL(self):
! 		default = self.get('Default', None)
! 		if default is not None:
  			default = str(default).strip()
  			if default.lower()=='none':  # kind of redundant
  				default = None
! 			return 'default ' + str(self.sqlForSampleInput(default))
  		else:
  			return ''
--- 483,496 ----
  
  	def createDefaultSQL(self):
! 		default = self.get('SQLDefault', None)
! 		if default is None:
! 			default = self.get('Default', None)
! 			if default is not None:
! 				default = self.sqlForSampleInput(str(default))
! 		if default:
  			default = str(default).strip()
  			if default.lower()=='none':  # kind of redundant
  				default = None
! 			return 'default ' + default
  		else:
  			return ''

Update of /cvsroot/webware/Webware/MiddleKit/Tests/MKEnums.mkmodel
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv15924/Tests/MKEnums.mkmodel

Modified Files:
	Settings2.config Settings3.config 
Log Message:
- revamp the enum settings


Index: Settings2.config
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Tests/MKEnums.mkmodel/Settings2.config,v
retrieving revision 1.1
retrieving revision 1.2
diff -C2 -d -r1.1 -r1.2
*** Settings2.config	16 Feb 2004 22:16:59 -0000	1.1
--- Settings2.config	28 Feb 2004 00:08:07 -0000	1.2
***************
*** 1,5 ****
  {
! 	# can use %(ClassName)s  %(AttrName)s  %(attrName)s
! 	'ExternalEnumsTableName': '%(ClassName)s%(AttrName)sEnum',
! 	'ExternalEnumsColumnName': 'name',
  }
--- 1,8 ----
  {
! 	'ExternalEnumsSQLNames': {
! 		'Enable': True,
! 		'TableName': '%(ClassName)s%(AttrName)sEnum',
! 		'ValueColName': 'value',
! 		'NameColName': 'name',
! 	},
  }

Index: Settings3.config
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Tests/MKEnums.mkmodel/Settings3.config,v
retrieving revision 1.1
retrieving revision 1.2
diff -C2 -d -r1.1 -r1.2
*** Settings3.config	16 Feb 2004 22:16:59 -0000	1.1
--- Settings3.config	28 Feb 2004 00:08:07 -0000	1.2
***************
*** 1,5 ****
  {
! 	# can use %(ClassName)s  %(AttrName)s  %(attrName)s
! 	'ExternalEnumsTableName': '%(ClassName)s%(AttrName)sEnum',
! 	'ExternalEnumsColumnName': '%(ClassName)sName',
  }
--- 1,8 ----
  {
! 	'ExternalEnumsSQLNames': {
! 		'Enable': True,
! 		'TableName': '%(ClassName)s%(AttrName)sEnum',
! 		'ValueColName': '%(ClassName)s%(AttrName)sId',
! 		'NameColName': '%(className)s_%(attrName)s_name',
! 	},
  }

Update of /cvsroot/webware/Webware/MiddleKit/Design
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv15924/Design

Modified Files:
	PythonGenerator.py SQLGenerator.py 
Log Message:
- revamp the enum settings


Index: PythonGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/PythonGenerator.py,v
retrieving revision 1.30
retrieving revision 1.31
diff -C2 -d -r1.30 -r1.31
*** PythonGenerator.py	24 Feb 2004 02:04:13 -0000	1.30
--- PythonGenerator.py	28 Feb 2004 00:08:07 -0000	1.31
***************
*** 356,360 ****
  
  	def stringToValue(self, string):
! 		if self.setting('ExternalEnumsTableName', None):
  			return self.intValueForString(string)
  		else:
--- 356,360 ----
  
  	def stringToValue(self, string):
! 		if self.usesExternalSQLEnums():
  			return self.intValueForString(string)
  		else:
***************
*** 368,372 ****
  			_%(name)sAttr = self.klass().lookupAttr('%(name)s')
  ''' % {'name': self.name()})
! 		if self.setting('ExternalEnumsTableName', None):
  			out.write('''
  		if value is not None:
--- 368,372 ----
  			_%(name)sAttr = self.klass().lookupAttr('%(name)s')
  ''' % {'name': self.name()})
! 		if self.usesExternalSQLEnums():
  			out.write('''
  		if value is not None:
***************
*** 410,413 ****
--- 410,424 ----
  
  
+ 	## Settings ##
+ 
+ 	def usesExternalSQLEnums(self):
+ 		# @@ 2004-02-25 ce: seems like this method and its use should be pushed down to SQLPythonGenerator.py
+ 		flag = getattr(self, '_usesExternalSQLEnums', None)
+ 		if flag is None:
+ 			flag = self.model().usesExternalSQLEnums()
+ 			self._usesExternalSQLEnums = flag
+ 		return flag
+ 
+ 
  class AnyDateTimeAttr:
  

Index: SQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/SQLGenerator.py,v
retrieving revision 1.40
retrieving revision 1.41
diff -C2 -d -r1.40 -r1.41
*** SQLGenerator.py	24 Feb 2004 02:04:13 -0000	1.40
--- SQLGenerator.py	28 Feb 2004 00:08:07 -0000	1.41
***************
*** 709,715 ****
  
  	def sqlType(self):
! 		if self.setting('ExternalEnumsTableName', None):
! 			tableName, columnName = self.externalEnumsTableAndColumnName()
! 			return 'int references %s(%sId)' % (tableName, tableName)
  		else:
  			return self.nativeEnumSQLType()
--- 709,715 ----
  
  	def sqlType(self):
! 		if self.usesExternalSQLEnums():
! 			tableName, valueColName, nameColName = self.externalEnumsSQLNames()
! 			return 'int references %s(%s)' % (tableName, valueColName)
  		else:
  			return self.nativeEnumSQLType()
***************
*** 720,724 ****
  
  	def sqlForNonNoneSampleInput(self, input):
! 		if self.setting('ExternalEnumsTableName', None):
  			return self.intValueForString(input)
  		else:
--- 720,724 ----
  
  	def sqlForNonNoneSampleInput(self, input):
! 		if self.usesExternalSQLEnums():
  			return self.intValueForString(input)
  		else:
***************
*** 726,745 ****
  			return repr(input)
  
- 	def externalEnumsTableAndColumnName(self):
- 		values = {
- 			'ClassName': self.klass().name(),
- 			'AttrName':  self.name()[0].upper() + self.name()[1:],
- 			'attrName':  self.name(),
- 		}
- 		tableName  = self.setting('ExternalEnumsTableName')  % values  # typical setting value is '%(ClassName)s%(AttrName)sEnum'. can use also use %(attrName)s for non-capped attr name
- 		columnName = self.setting('ExternalEnumsColumnName') % values  # typical setting value is simply 'name'
- 		return tableName, columnName
- 
  	def writeAuxiliaryCreateTable(self, generator, out):
! 		if self.setting('ExternalEnumsTableName', None):
! 			tableName, columnName = self.externalEnumsTableAndColumnName()
  			out.write('create table %s (\n' % tableName)
! 			out.write('\t%sId int not null primary key,\n' % tableName)
! 			out.write('\t%s varchar(255)\n' % columnName)
  			out.write(');\n')
  
--- 726,735 ----
  			return repr(input)
  
  	def writeAuxiliaryCreateTable(self, generator, out):
! 		if self.usesExternalSQLEnums():
! 			tableName, valueColName, nameColName = self.externalEnumsSQLNames()
  			out.write('create table %s (\n' % tableName)
! 			out.write('\t%s int not null primary key,\n' % valueColName)
! 			out.write('\t%s varchar(255)\n' % nameColName)
  			out.write(');\n')
  
***************
*** 747,755 ****
  			sep = ''
  			for enum in self.enums():
! 				out.write("insert into %(tableName)s (%(tableName)sId, %(columnName)s) values (%(i)i, '%(enum)s');\n" % locals())
  				i += 1
  			out.write('\n')
  
  
  class PrimaryKey:
  	"""
--- 737,776 ----
  			sep = ''
  			for enum in self.enums():
! 				out.write("insert into %(tableName)s values (%(i)i, '%(enum)s');\n" % locals())
  				i += 1
  			out.write('\n')
  
  
+ 	## Settings ##
+ 
+ 	def usesExternalSQLEnums(self):
+ 		flag = getattr(self, '_usesExternalSQLEnums', None)
+ 		if flag is None:
+ 			flag = self.model().usesExternalSQLEnums()
+ 			self._usesExternalSQLEnums = flag
+ 		return flag
+ 
+ 	def externalEnumsSQLNames(self):
+ 		"""
+ 		Returns the tuple (tableName, valueColName, nameColName)
+ 		derived from the model setting ExternalEnumsSQLNames.
+ 		"""
+ 		names = getattr(self, '_externalEnumsSQLNames', None)
+ 		if names is None:
+ 			_ClassName = self.klass().name()
+ 			ClassName  = _ClassName[0].upper() + _ClassName[1:]
+ 			className  = _ClassName[0].lower() + _ClassName[1:]
+ 			_AttrName  = self.name()
+ 			AttrName   = _AttrName[0].upper()  + _AttrName[1:]
+ 			attrName   = _AttrName[0].lower()  + _AttrName[1:]
+ 			values = locals()
+ 			names = self.setting('ExternalEnumsSQLNames')
+ 			names = [names['TableName'], names['ValueColName'], names['NameColName']]
+ 			for i in range(len(names)):
+ 				names[i] %= values
+ 			self._externalEnumsSQLNames = names
+ 		return names
+ 
+ 
  class PrimaryKey:
  	"""

