Compare commits

..

No commits in common. "main" and "parser2" have entirely different histories.

99 changed files with 4151 additions and 2238 deletions

1
.gitignore vendored
View File

@ -34,7 +34,6 @@ report.[0-9]_.[0-9]_.[0-9]_.[0-9]_.json
.DS_Store
/tmp
vscode-extension/tmp
/docs
*.vsix

View File

@ -42,7 +42,7 @@ function analyzeParser(filePath: string): Map<string, CallInfo> {
methods.set(currentMethod, {
method: currentMethod,
line: i + 1,
calls: new Set(),
calls: new Set()
})
}
}
@ -85,7 +85,7 @@ function buildTree(
indent = '',
isLast = true,
depth = 0,
maxDepth = 3,
maxDepth = 3
): string[] {
const lines: string[] = []
const info = callGraph.get(method)
@ -93,7 +93,7 @@ function buildTree(
if (!info) return lines
// Add current method
const prefix = depth === 0 ? '' : isLast ? '└─> ' : '├─> '
const prefix = depth === 0 ? '' : (isLast ? '└─> ' : '├─> ')
const suffix = info.isRecursive ? ' (recursive)' : ''
const lineNum = `[line ${info.line}]`
lines.push(`${indent}${prefix}${method}() ${lineNum}${suffix}`)
@ -116,9 +116,9 @@ function buildTree(
// Get sorted unique calls (filter out recursive self-calls for display)
const calls = Array.from(info.calls)
.filter((c) => callGraph.has(c)) // Only show parser methods
.filter((c) => c !== method) // Don't show immediate self-recursion
.filter((c) => !helperPatterns.test(c)) // Filter out helpers
.filter(c => callGraph.has(c)) // Only show parser methods
.filter(c => c !== method) // Don't show immediate self-recursion
.filter(c => !helperPatterns.test(c)) // Filter out helpers
.sort()
// Add children
@ -131,7 +131,7 @@ function buildTree(
newIndent,
idx === calls.length - 1,
depth + 1,
maxDepth,
maxDepth
)
lines.push(...childLines)
})
@ -163,11 +163,11 @@ console.log(` Entry point: parse()`)
// Find methods that are never called (potential dead code or entry points)
const allCalled = new Set<string>()
for (const info of callGraph.values()) {
info.calls.forEach((c) => allCalled.add(c))
info.calls.forEach(c => allCalled.add(c))
}
const uncalled = Array.from(callGraph.keys())
.filter((m) => !allCalled.has(m) && m !== 'parse')
.filter(m => !allCalled.has(m) && m !== 'parse')
.sort()
if (uncalled.length > 0) {

View File

@ -1,18 +0,0 @@
#!/usr/bin/env shrimp
# usage: dice <sides>
import math only=random
import list only=first
import str only=[replace starts-with?]
sides = $.args | first
sides ??= 20
if sides | starts-with? d:
sides = replace sides //\D// ''
end
sides = number sides
echo 'Rolling d$sides...'
random 1 sides | echo

View File

@ -1,31 +0,0 @@
#!/usr/bin/env shrimp
year = date.now | date.year
project = fs.pwd | fs.basename | str.titlecase
{
Copyright $year $project Authors
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the “Software”), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
IN THE SOFTWARE.
}
| str.trim
| echo

View File

@ -1,9 +0,0 @@
#!/usr/bin/env shrimp
if not fs.exists? 'package.json':
echo '🦐 package.json not found'
exit 1
end
package = fs.read 'package.json' | json.decode
package.scripts | dict.keys | list.sort | each do x: echo x end

View File

@ -1,16 +1,13 @@
{
"name": "shrimp",
"version": "0.1.0",
"exports": {
".": "./src/index.ts",
"./editor": "./src/editor/index.ts",
"./editor.css": "./src/editor/editor.css"
},
"exports": "./src/index.ts",
"private": true,
"type": "module",
"scripts": {
"editor": "bun --hot src/editor/example/server.tsx",
"repl": "bun bin/repl",
"dev": "bun generate-parser && bun --hot src/server/server.tsx",
"generate-parser": "lezer-generator src/parser/shrimp.grammar --typeScript -o src/parser/shrimp.ts",
"repl": "bun generate-parser && bun bin/repl",
"update-reef": "rm -rf ~/.bun/install/cache/ && rm bun.lock && bun update reefvm",
"cli:install": "ln -s \"$(pwd)/bin/shrimp\" ~/.bun/bin/shrimp",
"cli:remove": "rm ~/.bun/bin/shrimp",

