D2TxtAnalyser : debug your Mod easily

[SVR]

You can use my properties.txt from Udie2 as a starting point Paul.
I have each stat split.
I had property links in 1.09 for the stats but 1.10 changed the way properties linked so that info is invalid now. (though it's still in the file ;-)

You will run into problems with changed & added stats/properties though.

[Paul Siramy]

New version released (2.4).

It now test prop/param/min/max columns in all .txt, for at least basic errors (FoxBat should appreciate ). I still have to implement SVR idea about states.txt ID for instance (a column that uses ID should take the appropiate .txt, and check if the IDs are truly exists).

As always, if you have other ideas / needs for that tool, ask there.

[Sduibek]

Oh man, this tool just keeps getting better and better. What's next?!? Can't wait


Does this error "matter"? I swear i've used negative modifiers listed in that way before without them malfunctioning in-game. HMmm, could be negative modifiers are handled specially by the game. I'll look into it and let you know

* ERROR : Propertie (dmg-ac) : 'mod1min' (-25) should be <= than 'mod1max' (-40), row 47 (of Perforation)
* ERROR : Propertie (dmg-ac) : 'mod1min' (-10) should be <= than 'mod1max' (-20), row 48 (of Puncturing)




What about this one? -->

In Data\Global\Excel\Armor.txt :
* ERROR : Item 'code' (ci2) at row 117 (Tiara) has 'normcode' (ci1), expected (ci0)
* ERROR : Item 'code' (ci3) at row 118 (Diadem) has 'normcode' (ci1), expected (ci0)

I messed around with it, but can't seem to fix the error. Just wonder though if this error report simply reflects on how your program analyzes the txt files, or if in it's Vanilla form there's actually (for example) funky functionality when gambling for a circlet or something. I could i'll have to look into this one as well :p

[Paul Siramy]

[quote=Sduibek";p="184287"]Does this error "matter"? I swear i've used negative modifiers listed in that way before without them malfunctioning in-game. HMmm, could be negative modifiers are handled specially by the game. I'll look into it and let you know

* ERROR : Propertie (dmg-ac) : 'mod1min' (-25) should be <= than 'mod1max' (-40), row 47 (of Perforation)
* ERROR : Propertie (dmg-ac) : 'mod1min' (-10) should be <= than 'mod1max' (-20), row 48 (of Puncturing)[/quote]
The fact is I don't know what the game does in that case. On the pure mathematical side, -25 is greater than -40, so -40 should have been put into the MIN field, while -25 in the MAX field. Now, it depends how the game handle this case. It MAY detect that the MAX field is lower than the MIN field and therefore maybe it swap the values, making it correctly use the range [-40; -25]. I don't know, and I'm waiting for feedbacks / tests on this.

[quote=Sduibek";p="184287"]What about this one? -->

In Data\Global\Excel\Armor.txt :
* ERROR : Item 'code' (ci2) at row 117 (Tiara) has 'normcode' (ci1), expected (ci0)
* ERROR : Item 'code' (ci3) at row 118 (Diadem) has 'normcode' (ci1), expected (ci0)

I messed around with it, but can't seem to fix the error. Just wonder though if this error report simply reflects on how your program analyzes the txt files, or if in it's Vanilla form there's actually (for example) funky functionality when gambling for a circlet or something. I could i'll have to look into this one as well :p[/quote]
"Circlet" is 'code' ci0, and has norm/uber/ultra codes set to ci0, ci2 and ci3. So Circlet is the normal variation of that item (ci0 beeing in normcode), and the exceptionel & elite versions are Tiara (ci2) and Diadem (ci3).

If all goes right, normallly Tiara & Diadem rows should have ci0/ci2/ci3 in norm/uber/ultra codes as well but ... Tiara (ci2) has ci1/ci2/ci3 in norm/uber/ultra codes, so here's the explantion for the "* ERROR : Item 'code' (ci2) at row 117 (Tiara) has 'normcode' (ci1), expected (ci0)" error detected.

I don't know exactly for what norm/exceptionel/elite links are important for, but I guess that it has something to do with Vendors in nightmare/hell difficulties, Cube Recipes, and maybe Gambling too. Incorect linkage may bring problems in these areas.

Btw, thanks to have found that "HratliMagicLvl" bug for me Sduibek

[Sduibek]

Funny thing is, when I change ci1 to ci0 it gives me an error as well

Like I said though, i'll look into both and let ya know what I find. Another thing on the To Do list....

[Paul Siramy]

[quote=Sduibek";p="184443"]Funny thing is, when I change ci1 to ci0 it gives me an error as well [/quote]
But on another items.

Look : there are 4 items (Circlet, Coronet, Tiara, Diadem), that are linked together incorrectly.

Circlet < Tiara < Diadem
Coronet < Tiara < Diadem

If you remove the Circlet (ci0) item completly from Armor.txt, that should solve this problem (but introduce new ones )

[Joel]

the -25 -40 issue is handled correctly by the game.
It detects if value are < 0 and swap the value and test for absolute vlue.

