Diablo 2 Eastern Sun Rises Database Generator

A set of python scripts to generate a web based database for Diablo 2 mods.

Install/Requirements

1. Install Python3.12+ (https://www.python.org/downloads/)

Usage

1. Clone this repo (git clone https://github.com/d2esrdb/d2esrdb.github.io.git)
2. Install the package (cd d2esrdb.github.io; pip install .)
3. You might need to add Python312 Scripts directory to your PATH environment variable (there should be a warning indicating so output from step 2)
4. Make a new directory inside the repo folder data, name it the same as your mod
5. Copy all your txt files from global/data in to it, as well as all your .tbl files
6. Open data/config.yaml and add a new entry for your mod
7. Run the db-gen command (db-gen)
8. All the files should get generated in the output directory.
9. If you make changes to the db-gen code itself, you'll need to redo step 2 so the changes take effect.

Known issues

1. In-game, any weapon type that is a subtype of "blunt" will get 50% increased damage to undead. This is hard-coded in Diablo 2 and I'm probably never going to update the dbgen to account for this because it's a useless stat in 99% of cases and many mods even remove this "issue" with code edits.

2. Some mods format strangely because I don't really have a good way (or understanding) of going from properties param/min/max to stat strings.

3. Sometimes what's in the .txt files and what actually happens in game is different than it should be. For example in ESR there's a unique shield called Faith that has 0-2 sockets, but in game it never rolls 0 sockets, and seems to roll 1 socket more than 2 sockets. Another example: setting up an item with 1-10 %mag but using 1 in the param field and 10 in the min field still works in game for some reason, but the dbgen just shows 1% because of the invalid data in the txt.

4. Some of the column names are different between original d2 and resurrected. "Rune Name" -> "*Rune Name" in runes.txt "hcIdx" -> "*hcIdx" in monstat.txt "Id" -> "*Id" in skills.txt Are all the ones I know of so far. Since it looks like these are all mostly comment columns, we probably shouldn't be using them for the parser.