View File

@ -1,6 +1,9 @@
import { CompilerError } from '#compiler/compilerError.ts'
import { parse, setGlobals } from '#parser/parser2'
import { parse } from '#parser/parser2'
import { SyntaxNode, Tree } from '#parser/node'
import { parser } from '#parser/shrimp.ts'
import * as terms from '#parser/shrimp.terms'
import { setGlobals } from '#parser/tokenizer'
import { tokenizeCurlyString } from '#parser/curlyTokenizer'
import { assert, errorMessage } from '#utils/utils'
import { toBytecode, type Bytecode, type ProgramItem, bytecodeToString } from 'reefvm'
@ -57,12 +60,8 @@ export class Compiler {
loopLabelCount = 0
bytecode: Bytecode
pipeCounter = 0
pipeVarStack: string[] = [] // Stack of pipe variable names for nested pipes
constructor(
public input: string,
globals?: string[] | Record<string, any>,
) {
constructor(public input: string, globals?: string[] | Record<string, any>) {
try {
if (globals) setGlobals(Array.isArray(globals) ? globals : Object.keys(globals))
const ast = parse(input)
@ -92,7 +91,7 @@ export class Compiler {
}
#compileCst(cst: Tree, input: string) {
const isProgram = cst.topNode.type.is('Program')
const isProgram = cst.topNode.type.id === terms.Program
assert(isProgram, `Expected Program node, got ${cst.topNode.type.name}`)
let child = cst.topNode.firstChild
@ -108,20 +107,14 @@ export class Compiler {
const value = input.slice(node.from, node.to)
if (DEBUG) console.log(`🫦 ${node.name}: ${value}`)
switch (node.type.name) {
case 'Number':
switch (node.type.id) {
case terms.Number:
// Handle sign prefix for hex, binary, and octal literals
// Number() doesn't parse '-0xFF', '+0xFF', '-0o77', etc. correctly
let numberValue: number
if (
value.startsWith('-') &&
(value.includes('0x') || value.includes('0b') || value.includes('0o'))
) {
if (value.startsWith('-') && (value.includes('0x') || value.includes('0b') || value.includes('0o'))) {
numberValue = -Number(value.slice(1))
} else if (
value.startsWith('+') &&
(value.includes('0x') || value.includes('0b') || value.includes('0o'))
) {
} else if (value.startsWith('+') && (value.includes('0x') || value.includes('0b') || value.includes('0o'))) {
numberValue = Number(value.slice(1))
} else {
numberValue = Number(value)
@ -132,8 +125,9 @@ export class Compiler {
return [[`PUSH`, numberValue]]
case 'String': {
if (node.firstChild?.type.is('CurlyString')) return this.#compileCurlyString(value, input)
case terms.String: {
if (node.firstChild?.type.id === terms.CurlyString)
return this.#compileCurlyString(value, input)
const { parts, hasInterpolation } = getStringParts(node, input)
@ -149,19 +143,19 @@ export class Compiler {
parts.forEach((part) => {
const partValue = input.slice(part.from, part.to)
switch (part.type.name) {
case 'StringFragment':
switch (part.type.id) {
case terms.StringFragment:
// Plain text fragment - just push as-is
instructions.push(['PUSH', partValue])
break
case 'EscapeSeq':
case terms.EscapeSeq:
// Process escape sequence and push the result
const processed = processEscapeSeq(partValue)
instructions.push(['PUSH', processed])
break
case 'Interpolation':
case terms.Interpolation:
// Interpolation contains either Identifier or ParenExpr (the $ is anonymous)
const child = part.firstChild
if (!child) {
@ -175,7 +169,7 @@ export class Compiler {
throw new CompilerError(
`Unexpected string part: ${part.type.name}`,
part.from,
part.to,
part.to
)
}
})
@ -185,15 +179,15 @@ export class Compiler {
return instructions
}
case 'Boolean': {
case terms.Boolean: {
return [[`PUSH`, value === 'true']]
}
case 'Null': {
case terms.Null: {
return [[`PUSH`, null]]
}
case 'Regex': {
case terms.Regex: {
// remove the surrounding slashes and any flags
const [_, pattern, flags] = value.match(/^\/\/(.*)\/\/([gimsuy]*)$/) || []
if (!pattern) {
@ -210,15 +204,15 @@ export class Compiler {
return [['PUSH', regex]]
}
case 'Identifier': {
case terms.Identifier: {
return [[`TRY_LOAD`, value]]
}
case 'Word': {
case terms.Word: {
return [['PUSH', value]]
}
case 'DotGet': {
case terms.DotGet: {
// DotGet is parsed into a nested tree because it's hard to parse it into a flat one.
// However, we want a flat tree - so we're going to pretend like we are getting one from the parser.
//
@ -230,7 +224,7 @@ export class Compiler {
instructions.push(['TRY_LOAD', objectName])
const flattenProperty = (prop: SyntaxNode): void => {
if (prop.type.is('DotGet')) {
if (prop.type.id === terms.DotGet) {
const nestedParts = getDotGetParts(prop, input)
const nestedObjectValue = input.slice(nestedParts.object.from, nestedParts.object.to)
@ -239,7 +233,7 @@ export class Compiler {
flattenProperty(nestedParts.property)
} else {
if (prop.type.is('ParenExpr')) {
if (prop.type.id === terms.ParenExpr) {
instructions.push(...this.#compileNode(prop, input))
} else {
const propertyValue = input.slice(prop.from, prop.to)
@ -253,7 +247,7 @@ export class Compiler {
return instructions
}
case 'BinOp': {
case terms.BinOp: {
const { left, op, right } = getBinaryParts(node)
const instructions: ProgramItem[] = []
instructions.push(...this.#compileNode(left, input))
@ -301,7 +295,7 @@ export class Compiler {
return instructions
}
case 'Assign': {
case terms.Assign: {
const assignParts = getAssignmentParts(node)
const instructions: ProgramItem[] = []
@ -332,7 +326,7 @@ export class Compiler {
return instructions
}
case 'CompoundAssign': {
case terms.CompoundAssign: {
const { identifier, operator, right } = getCompoundAssignmentParts(node)
const identifierName = input.slice(identifier.from, identifier.to)
const instructions: ProgramItem[] = []
@ -383,7 +377,7 @@ export class Compiler {
throw new CompilerError(
`Unknown compound operator: ${opValue}`,
operator.from,
operator.to,
operator.to
)
}
@ -394,14 +388,14 @@ export class Compiler {
return instructions
}
case 'ParenExpr': {
case terms.ParenExpr: {
const child = node.firstChild
if (!child) return [] // I guess it is empty parentheses?
return this.#compileNode(child, input)
}
case 'FunctionDef': {
case terms.FunctionDef: {
const { paramNames, bodyNodes, catchVariable, catchBody, finallyBody } =
getFunctionDefParts(node, input)
const instructions: ProgramItem[] = []
@ -431,8 +425,8 @@ export class Compiler {
catchVariable,
catchBody,
finallyBody,
input,
),
input
)
)
} else {
instructions.push(...compileFunctionBody())
@ -447,8 +441,8 @@ export class Compiler {
return instructions
}
case 'FunctionCallOrIdentifier': {
if (node.firstChild?.type.is('DotGet')) {
case terms.FunctionCallOrIdentifier: {
if (node.firstChild?.type.id === terms.DotGet) {
const instructions: ProgramItem[] = []
const callLabel: Label = `.call_dotget_${++this.labelCount}`
const afterLabel: Label = `.after_dotget_${++this.labelCount}`
@ -490,8 +484,8 @@ export class Compiler {
PUSH 1 ; Named count
CALL
*/
case 'FunctionCall': {
case terms.FunctionCallWithNewlines:
case terms.FunctionCall: {
const { identifierNode, namedArgs, positionalArgs } = getFunctionCallParts(node, input)
const instructions: ProgramItem[] = []
instructions.push(...this.#compileNode(identifierNode, input))
@ -513,7 +507,7 @@ export class Compiler {
return instructions
}
case 'Block': {
case terms.Block: {
const children = getAllChildren(node)
const instructions: ProgramItem[] = []
@ -528,7 +522,7 @@ export class Compiler {
return instructions
}
case 'FunctionCallWithBlock': {
case terms.FunctionCallWithBlock: {
const [fn, _colon, ...block] = getAllChildren(node)
let instructions: ProgramItem[] = []
@ -541,18 +535,18 @@ export class Compiler {
...block
.filter((x) => x.type.name !== 'keyword')
.map((x) => this.#compileNode(x!, input))
.flat(),
.flat()
)
instructions.push(['RETURN'])
instructions.push([`${afterLabel}:`])
if (fn?.type.is('FunctionCallOrIdentifier')) {
