The fold already holds the name and the read input's finished layout, so
the offset falls out there rather than at construction: `RecordLayout`
carries the resolved numbers and `set_layout` copies them into the read
slots. Constructors are untouched, and census-marker reads keep their
current installation.
A read meets the value type the name was written at, so a disagreement
between a read here and a write upstream is the same graph error as two
writes disagreeing; the one-name-one-type check now spans reads and
writes together. An absent attribute stays absent and the read serves
the forced default rather than reporting it.
`read_attribute` is the catalog's get half, typed and never `Option` at
the kernel boundary, with the name declared exactly as the write side
declares it.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
`Named<X>` in parameter position declares where a placeholder's name is
wired, and the macro gives that input constant text; `Attr<Named<X, V>>`
in the return writes under it. The kernel is handed the bare placeholder,
since the name is spent resolving the layout and a folded offset is all
the write needs, so the hot path matches a marker node's exactly.
A name-generic write names its value type through the wired generic,
which only an implementations row resolves, so those nodes emit their
layout meta per row rather than sharing one across rows.
`write_attribute` is the catalog's set half, restoring the identifier
master's documents carry with its input positions. Its name input is a
constant, so those documents resolve without migration, and the reset
marker that stood in for the missing node retires with it.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
`compute_layouts` resolves every name-from-input write against the text
its name input carries, so a folded meta is indistinguishable from a
marker node's and the layout holds a `&'static str` from there on. That
leaves nowhere for a name to be computed, which is the whole of the
rule: a name input that is not a constant is refused right here, per
node, with the path the editor pins its diagnostic to.
One name carries one value type, checked across the census and the
names a node folds together, so no graph can declare one field at two
widths. `compute_layouts` returns those refusals rather than panicking.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* Revamp the syntax to #[hard(a..b)] and #[soft(a..b)] bounds on node definitions
* Address review feedback: error on empty bounds ranges and remove redundant count clamps
* Fix NaN transform from Repeat Array with a count of 1
* Remove redundant manual clamps already enforced by #[hard(...)] bounds
* Macro support
* Preprocessor support
* Allow Cow str as scope name
* Use Passthrough node to work around compiler auto wrapping the producer node and invalidating the node id
* Stable NodeIds
* Less hacky but with bad performace
* Remove Cow from RegistryValueSource::Scope
* Connect scope injections to the inner network export
---------
Co-authored-by: Dennis Kobert <dennis@kobert.dev>
* Rename the 'Identity' node to 'Passthrough' internally
* Rename the 'Memoize' node to 'Cache' internally
* Let skip_impl proto nodes auto-generate as document node definitions
* Remove the wrapper 'Passthrough' node from document_node_definitions.rs
* Remove the wrapper 'Cache' node from document_node_definitions.rs
* Remove the wrapper 'Monitor' node from document_node_definitions.rs
* Remove the wrapper 'Noise Pattern' node from document_node_definitions.rs
* Remove the wrapper 'Brush' node from document_node_definitions.rs
* Remove the wrapper 'Transform' node from document_node_definitions.rs
* Code review improvements
* Rename Cache node back to Memoize
* More code review
* Generate the MVP node catalog in the manual (with some placeholders)
* Implement nearly the rest of everything
* Move to the tools directory and make it generate nicer default values
* Add category descriptions
* Organize file structure and improve type naming
* Improve book table of contents code
* Add collapsing chapter navigation to the book template
* Add to build workflow
* Clean up site structure
* Add #[data] and #[serialize] attributes to node macro
- Add #[data] attribute for struct fields that aren't node parameters
- Data fields are initialized with Default::default()
- Passed as references to the underlying function
- Excluded from registry metadata (internal state)
- Generic types in data fields allowed without #[implementations]
- Add #[serialize] attribute for custom Node::serialize() implementation
- Receives references to all data fields
- Generates serialize() method in Node trait impl
- Conditional derives based on data field presence
- With data fields: Debug, Clone only
- Without data fields: Debug, Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Hash
* Refactor Memo and Monitor Node to use node macro
* Move Complex type into type alias
* Fix format
* Update node-graph/nodes/gcore/src/memo.rs
Co-authored-by: Keavon Chambers <keavon@keavon.com>
* Update node-graph/nodes/gcore/src/memo.rs
Co-authored-by: Keavon Chambers <keavon@keavon.com>
---------
Co-authored-by: Keavon Chambers <keavon@keavon.com>
* Rename VectorData to Vector
* Rename other VectorData* types to Vector*
* Move assorted data types out of vector_data.rs into misc.rs
* Rename vector_data.rs to vector_types.rs and remove the vector_types module folder
* Rename other references to "vector data"
* Remove label widgets for raster/vector/group to use "-" instead
* Rename RasterData to Raster
* Rename GraphicGroup to Group
* Fix migrations and rename graphic_element.rs -> graphic.rs
* Rename TaggedValue::ArtboardGroup -> TaggedValue::Artboard
* Instances -> Table
* instances.rs -> table.rs
* Rename occurrances of the word "instances"
* .instance -> .element
* Instance* -> TableRow*
* Rename Table and TableRow methods to not say "instance"
* Remove presumed unused serde defaults now that tables default to length 0 not 1
* Rename occurences of the word "instance"
* Un-alias the RasterDataTable<Storage>, VectorDataTable, GraphicGroupTable, ArtboardGroupTable typedefs
* Move artboard type and node code out of graphic_element.rs to a new artboard.rs
* Organize the TaggedValues
* Fix tests
* Fix prior regression with Image Value node not upgrading