All my affix file use stuff like :

hp -1 -5 and works correctly

[Paul Siramy]

Hmmm, and if only 1 value is negative ?

Min = -10, Max = 20, what becomes the range, [-10; 20] as expected ? What about Min = 20, Max = -10, and Min = -20, Max = 10 ?

[Sduibek]

[quote=Paul Siramy";p="184457"]If you remove the Circlet (ci0) item completly from Armor.txt, that should solve this problem (but introduce new ones )[/quote]
Yep.

[Joel]

[quote=Paul Siramy";p="184542"]Hmmm, and if only 1 value is negative ?

Min = -10, Max = 20, what becomes the range, [-10; 20] as expected ? What about Min = 20, Max = -10, and Min = -20, Max = 10 ?[/quote]

I don't know. For those kind of intrval, I've been a smartass.
I made two affix one with -1 -10 and one with 0 20 ^^

[Eltear]

Min = -10, Max = 20, what becomes the range, [-10; 20] as expected ? What about Min = 20, Max = -10,

Both those work as expected (no problems). I didn't bother to test the last one, since it should work the same as 'Min = 20, Max = -10'.

[Sduibek]

Seems like you fixed it for Misc.txt only to break Armor.txt and Weapons.txt... now I get:

* 104 : HraltiMagicLvl not found

On both of those files. I think it means misc.txt was a special case...



Suggestion:

When listing the files in log_extracted_txt, there could easily be some confusion. For example, right now i'm staring at the line:

* 16 : fInit not found


But both Mhyrginoc's FieldDataTypes list and the 1.10 MonuMod.txt guide say that this column is not read by the game at all. Shouldn't you only include columns in log_extracted_txt that the game actually reads? It seems kind of fruitless to say, "Uh oh, column not found" when the column in question is simply a comment field.. or not read by the game.. or both



One other thing... if I set the first 3 circlets to ci0, ci1, and ci2, then set the Diadam (circlet #4) to ci3, ci3, ci3.. then your program gives no complaints. Would this configuration cause any gameplay problems that you can foresee? Hmmm.

[Paul Siramy]

Seems like you fixed it for Misc.txt only to break Armor.txt and Weapons.txt... now I get:

* 104 : HraltiMagicLvl not found

On both of those files. I think it means misc.txt was a special case...

Hmmm, from Myhrginoc's rtf doc, I see that "HraltiMagicLvl" is used in the 3 files (weapons, armor, and misc). Further more, in D2Common.dll you can find "HraltiMagicLvl" but there's no "HratliMagicLvl". But it's true than in Armor of the 1.10 patch, you can find "HratliMagicLvl". So my best guess is that in Armor.txt and Weapons.txt, since the header names is incorrect (different than in dll) the game don't use it (someone, any experience here ?), or that these columns are accessed not by header but by column index (ignoring the name, whatever it can be)

Suggestion:

When listing the files in log_extracted_txt, there could easily be some confusion. For example, right now i'm staring at the line:

* 16 : fInit not found


But both Mhyrginoc's FieldDataTypes list and the 1.10 MonuMod.txt guide say that this column is not read by the game at all. Shouldn't you only include columns in log_extracted_txt that the game actually reads? It seems kind of fruitless to say, "Uh oh, column not found" when the column in question is simply a comment field.. or not read by the game.. or both

My bad. I generated part of my source code using a program I made, and it was parsing the columns it found in the .txt. I'll remove that column.

One other thing... if I set the first 3 circlets to ci0, ci1, and ci2, then set the Diadam (circlet #4) to ci3, ci3, ci3.. then your program gives no complaints. Would this configuration cause any gameplay problems that you can foresee? Hmmm.

Hmmm, I don't see any problems, but I'm not a specialist in this area.

[kingpin]

Hmmm, from Myhrginoc's rtf doc, I see that "HraltiMagicLvl" is used in the 3 files (weapons, armor, and misc). Further more, in D2Common.dll you can find "HraltiMagicLvl" but there's no "HratliMagicLvl". But it's true than in Armor of the 1.10 patch, you can find "HratliMagicLvl". So my best guess is that in Armor.txt and Weapons.txt, since the header names is incorrect (different than in dll) the game don't use it (someone, any experience here ?), or that these columns are accessed not by header but by column index (ignoring the name, whatever it can be)

You never read the column name, all tables is accessed index. There the next entry depends of the datatype for the column.

EDIT:
Forgott to mention that then the game compile the txt to bin it do read the column names. So, there is a chance that HratlyMagicLvl column isn't compiled. Need to check that to see if this is the case

[Sduibek]

Hmm.. I may have to take back what I said about unneeded columns. I'll do some testing today to make sure, but so far it seems that removing even totally useless columns can mess up the game. (Everything runs too fast). Wierd. o_O

[Myhrginoc]

These are from the d2common stack build for v1.10 final item files:

00099494 HraltiMin
00099488 HraltiMax
00099478 HraltiMagicMin
00099468 HraltiMagicMax
00099458 HraltiMagicLvl

If the column header is different in any of the item files (weapons, armor, misc) then that field will not load for any record in the file.