if (fn?.type.id === terms.FunctionCallOrIdentifier) {
instructions.push(['LOAD', input.slice(fn!.from, fn!.to)])
instructions.push(['MAKE_FUNCTION', [], fnLabel])
instructions.push(['PUSH', 1])
instructions.push(['PUSH', 0])
instructions.push(['CALL'])
} else if (fn?.type.is('FunctionCall')) {
} else if (fn?.type.id === terms.FunctionCall) {
let body = this.#compileNode(fn!, input)
const namedArgCount = (body[body.length - 2]![1] as number) * 2
const startSlice = body.length - namedArgCount - 3
@ -568,14 +562,14 @@ export class Compiler {
instructions.push(...body)
} else {
throw new Error(
`FunctionCallWithBlock: Expected FunctionCallOrIdentifier or FunctionCall`,
`FunctionCallWithBlock: Expected FunctionCallOrIdentifier or FunctionCall`
)
}
return instructions
}
case 'TryExpr': {
case terms.TryExpr: {
const { tryBlock, catchVariable, catchBody, finallyBody } = getTryExprParts(node, input)
return this.#compileTryCatchFinally(
@ -583,34 +577,32 @@ export class Compiler {
catchVariable,
catchBody,
finallyBody,
input,
input
)
}
case 'Throw':
case 'Not': {
const keyword = node.type.is('Throw') ? 'Throw' : 'Not'
case terms.Throw: {
const children = getAllChildren(node)
const [_throwKeyword, expression] = children
if (!expression) {
throw new CompilerError(
`${keyword} expected expression, got ${children.length} children`,
`Throw expected expression, got ${children.length} children`,
node.from,
node.to,
node.to
)
}
const instructions: ProgramItem[] = []
instructions.push(...this.#compileNode(expression, input))
instructions.push([keyword.toUpperCase()]) // THROW or NOT
instructions.push(['THROW'])
return instructions
}
case 'IfExpr': {
case terms.IfExpr: {
const { conditionNode, thenBlock, elseIfBlocks, elseThenBlock } = getIfExprParts(
node,
input,
input
)
const instructions: ProgramItem[] = []
instructions.push(...this.#compileNode(conditionNode, input))
@ -650,7 +642,7 @@ export class Compiler {
}
// - `EQ`, `NEQ`, `LT`, `GT`, `LTE`, `GTE` - Pop 2, push boolean
case 'ConditionalOp': {
case terms.ConditionalOp: {
const instructions: ProgramItem[] = []
const { left, op, right } = getBinaryParts(node)
const leftInstructions: ProgramItem[] = this.#compileNode(left, input)
@ -725,7 +717,7 @@ export class Compiler {
return instructions
}
case 'PipeExpr': {
case terms.PipeExpr: {
const { pipedFunctionCall, pipeReceivers } = getPipeExprParts(node)
if (!pipedFunctionCall || pipeReceivers.length === 0) {
throw new CompilerError('PipeExpr must have at least two operands', node.from, node.to)
@ -734,77 +726,67 @@ export class Compiler {
const instructions: ProgramItem[] = []
instructions.push(...this.#compileNode(pipedFunctionCall, input))
// Use a unique variable name for this pipe level to handle nested pipes correctly
this.pipeCounter++
const pipeVarName = `_pipe_${this.pipeCounter}`
this.pipeVarStack.push(pipeVarName)
const pipeValName = `_pipe_value_${this.pipeCounter}`
pipeReceivers.forEach((pipeReceiver) => {
// Store the piped value in the current pipe's variable
// Also store as `_` for direct access in simple cases
instructions.push(['DUP'])
instructions.push(['STORE', pipeVarName])
instructions.push(['STORE', '_'])
instructions.push(['STORE', pipeValName])
const isFunctionCall =
pipeReceiver.type.is('FunctionCall') || pipeReceiver.type.is('FunctionCallOrIdentifier')
if (isFunctionCall) {
// Function call receiver: check for explicit _ usage to determine arg handling
const { identifierNode, namedArgs, positionalArgs } = getFunctionCallParts(
pipeReceiver,
input,
input
)
instructions.push(...this.#compileNode(identifierNode, input))
const isUnderscoreInPositionalArgs = positionalArgs.some((arg) =>
arg.type.is('Underscore'),
const isUnderscoreInPositionalArgs = positionalArgs.some(
(arg) => arg.type.id === terms.Underscore
)
const isUnderscoreInNamedArgs = namedArgs.some((arg) => {
const { valueNode } = getNamedArgParts(arg, input)
return valueNode.type.is('Underscore')
return valueNode.type.id === terms.Underscore
})
const shouldPushPositionalArg = !isUnderscoreInPositionalArgs && !isUnderscoreInNamedArgs
// If no underscore is explicitly used, add the piped value as the first positional arg
if (shouldPushPositionalArg) {
instructions.push(['LOAD', pipeVarName])
instructions.push(['LOAD', pipeValName])
}
positionalArgs.forEach((arg) => {
if (arg.type.id === terms.Underscore) {
instructions.push(['LOAD', pipeValName])
} else {
instructions.push(...this.#compileNode(arg, input))
}
})
namedArgs.forEach((arg) => {
const { name, valueNode } = getNamedArgParts(arg, input)
instructions.push(['PUSH', name])
if (valueNode.type.id === terms.Underscore) {
instructions.push(['LOAD', pipeValName])
} else {
instructions.push(...this.#compileNode(valueNode, input))
}
})
instructions.push(['PUSH', positionalArgs.length + (shouldPushPositionalArg ? 1 : 0)])
instructions.push(['PUSH', namedArgs.length])
instructions.push(['CALL'])
} else {
// Non-function-call receiver (Array, ParenExpr, etc.): compile directly
// The `_` variable is available for use in nested expressions
instructions.push(...this.#compileNode(pipeReceiver, input))
}
})
this.pipeVarStack.pop()
return instructions
}
case 'Array': {
case terms.Array: {
const children = getAllChildren(node)
// We can easily parse [=] as an empty dict, but `[ = ]` is tougher.
// = can be a valid word, and is also valid inside words, so for now we cheat
// and check for arrays that look like `[ = ]` to interpret them as
// empty dicts
if (children.length === 1 && children[0]!.type.is('Word')) {
if (children.length === 1 && children[0]!.type.id === terms.Word) {
const child = children[0]!
if (input.slice(child.from, child.to) === '=') {
return [['MAKE_DICT', 0]]
@ -816,7 +798,7 @@ export class Compiler {
return instructions
}
case 'Dict': {
case terms.Dict: {
const children = getAllChildren(node)
const instructions: ProgramItem[] = []
@ -835,7 +817,7 @@ export class Compiler {
return instructions
}
case 'WhileExpr': {
case terms.WhileExpr: {
const [_while, test, _colon, block] = getAllChildren(node)
const instructions: ProgramItem[] = []
@ -853,15 +835,17 @@ export class Compiler {
return instructions
}
case 'Import': {
case terms.Import: {
const instructions: ProgramItem[] = []
const [_import, ...nodes] = getAllChildren(node)
const args = nodes.filter((node) => node.type.is('Identifier'))
const namedArgs = nodes.filter((node) => node.type.is('NamedArg'))
const args = nodes.filter(node => node.type.id === terms.Identifier)
const namedArgs = nodes.filter(node => node.type.id === terms.NamedArg)
instructions.push(['LOAD', 'import'])
args.forEach((dict) => instructions.push(['PUSH', input.slice(dict.from, dict.to)]))
args.forEach((dict) =>
instructions.push(['PUSH', input.slice(dict.from, dict.to)])
)
namedArgs.forEach((arg) => {
const { name, valueNode } = getNamedArgParts(arg, input)
@ -876,22 +860,15 @@ export class Compiler {
return instructions
}
case 'Comment': {
case terms.Comment: {
return [] // ignore comments
}
case 'Underscore': {
// _ refers to the piped value for the current (innermost) pipe
// Use the stack to handle nested pipes correctly
const pipeVar = this.pipeVarStack.at(-1) ?? '_'
return [['LOAD', pipeVar]]
}
default:
throw new CompilerError(
`Compiler doesn't know how to handle a "${node.type.name}" node.`,
`Compiler doesn't know how to handle a "${node.type.name}" (${node.type.id}) node.`,
node.from,
node.to,
node.to
)
}
}
@ -901,7 +878,7 @@ export class Compiler {
catchVariable: string | undefined,
catchBody: SyntaxNode | undefined,
finallyBody: SyntaxNode | undefined,
input: string,
input: string
): ProgramItem[] {
const instructions: ProgramItem[] = []
this.tryLabelCount++

