[rkward-cvs] SF.net SVN: rkward: [840] trunk/rkward

[tfry at users.sourceforge.net]

Revision: 840
          http://svn.sourceforge.net/rkward/?rev=840&view=rev
Author:   tfry
Date:     2006-10-11 03:15:22 -0700 (Wed, 11 Oct 2006)

Log Message:
-----------
Make sure to watch newly created symbols in the global env

Modified Paths:
--------------
    trunk/rkward/TODO
    trunk/rkward/rkward/core/renvironmentobject.cpp
    trunk/rkward/rkward/core/renvironmentobject.h

Modified: trunk/rkward/TODO
===================================================================
--- trunk/rkward/TODO	2006-10-09 15:44:54 UTC (rev 839)
+++ trunk/rkward/TODO	2006-10-11 10:15:22 UTC (rev 840)
@@ -68,6 +68,12 @@
 	- use RKGlobals::marginHint (), RKGlobals::spacingHint () everywhere
 
 Internal stuff:
+	RKVariable:
+		- the storage mechanism for invalid values (and the type detection) is flawed. The following should be done instead:
+			- R storage type is always == RKWard data type.
+			- Invalid values are stored in a list in an attribute of the variable in question
+			- The corresponding "real" fields are set to NA
+			- Need to add a conversion script. See .rk.get.meta (). Both should be changed at the same time.
 	- maybe we can do stuff like auto-printing, (toplevel handlers: no, not those, they are not called in R_ReplDLLdo1 ()), syntax error information by using R_ReplDLLdo1 () for user/console commands. The command-text would be fed in via R_ReadConsole (would need to keep a buffer).
 			- How to differentiate readline calls to get new command data from readline calls to get user info?! The prompt may not be realiable enough. Is there other info?
 				- the hist parameter may be a hint, but is also set in do_browser

Modified: trunk/rkward/rkward/core/renvironmentobject.cpp
===================================================================
--- trunk/rkward/rkward/core/renvironmentobject.cpp	2006-10-09 15:44:54 UTC (rev 839)
+++ trunk/rkward/rkward/core/renvironmentobject.cpp	2006-10-11 10:15:22 UTC (rev 840)
@@ -63,6 +63,19 @@
 	return (name + "$" + short_child_name);
 }
 
+RObject *REnvironmentObject::createNewChild (const QString &name, RKEditor *creator, bool container, bool data_frame) {
+	RK_TRACE (OBJECTS);
+
+	RObject *obj = RContainerObject::createNewChild (name, creator, container, data_frame);
+
+	if (type & GlobalEnv) {
+		RKGlobals::rInterface ()->issueCommand (".rk.watch.symbol (" + rQuote (name) + ")", RCommand::App | RCommand::Sync, QString::null, 0);
+	}
+
+	return obj;
+}
+
+
 void REnvironmentObject::writeMetaData (RCommandChain *chain) {
 	RK_TRACE (OBJECTS);
 

Modified: trunk/rkward/rkward/core/renvironmentobject.h
===================================================================
--- trunk/rkward/rkward/core/renvironmentobject.h	2006-10-09 15:44:54 UTC (rev 839)
+++ trunk/rkward/rkward/core/renvironmentobject.h	2006-10-11 10:15:22 UTC (rev 840)
@@ -38,6 +38,8 @@
 	QString makeChildName (const QString &short_child_name, bool misplaced=false);
 /** reimplemented from RContainerObject: If this is an environment var, call RContainerObject::writeMetaData (). Else, do nothing. An environment has no meta data. */
 	void writeMetaData (RCommandChain *chain);
+/** reimplemented from RContainerObject: If this is the global env, add a watch for the new child */
+	RObject *createNewChild (const QString &name, RKEditor *creator=0, bool container=false, bool data_frame=false);
 protected:
 	bool updateStructure (RData *new_data);
 /** reimplemented from RContainerObject to raise an assert if this is not the isGlobalEnv (). Otherwise calls "remove (objectname)" instead of objectname <- NULL" */


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.