This PR adds the ability to set `loadBefore` and `loadAfter` on a mod's package.json, this allows for modders to define an array of mods their current mod needs to load before or after.
Examples:
if we have <u>__MOD1__</u> that has `loadAfter` = `[ "MOD2" ]` the loading order would be:
1 - MOD2
2 - MOD1
if we have <u>__MOD2__</u> that has `loadBefore` = `[ "MOD1" ]` the loading order would also be:
1 - MOD2
2 - MOD1
Begone zzzzzz, name your mods the way you want to.
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/156
Co-authored-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
Co-committed-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
This PR makes it so typescript generates inline sourcemaps when transpiling mods to javascript.
This will make it so stacktraces originating in the mod source code point to the typescript files, lines and columns.
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/155
Co-authored-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
Co-committed-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
This PR fixes a long standing issue where stacktraces in the built executable didn't have line and column numbers, so you were left wondering where exactly in a given function an error occurred.
This also fixes source maps being generated but not actually included in the executable, this fix results in `source-map-support` actually doing it's job, so now stacktrace paths point to the typescript files, line and column number, instead of the transpiled javascript line and column number.
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/153
Co-authored-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
Co-committed-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
Don't rely on globals event value for event check, store event bools on SeasonaleventService class construction
Remove dupe function `getSeasonalEventItemsToBlock(), rely on `getAllSeasonalEventItems()`
This commit is my second go-around at refactoring the `InsuranceController`, attempting to improving the code's modularity, maintainability, and efficiency while squashing a few bugs along the way.
1. **InsuranceController.ts**
- Removed `ITemplateItem` import, as it is no longer used.
- Introduced the `adoptOrphanedItems` method to manage orphaned items in the insurance list.
- Since "normal" items are individually rolled for deletion, and can be nested within one another, there are situations where a parent item is deleted, leaving its children orphaned. This method moves those orphaned children from their missing parent into the root of the insurance container.
- Overhauled `findItemsToDelete` method to improve its efficiency and readability:
- Divided the original monolithic method into smaller, specialized methods like `populateItemsMap`, `populateParentAttachmentsMap`, `processRegularItems`, and `processAttachments`.
- Changed the return type to `Set<string>` for better performance.
- Introduced `EnrichedItem` interface (a simple extension of the `Item` interface) to add additional item data, like `name` and `maxPrice` to `Item` objects as they're processed throughout the class. This is done in place of repeatedly querying for this data, or complicating return types.
- Enhanced logging capabilities to debug the item deletion process. Due to the *current* lack of testing available I've stepped up the amount of debug logging that is done. This will hopefully help us find issues in the future.
- Modified the `rollForItemDelete` method, now renamed to `rollForDelete`, to include more detailed logging, return a boolean directly, and changed the `insuredItem` parameter to be optional.
- Added new methods for dealing with some of the particulars that arise from item adoption and creating item maps.
- Improved inline comments and documentation for better code maintainability.
2. **ItemHelper.ts**
- Added the `isRaidModdable` method to check if an item is *actually* modifiable in-raid, which takes into account not just the item, but the item that it's attached to.
- Added the `getAttachmentMainParent` method to fetch the main parent item of a given attachment, useful for item hierarchy traversal. For example, if you pass it an item ID of a suppressor, it will traverse up the muzzle brake, barrel, upper receiver, and return the gun that the suppressor is ultimately attached to, even if that gun is located within other multiple containers.
- Added the `isAttachmentAttached` method to check if an item is an attachment that is currently attached to its parent.
**Fixes:**
- Resolved an issue that caused item attachments from being property grouped together for deletion rolls. This issue prevented valuable attachments from being taken first.
- Resolved an issue that caused child items being orphaned when their parent was removed due to an insurance roll. Probable cause of the bug that made the client spaz out and send repeated insurance packages to the profile---Though I'm still unable to reproduce.
- Probably more...
Co-authored-by: Refringe <brownelltyler@gmail.com>
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/151
Co-authored-by: Refringe <refringe@noreply.dev.sp-tarkov.com>
Co-committed-by: Refringe <refringe@noreply.dev.sp-tarkov.com>
Initially this was going to be an update to dependencies but it seems i got a little carried away!