View File

@ -1,9 +1,5 @@
export class CompilerError extends Error {
constructor(
message: string,
private from: number,
private to: number,
) {
constructor(message: string, private from: number, private to: number) {
super(message)
if (from < 0 || to < 0 || to < from) {

View File

@ -112,12 +112,8 @@ describe('compiler', () => {
test('function call with no args', () => {
expect(`bloop = do: 'bleep' end; bloop`).toEvaluateTo('bleep')
expect(`bloop = [ go=do: 'bleep' end ]; bloop.go`).toEvaluateTo('bleep')
expect(`bloop = [ go=do: 'bleep' end ]; abc = do x: x end; abc (bloop.go)`).toEvaluateTo(
'bleep',
)
expect(`num = ((math.random) * 10 + 1) | math.floor; num >= 1 and num <= 10 `).toEvaluateTo(
true,
)
expect(`bloop = [ go=do: 'bleep' end ]; abc = do x: x end; abc (bloop.go)`).toEvaluateTo('bleep')
expect(`num = ((math.random) * 10 + 1) | math.floor; num >= 1 and num <= 10 `).toEvaluateTo(true)
})
test('function call with if statement and multiple expressions', () => {
@ -380,7 +376,7 @@ describe('default params', () => {
age: 60,
})
expect(
'make-person = do person=[name=Bob age=60]: person end; make-person [name=Jon age=21]',
'make-person = do person=[name=Bob age=60]: person end; make-person [name=Jon age=21]'
).toEvaluateTo({ name: 'Jon', age: 21 })
})
})
@ -412,9 +408,7 @@ describe('Nullish coalescing operator (??)', () => {
})
test('short-circuits evaluation', () => {
const throwError = () => {
throw new Error('Should not evaluate')
}
const throwError = () => { throw new Error('Should not evaluate') }
expect('5 ?? throw-error').toEvaluateTo(5, { 'throw-error': throwError })
})
@ -430,7 +424,7 @@ describe('Nullish coalescing operator (??)', () => {
// Use explicit call syntax to invoke the function
expect('(get-value) ?? (get-default)').toEvaluateTo(42, {
'get-value': getValue,
'get-default': getDefault,
'get-default': getDefault
})
})
})
@ -462,9 +456,7 @@ describe('Nullish coalescing assignment (??=)', () => {
})
test('short-circuits evaluation when not null', () => {
const throwError = () => {
throw new Error('Should not evaluate')
}
const throwError = () => { throw new Error('Should not evaluate') }
expect('x = 5; x ??= throw-error; x').toEvaluateTo(5, { 'throw-error': throwError })
})

View File

@ -10,16 +10,12 @@ describe('single line function blocks', () => {
})
test('work with named args', () => {
expect(
`attach = do signal fn: [ signal (fn) ] end; attach signal='exit': true end`,
).toEvaluateTo(['exit', true])
expect(`attach = do signal fn: [ signal (fn) ] end; attach signal='exit': true end`).toEvaluateTo(['exit', true])
})
test('work with dot-get', () => {
expect(`signals = [trap=do x y: [x (y)] end]; signals.trap 'EXIT': true end`).toEvaluateTo([
'EXIT',
true,
])
expect(`signals = [trap=do x y: [x (y)] end]; signals.trap 'EXIT': true end`).toEvaluateTo(['EXIT', true])
})
})
@ -48,6 +44,7 @@ attach signal='exit':
end`).toEvaluateTo(['exit', true])
})
test('work with dot-get', () => {
expect(`
signals = [trap=do x y: [x (y)] end]

View File

@ -219,7 +219,7 @@ describe('dict literals', () => {
describe('curly strings', () => {
test('work on one line', () => {
expect('{ one two three }').toEvaluateTo(' one two three ')
expect('{ one two three }').toEvaluateTo(" one two three ")
})
test('work on multiple lines', () => {
@ -227,7 +227,7 @@ describe('curly strings', () => {
one
two
three
}`).toEvaluateTo('\n one\n two\n three\n ')
}`).toEvaluateTo("\n one\n two\n three\n ")
})
test('can contain other curlies', () => {
@ -235,7 +235,7 @@ describe('curly strings', () => {
{ one }
two
{ three }
}`).toEvaluateTo('\n { one }\n two\n { three }\n ')
}`).toEvaluateTo("\n { one }\n two\n { three }\n ")
})
test('interpolates variables', () => {
@ -258,12 +258,12 @@ describe('curly strings', () => {
test('interpolation edge cases', () => {
expect(`{[a=1 b=2 c={wild}]}`).toEvaluateTo(`[a=1 b=2 c={wild}]`)
expect(`a = 1;b = 2;c = 3;{$a $b $c}`).toEvaluateTo(`1 2 3`)
expect(`a = 1;b = 2;c = 3;{$(a)$(b)$(c)}`).toEvaluateTo(`123`)
expect(`a = 1;b = 2;c = 3;{$a$b$c}`).toEvaluateTo(`123`)
})
})
describe('double quoted strings', () => {
test('work', () => {
test("work", () => {
expect(`"hello world"`).toEvaluateTo('hello world')
})
@ -272,11 +272,11 @@ describe('double quoted strings', () => {
expect(`"hello $(1 + 2)"`).toEvaluateTo('hello $(1 + 2)')
})
test('equal regular strings', () => {
test("equal regular strings", () => {
expect(`"hello world" == 'hello world'`).toEvaluateTo(true)
})
test('can contain newlines', () => {
test("can contain newlines", () => {
expect(`
"hello
world"`).toEvaluateTo('hello\n world')

View File

@ -40,7 +40,7 @@ describe('Native Function Exceptions', () => {
const vm = new VM(compiler.bytecode)
vm.set('async-fail', async () => {
await new Promise((resolve) => setTimeout(resolve, 1))
await new Promise(resolve => setTimeout(resolve, 1))
throw new Error('async error')
})
@ -237,7 +237,7 @@ describe('Native Function Exceptions', () => {
const result = await vm.run()
expect(result).toEqual({
type: 'string',
value: 'This is a very specific error message with details',
value: 'This is a very specific error message with details'
})
})

View File

@ -117,42 +117,4 @@ describe('pipe expressions', () => {
test('dict literals can be piped', () => {
expect(`[a=1 b=2 c=3] | dict.values | list.sort | str.join '-'`).toEvaluateTo('1-2-3')
})
test('pipe to array literal using _ in nested expressions', () => {
// _ should be accessible inside nested function calls within array literals
const code = `
double = do x: x * 2 end
triple = do x: x * 3 end
5 | [(double _) (triple _)]`
expect(code).toEvaluateTo([10, 15])
})
test('pipe to array literal using _ multiple times', () => {
expect(`10 | [_ _ _]`).toEvaluateTo([10, 10, 10])
})
test('pipe to parenthesized expression using _', () => {
const code = `
double = do x: x * 2 end
5 | (double _)`
expect(code).toEvaluateTo(10)
})
test('pipe chain with array literal receiver', () => {
// Pipe to array, then pipe that array to a function
const code = `
double = do x: x * 2 end
5 | [(double _) _] | list.sum`
expect(code).toEvaluateTo(15) // [10, 5] -> 15
})
test('_ in deeply nested expressions within pipe', () => {
// _ should work in nested function calls within function arguments
const code = `
add = do a b: a + b end
mul = do a b: a * b end
10 | add (mul _ 2) _`
// add(mul(10, 2), 10) = add(20, 10) = 30
expect(code).toEvaluateTo(30)
})
})

View File

@ -5,7 +5,7 @@ const buffer: string[] = []
const ribbitGlobals = {
ribbit: async (cb: Function) => {
await cb()
return buffer.join('\n')
return buffer.join("\n")
},
tag: async (tagFn: Function, atDefaults = {}) => {
return (atNamed = {}, ...args: any[]) => tagFn(Object.assign({}, atDefaults, atNamed), ...args)
@ -20,12 +20,10 @@ const ribbitGlobals = {
ul: (atNamed: {}, ...args: any[]) => tag('ul', atNamed, ...args),
li: (atNamed: {}, ...args: any[]) => tag('li', atNamed, ...args),
nospace: () => NOSPACE_TOKEN,
echo: (...args: any[]) => console.log(...args),
echo: (...args: any[]) => console.log(...args)
}
function raw(fn: Function) {
;(fn as any).raw = true
}
function raw(fn: Function) { (fn as any).raw = true }
const tagBlock = async (tagName: string, props = {}, fn: Function) => {
const attrs = Object.entries(props).map(([key, value]) => `${key}="${value}"`)
@ -41,13 +39,14 @@ const tagCall = (tagName: string, atNamed = {}, ...args: any[]) => {
const space = attrs.length ? ' ' : ''
const children = args
.reverse()
.map((a) => (a === TAG_TOKEN ? buffer.pop() : a))
.reverse()
.join(' ')
.map(a => a === TAG_TOKEN ? buffer.pop() : a)
.reverse().join(' ')
.replaceAll(` ${NOSPACE_TOKEN} `, '')
if (SELF_CLOSING.includes(tagName)) buffer.push(`<${tagName}${space}${attrs.join(' ')} />`)
else buffer.push(`<${tagName}${space}${attrs.join(' ')}>${children}</${tagName}>`)
if (SELF_CLOSING.includes(tagName))
buffer.push(`<${tagName}${space}${attrs.join(' ')} />`)
else
buffer.push(`<${tagName}${space}${attrs.join(' ')}>${children}</${tagName}>`)
}
const tag = async (tagName: string, atNamed = {}, ...args: any[]) => {
@ -61,25 +60,10 @@ const tag = async (tagName: string, atNamed = {}, ...args: any[]) => {
const NOSPACE_TOKEN = '!!ribbit-nospace!!'
const TAG_TOKEN = '!!ribbit-tag!!'
const SELF_CLOSING = [
'area',
'base',
'br',
'col',
'embed',
'hr',
'img',
'input',
'link',
'meta',
'param',
'source',
'track',
'wbr',
]
const SELF_CLOSING = ["area", "base", "br", "col", "embed", "hr", "img", "input", "link", "meta", "param", "source", "track", "wbr"]
describe('ribbit', () => {
beforeEach(() => (buffer.length = 0))
beforeEach(() => buffer.length = 0)
test('head tag', () => {
expect(`
@ -90,14 +74,11 @@ ribbit:
meta name=viewport content='width=device-width, initial-scale=1, viewport-fit=cover'
end
end
`).toEvaluateTo(
`<head>
`).toEvaluateTo(`<head>
<title>What up</title>
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1, viewport-fit=cover" />
</head>`,
ribbitGlobals,
)
</head>`, ribbitGlobals)
})
test('custom tags', () => {
@ -109,14 +90,11 @@ ribbit:
li two
li three
end
end`).toEvaluateTo(
`<ul class="list">
end`).toEvaluateTo(`<ul class="list">
<li border-bottom="1px solid black">one</li>
<li>two</li>
<li>three</li>
</ul>`,
ribbitGlobals,
)
</ul>`, ribbitGlobals)
})
test('inline expressions', () => {
@ -132,8 +110,6 @@ end`).toEvaluateTo(
<h1 class="bright" style="font-family: helvetica">Heya</h1>
<h2>man that is <b>wild</b>!</h2>
<p>Double the fun.</p>
</p>`,
ribbitGlobals,
)
</p>`, ribbitGlobals)
})
})

View File

@ -10,7 +10,8 @@ describe('while', () => {
a = false
b = done
end
b`).toEvaluateTo('done')
b`)
.toEvaluateTo('done')
})
test('basic expression', () => {
@ -19,7 +20,8 @@ describe('while', () => {
while a < 10:
a += 1
end
a`).toEvaluateTo(10)
a`)
.toEvaluateTo(10)
})
test('compound expression', () => {
@ -29,7 +31,8 @@ describe('while', () => {
while a > 0 and b < 100:
b += 1
end
b`).toEvaluateTo(100)
b`)
.toEvaluateTo(100)
})
test('returns value', () => {
@ -39,6 +42,7 @@ describe('while', () => {
a += 1
done
end
ret`).toEvaluateTo('done')
ret`)
.toEvaluateTo('done')
})
})