Update of /cvsroot/webware/Webware/MiddleKit/Core
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv15924/Core

Modified Files:
	Model.py 
Log Message:
- revamp the enum settings


Index: Model.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/Model.py,v
retrieving revision 1.23
retrieving revision 1.24
diff -C2 -d -r1.23 -r1.24
*** Model.py	24 Feb 2004 02:04:13 -0000	1.23
--- Model.py	28 Feb 2004 00:08:07 -0000	1.24
***************
*** 328,332 ****
  			'PostSQL': '',
  			'DropStatements': 'database',  # database, tables
! 			'ExternalEnumsTableName': None, # '%(ClassName)s%(AttrName)sEnum'
! 			'ExternalEnumsColumnName': 'name',  # could use %(ClassName)s and %(AttrName)s
  		}
--- 328,345 ----
  			'PostSQL': '',
  			'DropStatements': 'database',  # database, tables
! 			'ExternalEnumsSQLNames': {
! 				'Enable': False,
! 				'TableName': '%(ClassName)s%(AttrName)sEnum',
! 				'ValueColName': 'value',
! 				'NameColName': 'name',
! 			},
! 				# can use: [cC]lassName, _ClassName, [aA]ttrName, _AttrName.
! 				# "_" prefix means "as is", the others control the case of the first character.
  		}
+ 
+ 	def usesExternalSQLEnums(self):
+ 		flag = getattr(self, '_usesExternalSQLEnums', None)
+ 		if flag is None:
+ 			flag = self.setting('ExternalEnumsSQLNames')['Enable']
+ 			self._usesExternalSQLEnums = flag
+ 		return flag

Update of /cvsroot/webware/Webware/MiscUtils
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv27397

Modified Files:
	MixIn.py 
Log Message:
Added support for calling the original/parent method from a mixed-in method.



Index: MixIn.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiscUtils/MixIn.py,v
retrieving revision 1.2
retrieving revision 1.3
diff -C2 -d -r1.2 -r1.3
*** MixIn.py	20 Jun 2002 17:47:59 -0000	1.2
--- MixIn.py	26 Feb 2004 16:15:46 -0000	1.3
***************
*** 3,7 ****
  
  if hasattr(sys, 'version_info') and sys.version_info[0]>=2:
! 	def MixIn(pyClass, mixInClass, makeAncestor=0):
  		"""
  		Mixes in the attributes of the mixInClass into the pyClass. These attributes are typically methods (but don't have to be). Note that private attributes, denoted by a double underscore, are not mixed in. Collisions are resolved by the mixInClass' attribute overwriting the pyClass'. This gives mix-ins the power to override the behavior of the pyClass.
--- 3,7 ----
  
  if hasattr(sys, 'version_info') and sys.version_info[0]>=2:
! 	def MixIn(pyClass, mixInClass, makeAncestor=0, mixInSuperMethods=0):
  		"""
  		Mixes in the attributes of the mixInClass into the pyClass. These attributes are typically methods (but don't have to be). Note that private attributes, denoted by a double underscore, are not mixed in. Collisions are resolved by the mixInClass' attribute overwriting the pyClass'. This gives mix-ins the power to override the behavior of the pyClass.
***************
*** 21,24 ****
--- 21,32 ----
  		If makeAncestor is 1, then a different technique is employed: the mixInClass is made the first base class of the pyClass. You probably don't need to use this and if you do, be aware that your mix-in can no longer override attributes/methods in pyClass.
  
+ 		If mixInSuperMethods is 1, then support will be enabled for you to be able to call the original or 
+ 		"parent" method from the mixed-in method.  This is done like so:
+ 
+ 		    class MyMixInClass:
+ 			def foo(self):
+ 			    MyMixInClass.mixInSuperFoo(self)	# call the original method
+ 			    # now do whatever you want
+ 
  		This function only exists if you are using Python 2.0 or later. Python 1.5.2 has a problem where functions (as in aMethod.im_func) are tied to their class, when in fact, they should be totally generic with only the methods being tied to their class. Apparently this was fixed in Py 2.0.
  		"""
***************
*** 53,56 ****
--- 61,69 ----
  				if not name.startswith('__'): # skip private members
  					member = getattr(mixInClass, name)
+ 
+ 					if type(member) is MethodType and mixInSuperMethods:
+ 						if hasattr(pyClass, name):
+ 							origmember = getattr(pyClass, name)
+ 							setattr(mixInClass, 'mixInSuper' + name[0].upper() + name[1:], origmember)
  					if type(member) is MethodType:
  						member = member.im_func

Update of /cvsroot/webware/Webware/MiddleKit/Run
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv18927/Run

Modified Files:
	MSSQLObjectStore.py MiddleObject.py MySQLObjectStore.py 
	ObjectStore.py PostgreSQLObjectStore.py SQLObjectStore.py 
Log Message:
- split obj ref SQL columns in two (class id and obj id)
- merge SQLGenerater.config into Settings.config
- see release notes for details


Index: MSSQLObjectStore.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/MSSQLObjectStore.py,v
retrieving revision 1.8
retrieving revision 1.9
diff -C2 -d -r1.8 -r1.9
*** MSSQLObjectStore.py	18 Dec 2003 06:17:02 -0000	1.8
--- MSSQLObjectStore.py	24 Feb 2004 02:04:13 -0000	1.9
***************
*** 93,96 ****
--- 93,97 ----
  		return '[%s]' % self.name()
  
+ 
  class Attr:
  
***************
*** 98,112 ****
  		""" Returns the SQL column name corresponding to this attribute, consisting of self.name() + self.sqlTypeSuffix(). """
  		if not self._sqlColumnName:
! 			self._sqlColumnName = self.name() + self.sqlTypeSuffix()
! 		return '[' + self._sqlColumnName + ']'
  
  
  class StringAttr:
  
! 	def sqlValue(self, value):
! 		if value is None:
! 			return 'NULL'
! 		else:
! 			# do the right thing
! 			value = value.replace("'","''")
! 			return "'" + value + "'"
--- 99,122 ----
  		""" Returns the SQL column name corresponding to this attribute, consisting of self.name() + self.sqlTypeSuffix(). """
  		if not self._sqlColumnName:
! 			self._sqlColumnName = '[' + self.name() + ']'
! 		return self._sqlColumnName
! 
! 
! class ObjRefAttr:
! 
! 	def sqlColumnName(self):
! 		if not self._sqlColumnName:
! 			if self.setting('UseBigIntObjRefColumns', False):
! 				self._sqlColumnName = '[' + self.name() + 'Id' + ']'  # old way: one 64 bit column
! 			else:
! 				# new way: 2 int columns for class id and obj id
! 				self._sqlColumnName = '[%s],[%s]' % self.sqlColumnNames()
! 		return self._sqlColumnName
  
  
  class StringAttr:
  
! 	def sqlForNonNone(self, value):
! 		# do the right thing
! 		value = value.replace("'","''")
! 		return "'" + value + "'"

Index: MiddleObject.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/MiddleObject.py,v
retrieving revision 1.27
retrieving revision 1.28
diff -C2 -d -r1.27 -r1.28
*** MiddleObject.py	8 Dec 2003 06:08:40 -0000	1.27
--- MiddleObject.py	24 Feb 2004 02:04:13 -0000	1.28
***************
*** 60,94 ****
  		else:
  			self.setStore(store)
! 		fullClassName = self.__class__.__module__ + '.' + self.__class__.__name__
! 		cache = self._mk_setCache.setdefault(fullClassName, [])
! 		if not cache:
! 			allAttrs = self.klass().allDataAttrs()
! 			# @@ 2000-10-29 ce: next line is major hack: hasSQLColumn()
! 			attrs = [attr for attr in allAttrs if attr.hasSQLColumn()]
! 			attrNames = [attr.name() for attr in attrs]
! 			assert len(attrNames)+1==len(row)  # +1 because row has serialNumber
! 			for name in attrNames:
! 				setMethodName = 'set' + name[0].upper() + name[1:]
! 				setMethod = getattr(self.__class__, setMethodName, '_'+name)
! 				cache.append(setMethod)
  
! 		assert len(cache)+1==len(row)
! 		dict = self.__dict__ # we use this to bypass our own __setattr__
! 		dict['_mk_initing'] = 1
! 		if self._mk_serialNum==0:
! 			self.setSerialNum(row[0])
! 		else:
! 			assert self._mk_serialNum==row[0]
! 		# Set all of our attributes with setFoo() or by assigning to _foo
! 		for i in xrange(len(cache)):
! 			value = row[i+1]
! 			setter = cache[i]
! 			if isinstance(setter, StringTypes):
! 				dict['_'+setter] = value
  			else:
! 				# a method
  				setter(self, value)
! 		dict['_mk_initing'] = 0
! 		dict['_mk_inStore'] = 1
  		return self
  
--- 60,101 ----
  		else:
  			self.setStore(store)
! 		if store.setting('UseBigIntObjRefColumns', False):
! 			fullClassName = self.__class__.__module__ + '.' + self.__class__.__name__
! 			cache = self._mk_setCache.setdefault(fullClassName, [])
! 			if not cache:
! 				allAttrs = self.klass().allDataAttrs()
! 				# @@ 2000-10-29 ce: next line is major hack: hasSQLColumn()
! 				attrs = [attr for attr in allAttrs if attr.hasSQLColumn()]
! 				attrNames = [attr.name() for attr in attrs]
! 				assert len(attrNames)+1==len(row)  # +1 because row has serialNumber
! 				for name in attrNames:
! 					setMethodName = 'set' + name[0].upper() + name[1:]
! 					setMethod = getattr(self.__class__, setMethodName, '_'+name)
! 					cache.append(setMethod)
  
