SjASMPlus от z00m

[Ped7g]

а на каком pass-е уже будет результат?
так как полученные цифры еще будут передаваться lua для расчета таблиц

Macros are like "includes", expanded upon instance in every pass (becoming part of source). Notice how inside my example macro "CN=CN+(cycles?)" -> the argument is in parentheses - that's because the macro arguments are substituted in C-preprocessor way, substituting the strings, doing the value evaluation as last step after all substitutions, so without parentheses the usage `countT (26-CN+3)&-4` does expand to `CN=CN+(26-CN+3)&-4` which will evaluate to different result as `CN=CN+((26-CN+3)&-4)`! So it's similar landmine field like C #defines ...

But the value of symbol CN is updated in every pass, just as if you would do the CN=CN+4 at beginning of line (but it happens after instruction, because macro is expanded after instruction).

---------------

listing disabling:
http://z00m128.github.io/sjasmplus/d...on.html#po_opt

    OPT push listoff    ; preserve current state of listing, and switch listing off
      ; some code without listing (only marked by "# listing disabled" and "# listing enabled" lines in listing)
      ...
    OPT pop              ; restore the listing to original state

(the disabling/enabling of listing is announced in the listing file, so this shortens listing for blocks of code, for single line this will explode listing even more)

the OPT state can be push/pop-ed, so you can even nest these in includes/etc... if the top level has listing enabled, and it will go through multiple nested blocks having this switch off, then the first will switch listing off, nested ones will make no visible change, and the final last `opt pop` will restore listing back to "on".

Similarly if you are doing sjasmplus library to be included as source code, and you want customize the syntax, but you don't want to affect project which is including your source, you can do `OPT push reset --syntax=aBf` to have your source going by these rules, then `OPT pop` at end of your library to restore the syntax of original project...

Sorry for spamming you with this info you didn't ask for, but I'm kinda proud of the OPT design... :P

[Advertiser]

[NEO SPECTRUMAN]

Ped7g, а сейчас как то можно организовать синонимы для макросов?

есть куча макросов которые для сохранения совместимости переименовать нельзя
ну и делать копию тоже нельзя


по смыслу должно работать так

	macro aabb123 || blahblahblah_aa_bb_123_NOW111 aaa 
	defb aaa
	...
	100500 lines of code
	...
	endm

или так

	macro aabb123 || blahblahblah_aa_bb_123_NOW111 aaa 
	defb aaa
	...
	100500 lines of code
	...
	endm

	macro blahblahblah_aa_bb_123_NOW111 aaa 
	aabb123 aaa
	endm

с результатом

	aabb123 1

	blahblahblah_aa_bb_123_NOW111 2

	defb 1
	...
	100500 lines of code
	...
	defb 2
	...
	100500 lines of code
	...

при этом на всех PASS-ах
синонимы должны вести себя так же как и оригинальный макрос

define-ом наверно будет опасно пользоваться для такого
хотя надо будет попробовать

[Grand]

Полезно и в восьмеричном виде выводить числовую информацию. Вспомните, байты атрибутов экрана нагляднее именно в восьмеричном виде.

[NEO SPECTRUMAN]

так намного лучше

[Black Cat / Era CG]

А еще язык славянский.

[Ped7g]

So you want the same macro, but with the two different names?

The second way should work quite ok?

	macro aabb123 aaa ; "main" version 
	defb aaa
	...
	100500 lines of code
	...
	endm

	macro blahblahblah_aa_bb_123_NOW111 aaa ; "alias" version 
	aabb123 aaa ; call the "main" version of macro
	endm

	; usage
	aabb123 1
	blahblahblah_aa_bb_123_NOW111 2

Is this broken for you?

DEFINE is of course another option, but that works best with distinct names, like `DEFINE blahblahblah_aa_bb_123_NOW111 aabb123` .. seems unlikely you will have collision with such long string like "blahblahblah_aa_bb_123_NOW111" (although may happen, if you have "blahblahblah_aa_bb_123_NOW111_2" -> "aabb123_2" will be substituted.

[NEO SPECTRUMAN]

The second way should work quite ok?

еще не проверил
додумался до такого варианта только когда уже написал сообщение

да и есть некоторые сомнения
так как на PASS 1 уже в макросах прицеплена lua

[Ped7g]

The nested macros are possible and should generally work, although they will exercise some not pretty parts of the code, and the maximum depth of nesting is limited (like 20 probably). So I can't promise there's no hidden bug nobody reported yet, but try it, it should work.

[NEO SPECTRUMAN]

The nested macros are possible and should generally work, although they will exercise some not pretty parts of the code, and the maximum depth of nesting is limited (like 20 probably). So I can't promise there's no hidden bug nobody reported yet, but try it, it should work.

интересная информация

а так уже применил
сейчас тестирую

[NEO SPECTRUMAN]

меня периодически достает такая бага
или может быть я чего то не вижу?

SjASMPlus Z80 Cross-Assembler v1.17.0 (https://github.com/z00m128/sjasmplu
Pass 1 complete (0 errors)
Pass 2 complete (0 errors)
test.asm(34): error: [IF] No ENDIF
Pass 3 complete
Errors: 1, warnings: 0, compiled: 14 lines, work time: 0.000 seconds

mode = 1

	if mode = 6
		lua allpass	
			for temp_cnt = 0,255,2 do

				var = 0

				if var > 65535 then
				var = 65535
				end
			end	
		endlua
	endif

в исходнике из которого я эту конструкция скопировал (упрощенно)
ошибка происходит уже на pass 3

- - - Добавлено - - -

при этом если сделать

mode = 1

	if mode = 6
	include "fix.asm
	endif

fix.asm

		lua allpass	
			for temp_cnt = 0,255,2 do

				var = 0

				if var > 65535 then
				var = 65535
				end
			end	
		endlua

то все работает...

SjASMPlus Z80 Cross-Assembler v1.17.0 (https://github.com/z00m128/sjasmplu
Pass 1 complete (0 errors)
Pass 2 complete (0 errors)
Pass 3 complete
Errors: 0, warnings: 0, compiled: 24 lines, work time: 0.000 seconds

так тоже без ошибок

mode = 1

	if mode = 6
	
		lua allpass	
			for temp_cnt = 0,255,2 do

				var = 0

				--if var > 65535 then
				--var = 65535
				--end
			end	
		endlua

	endif