Anyways this PR removes 2 unused dependencies (`jshint` and `utf-8-validate`), and 2 other, `del` and `fs-extra` that were replaced by the built-in `fs/promises`.
It also renames all `tsconfig` and `Dockerfile` files, in a way that when viewed in a file tree sorted alphabetically they will be next to each other.
It also updates the typescript target to `ES2022`, and changes moduleResolution from `Node` to `Node10` (this isn't an update, they are the same thing but `Node` is now deprecated).
It also adds the `node:` discriminator to every import from built-in modules.
It also has major changes to the build script, `del` and `fs-extra` were only being used in the build script, it's now using `fs/promises` instead, cleaned up the code from some functions, adds better documentation to a few functions, and renames some gulp tasks and npm scripts to better represent what they actually do.
And finally it updates dependencies, except for `atomically` which can't be updated unless the project switches to ESM.
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/150
Co-authored-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
Co-committed-by: TheSparta <thesparta@noreply.dev.sp-tarkov.com>
That error caused game freeze because of server exception after played scav death or after successful extraction of played scav
Co-authored-by: ElRabbito <you@example.com>
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/149
Co-authored-by: ElRabbito <elrabbito@noreply.dev.sp-tarkov.com>
Co-committed-by: ElRabbito <elrabbito@noreply.dev.sp-tarkov.com>
Added a new range for weekly elimination, intead of having two, `1-15, 16+`, we have` 1-15, 16-40, 17+`
weekly elimination kill count requirements went way up
body part targeting is lower for low level weeklies, no change at high level
disabled specific weapon requirement for all elimination quests
level 1-15 weekly quests will target scavs more
This pull request aims to refactor the InsuranceController class to improve its code quality and maintainability. The changes include restructuring methods, adding detailed comments, and enhancing the overall logic for processing insured items.
Specific changes include:
- Updating the `findItemstoDelete()` method (and the entire class, really) to keep track of which items are tagged using a Set. This *ensures* that an item can only be attempted to be deleted once, as Sets cannot contain duplicates.
- Changing the method in which we remove insurance packages from the profile. We were iterating over them using a `for` in reverse order and then removing them with a splice and the current index. On paper this should work, but funny things were happening... Sometimes. I've created a new method that removes packages from a profile by matching against the package systemData date, time, and location. This should give us a specific insurance package (as we don't have an ID to use). In addition to this we're fetching a new instance of the profile to edit, instead of modifying packages that are being iterated over as they're being iterated over.
-----
I was unable to reproduce the issue where insurance packages were not being removed from the profile, but hopefully these simplified approaches lead to less funny business.
Co-authored-by: Refringe <brownelltyler@gmail.com>
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/148
Co-authored-by: Refringe <refringe@noreply.dev.sp-tarkov.com>
Co-committed-by: Refringe <refringe@noreply.dev.sp-tarkov.com>
Signed-off-by: bili@育碧苏联Ubisoviet
server strings are translated. player chat message WIP
Co-authored-by: Your Name <you@example.com>
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/131
Co-authored-by: ghost114514 <ghost114514@noreply.dev.sp-tarkov.com>
Co-committed-by: ghost114514 <ghost114514@noreply.dev.sp-tarkov.com>
The `BotEquipmentModPoolService.generatePool()` method is recursive, but it does not check if the pool it creates already contains data for sub items when it finds them and calls itself on them. This creates substantial bloat in the function that is undetectable to end users, but causes some additional delay when it is called.
I noticed this when using a server mod that allows for less restrictions on mod combinations (as it would hit the system recursion limit and error out) but the bloat still affects vanilla - it takes ~1 minute for all the recursive calls to end, whereas this fix reduces that to <10 seconds.
To fix this, I've moved a block of code into the conditional block that checks if the item we're adding (and its' sub items) are already in the pool before it calls itself.
This is my first pull request, so let me know if I've fudged up somewhere and I'll do my best to fix it.
I submitted a previous request about this, but I removed my name from the commits so I had to delete the branch.
Co-authored-by: gasmo <burr.cameron@gmail.com>
Reviewed-on: https://dev.sp-tarkov.com/SPT-AKI/Server/pulls/120
Co-authored-by: gasmo <gasmo@noreply.dev.sp-tarkov.com>
Co-committed-by: gasmo <gasmo@noreply.dev.sp-tarkov.com>