! 			assert len(cache)+1==len(row)
! 			self._mk_initing = 1
! 			if self._mk_serialNum==0:
! 				self.setSerialNum(row[0])
  			else:
! 				assert self._mk_serialNum==row[0]
! 			# Set all of our attributes with setFoo() or by assigning to _foo
! 			for i in xrange(len(cache)):
! 				value = row[i+1]
! 				setter = cache[i]
  				setter(self, value)
! 		else:
! 			self._mk_initing = 1
! 			if self._mk_serialNum==0:
! 				self.setSerialNum(row[0])
! 			else:
! 				assert self._mk_serialNum==row[0]
! 			allAttrs = self.klass().allDataAttrs()
! 			i = 1
! 			for attr in allAttrs:
! 				i = attr.readStoreDataRow(self, row, i)
! 
! 		self._mk_initing = 0
! 		self._mk_inStore = 1
  		return self
  

Index: MySQLObjectStore.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/MySQLObjectStore.py,v
retrieving revision 1.6
retrieving revision 1.7
diff -C2 -d -r1.6 -r1.7
*** MySQLObjectStore.py	12 Feb 2002 19:29:20 -0000	1.6
--- MySQLObjectStore.py	24 Feb 2004 02:04:13 -0000	1.7
***************
*** 52,59 ****
  
  class StringAttr:
! 	def sqlValue(self, value):
! 		""" MySQL provides a quoting function for string -- use it. """
! 		if value is None:
! 			return 'NULL'
! 		else:
! 			return "'" + MySQLdb.escape_string(value) + "'"
--- 52,56 ----
  
  class StringAttr:
! 	def sqlForNonNone(self, value):
! 		""" MySQL provides a quoting function for string -- this method uses it. """
! 		return "'" + MySQLdb.escape_string(value) + "'"

Index: ObjectStore.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/ObjectStore.py,v
retrieving revision 1.29
retrieving revision 1.30
diff -C2 -d -r1.29 -r1.30
*** ObjectStore.py	8 Dec 2003 06:08:40 -0000	1.29
--- ObjectStore.py	24 Feb 2004 02:04:13 -0000	1.30
***************
*** 97,109 ****
  			self._changedObjects = NonThreadedDict()
  
- 	## Settings ##
- 
- 	def setting(self, name, default=NoDefault):
- 		"""
- 		Returns the given setting for the store, which is actually
- 		just taken from the model.
- 		"""
- 		return self._model.setting(name, default)
- 
  
  	## Manipulating the objects in the store ##
--- 97,100 ----

Index: PostgreSQLObjectStore.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/PostgreSQLObjectStore.py,v
retrieving revision 1.1
retrieving revision 1.2
diff -C2 -d -r1.1 -r1.2
*** PostgreSQLObjectStore.py	6 Mar 2003 18:06:32 -0000	1.1
--- PostgreSQLObjectStore.py	24 Feb 2004 02:04:13 -0000	1.2
***************
*** 24,28 ****
  		args['database'] = self._model.sqlDatabaseName()
  		return self.dbapiModule().connect(**args)
! 	
  	def _insertObject(self, object, unknownSerialNums):
  		# basically the same as the SQLObjectStore-
--- 24,28 ----
  		args['database'] = self._model.sqlDatabaseName()
  		return self.dbapiModule().connect(**args)
! 
  	def _insertObject(self, object, unknownSerialNums):
  		# basically the same as the SQLObjectStore-
***************
*** 58,62 ****
  		conn, cur = self.connectionAndCursor()
  		try:
! 			
  			SQLObjectStore.saveChanges(self)
  		except DatabaseError:
--- 58,62 ----
  		conn, cur = self.connectionAndCursor()
  		try:
! 
  			SQLObjectStore.saveChanges(self)
  		except DatabaseError:
***************
*** 74,81 ****
  
  class StringAttr:
! 	def sqlValue(self, value):
  		# Postgres provides a quoting function for string -- use it.
! 		if value is None:
! 			return 'NULL'
! 		else:
! 			return str(dbi._quote(value))
--- 74,79 ----
  
  class StringAttr:
! 
! 	def sqlForNonNone(self, value):
  		# Postgres provides a quoting function for string -- use it.
! 		return str(dbi._quote(value))

Index: SQLObjectStore.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Run/SQLObjectStore.py,v
retrieving revision 1.59
retrieving revision 1.60
diff -C2 -d -r1.59 -r1.60
*** SQLObjectStore.py	18 Dec 2003 06:17:02 -0000	1.59
--- SQLObjectStore.py	24 Feb 2004 02:04:13 -0000	1.60
***************
*** 42,45 ****
--- 42,51 ----
  
  class UnknownSerialNumInfo:
+ 	"""
+ 	Attrs assigned externally are:
+ 		sourceObject
+ 		sourceAttr
+ 		targetObject
+ 	"""
  
  	def updateStmt(self):
***************
*** 50,55 ****
  		sourceTableName = sourceKlass.sqlTableName()
  		sourceSqlIdName = sourceKlass.sqlIdName()
! 		return 'update %s set %s=%s where %s=%s;' % (
! 			sourceTableName, self.fieldName, self.targetObject.sqlObjRef(), sourceSqlIdName, self.sourceObject.serialNum())
  
  	def __repr__(self):
--- 56,61 ----
  		sourceTableName = sourceKlass.sqlTableName()
  		sourceSqlIdName = sourceKlass.sqlIdName()
! 		return 'update %s set %s where %s=%s;' % (
! 			sourceTableName, self.sourceAttr.sqlUpdateExpr(self.targetObject), sourceSqlIdName, self.sourceObject.serialNum())
  
  	def __repr__(self):
***************
*** 109,112 ****
--- 115,119 ----
  			for attr in klass.allDataAttrs():
  				attr._sqlColumnName = None
+ 				attr._sqlColumnNames = None
  
  		# Connect
***************
*** 552,555 ****
--- 559,566 ----
  
  	def sqlObjRef(self):
+ 		"""
+ 		Returns the 64-bit integer value that refers to self in a SQL database.
+ 		This only makes sense if the UseBigIntObjRefColumns setting is True.
+ 		"""
  		return objRefJoin(self.klass().id(), self.serialNum())
  
***************
*** 566,569 ****
--- 577,581 ----
  		values = []
  		append = values.append
+ 		extend = values.extend
  		for attr in sqlAttrs:
  			try:
***************
*** 572,577 ****
  				exc.info.sourceObject = self
  				unknowns.append(exc.info)
! 				value = 'NULL'
! 			append(value)
  		if len(values)==0:
  			values = ['0']
--- 584,595 ----
  				exc.info.sourceObject = self
  				unknowns.append(exc.info)
! 				if self.store().model().setting('UseBigIntObjRefColumns', False):
! 					value = 'NULL'
! 				else:
! 					value = ('NULL', 'NULL')
! 			if isinstance(value, str):
! 				append(value)
! 			else:
! 				extend(value)  # value could be sequence for attrs that require multiple SQL columns
  		if len(values)==0:
  			values = ['0']
***************
*** 581,585 ****
  	def sqlUpdateStmt(self):
  		"""
! 		Returns the SQL update statement for MySQL of the form:
  			update table set name=value, ... where idName=idValue;
  		Installed as a method of MiddleObject.
--- 599,603 ----
  	def sqlUpdateStmt(self):
  		"""
! 		Returns the SQL update statement of the form:
  			update table set name=value, ... where idName=idValue;
  		Installed as a method of MiddleObject.
***************
*** 589,594 ****
  		res = []
  		for attr in self._mk_changedAttrs.values():
! 			colName = attr.sqlColumnName()
! 			res.append(colName+'='+attr.sqlValue(self.valueForAttr(attr)))
  		res = ','.join(res)
  		res = ('update ', klass.sqlTableName(), ' set ', res, ' where ', klass.sqlIdName(), '=', str(self.serialNum()))
--- 607,611 ----
  		res = []
  		for attr in self._mk_changedAttrs.values():
! 			res.append(attr.sqlUpdateExpr(self.valueForAttr(attr)))
  		res = ','.join(res)
  		res = ('update ', klass.sqlTableName(), ' set ', res, ' where ', klass.sqlIdName(), '=', str(self.serialNum()))
***************
*** 611,615 ****
  
  	def referencingObjectsAndAttrsFetchKeywordArgs(self, backObjRefAttr):
! 		return {'refreshAttrs': 1, 'clauses': 'WHERE %s=%s' % (backObjRefAttr.sqlColumnName(), self.sqlObjRef())}
  
  MixIn(MiddleObject, MiddleObjectMixIn)
--- 628,642 ----
  
  	def referencingObjectsAndAttrsFetchKeywordArgs(self, backObjRefAttr):
! 		if self.store().setting('UseBigIntObjRefColumns'):
! 			return {
! 				'refreshAttrs': 1,
! 				'clauses': 'WHERE %s=%s' % (backObjRefAttr.sqlColumnName(), self.sqlObjRef())
! 			}
! 		else:
! 			classIdName, objIdName = backObjRefAttr.sqlColumnNames()
! 			return {
! 				'refreshAttrs': 1,
! 				'clauses': 'WHERE (%s=%s AND %s=%s)' % (classIdName, self.klass().id(), objIdName, self.serialNum())
! 			}
  
