You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Seemingly, everything very simple, and as static as can be. However, the stack trace is like:
ERROR: cannot parse file name
called from <anonymous>
called from <anonymous> on line 1268 of file /usr/share/chibi-scheme/init-7.scm
called from <anonymous> on line 800 of file /usr/share/chibi-scheme/init-7.scm
The script file name is not in the stack trace, which is kind of weird? Well, the function call may be eliminated due to tail recursion, but "error", I guess, can save source line information somewhere?
ERROR: cannot parse file name
called from <anonymous> on line 1268 of file /usr/share/chibi-scheme/init-7.scm
called from <anonymous> on line 800 of file /usr/share/chibi-scheme/init-7.scm
So the difference is a missing (newline), which, I guess, is a syntax error, but having it influence the stack trace is confusing as well.
If that "always failing" procedure is removed, but the erroneous (?) (newline) is present:
ERROR: cannot parse file name
called from parse-url-into-website-program_name-version on line 4 of file ./test-error-file-line-info.scm
called from <anonymous> on line 1268 of file /usr/share/chibi-scheme/init-7.scm
called from <anonymous> on line 800 of file /usr/share/chibi-scheme/init-7.scm
Suddenly we are getting line information correctly!
However, removing that "erroneous" (newline) removes line information once again:
./test-error-file-line-info.scm
ERROR: cannot parse file name
called from <anonymous> on line 1268 of file /usr/share/chibi-scheme/init-7.scm
called from <anonymous> on line 800 of file /usr/share/chibi-scheme/init-7.scm
Maybe some syntactic expanders "forget" to attach source line information to sexps? Or, maybe, it would be worth attaching source line information to error objects, regardless of the stack trace?
The text was updated successfully, but these errors were encountered:
This might be a side-effect of #837 , although I am not very sure about that.
I have two MWEs:
One:
Seemingly, everything very simple, and as static as can be. However, the stack trace is like:
The script file name is not in the stack trace, which is kind of weird? Well, the function call may be eliminated due to tail recursion, but "error", I guess, can save source line information somewhere?
The other one is :
And the stack trace is:
So the difference is a missing
(newline)
, which, I guess, is a syntax error, but having it influence the stack trace is confusing as well.If that "always failing" procedure is removed, but the erroneous (?)
(newline)
is present:The stack trace is:
Suddenly we are getting line information correctly!
However, removing that "erroneous"
(newline)
removes line information once again:Maybe some syntactic expanders "forget" to attach source line information to sexps? Or, maybe, it would be worth attaching source line information to error objects, regardless of the stack trace?
The text was updated successfully, but these errors were encountered: