Getting part of string with sscanf()

[AndiPersti]

I'm currently working through the book "C Programming: A Modern Approach" (by K.N.King) and I'm having some problems with the following exercise:

Assume that "str" is a string that contains a "sales rank" immediately preceded by the # symbol (other characters may precede the # and/or follow the sales rank). A sales rank is a series of decimal digits possibly containing commas, such as the following examples:
989
24,675
1,162,620
Write a call of "sscanf" that extracts the sales rank (but not the # symbol) and stores it in a string variable named "sales_rank".

My solution is

sscanf(str, "%*[^#]#%[0-9,]", sales_rank);

which works except for the case where "str" starts with the # symbol (without any characters preceding it).
After rereading the corresponding parts in the book and of the man page of sscanf I don't see any way to have a single call of sscanf so that it works both for any characters before # and no character before #.
Am I missing something or misinterpreting the exercise?

Bye, Andreas

[oogabooga]

Good question. It doesn't seem possible since %*[^#] says to read at least one non # character. Can you do it with more than one sscanf?

[oogabooga]

Of course right after I say it can't be done I realize that it can. Consider a format string like this:

(*str=='#') ? "#%[0-9,]" : "%*[^#]#%[0-9,]"

[AndiPersti]

Thanks, that's probably the solution the author thought about.

Bye, Andreas

[christop]

Of course right after I say it can't be done I realize that it can. Consider a format string like this:

(*str=='#') ? "#%[0-9,]" : "%*[^#]#%[0-9,]"

Or more succinctly:

"%*[^#]#%[0-9,]"+6*(*str=='#')

But I recommend a solution like that only for the IOCCC.