  MixIn(MiddleObject, MiddleObjectMixIn)
***************
*** 663,666 ****
--- 690,694 ----
  		return self._insertSQLStart, self._sqlAttrs
  
+ 
  class Attr:
  
***************
*** 676,734 ****
  		""" Returns the SQL column name corresponding to this attribute, consisting of self.name() + self.sqlTypeSuffix(). """
  		if not self._sqlColumnName:
! 			self._sqlColumnName = self.name() + self.sqlTypeSuffix()
  		return self._sqlColumnName
  
- 	def sqlTypeSuffix(self):
- 		""" Returns a string to be used as a suffix for sqlColumnName(). Returns an empty string. Occasionally, a subclass will override this to help clarify SQL column names of their type. """
- 		return ''
- 
  	def sqlValue(self, value):
! 		""" For a given Python value, this returns the correct string for use in a SQL INSERT statement. Subclasses should override if this implementation, which returns repr(value), doesn't work for them. This method is responsible for returning 'NULL' if the value is None. """
  		if value is None:
! 			return 'NULL'
  		else:
! 			return repr(value)
  
  
  class IntAttr:
  
! 	def sqlValue(self, value):
! 		if value is None:
! 			return 'NULL'
! 		else:
! 			return str(value)
! 			# it's important to use str() since an int might
! 			# point to a long (whose repr() would be suffixed
! 			# with an 'L')
  
  
  class LongAttr:
  
! 	def sqlValue(self, value):
! 		if value is None:
! 			return 'NULL'
! 		else:
! 			return str(value)
  
  
  class ObjRefAttr:
  
! 	def sqlTypeSuffix(self):
! 		return 'Id'
  
! 	def sqlValue(self, value):
! 		if value is None:
  			return 'NULL'
  		else:
! 			assert type(value) is InstanceType
! 			assert isinstance(value, MiddleObject)
! 			if value.serialNum()==0:
! 				info = UnknownSerialNumInfo()
! 				info.fieldName = self.sqlColumnName()
! 				info.targetObject = value
! 				raise UnknownSerialNumberError(info)
  			else:
! 				value = value.sqlObjRef()
! 				return str(value)
  
  
--- 704,839 ----
  		""" Returns the SQL column name corresponding to this attribute, consisting of self.name() + self.sqlTypeSuffix(). """
  		if not self._sqlColumnName:
! 			self._sqlColumnName = self.name()
  		return self._sqlColumnName
  
  	def sqlValue(self, value):
! 		"""
! 		For a given Python value, this returns the correct string for
! 		use in a SQL statement. Subclasses will typically *not*
! 		override this method, but instead, sqlForNonNone() and on
! 		rare occasions, sqlForNone().
! 		"""
  		if value is None:
! 			return self.sqlForNone()
  		else:
! 			return self.sqlForNonNone(value)
! 
! 	def sqlForNone(self):
! 		return 'NULL'
! 
! 	def sqlForNonNone(self, value):
! 		return repr(value)
! 
! 	def sqlUpdateExpr(self, value):
! 		"""
! 		Returns the assignment portion of an UPDATE statement such as:
! 			"foo='bar'"
! 		Using sqlColumnName() and sqlValue(). Subclasses only need to
! 		override this if they have a special need (such as multiple
! 		columns, see ObjRefAttr).
! 		"""
! 		colName = self.sqlColumnName()
! 		return colName+'='+self.sqlValue(value)
! 
! 	def readStoreDataRow(self, obj, row, i):
! 		"""
! 		By default, an attr reads one data value out of the row.
! 		"""
! 		value = row[i]
! 		obj.setValueForAttr(self, value)
! 		return i+1
! 
! 
! class BasicTypeAttr:
! 
! 	pass
  
  
  class IntAttr:
  
! 	def sqlForNonNone(self, value):
! 		return str(value)
! 		# it's important to use str() since an int might
! 		# point to a long (whose repr() would be suffixed
! 		# with an 'L')
  
  
  class LongAttr:
  
! 	def sqlForNonNone(self, value):
! 		return str(value)
  
  
  class ObjRefAttr:
  
! 	def sqlColumnName(self):
! 		if not self._sqlColumnName:
! 			if self.setting('UseBigIntObjRefColumns', False):
! 				self._sqlColumnName = self.name() + 'Id'  # old way: one 64 bit column
! 			else:
! 				# new way: 2 int columns for class id and obj id
! 				self._sqlColumnName = '%s,%s' % self.sqlColumnNames()
! 		return self._sqlColumnName
  
! 	def sqlColumnNames(self):
! 		if not self._sqlColumnNames:
! 			assert not self.setting('UseBigIntObjRefColumns', False)
! 			name = self.name()
! 			classIdName, objIdName = self.setting('ObjRefSuffixes')
! 			classIdName = name + classIdName
! 			objIdName = name + objIdName
! 			self._sqlColumnNames = (classIdName, objIdName)
! 		return self._sqlColumnNames
! 
! 	def sqlForNone(self):
! 		if self.setting('UseBigIntObjRefColumns', False):
  			return 'NULL'
  		else:
! 			return 'NULL,NULL'
! 
! 	def sqlForNonNone(self, value):
! 		assert type(value) is InstanceType
! 		assert isinstance(value, MiddleObject)
! 		if value.serialNum()==0:
! 			info = UnknownSerialNumInfo()
! 			info.sourceAttr = self
! 			info.targetObject = value
! 			raise UnknownSerialNumberError(info)
! 		else:
! 			if self.setting('UseBigIntObjRefColumns', False):
! 				return str(value.sqlObjRef())
  			else:
! 				return (str(value.klass().id()), str(value.serialNum()))
! 
! 	def sqlUpdateExpr(self, value):
! 		"""
! 		Returns the assignment portion of an UPDATE statement such as:
! 			"foo='bar'"
! 		Using sqlColumnName() and sqlValue(). Subclasses only need to
! 		override this if they have a special need (such as multiple
! 		columns, see ObjRefAttr).
! 		"""
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			colName = self.sqlColumnName()
! 			return colName+'='+self.sqlValue(value)
! 		else:
! 			classIdName, objIdName = self.sqlColumnNames()
! 			if value is None:
! 				classId = objId = 'NULL'
! 			else:
! 				classId = value.klass().id()
! 				objId = value.serialNum()
! 			return '%s=%s,%s=%s' % (classIdName, classId, objIdName, objId)
! 
! 	def readStoreDataRow(self, obj, row, i):
! 		# this does *not* get called under the old approach of single obj ref columns. see MiddleObject.readStoreData
! 		classId, objId = row[i], row[i+1]
! 		if classId is None:
! 			value = None
! 		else:
! 			value = objRefJoin(classId, objId)
! 			# @@ 2004-20-02 ce ^ that's wasteful to join them just so they can be split later, but it works well with the legacy code
! 		obj.setValueForAttr(self, value)
! 		return i+2
  
  
***************
*** 738,767 ****
  		return 0
  
  
  class AnyDateTimeAttr:
  
! 	def sqlValue(self, value):
! 		if value is None:
! 			return 'NULL'
! 		else:
! 			# Chop off the milliseconds -- SQL databases seem to dislike that.
! 			return "'%s'" % str(value).split('.')[0]
  
  
  class DateAttr:
  
! 	def sqlValue(self, value):
! 		if value is None:
! 			return 'NULL'
! 		else:
! 			# We often get "YYYY-MM-DD HH:MM:SS" from mx's DateTime
! 			# so we split on space and take the first value to
! 			# work around that.
! 			if 0:
! 				print
! 				print '>> type of value =', type(value)
! 				print '>> value = %r' % value
! 				print
! 			if not isinstance(value, StringTypes):
! 				value = str(value).split()[0]
! 			return "'%s'" % value
--- 843,869 ----
  		return 0
  
+ 	def readStoreDataRow(self, obj, row, i):
+ 		return i
+ 
  
  class AnyDateTimeAttr:
  
! 	def sqlForNonNone(self, value):
! 		# Chop off the milliseconds -- SQL databases seem to dislike that.
! 		return "'%s'" % str(value).split('.')[0]
  
  
  class DateAttr:
  
! 	def sqlForNonNone(self, value):
! 		# We often get "YYYY-MM-DD HH:MM:SS" from mx's DateTime
! 		# so we split on space and take the first value to
! 		# work around that.
! 		if 0:
! 			print
! 			print '>> type of value =', type(value)
! 			print '>> value = %r' % value
! 			print
! 		if not isinstance(value, StringTypes):
! 			value = str(value).split()[0]
! 		return "'%s'" % value

Update of /cvsroot/webware/Webware/MiddleKit/Tests/MKObjRef.mkmodel
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv18927/Tests/MKObjRef.mkmodel

Modified Files:
	TestEmpty1.py 
Added Files:
	Settings1.config Settings2.config Settings3.config 
Log Message:
- split obj ref SQL columns in two (class id and obj id)
- merge SQLGenerater.config into Settings.config
- see release notes for details


--- NEW FILE: Settings1.config ---
{
	'UseBigIntObjRefColumns': False,
#	'SQLLog': { 'File': 'stdout', },
}

