[antlr-interest] [CSharp3] TraceIn/TraceOut and ANTLR_TRACE conditional

[Sam Harwell]

Hi Peter,

The TraceIn/TraceOut methods in Parser and BaseRecognizer have different
signatures (2 parameters vs. 3). The code generator emits calls to the one
in Parser, which is supposed to forward the calls to BaseRecognizer. From
what I can tell from the code generation templates, the methods in
BaseRecognizer should not have the ConditionalAttribute applied to them.
I'll remove the attribute for the next build, which will restore the
expected behavior to the methods in Parser and Lexer.

Thanks for the report. :)

Sam

-----Original Message-----
From: antlr-interest-bounces at antlr.org
[mailto:antlr-interest-bounces at antlr.org] On Behalf Of Peter Kooiman
Sent: Saturday, April 02, 2011 11:42 AM
To: antlr-interest at antlr.org
Subject: [antlr-interest] [CSharp3] TraceIn/TraceOut and ANTLR_TRACE
conditional

After struggling a bit to get tracing working in my parser by defining
ANTLR_TRACE and getting nowhere, I realized the following, which may or may
not be intended: 

Both the Parser and Lexer classes inherit from BaseRecognizer and override
(well, hide) the BaseRecognizer TraceIn and TraceOut methods. These
overrides call base.Traceout / base.TraceIn, BUT, since during library
compilation ANTLR_TRACE is not defined, these base method calls become
no-ops. The result is, that while BaseRecognizer has usable TraceIn and
TraceOut methods, the TraceIn and TraceOut methods of both Parser and Lexer
become essentially no-ops. 

This is the MSIL for Parser::TraceIn in Antlr3.Runtime.dll: 

.method public hidebysig newslot virtual instance void TraceIn(string
ruleName,
int32 ruleIndex) cil managed
{
.custom instance void
[mscorlib]System.Diagnostics.ConditionalAttribute::.ctor(string) = ( 01 00
0B 41 4E 54 4C 52 5F 54 52 41 43 45 00 00 ) // ...ANTLR_TRACE.. 
// Code size 1 (0x1)
.maxstack 8
IL_0000: ret
} // end of method Parser::TraceIn 

Fixed easily enough by overriding TraceIn/TraceOut in my parser, but a bit
of a pain to figure out, so I thought I'd at least document it here. 

Regards
Peter 

List: http://www.antlr.org/mailman/listinfo/antlr-interest
Unsubscribe:
http://www.antlr.org/mailman/options/antlr-interest/your-email-address