Salmelo:"Here is another patch, this time adding counters to target specifications, so now you should be able to target things with counters on them as well as use counters for criterion for Lord, aslongas, foreach, etc.
I had to muddle the syntax a little bit though, so heres an example, taken from gwafa hazid, whom I added to the primitives to test with.
lord(creature[counter{0/0.1.Bribe}]|opponentbattlefield) cantattack
obviously the part we are worried about is the counter part, note that those are curly braces { } and periods . instead of parentheses ( ) and commas , this is so that it does not conflict with how targets are normally parsed, which it did before I changed those. Counters still work the same way everywhere else though, you only need to use { } and periods in target code.
Also, you can use counter{any} to specify that it should look for things with any kind of counter on them, not just specific ones, this is used by Kulrath Knight, for example.
I also consolidated most of the code used to parse counter specifications, as it seemed unnecessary duplicating it three times.
Anyway, aside from this major addition, I also added Gwafa Hazid, Profiteer; and Kulrath Knight to the primitives, as well as a test file for each to make sure it worked.
Hopefully y'all will find this patch useful.
Oh, and I should point out that it only works with one Counter in the target specification, i imagine if you put two in the same one then it would either overwrite the first with the second or merge them into some sort of hybrid counter monstrosity, either way, I wouldn't suggest it. "
Cards with vanishing are also codable now. Have a look at the comments for explanation.
daily_build.
- Added remove/add counters as a cost, patch by Salmelo, thanks man!. See primitives/mtg.txt -> Thallid to see how it works
- added test for i286 by salmelo
- added basic display for counters (this needs improvement) by salmelo
* Fixes for compilation on 64-bit architectures :
string position-returning functions return size_t, not uint.
Fix that.
* Fixes warnings with new gcc about non-fixed string AND no arguments.
NOTE : 64-bit architectures still do not compile. There are 2 different places
where printf is used with a specifier that is not large enough on 64-bit because
size_t is now a ulong and not a uint. The solution on GNU systems is the %zu
specifier, but as I'm not sure it is supported on windows I don't want to
introduce a huge hard-to-notice bug, so I'll do it at a time when I got a
windowser to back me up.
- fix for issue 301 (creatures go to graveyard)
- This adds an important change to "movedTo", which now accepts a "full" target description in both "from" and "movedTo". The point is that a card does not move to "creature|graveyard" from "battlefield" but moves from "creature|battlefield" to "graveyard".
Notes to programmers:
1. This feature uses an awful lot of copy-pasta, using "protetcion from()" as a base. While I'm learning the architecture, it's easier for me to have a dedicated single-purpose piece of code to work with, than trying to create multi-purpose code. I'm aware that this isn't a very elegant approach though, and I hope to be able to refactor additions like this into multi-purpose code once I have a better understanding of the architecture as a whole.
2. Please check the questions I'll add in my next code comments, specifically about AI integration and activated abilities / instants / sorceries.
3. I did have a crash (with the debugger complaining about stack corruption around the cd variable) during testing. I wasn't able to reproduce it though. I did have to clean the solution in-between when I updated to the primitives system, so perhaps there was this cleaning solved whatever corruption was in my files. I'm mentioning the issue in case someone has an idea on where my code might be risky (I'm still probne to making beginners mistakes).
- Card Primitives system. Check Royal Assassin in RV, 10E, M10
- Please review, is sets/primitives a good directory? Should we rename MTGCard into "CardPrint"?
- Unfortunately for now it is not possible to "override" a Primitive. A card that links to a primitive but also defines new "values" will create its own data and ignore the data in the "linked" primitive for the time being. I hope to solve that at some point...
Limitations:
- Operators for "greater than", "less than", "unequal" have not been implemented, but if a card actually needs them, you can use a preceding minus sign to negate a comparison. Example: -power=3 means "power not equal to 3", -toughness<=3 means "toughness>3".
- You can't use spaces when specifying such restrictions. Write "power<=3" instead of "power <= 3"
- You now need to use a space before the "<" and ">" commands that count the matches for lord(), foreach(), all() and aslongas(). So far we always did use spaces in front of them without actually needing to, now we need to.
- manacost restrictions don't take "X" costs into account. Example: Mistmeadow Skulkin (FUT) has protection from manacost>=3. Blaze has a converted manacost of 1, but when you cast it with an X of 2, then it actually has a converted manacost of 3 while on the stack, and Mistmeadow Skulkin would be protected from it, but currently it isn't.
Please review the code, I'll add a few remarks/questions of my own.
-fix issue 161 (brass man can untap outside of upkeep). Use "myUpkeepOnly" keyword. Other keywords such as "myDrawOnly" of course work. Haven't implemented "opponent" yet, is it needed ?