--- NEW FILE: Settings2.config ---
{
	'UseBigIntObjRefColumns': False,
	'ObjRefSuffixes': ('ClassSerialNumber', 'ObjectSerialNumber'),
}

--- NEW FILE: Settings3.config ---
{
	'UseBigIntObjRefColumns': True,
}

Index: TestEmpty1.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Tests/MKObjRef.mkmodel/TestEmpty1.py,v
retrieving revision 1.2
retrieving revision 1.3
diff -C2 -d -r1.2 -r1.3
*** TestEmpty1.py	3 May 2002 14:26:31 -0000	1.2
--- TestEmpty1.py	24 Feb 2004 02:04:14 -0000	1.3
***************
*** 23,26 ****
--- 23,27 ----
  	f1 = b.foo()
  	f2 = b.foo()
+ 	assert f1 is not None, 'got None instead of a Foo'
  	assert f1 is f2  # test uniqueness
  	assert b.foo().x()==3

Update of /cvsroot/webware/Webware/MiddleKit/Design
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv18927/Design

Modified Files:
	MSSQLSQLGenerator.py MySQLSQLGenerator.py PythonGenerator.py 
	SQLGenerator.py SQLPythonGenerator.py 
Log Message:
- split obj ref SQL columns in two (class id and obj id)
- merge SQLGenerater.config into Settings.config
- see release notes for details


Index: MSSQLSQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/MSSQLSQLGenerator.py,v
retrieving revision 1.15
retrieving revision 1.16
diff -C2 -d -r1.15 -r1.16
*** MSSQLSQLGenerator.py	16 Feb 2004 22:14:52 -0000	1.15
--- MSSQLSQLGenerator.py	24 Feb 2004 02:04:13 -0000	1.16
***************
*** 85,89 ****
  
  create table _MKClassIds (
! 	id bigint not null primary key,
  	name varchar(100)
  )\ngo
--- 85,89 ----
  
  create table _MKClassIds (
! 	id int not null primary key,
  	name varchar(100)
  )\ngo
***************
*** 91,96 ****
  		wr('delete from _MKClassIds\n\n')
  		for klass in self._klasses:
! 			wr('insert into _MKClassIds (id, name) values\n')
! 			wr("\t(%s, '%s');\n" % (klass.id(), klass.name()))
  		wr('\ngo\n\n')
  
--- 91,96 ----
  		wr('delete from _MKClassIds\n\n')
  		for klass in self._klasses:
! 			wr('insert into _MKClassIds (id, name) values ')
! 			wr("(%s, '%s');\n" % (klass.id(), klass.name()))
  		wr('\ngo\n\n')
  
***************
*** 143,147 ****
  #			wr("print 'Creating table %s'\n" % name)
  #			wr('create table [%s] (\n' % name)
! #			wr('	%s bigint primary key not null IDENTITY (1, 1),\n' % ljust(sqlIdName, self.maxNameWidth()))
  
  #			for attr in self.allAttrs():
--- 143,147 ----
  #			wr("print 'Creating table %s'\n" % name)
  #			wr('create table [%s] (\n' % name)
! #			wr('	%s int primary key not null IDENTITY (1, 1),\n' % ljust(sqlIdName, self.maxNameWidth()))
  
  #			for attr in self.allAttrs():
***************
*** 167,171 ****
  #		print("print 'Creating table %s'\n" % name)
  #		print('create table [%s] (\n' % name)
! 		z = '	%s bigint primary key not null IDENTITY (1, 1),\n' % self.sqlIdName().ljust(self.maxNameWidth())
  #		print(z)
  		return z
--- 167,171 ----
  #		print("print 'Creating table %s'\n" % name)
  #		print('create table [%s] (\n' % name)
! 		z = '	%s int primary key not null IDENTITY (1, 1),\n' % self.sqlIdName().ljust(self.maxNameWidth())
  #		print(z)
  		return z
***************
*** 268,276 ****
  
  	def sqlType(self):
! 		if self.get('Ref',None):
! 			return 'bigint foreign key references %(Type)s(%(Type)sID) ' % self
  		else:
! 			return 'bigint /* relates to %s */ ' % self['Type']
! 
  
  
--- 268,278 ----
  
  	def sqlType(self):
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			if self.get('Ref', None):
! 				return 'bigint foreign key references %(Type)s(%(Type)sId) ' % self
! 			else:
! 				return 'bigint /* relates to %s */ ' % self['Type']
  		else:
! 			return 'int'
  
  
***************
*** 284,295 ****
  
  	def sqlType(self):
! 		return 'decimal(16,8)'
! 		# @@ 2001-04-26 dr: this (16,8) should be doable through your model
! 		# you should use the decimal attribute and max/numDecimalPlaces
! 		# float is supported for mySQL test compatibility
  
  	def sampleValue(self, value):
  		float(value) # raises exception if value is invalid
  		return value
- 
- 
--- 286,294 ----
  
  	def sqlType(self):
! 		return 'float'
! 		# return 'decimal(16,8)'
! 		# ^ use the decimal type instead
  
  	def sampleValue(self, value):
  		float(value) # raises exception if value is invalid
  		return value

Index: MySQLSQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/MySQLSQLGenerator.py,v
retrieving revision 1.11
retrieving revision 1.12
diff -C2 -d -r1.11 -r1.12
*** MySQLSQLGenerator.py	16 Feb 2004 22:14:52 -0000	1.11
--- MySQLSQLGenerator.py	24 Feb 2004 02:04:13 -0000	1.12
***************
*** 71,74 ****
  
  	def sqlType(self):
! 		# @@ 2001-02-04 ce: Is this standard SQL? If so, it can be moved up.
! 		return 'bigint unsigned /* %s */' % self['Type']
--- 71,76 ----
  
  	def sqlType(self):
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			return 'bigint unsigned /* %s */' % self['Type']
! 		else:
! 			return 'int unsigned'

Index: PythonGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/PythonGenerator.py,v
retrieving revision 1.29
retrieving revision 1.30
diff -C2 -d -r1.29 -r1.30
*** PythonGenerator.py	16 Feb 2004 22:14:52 -0000	1.29
--- PythonGenerator.py	24 Feb 2004 02:04:13 -0000	1.30
***************
*** 16,24 ****
  class PythonGenerator(CodeGenerator):
  
- 	def defaultConfig(self):
- 		return {
- 			'DestDir': 'GeneratedPy',
- 		}
- 
  	def generate(self, dirname):
  		self.requireDir(dirname)
--- 16,19 ----

Index: SQLGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/SQLGenerator.py,v
retrieving revision 1.39
retrieving revision 1.40
diff -C2 -d -r1.39 -r1.40
*** SQLGenerator.py	16 Feb 2004 22:42:13 -0000	1.39
--- SQLGenerator.py	24 Feb 2004 02:04:13 -0000	1.40
***************
*** 51,71 ****
  		return self.model().sqlDatabaseName()
  
- 	def configFilename(self):
- 		filename = self.model().filename()
- 		if filename is not None:
- 			filename = os.path.join(filename, 'SQLGenerator.config')
- 		return filename
- 
- 	def defaultConfig(self):
- 		config = CodeGenerator.defaultConfig(self)
- 		config.update({
- 			'PreSQL': '',
- 			'PostSQL': '',
- 			'DropStatements': 'database',  # database, tables
- 			'ExternalEnumsTableName': None, # '%(ClassName)s%(AttrName)sEnum'
- 			'ExternalEnumsColumnName': 'name',  # could use %(ClassName)s and %(AttrName)s
- 		})
- 		return config
- 
  	def generate(self, dirname):
  		self.requireDir(dirname)
--- 51,54 ----
***************
*** 180,205 ****
  					except IndexError:
  						raise SampleError( linenum, "Couldn't find value for attribute '%s'" % attr.name() )
! 					value = value.strip()
! 					#print '>> (%s, %s)' % (value, attr)
! 					if value=='':
! 						value = attr.get('Default', None)
! 						if value is None:
! 							value = 'NULL'
! 						else:
! 							value = attr.sampleValue(value)
! 					elif value.lower()=='none':
! 						value = 'NULL'
! 					else:
! 						value = attr.sampleValue(value)
! #						print 'Attr: %s, Value: %s' % (attr,value)
! 					value = str(value)
! 					if 0:
! 						if not isinstance(value, StringTypes):
! 							print 'attr:', attr
! 							print 'value:', value
! 							print 'type of value:', type(value)
! 							assert isinstance(value, StringTypes), value
! 					assert value  # value cannot be blank
! 					values[i] = value
  					i += 1
  				#print
--- 163,168 ----
  					except IndexError:
  						raise SampleError( linenum, "Couldn't find value for attribute '%s'" % attr.name() )
! 					values[i] = attr.sqlForSampleInput(value)
! 					assert len(values[i])>0, repr(values[i])  # SQL value cannot be blank
  					i += 1
  				#print
***************
*** 326,332 ****
  		# assign the class ids up-front, so that we can resolve forward object references
  		self.assignClassIds(generator)
  		for klass in self._model._allKlassesInOrder:
  			klass.writeCreateSQL(self._sqlGenerator, out)
- 		self.writeClassIdsSQL(generator, out)
  
  	def writeClassIdsSQL(self, generator, out):
--- 289,295 ----
  		# assign the class ids up-front, so that we can resolve forward object references
  		self.assignClassIds(generator)
+ 		self.writeClassIdsSQL(generator, out)
  		for klass in self._model._allKlassesInOrder:
  			klass.writeCreateSQL(self._sqlGenerator, out)
  
  	def writeClassIdsSQL(self, generator, out):
***************
*** 465,493 ****
  		return not self.get('isDerived', 0)
  
