Junio C Hamano writes:
>> ...
>> +static int match_placeholder_bool_arg(const char *to_parse, const char
>> *candidate,
>> + const char **end, int *val)
>> +{
>> + char buf[8];
>> + const char *strval;
>> + size_t len;
>> + int v;
>> +
>> + if (!match_placeholder_arg_value(to_parse, candidate, end, &strval,
>> &len))
>> + return 0;
>> +
>> + if (!strval) {
>> + *val = 1;
>> + return 1;
>> + }
>> +
>> + strlcpy(buf, strval, sizeof(buf));
>> + if (len < sizeof(buf))
>> + buf[len] = 0;
>
> Doesn't strlcpy() terminate buf[len] if len is short enough?
> Even if the strval is longer than buf[], strlcpy() would truncate
> and make sure buf[] is NUL terminated, no?
Yes, but no. strval is not NUL-terminated at len. E.g strval would point
to "false,something=true". `buf[len] = 0` makes sure it becomes "false".
> Instead of using "char buf[8]", just using a strbuf and avoidng
> strlcpy() would make the code much better, I would think.
Yes, taking the heap allocation hit would most likely make the intent
clearer.
