Remove type check when iterating upstream

[adamgerhant]

I changed the upstream traversal for the tools to only iterate over the same type so that modifications to the output of a Boolean were applied to the vector data after the operation, not an upstream path node. This was a hacky solution, and we need better support for selection which path node to modify. This can be done with dots in the layer panel. More information in comment above existing_node_id

[github-actions]

Found Clippy warnings

Clippy Warnings/Errors

warning: you seem to be trying to use `match` for destructuring a single pattern. Consider using `if let`
   --> node-graph/gcore/src/raster/adjustments.rs:813:3
    |
813 | /         match self {
814 | |             Some(ref mut v) => *v = map_fn(v),
815 | |             None => (),
816 | |         }
    | |_________^ help: try: `if let Some(ref mut v) = self { *v = map_fn(v) }`
    |
    = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#single_match
    = note: `-W clippy::single-match` implied by `-W clippy::all`
    = help: to override `-W clippy::all` add `#[allow(clippy::single_match)]`

warning: `graphene-core` (lib) generated 1 warning (run `cargo clippy --fix --lib -p graphene-core` to apply 1 suggestion)
warning: large size difference between variants
   --> node-graph/graph-craft/src/document/value.rs:28:3
    |
28  | /         pub enum TaggedValue {
29  | |             None,
30  | |             $( $(#[$meta] ) *$identifier( $ty ), )*
    | |                              ------------------ the largest variant contains at least 648 bytes
31  | |             RenderOutput(RenderOutput),
    | |             -------------------------- the second-largest variant contains at least 216 bytes
...   |
34  | |             EditorApi(Arc<WasmEditorApi>)
35  | |         }
    | |_________^ the entire enum is at least 0 bytes
...
116 | / tagged_value! {
117 | |     String(String),
118 | |     U32(u32),
119 | |     U64(u64),
...   |
181 | |     FontCache(Arc<graphene_core::text::FontCache>),
182 | | }
    | |_- in this macro invocation
    |
    = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#large_enum_variant
    = note: `-W clippy::large-enum-variant` implied by `-W clippy::all`
    = help: to override `-W clippy::all` add `#[allow(clippy::large_enum_variant)]`
    = note: this warning originates in the macro `tagged_value` (in Nightly builds, run with -Z macro-backtrace for more info)
help: consider boxing the large fields to reduce the total size of the enum
    |
138 |     VectorData(Box<graphene_core::vector::VectorData>),
    |                ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

warning: `graph-craft` (lib) generated 1 warning
warning: `graphene-core` (lib) generated 1 warning (1 duplicate)
warning: `graph-craft` (lib) generated 1 warning (1 duplicate)
warning: `graph-craft` (lib test) generated 1 warning (1 duplicate)
warning: use of deprecated type alias `std::panic::PanicInfo`: use `PanicHookInfo` instead
  --> website/other/bezier-rs-demos/wasm/src/lib.rs:22:36
   |
22 |         fn panic_hook(info: &std::panic::PanicInfo<'_>) {
   |                                          ^^^^^^^^^
   |
   = note: `#[warn(deprecated)]` on by default

warning: `bezier-rs-wasm` (lib) generated 1 warning
warning: `bezier-rs-wasm` (lib test) generated 1 warning (1 duplicate)
warning: `graphene-core` (lib test) generated 1 warning (1 duplicate)
    Finished `dev` profile [optimized + debuginfo] target(s) in 4m 58s

[0HyperCube]

The code looks good.

Automatically inserting the flatten elements before the path node foregrounds an issue with the ID collisions. If you have a group with two shapes with the same point ids (e.g. two rectangles) and flatten them, the ids for the second rectangle differ each time, meaning that it is not possible to drag them around.

https://github.com/GraphiteEditor/Graphite/blob/c3a3c4c907419e2ebb9a382546a1d7b66f57cdc0/node-graph/gcore/src/vector/vector_data/attributes.rs#L762C78-L763C1

[adamgerhant]

Yes, that is why I created the generate_from_hash, which generates stable yet unique ids based on the NodeId of the merge node that the vector data flows through. It is implemented for the flatten vector elements in #2049, here where I needed the functionality to add the joining of vector data with the pen tool (this should be split into a separate PR).

It is also currently used by the repeat and copy to point nodes so that each new vector data has stable and unique ids.

This is still a work around to allowing the path tool to operate directly on vector data. A more robust method is necessary for the path node to input and output a GraphicGroup. I suppose each vector data will need a unique ID that the path node can reference, since the merge node id will change when it is copy and pasted.

[adamgerhant]

@Keavon Noted a crash on this branch, but I have not been able to reproduce it