Implicit Function Declaration Issue

[mike-ward]

Describe the bug

This may or may not be a bug. Posting for discussion.

Reproduction Steps

Started out with this function which compiles and runs successfully

fn sort_dictionary_order(mut lines []string, options Options) {
	lines.sort_with_compare(fn (a &string, b &string) int {
		aa := a.bytes().map(fn (e u8) u8 {
			return if e.is_digit() || e.is_letter() || e == ` ` { e } else { ` ` }
		}).bytestr()
		bb := b.bytes().map(fn (e u8) u8 {
			return if e.is_digit() || e.is_letter() || e == ` ` { e } else { ` ` }
		}).bytestr()
		return compare_strings(aa, bb)
	})
}

To eliminate the repeated code in the map functions I refactored as follows:

fn sort_dictionary_order(mut lines []string, options Options) {
	map_fn := fn (e u8) u8 {
		return if e.is_digit() || e.is_letter() || e == ` ` { e } else { ` ` }
	}
	lines.sort_with_compare(fn [map_fn] (a &string, b &string) int {
		aa := a.bytes().map(map_fn).bytestr()
		bb := b.bytes().map(map_fn).bytestr()
		return compare_strings(aa, bb)
	})
}

The refactored code generated an error message.

/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:4103:12: error: call to undeclared function 'map_fn'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
                u8 _t2 = map_fn(it);
                         ^
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:4114:12: error: call to undeclared function 'map_fn'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
                u8 _t5 = map_fn(it);
                         ^
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:15615:4: warning: expression result unused [-Wunused-value]
  (*(flag__Flag*)_t1.data);
   ^~~~~~~~~~~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:15622:4: warning: expression result unused [-Wunused-value]
  (*(flag__Flag*)_t2.data);
   ^~~~~~~~~~~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:15777:4: warning: expression result unused [-Wunused-value]
  (*(string*)_t3.data);
   ^~~~~~~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:20460:4: warning: expression result unused [-Wunused-value]
  (*(int*)_t4.data);
   ^~~~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:20881:4: warning: expression result unused [-Wunused-value]
  (*(string*)_t1.data);
   ^~~~~~~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:22372:4: warning: expression result unused [-Wunused-value]
  (*(os__SignalHandler*)_t3.data);
   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:22560:20: warning: passing 'u8 *' (aka 'unsigned char *') to parameter of type 'const char *' converts between pointers to integer types where one is of the unique plain 'char' type and the other is not [-Wpointer-sign]
        setlocale(LC_ALL, _SLIT("").str);
                          ^~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:244:18: note: expanded from macro '_SLIT'
#define _SLIT(s) ((string){.str=(byteptr)("" s), .len=(sizeof(s)-1), .is_lit=1})
                 ^
/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include/locale.h:53:35: note: passing argument to parameter here
char            *setlocale(int, const char *);
                                            ^
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:22817:4: warning: expression result unused [-Wunused-value]
  (*(os__File*)_t1.data);
   ^~~~~~~~~~~~~~~~~~~~
/tmp/v_501/sort.01J1G0G6TX6WE0YT9J99TFE09D.tmp.c:22841:4: warning: expression result unused [-Wunused-value]
  (*(int*)_t3.data);
   ^~~~~~~~~~~~~~~
9 warnings and 2 errors generated.

However, if I lift the map_fn out of the function, it compiles and works fine.

fn map_fn(e u8) u8 {
	return if e.is_digit() || e.is_letter() || e == ` ` { e } else { ` ` }
}

fn sort_dictionary_order(mut lines []string, options Options) {
	lines.sort_with_compare(fn (a &string, b &string) int {
		aa := a.bytes().map(map_fn).bytestr()
		bb := b.bytes().map(map_fn).bytestr()
		return compare_strings(aa, bb)
	})
}

Expected Behavior

Should the second method compile?

Current Behavior

See error message above

Possible Solution

No response

Additional Information/Context

No response

V version

V 0.4.6 fccd7cd

Environment details (OS name and version, etc.)

V full version: V 0.4.6 8215f21.fccd7cd
OS: macos, macOS, 14.5, 23F79
Processor: 8 cpus, 64bit, little endian, Apple M2

getwd: /Users/mike/Documents/github/coreutils/src/sort
vexe: /Users/mike/Documents/github/v/v
vexe mtime: 2024-06-27 19:39:41

vroot: OK, value: /Users/mike/Documents/github/v
VMODULES: OK, value: /Users/mike/.vmodules
VTMP: OK, value: /tmp/v_501

Git version: git version 2.45.2
Git vroot status: weekly.2024.26-11-gfccd7cd0 (3 commit(s) behind V master)
.git/config present: true

CC version: Apple clang version 15.0.0 (clang-1500.3.9.4)
thirdparty/tcc status: thirdparty-macos-arm64 5c1d002f

Note

You can use the 👍 reaction to increase the issue's priority for developers.

Please note that only the 👍 reaction to the issue itself counts as a vote.
Other reactions and those to comments will not be taken into account.

[JalonSolov]

Well, V should be at least giving a better error message.

Seeing code like this makes me wonder... why? Why not just declare a normal function, and call it?

The only difference in doing that is that you have to name the function, instead of V coming up with a dummy name.

And that is exactly what you did to make it work.

All that said, it should probably work with the anonymous function. Or, as I said, give a nice error message telling you why not.

(it's just too FP for me. :-) )

[mike-ward]

It's why I said it may be bug. I would be fine with a better error message. On the other hand, it might point to an issue in the compiler. Not my area of expertise.