! 	def sampleValue(self, value):
! 		""" Returns a string suitable for a SQL insert statement including any necessary SQL syntax. Subclasses should override to perform type checking and handle any special capabilities.
! 		The invoker of this method already strips preceding and trailing whitespace, as well detects blanks as NULLs.
! 		"""
! 		# @@ 2001-02-20 ce: restructure this
! 			# sqlValue() instead of sampleValue()
! 			# w.s. stripping, blanks as default value, none as NULL
! 			# make _sqlValue()
! 		return value
  
  	def writeCreateSQL(self, generator, out):
  		try:
  			if self.hasSQLColumn():
! 				name = self.sqlName().ljust(self.maxNameWidth())
! 				if self.isRequired():
! 					notNullSQL = ' not null'
! 				else:
! 					notNullSQL = self.sqlNullSpec()
! 				if generator.sqlSupportsDefaultValues():
! 					defaultSQL = self.createDefaultSQL()
! 					if defaultSQL:
! 						defaultSQL = ' ' + defaultSQL
! 				else:
! 					defaultSQL = ''
! 				out.write('\t%s %s%s%s' % (name, self.sqlType(), notNullSQL, defaultSQL))
  			else:
  				out.write('\t/* %(Name)s %(Type)s - not a SQL column */' % self)
--- 428,450 ----
  		return not self.get('isDerived', 0)
  
! 	def sqlForSampleInput(self, input):
! 		input = input.strip()
! 		if input=='':
! 			input = self.get('Default', '')
! 		if input in (None, 'None', 'none'):
! 			return self.sqlForNone()
! 		else:
! 			return str(self.sqlForNonNoneSampleInput(input))
! 
! 	def sqlForNone(self):
! 		return 'NULL'
! 
! 	def sqlForNonNoneSampleInput(self, input):
! 		return input
  
  	def writeCreateSQL(self, generator, out):
  		try:
  			if self.hasSQLColumn():
! 				self.writeRealCreateSQLColumn(generator, out)
  			else:
  				out.write('\t/* %(Name)s %(Type)s - not a SQL column */' % self)
***************
*** 504,507 ****
--- 461,478 ----
  			raise
  
+ 	def writeRealCreateSQLColumn(self, generator, out):
+ 		name = self.sqlName().ljust(self.maxNameWidth())
+ 		if self.isRequired():
+ 			notNullSQL = ' not null'
+ 		else:
+ 			notNullSQL = self.sqlNullSpec()
+ 		if generator.sqlSupportsDefaultValues():
+ 			defaultSQL = self.createDefaultSQL()
+ 			if defaultSQL:
+ 				defaultSQL = ' ' + defaultSQL
+ 		else:
+ 			defaultSQL = ''
+ 		out.write('\t%s %s%s%s' % (name, self.sqlType(), notNullSQL, defaultSQL))
+ 
  	def writeAuxiliaryCreateTable(self, generator, out):
  		# most attribute types have no such beast
***************
*** 517,521 ****
  			if default.lower()=='none':  # kind of redundant
  				default = None
! 			return 'default ' + str(self.sampleValue(default))
  		else:
  			return ''
--- 488,492 ----
  			if default.lower()=='none':  # kind of redundant
  				default = None
! 			return 'default ' + str(self.sqlForSampleInput(default))
  		else:
  			return ''
***************
*** 528,534 ****
  
  	def sqlColumnName(self):
! 		""" Returns the SQL column name corresponding to this attribute, consisting of self.name() + self.sqlTypeSuffix(). """
  		if not self._sqlColumnName:
! 			self._sqlColumnName = self.name() # + self.sqlTypeSuffix()
  		return self._sqlColumnName
  
--- 499,509 ----
  
  	def sqlColumnName(self):
! 		"""
! 		Returns the SQL column name corresponding to this attribute
! 		which simply defaults to the attribute's name. Subclasses may
! 		override to customize.
! 		"""
  		if not self._sqlColumnName:
! 			self._sqlColumnName = self.name()
  		return self._sqlColumnName
  
***************
*** 540,546 ****
  		return 'bool'
  
! 	def sampleValue(self, value):
  		try:
! 			value = value.upper()
  		except:
  			pass
--- 515,521 ----
  		return 'bool'
  
! 	def sqlForNonNoneSampleInput(self, input):
  		try:
! 			value = input.upper()
  		except:
  			pass
***************
*** 558,564 ****
  		return 'int'
  
! 	def sampleValue(self, value):
! 		if not isinstance(value, int):
! 			value = str(value)
  			if value.endswith('.0'):
  				# numeric values from Excel-based models are always float
--- 533,539 ----
  		return 'int'
  
! 	def sqlForNonNoneSampleInput(self, input):
! 		if not isinstance(input, int):
! 			value = str(input)
  			if value.endswith('.0'):
  				# numeric values from Excel-based models are always float
***************
*** 574,580 ****
  		return 'bigint'
  
! 	def sampleValue(self, value):
! 		long(value) # raises exception if value is invalid
! 		return value
  
  
--- 549,555 ----
  		return 'bigint'
  
! 	def sqlForNonNoneSampleInput(self, input):
! 		long(input) # raises exception if value is invalid
! 		return input
  
  
***************
*** 583,592 ****
  	def sqlType(self):
  		return 'double precision'
- 		# @@ 2001-02-04 ce: this (8,8) stuff is bad,]
- 		# but haven't come up with solution yet
  
! 	def sampleValue(self, value):
! 		float(value) # raises exception if value is invalid
! 		return value
  
  
--- 558,565 ----
  	def sqlType(self):
  		return 'double precision'
  
! 	def sqlForNonNoneSampleInput(self, input):
! 		float(input) # raises exception if value is invalid
! 		return input
  
  
***************
*** 609,614 ****
  		return 'decimal(%s,%s)' % (precision, scale)
  
! 	def sampleValue(self, value):
! 		return value
  
  
--- 582,587 ----
  		return 'decimal(%s,%s)' % (precision, scale)
  
! 	def sqlForNonNoneSampleInput(self, input):
! 		return input
  
  
***************
*** 623,627 ****
  		raise AbstractError, self.__class__
  
! 	def sampleValue(self, value):
  		if value=="''":
  			value = ''
--- 596,601 ----
  		raise AbstractError, self.__class__
  
! 	def sqlForNonNoneSampleInput(self, input):
! 		value = input
  		if value=="''":
  			value = ''
***************
*** 645,650 ****
  		return self['Type']  # e.g., date, time and datetime
  
! 	def sampleValue(self, value):
! 		return repr(value)
  
  
--- 619,624 ----
  		return self['Type']  # e.g., date, time and datetime
  
! 	def sqlForNonNoneSampleInput(self, input):
! 		return repr(input)
  
  
***************
*** 652,664 ****
  
  	def sqlName(self):
! 		return self.name() + 'Id'
  
! 	# @@ 2001-02-04 ce: Is this standard SQL?
! 	#def sqlType(self):
! 	#	return 'bigint unsigned /* %s */' % self['Type']
  
! 	def sampleValue(self, value):
! 		""" Obj ref sample data format is "Class.serialNum", such as "Thing.3". If the Class and period are missing, then the obj ref's type is assumed. """
! 		parts = value.split('.')
  		if len(parts)==2:
  			className = parts[0]
--- 626,682 ----
  
  	def sqlName(self):
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			return self.name() + 'Id'  # old way: one 64 bit column
! 		else:
! 			# new way: 2 int columns for class id and obj id
! 			name = self.name()
! 			classIdName, objIdName = self.setting('ObjRefSuffixes')
! 			classIdName = name + classIdName
! 			objIdName = name + objIdName
! 			return '%s,%s' % (classIdName, objIdName)
  
! 	def writeRealCreateSQLColumn(self, generator, out):
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			# the old technique of having both the class id and the obj id in one 64 bit reference
! 			name = self.sqlName().ljust(self.maxNameWidth())
! 			if self.get('Ref', None):
! 				refs = ' references %(Type)s(%(Type)sId)' % self
! 			else:
! 				refs = ''
! 			if self.isRequired():
! 				notNull = ' not null'
! 			else:
! 				notNull = self.sqlNullSpec()
! 			out.write('\t%s %s%s%s' % (name, self.sqlType(), refs, notNull))
! 		else:
! 			# the new technique uses one column for each part of the obj ref: class id and obj id
! 			classIdName = self.name()+self.setting('ObjRefSuffixes')[0]
! 			classIdName = classIdName.ljust(self.maxNameWidth())
! 			objIdName = self.name()+self.setting('ObjRefSuffixes')[1]
! 			objIdName = objIdName.ljust(self.maxNameWidth())
! 			if self.isRequired():
! 				notNull = ' not null'
! 			else:
! 				notNull = self.sqlNullSpec()
! 			if self.get('Ref', None):
! 				objIdRef = ' references %(Type)s(%(Type)sId) ' % self
! 			else:
! 				objIdRef = ''
! 			out.write('\t%s %s%s references _MKClassIds,\n' % (classIdName, self.sqlType(), notNull))
! 			out.write('\t%s %s%s%s' % (objIdName, self.sqlType(), notNull, objIdRef))
  
