Re: [GSoC PATCH] rm: fix sign comparison warnings

[Arnav Bhate <bhatearnav@gmail.com>]

Junio C Hamano <gitster@pobox.com> writes:
> Arnav Bhate <bhatearnav@gmail.com> writes:
> 
>>  static int get_ours_cache_pos(const char *path, int pos)
>>  {
>> -	int i = -pos - 1;
>> +	/*
>> +	 * This function is only called when pos < 0, so -pos - 1 is
>> +	 * greater than or equal to 0, so it can be safely be stored in
>> +	 * an unsigned int.
>> +	 */
>> +	unsigned int i = -pos - 1;
> 
> "Can be safely stored", sure.
> 
> But so is "int i" perfectly adequate to hold such a value, no?
> 
> This is one of the many instances that demonstrate why the
> "-Wsign-compare" warning is of dubious value, and invites worse code
> than necessary.

While "int i" may be adequate, this, in my opinion, emphasises the fact
that we do not want i to take negative values, so I do not think this is
worse. Karthik's suggestion in another reply is also an alternative.
 
>> @@ -58,7 +62,7 @@ static void print_error_files(struct string_list *files_list,
>>  			      int *errs)
>>  {
>>  	if (files_list->nr) {
>> -		int i;
>> +		unsigned int i;
>>  		struct strbuf err_msg = STRBUF_INIT;
>>  
>>  		strbuf_addstr(&err_msg, main_msg);
>> @@ -271,6 +275,7 @@ int cmd_rm(int argc,
>>  {
>>  	struct lock_file lock_file = LOCK_INIT;
>>  	int i, ret = 0;
>> +	unsigned int j;
>>  	struct pathspec pathspec;
>>  	char *seen;
>>  
>> @@ -314,8 +319,8 @@ int cmd_rm(int argc,
>>  	if (pathspec_needs_expanded_index(the_repository->index, &pathspec))
>>  		ensure_full_index(the_repository->index);
>>  
>> -	for (i = 0; i < the_repository->index->cache_nr; i++) {
>> -		const struct cache_entry *ce = the_repository->index->cache[i];
>> +	for (j = 0; j < the_repository->index->cache_nr; j++) {
>> +		const struct cache_entry *ce = the_repository->index->cache[j];
>>  
>>  		if (!include_sparse &&
>>  		    (ce_skip_worktree(ce) ||

-- 
Regards,
Arnav Bhate
(He/Him)