Introduce `scanDeclareBlocks` in `declare-parser.ts` to provide a
single source of truth for parsing `role=declare` blocks. This
replaces redundant scanning logic in both the CLI and the journal
completions.
Also remove unused `dice.ts` and `spark-scanner.ts` files.
Add `rebuildReactivityFromStore` to synchronize local reactivity
state with the hydrated variable store. This ensures tag activations
are correctly tracked following a reducer replay.
Also remove the unused `invalidateCompletions` function.
Remove the mutable `streamFallback` global state and instead pass the
`VariableStore` as an explicit argument to `getCombined` and related
functions. This ensures correctness by avoiding reliance on mutable
module state during reactivity calculations.
Introduce a distinction between base values (set via `/set` or
declarations) and active modifiers (applied via tags). This ensures
tag transitions and numeric modifiers are calculated correctly during
cascades.
- Refactor `VariableView` to use a flat list with hover popups for
declarations and active modifiers.
- Update `command-dispatcher` to use `getCombined` for accurate
value lookups and `setBase` for variable updates.
- Update `var-reactivity` to manage `baseValues`, `activeMods`, and
`sourceActivations` separately.
- Export new utility functions for accessing combined values,
modifiers, and declaration expressions.
Replace manual CSV splitting with `csv-parse` to allow for
order-agnostic columns and better handling of quoted values.
The `tag` column is now optional.
Implement `computeInitialValues` to calculate the starting state of
declared variables and seed them into the journal stream store during
the completions initialization process.
Replace the old stat-parser logic with a new declare-parser to handle
variable declarations and tag modifiers. This includes moving the
declare-parser logic to a shared location in cli/completions to avoid
duplication between the CLI and the journal component.
Replaces the specialized `stat` system with a more general-purpose
variable system. This includes:
- Renaming `ReactiveStatManager` to `ReactiveVariableManager`
- Changing template syntax from `${key}` to `{{$key}}`
- Replacing `StatsView` with `VariableView` to show declared, plain,
and tag-typed variables
- Updating command `/stat` to `/set`
- Refactoring the reactivity engine to support variable declarations
and tag modifiers via `csv role=declare` blocks
Add support for `role=declare` blocks in markdown to allow for
dynamic variable declarations and tag-based modifiers.
- Implement `declare-parser` to parse CSV-formatted declaration blocks.
- Implement `var-reactivity` to manage dependency graphs, topological
sorting, and cascading updates when variables change.
- Implement `variable-expression` to evaluate arithmetic, dice rolls,
and math functions within expressions.
Introduce a shared `dispatchError` signal in the command dispatcher
to unify error reporting across different components. This allows
`CommandLinkManager` to surface errors through the `JournalInput`
UI, ensuring that errors from `:cmd[]` directive clicks are visible
to the user.
Move command parsing and dispatching logic from `CommandLinkManager`
and `JournalInput` into a new `command-dispatcher.ts` module. This
unifies how commands are handled whether they are typed manually or
triggered via `:cmd[]` directive clicks.
- Simplify `dispatchCommand` by removing role-based logic and ensuring
commands are prefixed with a slash.
- Change `:cmd[]` directive output from a `span` to an `a` tag.
- Update `.cmd-link` styles to improve link appearance and prevent
text selection.
Replace the previous mechanism of intercepting specific command-prefixed
anchor tags with a dedicated `:cmd[]` markdown directive.
- Implement `cmdDirective` for `marked-directive` support.
- Update `CommandLinkManager` to listen for clicks on `[data-cmd]` spans
instead of parsing `href` attributes.
- Add CSS styles for the new `.cmd-link` class.
Implement a component that uses event delegation to intercept clicks
on markdown links starting with command prefixes (e.g., >roll, >spark).
It dispatches these as journal stream messages based on the user's
role (Observer, Player, or GM).
Introduce ReactiveStatManager to allow dynamic stat replacement within
rendered markdown content. This component scans the article DOM for
`${key}` patterns and reactively updates them based on the journal
stream.
As part of this change, the SVG-based StatSheet system is removed in
favor of this more flexible text-based approach.
Convert div elements to anchor tags in FileTree and HeadingNode to
enable native browser features like middle-click to open in new tabs.
The implementation preserves SPA navigation for left-clicks while
ensuring search parameters are preserved in the href.
Persist name and role to localStorage and URL parameters when
received via patch to ensure synchronization. Broadcast full state
upon successful connection.
Move MQTT connection and message log management to a Shared Worker
using Comlink. This allows multiple browser tabs to share a single
connection and synchronized state.
Introduces a `remix` option that allows each data column in a spark
table to be rolled independently. This is enabled via a `remix=true`
attribute in the directive's extra attributes.
Also refactors spark table scanning by moving markdown parsing logic
from the frontend to a new CLI-side `spark-scanner.ts` utility,
improving separation of concerns.
Switch the spark table resolution logic to use the direct path to the
backing CSV file rather than attempting to parse the containing
markdown file. This simplifies the lookup process and improves
reliability.
Also refactor `SparkTableMeta` to store rows as objects keyed by
header instead of raw string arrays.
Update `createSession` to return the session ID, allowing the UI to
automatically set the active session and trigger a server hydration
immediately after creation. This ensures the UI reflects the new
session state without a manual refresh.
Change the proxy path from `/journal` to `/.ttrpg` and ensure the
MQTT broker URL points to localhost:3000 during development to
match the CLI server configuration.
Remove the webpack-based file indexing strategy in favor of a proxy
approach. The dev server now proxies content requests to a local CLI
server, simplifying the data loading logic and removing webpack-specific
types and loaders.
Relocate spark table logic from the block processor to a dedicated
directive scanner. This allows spark tables to be treated as
`:md-table` directives and ensures they are processed in a second pass
after the content index is fully populated.
- Update package name and binary from ttrpg-tools/ttrpg to tttk
- Update UI titles in App.tsx and index.html
- Add Node.js engine requirement to package.json
- Update package name and CLI binary to 'tttk'
- Revise documentation to reflect new name
- Add default serve behavior when no subcommand is given
- Add directory existence check in serve command
- Refactor inline comments to Chinese
- Update SheetView to use full-width SVG and auto-height for better
scaling
- Center template nodes in SheetView to maintain text alignment
- Enable overflow-y-auto in SheetView to support long sheets
- Persist selected sheet ID in URL search parameters in StatsView
Update the stat sheet parser to scan <tspan> elements within <text>
nodes. This allows for individual binding of template patterns to
specific tspans, preserving their position attributes.
Introduces the ability to discover and render `*.sheet.svg` files.
These files can contain `<text>` elements with `${key}` templates
that reactively update based on the current journal stats.
- Add `StatSheet` type and CLI scanning logic
- Implement `SheetView` for rendering SVGs with live bindings
- Add `parseSheet` utility to extract template patterns from SVG
- Update file indexer to include `.svg` files
Allows passing an optional label to `parseStatModifiers` to override
the default ID-based label. This enables more descriptive names in
the UI and improves how modifier labels are derived in the journal
view.
Introduces a new `stat-modifiers` CSV role that simplifies the
creation of template stats and their associated modifier stats.
A single CSV block now automatically generates:
- A template stat definition.
- Multiple modifier stat definitions (prefixed with the template ID).
- The corresponding template table.
This reduces the need for manual YAML definitions for every
modifier associated with a template.
Introduce a centralized `block-processor` and `block-scanner` to
handle markdown fenced code blocks. This replaces the previous
`inline-blocks.ts` and individual completion sources with a unified
approach that handles block stripping, directive replacement, and
metadata extraction (stats, templates, and spark tables) in a single
pass.
hypercross
hyper