! 	def sqlForNone(self):
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			return 'NULL'
! 		else:
! 			return 'NULL,NULL'
! 
! 	def sqlForNonNoneSampleInput(self, input):
! 		"""
! 		Obj ref sample data format is "Class.serialNum", such as
! 		"Thing.3". If the Class and period are missing, then the obj
! 		ref's type is assumed.
! 		"""
! 		parts = input.split('.')
  		if len(parts)==2:
  			className = parts[0]
***************
*** 666,675 ****
  		else:
  			className = self.className()
! 			objSerialNum = value
! 		# @@ 2000-11-24 ce: check that we're pointing to a legal class
  		klass = self.klass().klasses()._model.klass(className)
  		klassId = klass.id()
! 		objRef = objRefJoin(klassId, objSerialNum)
! 		return str(objRef)
  
  
--- 684,695 ----
  		else:
  			className = self.className()
! 			objSerialNum = input
  		klass = self.klass().klasses()._model.klass(className)
  		klassId = klass.id()
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			objRef = objRefJoin(klassId, objSerialNum)
! 			return str(objRef)
! 		else:
! 			return '%s,%s' % (klassId, objSerialNum)
  
  
***************
*** 682,688 ****
  		return 0
  
! 	def sampleValue(self, value):
! 		# @@ 2000-11-24 ce: need specific extension?
! 		# @@ 2001-02-04 ce: ^^^ what does that mean???
  		raise Exception, 'Lists are implicit. They cannot have sample values.'
  
--- 702,706 ----
  		return 0
  
! 	def sqlForSampleInput(self, input):
  		raise Exception, 'Lists are implicit. They cannot have sample values.'
  
***************
*** 701,710 ****
  		return 'varchar(%s)' % maxLen
  
! 	def sampleValue(self, value):
  		if self.setting('ExternalEnumsTableName', None):
! 			return self.intValueForString(value)
  		else:
! 			assert value in self._enums, 'value = %r, enums = %r' % (value, self._enums)
! 			return repr(value)
  
  	def externalEnumsTableAndColumnName(self):
--- 719,728 ----
  		return 'varchar(%s)' % maxLen
  
! 	def sqlForNonNoneSampleInput(self, input):
  		if self.setting('ExternalEnumsTableName', None):
! 			return self.intValueForString(input)
  		else:
! 			assert input in self._enums, 'input=%r, enums=%r' % (input, self._enums)
! 			return repr(input)
  
  	def externalEnumsTableAndColumnName(self):
***************
*** 755,758 ****
  		return self._props.get(key,default)
  
! 	def sampleValue(self,value):
! 		return value
--- 773,776 ----
  		return self._props.get(key,default)
  
! 	def sqlForNonNoneSampleInput(self, input):
! 		return input

Index: SQLPythonGenerator.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Design/SQLPythonGenerator.py,v
retrieving revision 1.3
retrieving revision 1.4
diff -C2 -d -r1.3 -r1.4
*** SQLPythonGenerator.py	5 Nov 2002 07:39:28 -0000	1.3
--- SQLPythonGenerator.py	24 Feb 2004 02:04:13 -0000	1.4
***************
*** 36,40 ****
  
  	def writePyGet(self, out, names):
! 		out.write('''
  	def %(pyGetName)s(self):
  		if self._%(name)s is None:
--- 36,41 ----
  
  	def writePyGet(self, out, names):
! 		if self.setting('UseBigIntObjRefColumns', False):
! 			out.write('''
  	def %(pyGetName)s(self):
  		if self._%(name)s is None:
***************
*** 43,44 ****
--- 44,55 ----
  		return self._%(name)s
  ''' % names)
+ 		else:
+ 			classIdSuffix, objIdSuffix = self.setting('ObjRefSuffixes')
+ 			names.update(locals())
+ 			out.write('''
+ 	def %(pyGetName)s(self):
+ 		if self._%(name)s is None:
+ 			from %(package)s%(targetClassName)s import %(targetClassName)s
+ 			self._%(name)s = self._mk_store.fetchObjectsOfClass(%(targetClassName)s, clauses='where %(lowerSourceClassName)s%(classIdSuffix)s=%%i and %(lowerSourceClassName)s%(objIdSuffix)s=%%i' %% (self.klass().id(), self.serialNum()))
+ 		return self._%(name)s
+ ''' % names)

Update of /cvsroot/webware/Webware/MiddleKit/Docs
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv18927/Docs

Modified Files:
	RelNotes-X.Y.html UsersGuide.html 
Log Message:
- split obj ref SQL columns in two (class id and obj id)
- merge SQLGenerater.config into Settings.config
- see release notes for details


Index: RelNotes-X.Y.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/RelNotes-X.Y.html,v
retrieving revision 1.3
retrieving revision 1.4
diff -C2 -d -r1.3 -r1.4
*** RelNotes-X.Y.html	4 Dec 2003 06:58:57 -0000	1.3
--- RelNotes-X.Y.html	24 Feb 2004 02:04:13 -0000	1.4
***************
*** 18,22 ****
  		not be interpreted as comments in Mozilla 1.2.1
  	-->
! 	<title>CGIWrapper Release Notes</title>
  </head>
  
--- 18,22 ----
  		not be interpreted as comments in Mozilla 1.2.1
  	-->
! 	<title>MiddleKit Release Notes</title>
  </head>
  
***************
*** 29,41 ****
  
  
! <a name=Introduction<h2>Introduction</h2></a>
  <ul>
! 	<li> *
  </ul>
  
! <h2>Major Changes</h2>
! <ul>
! 	<li> *
! </ul>
  
  
--- 29,56 ----
  
  
! <a name=Incompatible><h2>Incompatible Changes</h2></a>
  <ul>
! 	<li> <font color=red>There are some MiddleKit improvements that break compatibility with previous versions of MiddleKit, although all are easy to address. Read below.</font>
  </ul>
  
! <h3>Obj Ref Columns</h3>
! 
! <p> A long standing flaw in MiddleKit was the use of 64-bit integer SQL columns to store references/pointers to objects. The first 32-bits were for the class id and the second 32-bits were for the object id. These are difficult to work with from SQL. For example, an object reference to (classId=1, objId=1) showed as 4294967297 in SQL and couldn't not be readily joined with other tables.
! 
! <p> The new approach is to break these values into two SQL columns. If the attribute name is "foo" then the SQL columns will be named "fooClassId" and "fooObjId". If, for some reason, you don't like those suffixes, you can customize them through the setting <a href=UsersGuide.html#Configuration_ObjRefSuffixes >ObjRefSuffixes</a>. As always, the class ids are stored in the _MKClassIds table that MiddleKit automatically generates.
! 
! <p> <font color=red>If you are upgrading Webware then your database schema will no longer be compatible with MiddleKit.</font> The easy way to fix this is to set UseBigIntObjRefColumns to True in Settings.config (see <a href=UsersGuide.html#Configuration_UseBigIntObjRefColumns>User's Guide - Configuration</a> for more information). This will give you the old behavior. The hard way is to fix up your schema and migrate the data from the old columns to the new.
! 
! 
! <h3>SQLGenerator.config</h3>
! 
! <p> There used to be another config file, besides Settings.config, named SQLGenerator.config, with a single setting, DropStatements. That file is no longer used and the setting has been moved to Settings.config. If you have a SQLGenerator.config in your model directories, move the setting over.
! 
! 
! <h3>MS SQL Support</h3>
! 
! <p> The <span class=name>float</span> type now yields a SQL type of <span class=name>float</span> rather than <span class=name>decimal</span>, because that matches the semantics of Python's float type more closely (perhaps identically). If you want decimal, use decimal; it is a valid MiddleKit datatype.
! 
! <p> The serial primary key field is now typed as int instead of bigint. This matches the MiddleKit approach for other databases and also allows object id columns to be foreign keys to this primary key (MS SQL will give an error if an int column tries to reference a bigint column).
  
  
***************
*** 48,53 ****
  <h2>Improvements and Refinements</h2>
  <ul>
! 	<li>Improved error message when there are errors in the model (instead of 
! 		cryptic tracebacks).  In many cases the filename of the model and the 
  	line number are printed, along with the error message.
  </ul>
--- 63,68 ----
  <h2>Improvements and Refinements</h2>
  <ul>
! 	<li>Improved error message when there are errors in the model (instead of
! 		cryptic tracebacks).  In many cases the filename of the model and the
  	line number are printed, along with the error message.
  </ul>

Index: UsersGuide.html
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Docs/UsersGuide.html,v
retrieving revision 1.24
retrieving revision 1.25
diff -C2 -d -r1.24 -r1.25
*** UsersGuide.html	12 Feb 2003 06:27:59 -0000	1.24
--- UsersGuide.html	24 Feb 2004 02:04:13 -0000	1.25
***************
*** 258,262 ****
  <p> <span class=filename>Settings.config</span> is the primary configuration file.
  
! <p> The <span class=name>Package</span> setting can be used to declare the package that your set of middle objects are contained by. This is useful for keeping your middle objects packaged away from other parts of your programs, thereby reducing the chances of a name conflict. This is the recommended way of using MK.
  
  <p> An example <a name=filename>Settings.config</a>:
--- 258,262 ----
  <p> <span class=filename>Settings.config</span> is the primary configuration file.
  