View File

@ -1,4 +1,5 @@
import { CompilerError } from '#compiler/compilerError.ts'
import * as terms from '#parser/shrimp.terms'
import type { SyntaxNode, Tree } from '#parser/node'
export const checkTreeForErrors = (tree: Tree): CompilerError[] => {
@ -23,7 +24,7 @@ export const getAllChildren = (node: SyntaxNode): SyntaxNode[] => {
child = child.nextSibling
}
return children.filter((n) => !n.type.is('Comment'))
return children.filter((n) => n.type.id !== terms.Comment)
}
export const getBinaryParts = (node: SyntaxNode) => {
@ -45,23 +46,22 @@ export const getAssignmentParts = (node: SyntaxNode) => {
throw new CompilerError(
`Assign expected 3 children, got ${children.length}`,
node.from,
node.to,
node.to
)
}
// array destructuring
if (left && left.type.is('Array')) {
const identifiers = getAllChildren(left).filter((child) => child.type.is('Identifier'))
if (left && left.type.id === terms.Array) {
const identifiers = getAllChildren(left).filter((child) => child.type.id === terms.Identifier)
return { arrayPattern: identifiers, right }
}
if (!left || !left.type.is('AssignableIdentifier')) {
if (!left || left.type.id !== terms.AssignableIdentifier) {
throw new CompilerError(
`Assign left child must be an AssignableIdentifier or Array, got ${
left ? left.type.name : 'none'
`Assign left child must be an AssignableIdentifier or Array, got ${left ? left.type.name : 'none'
}`,
node.from,
node.to,
node.to
)
}
@ -72,19 +72,18 @@ export const getCompoundAssignmentParts = (node: SyntaxNode) => {
const children = getAllChildren(node)
const [left, operator, right] = children
if (!left || !left.type.is('AssignableIdentifier')) {
if (!left || left.type.id !== terms.AssignableIdentifier) {
throw new CompilerError(
`CompoundAssign left child must be an AssignableIdentifier, got ${
left ? left.type.name : 'none'
`CompoundAssign left child must be an AssignableIdentifier, got ${left ? left.type.name : 'none'
}`,
node.from,
node.to,
node.to
)
} else if (!operator || !right) {
throw new CompilerError(
`CompoundAssign expected 3 children, got ${children.length}`,
node.from,
node.to,
node.to
)
}
@ -99,16 +98,16 @@ export const getFunctionDefParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`FunctionDef expected at least 4 children, got ${children.length}`,
node.from,
node.to,
node.to
)
}
const paramNames = getAllChildren(paramsNode).map((param) => {
if (!param.type.is('Identifier') && !param.type.is('NamedParam')) {
if (param.type.id !== terms.Identifier && param.type.id !== terms.NamedParam) {
throw new CompilerError(
`FunctionDef params must be Identifier or NamedParam, got ${param.type.name}`,
param.from,
param.to,
param.to
)
}
return input.slice(param.from, param.to)
@ -123,7 +122,7 @@ export const getFunctionDefParts = (node: SyntaxNode, input: string) => {
let finallyBody: SyntaxNode | undefined
for (const child of rest) {
if (child.type.is('CatchExpr')) {
if (child.type.id === terms.CatchExpr) {
catchExpr = child
const catchChildren = getAllChildren(child)
const [_catchKeyword, identifierNode, _colon, body] = catchChildren
@ -131,12 +130,12 @@ export const getFunctionDefParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`CatchExpr expected identifier and body, got ${catchChildren.length} children`,
child.from,
child.to,
child.to
)
}
catchVariable = input.slice(identifierNode.from, identifierNode.to)
catchBody = body
} else if (child.type.is('FinallyExpr')) {
} else if (child.type.id === terms.FinallyExpr) {
finallyExpr = child
const finallyChildren = getAllChildren(child)
const [_finallyKeyword, _colon, body] = finallyChildren
@ -144,7 +143,7 @@ export const getFunctionDefParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`FinallyExpr expected body, got ${finallyChildren.length} children`,
child.from,
child.to,
child.to
)
}
finallyBody = body
@ -165,9 +164,9 @@ export const getFunctionCallParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(`FunctionCall expected at least 1 child, got 0`, node.from, node.to)
}
const namedArgs = args.filter((arg) => arg.type.is('NamedArg'))
const namedArgs = args.filter((arg) => arg.type.id === terms.NamedArg)
const positionalArgs = args
.filter((arg) => arg.type.is('PositionalArg'))
.filter((arg) => arg.type.id === terms.PositionalArg)
.map((arg) => {
const child = arg.firstChild
if (!child) throw new CompilerError(`PositionalArg has no child`, arg.from, arg.to)
@ -199,7 +198,7 @@ export const getIfExprParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`IfExpr expected at least 4 children, got ${children.length}`,
node.from,
node.to,
node.to
)
}
@ -208,13 +207,13 @@ export const getIfExprParts = (node: SyntaxNode, input: string) => {
rest.forEach((child) => {
const parts = getAllChildren(child)
if (child.type.is('ElseExpr')) {
if (child.type.id === terms.ElseExpr) {
if (parts.length !== 3) {
const message = `ElseExpr expected 1 child, got ${parts.length}`
throw new CompilerError(message, child.from, child.to)
}
elseThenBlock = parts.at(-1)
} else if (child.type.is('ElseIfExpr')) {
} else if (child.type.id === terms.ElseIfExpr) {
const [_else, _if, conditional, _colon, thenBlock] = parts
if (!conditional || !thenBlock) {
const names = parts.map((p) => p.type.name).join(', ')
@ -249,32 +248,35 @@ export const getStringParts = (node: SyntaxNode, input: string) => {
// The text is just between the quotes
const parts = children.filter((child) => {
return (
child.type.is('StringFragment') ||
child.type.is('Interpolation') ||
child.type.is('EscapeSeq') ||
child.type.is('CurlyString')
child.type.id === terms.StringFragment ||
child.type.id === terms.Interpolation ||
child.type.id === terms.EscapeSeq ||
child.type.id === terms.CurlyString
)
})
// Validate each part is the expected type
parts.forEach((part) => {
if (
part.type.is('StringFragment') &&
part.type.is('Interpolation') &&
part.type.is('EscapeSeq') &&
part.type.is('CurlyString')
part.type.id !== terms.StringFragment &&
part.type.id !== terms.Interpolation &&
part.type.id !== terms.EscapeSeq &&
part.type.id !== terms.CurlyString
) {
throw new CompilerError(
`String child must be StringFragment, Interpolation, or EscapeSeq, got ${part.type.name}`,
part.from,
part.to,
part.to
)
}
})
// hasInterpolation means the string has interpolation ($var) or escape sequences (\n)
// A simple string like 'hello' has one StringFragment but no interpolation
const hasInterpolation = parts.some((p) => p.type.is('Interpolation') || p.type.is('EscapeSeq'))
const hasInterpolation = parts.some(
(p) => p.type.id === terms.Interpolation || p.type.id === terms.EscapeSeq
)
return { parts, hasInterpolation }
}
@ -286,23 +288,23 @@ export const getDotGetParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`DotGet expected 2 identifier children, got ${children.length}`,
node.from,
node.to,
node.to
)
}
if (!object.type.is('IdentifierBeforeDot')) {
if (object.type.id !== terms.IdentifierBeforeDot && object.type.id !== terms.Dollar) {
throw new CompilerError(
`DotGet object must be an IdentifierBeforeDot, got ${object.type.name}`,
object.from,
object.to,
object.to
)
}
if (!['Identifier', 'Number', 'ParenExpr', 'DotGet'].includes(property.type.name)) {
if (![terms.Identifier, terms.Number, terms.ParenExpr, terms.DotGet].includes(property.type.id)) {
throw new CompilerError(
`DotGet property must be an Identifier, Number, ParenExpr, or DotGet, got ${property.type.name}`,
property.from,
property.to,
property.to
)
}
@ -321,7 +323,7 @@ export const getTryExprParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`TryExpr expected at least 3 children, got ${children.length}`,
node.from,
node.to,
node.to
)
}
@ -332,7 +334,7 @@ export const getTryExprParts = (node: SyntaxNode, input: string) => {
let finallyBody: SyntaxNode | undefined
rest.forEach((child) => {
if (child.type.is('CatchExpr')) {
if (child.type.id === terms.CatchExpr) {
catchExpr = child
const catchChildren = getAllChildren(child)
const [_catchKeyword, identifierNode, _colon, body] = catchChildren
@ -340,12 +342,12 @@ export const getTryExprParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`CatchExpr expected identifier and body, got ${catchChildren.length} children`,
child.from,
child.to,
child.to
)
}
catchVariable = input.slice(identifierNode.from, identifierNode.to)
catchBody = body
} else if (child.type.is('FinallyExpr')) {
} else if (child.type.id === terms.FinallyExpr) {
finallyExpr = child
const finallyChildren = getAllChildren(child)
const [_finallyKeyword, _colon, body] = finallyChildren
@ -353,7 +355,7 @@ export const getTryExprParts = (node: SyntaxNode, input: string) => {
throw new CompilerError(
`FinallyExpr expected body, got ${finallyChildren.length} children`,
child.from,
child.to,
child.to
)
}
finallyBody = body

