[Rd] Why does the lexical analyzer drop comments ?

Romain Francois romain.francois at dbmail.com
Tue Mar 31 16:39:04 CEST 2009

hadley wickham wrote:
>> At the moment, I am concentrating efforts deep down in the parser code, but
>> there are other challenges:
>> - once the expressions are parsed, we will need something that investigates
>> to find evidence about function calls, to get an idea of where the function
>> is defined (by the user, in a package, ...) . This is tricky, and unless you
>> actually evaluate the code, there will be some errors made.
> Are you aware of Luke Tierney's codetools package?  That would seem to
> be the place to start.
Yep. Plan to combine the more verbose information out of the modified 
parser with the same guess machine that checkUsage uses.
Another side effect is that we could imagine to link error patterns 
identified by checkUsage (no visible binding for global variable "y", 
...) to actual locations on the file (for example the place where the 
variable y is used in that case ), which at the moment is not possible 
because the parser only locates entire expression (semantic groupings) 
and not tokens.

 > f <- function( x = 2) {
+ y + 2
+ }
 > checkUsage( f )
<anonymous>: no visible binding for global variable ‘y’

> Hadley

Romain Francois
Independent R Consultant
+33(0) 6 28 91 30 30

More information about the R-devel mailing list