Fluid pipe tries to insert stack with size 0
raoulvdberge opened this issue · 4 comments
Hello there,
In a Refined Storage network, try to attach a fluid pipe to an Refined Storage Interface.
Then, configure a fluid filter in the pipe.
Following crash happens:
java.lang.IllegalArgumentException: Amount must be larger than 0
at TRANSFORMER/[email protected]/com.refinedmods.refinedstorage.api.core.CoreValidations.validateLargerThanZero(CoreValidations.java:55) ~[%23186!/:?] {re:classloading}
at TRANSFORMER/[email protected]/com.refinedmods.refinedstorage.api.resource.ResourceAmount.validate(ResourceAmount.java:21) ~[%23186!/:?] {re:classloading}
at TRANSFORMER/[email protected]/com.refinedmods.refinedstorage.api.resource.ResourceAmount.<init>(ResourceAmount.java:17) ~[%23186!/:?] {re:classloading}
at TRANSFORMER/[email protected]/com.refinedmods.refinedstorage.common.support.resource.ResourceContainerImpl.insertIntoEmptySlot(ResourceContainerImpl.java:326) ~[%23186!/:?] {re:classloading}
at TRANSFORMER/[email protected]/com.refinedmods.refinedstorage.common.support.resource.ResourceContainerImpl.insert(ResourceContainerImpl.java:303) ~[%23186!/:?] {re:classloading}
at TRANSFORMER/[email protected]/com.refinedmods.refinedstorage.neoforge.support.resource.ResourceContainerFluidHandlerAdapter.fill(ResourceContainerFluidHandlerAdapter.java:55) ~[%23186!/:?] {re:classloading}
at TRANSFORMER/[email protected]/aztech.modern_industrialization.util.IOFluidHandler.fill(IOFluidHandler.java:61) ~[Modern-Industrialization-2.3.10.jar%23185!/:?] {re:classloading}
at TRANSFORMER/[email protected]/aztech.modern_industrialization.pipes.fluid.FluidNetwork$TransferOperation$1.internalTransfer(FluidNetwork.java:176) ~[Modern-Industrialization-2.3.10.jar%23185!/:?] {re:classloading}
It appears like the fluid pipe tries to insert a stack with size 0:
Modern Industrialization v2.3.10
Relevant RS issue: refinedmods/refinedstorage2#997
Seems like it's allowed by API indeed, because of lack of specification.
However, is there any use in sending an empty stack to fill?
Hey that's actually allowed by the API as far as I know. Most of the base implementations have an early return for empty stacks that catches this case at least. Only strictly negative amounts are problematic.
I've added checks in my code to check for empty stacks, should be more resilient now.
If you want, you can close this.
However, I'd rather not waste CPU cycles calling methods without any relevant data :-)