261
src/editor/commands.ts Normal file
View File

@ -0,0 +1,261 @@
export type CommandShape = {
command: string
description?: string
args: ArgShape[]
execute: string | ((...args: any[]) => any)
}
type ArgShape<T extends keyof ArgTypeMap = keyof ArgTypeMap> = {
name: string
type: T
description?: string
optional?: boolean
default?: ArgTypeMap[T]
}
type ArgTypeMap = {
string: string
number: number
boolean: boolean
}
const commandShapes: CommandShape[] = [
{
command: 'ls',
description: 'List the contents of a directory',
execute: './commands/ls.ts',
args: [
{ name: 'path', type: 'string', description: 'The path to list' },
{ name: 'all', type: 'boolean', description: 'Show hidden files', default: false },
{ name: 'long', type: 'boolean', description: 'List in long format', default: false },
{
name: 'short-names',
type: 'boolean',
description: 'Only print file names',
default: false,
},
{ name: 'full-paths', type: 'boolean', description: 'Display full paths', default: false },
],
},
{
command: 'cd',
description: 'Change the current working directory',
execute: './commands/cd.ts',
args: [{ name: 'path', type: 'string', description: 'The path to change to' }],
},
{
command: 'cp',
description: 'Copy files or directories',
execute: './commands/cp.ts',
args: [
{ name: 'source', type: 'string', description: 'Source file or directory' },
{ name: 'destination', type: 'string', description: 'Destination path' },
{ name: 'recursive', type: 'boolean', description: 'Copy recursively', default: false },
{ name: 'verbose', type: 'boolean', description: 'Verbose output', default: false },
],
},
{
command: 'mv',
description: 'Move files or directories',
execute: './commands/mv.ts',
args: [
{ name: 'source', type: 'string', description: 'Source file or directory' },
{ name: 'destination', type: 'string', description: 'Destination path' },
{ name: 'verbose', type: 'boolean', description: 'Verbose output', default: false },
],
},
{
command: 'rm',
description: 'Remove files or directories',
execute: './commands/rm.ts',
args: [
{ name: 'path', type: 'string', description: 'Path to remove' },
{ name: 'recursive', type: 'boolean', description: 'Remove recursively', default: false },
{ name: 'force', type: 'boolean', description: 'Force removal', default: false },
{ name: 'verbose', type: 'boolean', description: 'Verbose output', default: false },
],
},
{
command: 'mkdir',
description: 'Create directories',
execute: './commands/mkdir.ts',
args: [
{ name: 'path', type: 'string', description: 'Directory path to create' },
{ name: 'verbose', type: 'boolean', description: 'Verbose output', default: false },
],
},
{
command: 'touch',
description: 'Create empty files or update timestamps',
execute: './commands/touch.ts',
args: [
{ name: 'path', type: 'string', description: 'File path to touch' },
{ name: 'access', type: 'boolean', description: 'Update access time only', default: false },
{
name: 'modified',
type: 'boolean',
description: 'Update modified time only',
default: false,
},
],
},
{
command: 'echo',
description: 'Display a string',
execute: './commands/echo.ts',
args: [
{ name: 'text', type: 'string', description: 'Text to display' },
{ name: 'no-newline', type: 'boolean', description: "Don't append newline", default: false },
],
},
{
command: 'cat',
description: 'Display file contents',
execute: './commands/cat.ts',
args: [
{ name: 'path', type: 'string', description: 'File to display' },
{ name: 'numbered', type: 'boolean', description: 'Show line numbers', default: false },
],
},
{
command: 'head',
description: 'Show first lines of input',
execute: './commands/head.ts',
args: [
{ name: 'path', type: 'string', description: 'File to read from' },
{ name: 'lines', type: 'number', description: 'Number of lines', default: 10 },
],
},
{
command: 'tail',
description: 'Show last lines of input',
execute: './commands/tail.ts',
args: [
{ name: 'path', type: 'string', description: 'File to read from' },
{ name: 'lines', type: 'number', description: 'Number of lines', default: 10 },
{ name: 'follow', type: 'boolean', description: 'Follow file changes', default: false },
],
},
{
command: 'grep',
description: 'Search for patterns in text',
execute: './commands/grep.ts',
args: [
{ name: 'pattern', type: 'string', description: 'Pattern to search for' },
{
name: 'ignore-case',
type: 'boolean',
description: 'Case insensitive search',
default: false,
},
{ name: 'invert-match', type: 'boolean', description: 'Invert match', default: false },
{ name: 'line-number', type: 'boolean', description: 'Show line numbers', default: false },
],
},
{
command: 'sort',
description: 'Sort input',
execute: './commands/sort.ts',
args: [
{ name: 'reverse', type: 'boolean', description: 'Sort in reverse order', default: false },
{
name: 'ignore-case',
type: 'boolean',
description: 'Case insensitive sort',
default: false,
},
{ name: 'numeric', type: 'boolean', description: 'Numeric sort', default: false },
],
},
{
command: 'uniq',
description: 'Filter out repeated lines',
execute: './commands/uniq.ts',
args: [
{ name: 'count', type: 'boolean', description: 'Show count of occurrences', default: false },
{
name: 'repeated',
type: 'boolean',
description: 'Show only repeated lines',
default: false,
},
{ name: 'unique', type: 'boolean', description: 'Show only unique lines', default: false },
],
},
{
command: 'select',
description: 'Select specific columns from data',
execute: './commands/select.ts',
args: [{ name: 'columns', type: 'string', description: 'Columns to select' }],
},
{
command: 'where',
description: 'Filter data based on conditions',
execute: './commands/where.ts',
args: [{ name: 'condition', type: 'string', description: 'Filter condition' }],
},
{
command: 'group-by',
description: 'Group data by column values',
execute: './commands/group-by.ts',
args: [{ name: 'column', type: 'string', description: 'Column to group by' }],
},
{
command: 'ps',
description: 'List running processes',
execute: './commands/ps.ts',
args: [
{ name: 'long', type: 'boolean', description: 'Show detailed information', default: false },
],
},
{
command: 'sys',
description: 'Show system information',
execute: './commands/sys.ts',
args: [],
},
{
command: 'which',
description: 'Find the location of a command',
execute: './commands/which.ts',
args: [
{ name: 'command', type: 'string', description: 'Command to locate' },
{ name: 'all', type: 'boolean', description: 'Show all matches', default: false },
],
},
] as const
let commandSource = () => commandShapes
export const setCommandSource = (fn: () => CommandShape[]) => {
commandSource = fn
}
export const resetCommandSource = () => {
commandSource = () => commandShapes
}
export const matchingCommands = (prefix: string) => {
const match = commandSource().find((cmd) => cmd.command === prefix)
const partialMatches = commandSource().filter((cmd) => cmd.command.startsWith(prefix))
return { match, partialMatches }
}

