[rkward-cvs] SF.net SVN: rkward: [1625] trunk/rkward
tfry at users.sourceforge.net
tfry at users.sourceforge.net
Sun Mar 18 23:33:58 UTC 2007
Revision: 1625
http://svn.sourceforge.net/rkward/?rev=1625&view=rev
Author: tfry
Date: 2007-03-18 16:33:55 -0700 (Sun, 18 Mar 2007)
Log Message:
-----------
Reinstantiate display of parse error messages, as much as possible
Modified Paths:
--------------
trunk/rkward/TODO
trunk/rkward/rkward/rbackend/rembedinternal.cpp
Modified: trunk/rkward/TODO
===================================================================
--- trunk/rkward/TODO 2007-03-18 21:54:13 UTC (rev 1624)
+++ trunk/rkward/TODO 2007-03-18 23:33:55 UTC (rev 1625)
@@ -106,10 +106,6 @@
- show current wd in <browser>, somewhere
Internal stuff:
- - Handling locales:
- - RThread should define an own QTextCodec.
- - this should be updated whenever Sys.setlocale() is called
- - Should be used for all conversion to and from local8Bit for data passed to / from the backend
- Handling fonts:
- http://sourceforge.net/mailarchive/forum.php?thread_id=31631211&forum_id=12970
- Look at package odfWeave, and find out, how to use this
Modified: trunk/rkward/rkward/rbackend/rembedinternal.cpp
===================================================================
--- trunk/rkward/rkward/rbackend/rembedinternal.cpp 2007-03-18 21:54:13 UTC (rev 1624)
+++ trunk/rkward/rkward/rbackend/rembedinternal.cpp 2007-03-18 23:33:55 UTC (rev 1625)
@@ -795,7 +795,10 @@
if (*error == NoError) runCommandInternalBase (parsed, error);
} else { // run a user command
SEXP parsed = parseCommand (command_qstring, error);
- if ((*error != NoError)) return;
+ // do not run incomplete commands, but *do* run commands with syntax errors if USE_R_REPLDLLDO1. Why? Because this is the only way to get a syntax error messages, so far.
+ if ((*error != NoError)) {
+ if (*error != SyntaxError) return;
+ }
#ifdef USE_R_REPLDLLDO1
/* Using R_ReplDLLdo1 () is a pain, but it seems to be the only entry point for evaluating a command as if it had been entered on a plain R console (with auto-printing if not invisible, etc.). Esp. since R_Visible is no longer exported in R 2.5.0, as it seems as of today (2007-01-17).
@@ -845,6 +848,9 @@
repldlldo1_wants_code = false; // make sure we don't get confused in RReadConsole
#else
+ // in the non USE_R_REPLDLLDO1 case, any error is fatal, and we should not try running the command at all.
+ if ((*error) != NoError) return;
+
R_Visible = (Rboolean) 0;
SEXP exp;
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
More information about the rkward-tracker
mailing list