And yes, as kingpin found out, when the game compiles txt to bin it does refer to the column headers (Fog.10208). It ignores any field whose column header is not in the stack build. It doesn't like not finding any field whose column is listed in the stack build, so removing columns can be dangerous.

[Paul Siramy]

So, just for my general knowledge, this means than in Single Player, Hratli always use a MagicLevel of zero for his Weapons and Armors (typo with header names in weapons.txt & armor.txt), or something worst ?

[Myhrginoc]

That is correct. I checked armor.bin against armor.txt and the first four records are 255 in the txt but 00 in the bin. After that the field value in the txt changes, but the compiled data stays resolutely at 0.

To enable the field, just change the header from "HratliMagicLvl" to "HraltiMagicLvl" in both Weapons.txt and Armor.txt. And Paul's utility could check the field header, and then flag it or correct it, if he sets it up for that. (The other four Hratli fields are okay.)

[Volf]

Just been rearrangeing my misc.txt to keep things more organised,, added a bunsh of categories and empty lines

Analyzer gives duplicated code error for this.. My suggestion is to make it ignore empty lines or lines with only a title in first column in misc.txt.. Dont know if thats possible to do thou.

[Sduibek]

[quote=Volf";p="190745"]Just been rearrangeing my misc.txt to keep things more organised,, added a bunsh of categories and empty lines

Analyzer gives duplicated code error for this.. My suggestion is to make it ignore empty lines or lines with only a title in first column in misc.txt.. Dont know if thats possible to do thou.[/quote]Isn't the first column read by the game? if so i wouldnt put comments in the first column, just add another column for your comments

[Volf]

The comments in first column are there by default, i have just added enpty rows as markers to make the file easier to manage.

[Nemofish_5]

[quote=Volf";p="225622"]The comments in first column are there by default, i have just added enpty rows as markers to make the file easier to manage.[/quote]

just something to keep in mind when adding empty lines in misc.txt, armor.txt and weapons.txt: this WILL make the RUNEWORDS not to work. I don't know why, but i had the problem many times. Didn't find a solution tho.

[SVR]

Nemo - the cause of that has been found.

The solution is to put xxx into the code columns in the blank lines.
That way D2 doesn't add a "blank" item into it's lookup lists causing empty rune columns to require this blank gem.

[Ulmo]

There is a real need for testing mismatching brackets in calc areas. Currently you don't get a simple warning while some field are lost !
This means :

Code: Select all

Missiles :
	SrvCalc1
	CltCalc1
	{filtered}
	CHitCalc1
	DmgCalc1
	DmgSymPerCalc
	EDmgSymPerCalc

SkillDesc :
	ddam calc1
	ddam calc2
	p1dmmin
	p1dmmax
	p2dmmin
	p2dmmax
	p3dmmin
	p3dmmax
	desccalca1
	desccalcb1
	desccalca2
	desccalcb2
	desccalca3
	desccalcb3
	desccalca4
	desccalcb4
	desccalca5
	desccalcb5
	desccalca6
	desccalcb6
	dsc2calca1
	dsc2calcb1
	dsc2calca2
	dsc2calcb2
	dsc2calca3
	dsc2calcb3
	dsc2calca4
	dsc2calcb4
	dsc3calca1
	dsc3calcb1
	dsc3calca2
	dsc3calcb2
	dsc3calca3
	dsc3calcb3
	dsc3calca4
	dsc3calcb4
	dsc3calca5
	dsc3calcb5
	dsc3calca6
	dsc3calcb6
	dsc3calca7
	dsc3calcb7

Armor/Misc/Weapons :
	len
	calc1
	calc2
	calc3
	spelldesccalc

Skills :
	prgcalc1
	prgcalc2
	prgcalc3
	auralencalc
	aurarangecalc
	aurastatcalc1
	aurastatcalc2
	aurastatcalc3
	aurastatcalc4
	aurastatcalc5
	aurastatcalc6
	passivecalc1
	passivecalc2
	passivecalc3
	passivecalc4
	passivecalc5
	petmax
	sumsk1calc
	sumsk2calc
	sumsk3calc
	sumsk4calc
	sumsk5calc
	cltcalc1
	cltcalc2
	cltcalc3
	skpoints
	delay
	perdelay
	calc1
	calc2
	calc3
	calc4
	ToHitCalc
	DmgSymPerCalc
	EDmgSymPerCalc
	ELenSymPerCalc

I hope you still have time to do this.

[Paul Siramy]

Will a simple count of '(' and ')' characters in each formula field will be enough ? If yes, then it's really simple to do. It won't support the ternary "? :" operator tough.