! <p> <a name=Configuration_Package></a>The <span class=name>Package</span> setting can be used to declare the package that your set of middle objects are contained by. This is useful for keeping your middle objects packaged away from other parts of your programs, thereby reducing the chances of a name conflict. This is the recommended way of using MK.
  
  <p> An example <a name=filename>Settings.config</a>:
***************
*** 274,278 ****
  
  
! <p> The <span class=name>SQLLog</span> setting can be used to get MiddleKit to echo all SQL statements to 'stdout', 'stderr' or a filename. For filenames, an optional 'Mode' setting inside SQLLog can be used to write over or append to an existing file. The default is write. Here are some examples:
  
  <pre>{
--- 274,278 ----
  
  
! <p> <a name=Configuration_SQLLog></a> The <span class=name>SQLLog</span> setting can be used to get MiddleKit to echo all SQL statements to 'stdout', 'stderr' or a filename. For filenames, an optional 'Mode' setting inside SQLLog can be used to write over or append to an existing file. The default is write. Here are some examples:
  
  <pre>{
***************
*** 289,293 ****
  
  
! <p> The <span class=name>Database</span> setting overrides the database name, which is otherwise assumed to be same name as the model. This is particularly useful if you are running two instances of the same application on one host.
  
  <pre>{
--- 289,293 ----
  
  
! <p> <a name=Configuration_Database></a> The <span class=name>Database</span> setting overrides the database name, which is otherwise assumed to be same name as the model. This is particularly useful if you are running two instances of the same application on one host.
  
  <pre>{
***************
*** 296,300 ****
  
  
! <p> The <span class=name>DeleteBehavior</span> setting can be used to change what MiddleKit does when you delete objects.
  The default behavior is "delete" which means that objects are deleted from the SQL database when they are deleted from the MiddleKit object store.
  But setting DeleteBehavior to "mark" causes an extra SQL datetime column called "deleted" to be added to each SQL table, and records that are deleted from the object store in MiddleKit are kept in SQL tables with the deleted field set to the date/time when the object was deleted.
--- 296,300 ----
  
  
! <p> <a name=Configuration_DeleteBehavior></a> The <span class=name>DeleteBehavior</span> setting can be used to change what MiddleKit does when you delete objects.
  The default behavior is "delete" which means that objects are deleted from the SQL database when they are deleted from the MiddleKit object store.
  But setting DeleteBehavior to "mark" causes an extra SQL datetime column called "deleted" to be added to each SQL table, and records that are deleted from the object store in MiddleKit are kept in SQL tables with the deleted field set to the date/time when the object was deleted.
***************
*** 305,309 ****
  }</pre>
  
! <p> The <span class=name>SQLConnectionPoolSize</span> setting is used to create a <a href=../../MiscUtils/Docs/index.html>MiscUtils</a>.DBPool instance for use by the store. For DB API modules with a threadsafety of only 1 (such as MySQLdb), this is particularly useful (in one benchmark, the speed up was 15 - 20%). Simply set the size of the pool in order to have one created and used:
  
  <pre>{
--- 305,309 ----
  }</pre>
  
! <p> <a name=Configuration_SQLConnectionPoolSize></a> The <span class=name>SQLConnectionPoolSize</span> setting is used to create a <a href=../../MiscUtils/Docs/index.html>MiscUtils</a>.DBPool instance for use by the store. For DB API modules with a threadsafety of only 1 (such as MySQLdb), this is particularly useful (in one benchmark, the speed up was 15 - 20%). Simply set the size of the pool in order to have one created and used:
  
  <pre>{
***************
*** 311,317 ****
  }</pre>
  
- <p> The <span class=name>UsePickledClassesCache</span> setting defaults to 1 and causes MiddleKit to cache the <span class=filename>Classes.csv</span> text file as a binary pickle file named <span class=filename>Classes.pickle.cache</span>. This reduces subsequent load times by about 40%. The cache will be ignored if it can't be read, is older than the CSV file, has a different Python version, etc. You don't normally even need to think about this, but if for some reason you would like to turn off the use of the cache, you can do so through this setting.
  
! <p> <span class=filename>SQLGenerator.config</span> has one setting: <span class=name>DropStatements</span> whose potential values are:
  
  <ul>
--- 311,333 ----
  }</pre>
  
  
! <p> <a name=Configuration_ObjRefSuffixes></a> The <span class=name>ObjRefSuffixes</span> controls the suffixes that are appended for the names of the two SQL column that are created for each obj ref attribute. The suffixes must be different from each other.
! 
! <pre>{
!     'ObjRefSuffixes': ('ClassId', 'ObjId'),  # the default
! }</pre>
! 
! 
! <p> <a name=Configuration_UseBigIntObjRefColumns></a> The <span class=name>UseBigIntObjRefColumns</span> causes MiddleKit to store object references in 64-bit fields, instead of in two fields (one for the class id and one for the obj id). You would only do this for a legacy MiddleKit application. Turning this on obsoletes the ObjRefSuffixes setting.
! 
! <pre>{
!     'UseBigIntObjRefColumns': True,  # use single 64-bit obj ref fields
! }</pre>
! 
! 
! 
! <p> <a name=Configuration_UsePickledClassesCache></a> The <span class=name>UsePickledClassesCache</span> setting defaults to 1 and causes MiddleKit to cache the <span class=filename>Classes.csv</span> text file as a binary pickle file named <span class=filename>Classes.pickle.cache</span>. This reduces subsequent load times by about 40%. The cache will be ignored if it can't be read, is older than the CSV file, has a different Python version, etc. You don't normally even need to think about this, but if for some reason you would like to turn off the use of the cache, you can do so through this setting.
! 
! <p> <a name=Configuration_DropStatements></a> The <span class=name>DropStatements</span> setting has these potential values:
  
  <ul>
***************
*** 320,325 ****
  </ul>
  
- <p> An example <a name=filename>SQLGenerator.config</a>:
- 
  <pre>{
      'DropStatements': 'database',  # database, tables
--- 336,339 ----

Update of /cvsroot/webware/Webware/MiddleKit
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv18927

Modified Files:
	.cvsignore 
Log Message:
- split obj ref SQL columns in two (class id and obj id)
- merge SQLGenerater.config into Settings.config
- see release notes for details


Index: .cvsignore
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/.cvsignore,v
retrieving revision 1.1
retrieving revision 1.2
diff -C2 -d -r1.1 -r1.2
*** .cvsignore	25 May 2000 03:18:32 -0000	1.1
--- .cvsignore	24 Feb 2004 02:04:13 -0000	1.2
***************
*** 1 ****
--- 1,2 ----
  *.pyc
+ *.cache

Update of /cvsroot/webware/Webware/MiddleKit/Core
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv18927/Core

Modified Files:
	Model.py ModelUser.py 
Log Message:
- split obj ref SQL columns in two (class id and obj id)
- merge SQLGenerater.config into Settings.config
- see release notes for details


Index: Model.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/Model.py,v
retrieving revision 1.22
retrieving revision 1.23
diff -C2 -d -r1.22 -r1.23
*** Model.py	16 Feb 2004 22:14:52 -0000	1.22
--- Model.py	24 Feb 2004 02:04:13 -0000	1.23
***************
*** 322,324 ****
--- 322,332 ----
  		return {
  			'Threaded': 1,
+ 			'ObjRefSuffixes': ('ClassId', 'ObjId'),
+ 			'UseBigIntObjRefColumns': False,
+ #			'SQLLog': { 'File': 'stdout', },
+ 			'PreSQL': '',
+ 			'PostSQL': '',
+ 			'DropStatements': 'database',  # database, tables
+ 			'ExternalEnumsTableName': None, # '%(ClassName)s%(AttrName)sEnum'
+ 			'ExternalEnumsColumnName': 'name',  # could use %(ClassName)s and %(AttrName)s
  		}

Index: ModelUser.py
===================================================================
RCS file: /cvsroot/webware/Webware/MiddleKit/Core/ModelUser.py,v
retrieving revision 1.4
retrieving revision 1.5
diff -C2 -d -r1.4 -r1.5
*** ModelUser.py	20 Jun 2002 17:47:44 -0000	1.4
--- ModelUser.py	24 Feb 2004 02:04:13 -0000	1.5
***************
*** 1,24 ****
- from MiscUtils.Configurable import Configurable
  from MiscUtils.MixIn import MixIn
  import sys
  from types import *
  
  
! class ModelUser(Configurable):
  
  	## Init ##
  
  	def __init__(self):
- 		Configurable.__init__(self)
  		self._model = None
  
  
! 	## Configuration ##
! 
! 	def configFilename(self):
! 		return self.name() + '.config'
  
! 	def defaultConfig(self):
! 		return {}
  
  
--- 1,24 ----
  from MiscUtils.MixIn import MixIn
+ from MiscUtils import NoDefault
  import sys
  from types import *
  
  
! class ModelUser:
  
  	## Init ##
  
  	def __init__(self):
  		self._model = None
  
  
! 	## Settings ##
  
! 	def setting(self, name, default=NoDefault):
! 		"""
! 		Returns the given setting which is actually just taken from
! 		the model.
! 		"""
! 		return self._model.setting(name, default)
  
  
***************
*** 96,100 ****
  		and potentially customize the behavior, including
  		providing more debugging information.
! 		
  		This implementation writes the msg to stdout.
  		"""
--- 96,100 ----
  		and potentially customize the behavior, including
  		providing more debugging information.
! 
  		This implementation writes the msg to stdout.
  		"""