- fix issue 168 (kudzu)
- Fix issue 162 (copy VS shroud). This fix introduces a new keyword: NotATarget() instead of Target(). (Which was the best way for me to keep some kind of backward compatibility/ not alter the code too much / fix the bug)
- fixed divergent growth (issue 243). Unfortunately the underlying issue is quite complex, and Wagic doesn't handle it very well. Basically, all "until end of turn" effects run the risk of being partially deleted before the end of the turn, leading to segfaults or memory leaks. This is because some abilities' "clone" method doesn't actually clone them entirely... The fix for divergent growth is a hack that works only for lords+mana producer combination, but I need a more "generic fix" in the long run...
- fix a memory leak with "Prevent All combat damages" ability
- Fix issue 242 (Equip can be used outside of the main phases)
- introducing "attach" keyword. Same as equip but can be used anytime. Untested
- introducing "asSorcery" keyword. Can be used the same way as "myTurnOnly" on activated abilities to restrict their usage. Untested. Other similar keywords will follow, please let me know which ones would be useful
- Adding cycling. Check Akroma's vengeance in ONS for an example. Note that this uses autohand instead of auto, this is important! You can also use autograveyard.
- All "auto" activated abilities should work with autohand, so this is not only for cycling, but could be used for other abilities as well. For example autohand={3}:cycling can also be written autohand={3}{S}:Draw:1
- Adding equipments. They work like auras, except you have to add an "auto={cost}:equip" line. See Behemoth sledge in ARB for an example. Please test a lot before committing, thanks :)
-added WParsedInt (X, p, t, manacost) for "Draw" effects. See prosperity
- Fixed denizen of the deep (P02), it had broken the test suite, sorry for that!
- Code cleanup
- fix issue 142 . This will lead to other issues for some cards. These issues can be fixed by using targetcontroller/controller...etc
- Added a few cards
- removed some unused code. Please review!
- Added protection from() auto keyword. It is still possible to use protection from [color] in abilities, but when it is not possible, please use protection from([target]) in auto=
-fixed a memory leak
- Added P02 and PTK
- New way to create tokens in the parser, much more flexible, see the Hive in RV. Tokens can now be written as other cards, with a rarity of "T". I suggest their id to be the negative value of the card that generates them when possible. Naming convention for images is the same as before: a negative id such as -1138 will need a [id]t.jpg image (1138t.jpg). Positive ids work as "normal" pictures
- Added "PreventAllCombatDamage" [from(...)] [to(...)] keyword. Please test it on a few cards before we "mass" use it. As a side effect, fixed issue 155 (ebony horse target).
-reduced SFX Quality. This seems to fix the Bug with samples getting silent on the PSP (issue 112), and improves loading times. If you think this is unacceptable, let's discuss it
- AI now should play fetchlands' ability, although still not "efficiently"
- Fix a bug where the game would crash if a card has a valid alias to a hardcoded card that is in no _cards.dat (starwars mod cards aliasing a MTG card)
Please note that I don't plan to update this file for future Wagic versions, so if anybody wants to take over, please do. :)
Some files were updated in the process:
- CardGui.cpp - exposed the card rarity info on the alternate render cards to the translation engine. Also switched the color of said info to white on green and blue cards, imho they are much easier to read now.
- GameStateDeckViewer.cpp - exposed the term "Collection" (shown under the scrollbar) to the translation engine.
- GuiPhaseBar.cpp - exposed the strings denoting "your turn", "opponent's turn", "you play", "opponent plays", to the translation engine.
- SimplePad.cpp - exposed the special keys ("Spacebar", "Confirm" etc.) to the translation engine. Had to increase the keypad width to make room for "Abbrechen" (Cancel). I couldn't translate it differently because there's also a *spell* called cancel, and to have a different translation in SimplePad, I would have had to change the official German name of the Cancel spell, which wasn't an option.
- MTGAbility.cpp - exposed the words from which the text of mana-producing abilities is constructed to the translation engine. Please review this one and suggest a better solution - the one I chose is somewhat awkward. The translation works, but when the whole translated sentence is constructed and gets returned, the calling procedure tries to translate it again (all other ability descriptions work this way, translation takes place *after* the string has been returned). However, for mana producing abilities this doesn't work, since the costructed string can take an infinite number of diffeent forms, depending on the mana produced, and we can't translate that. Hence I chose to translate the individual words during construction, with the described side effect that the engine now tries to translate them again later in the process (and the sentence ends up in missing.txt).
- Also, cheat mode menu items now aren't denoted by the ugly "(cheat)" appended to them, instead they are enclosed by asterisks. Uses less space.
salmelo16
solo81@web.de
wagic.the.homebrew@gmail.com
jean.chalard
Psyyringe
wagic.laurent
d32.wagic