View File

@ -1,87 +0,0 @@
import { autocompletion, type CompletionContext, type Completion } from '@codemirror/autocomplete'
import { Shrimp } from '#/index'
const keywords = [
'import',
'end',
'do',
'if',
'else',
'while',
'try',
'catch',
'finally',
'throw',
'not',
'and',
'or',
'true',
'false',
'null',
]
const keywordCompletions: Completion[] = keywords.map((k) => ({
label: k,
type: 'keyword',
boost: -1,
}))
export const createShrimpCompletions = (shrimp: Shrimp) => {
// Build completions from all names in the shrimp scope
const scopeNames = shrimp.vm.scope.vars()
console.log(`🌭`, shrimp.vm.vars())
const functionCompletions: Completion[] = scopeNames.map((name) => {
const value = shrimp.vm.scope.get(name)
let type
if (value?.type === 'function' || value?.type === 'native') {
type = 'function'
} else if (value?.type === 'dict') {
type = 'namespace'
} else {
type = 'variable'
}
return { type, label: name }
})
const allCompletions = [...keywordCompletions, ...functionCompletions]
// Get methods for a module (e.g., math, str, list)
const getModuleMethods = (moduleName: string): Completion[] => {
const module = shrimp.get(moduleName)
if (!module || typeof module !== 'object') return []
return Object.keys(module).map((m) => ({ label: m, type: 'function' }))
}
const shrimpCompletionSource = (context: CompletionContext) => {
// Check for module.method pattern (e.g., "math.")
const dotMatch = context.matchBefore(/[\w\-]+\.[\w\-\?]*/)
if (dotMatch) {
const [moduleName, methodPrefix] = dotMatch.text.split('.')
const methods = getModuleMethods(moduleName!)
if (methods.length > 0) {
const dotPos = dotMatch.from + moduleName!.length + 1
return {
from: dotPos,
options: methods.filter((m) => m.label.startsWith(methodPrefix ?? '')),
}
}
}
// Regular completions
const word = context.matchBefore(/[\w\-\?\$]+/)
if (!word || (word.from === word.to && !context.explicit)) return null
return {
from: word.from,
options: allCompletions.filter((c) => c.label.startsWith(word.text)),
}
}
return autocompletion({
override: [shrimpCompletionSource],
})
}

View File

@ -1,24 +0,0 @@
import { linter, type Diagnostic } from '@codemirror/lint'
import { Shrimp } from '#/index'
import { CompilerError } from '#compiler/compilerError'
export const createShrimpDiagnostics = (shrimp: Shrimp) =>
linter((view) => {
const code = view.state.doc.toString()
const diagnostics: Diagnostic[] = []
try {
shrimp.parse(code)
} catch (err) {
if (err instanceof CompilerError) {
diagnostics.push({
from: err.from,
to: err.to,
severity: 'error',
message: err.message,
})
}
}
return diagnostics
})

View File

@ -1,78 +1,53 @@
:root {
/* Background colors */
--bg-editor: #011627;
--bg-output: #40318D;
--bg-status-bar: #1E2A4A;
--bg-status-border: #0E1A3A;
--bg-selection: #1D3B53;
/* Text colors */
--text-editor: #D6DEEB;
--text-output: #7C70DA;
--text-status: #B3A9FF55;
--caret: #80A4C2;
/* Syntax highlighting colors */
--color-keyword: #C792EA;
--color-function: #82AAFF;
--color-string: #C3E88D;
--color-number: #F78C6C;
--color-bool: #FF5370;
--color-operator: #89DDFF;
--color-paren: #676E95;
--color-function-call: #FF9CAC;
--color-variable-def: #FFCB6B;
--color-error: #FF6E6E;
--color-regex: #E1ACFF;
/* ANSI terminal colors */
--ansi-black: #011627;
--ansi-red: #FF5370;
--ansi-green: #C3E88D;
--ansi-yellow: #FFCB6B;
--ansi-blue: #82AAFF;
--ansi-magenta: #C792EA;
--ansi-cyan: #89DDFF;
--ansi-white: #D6DEEB;
/* ANSI bright colors (slightly more vibrant) */
--ansi-bright-black: #676E95;
--ansi-bright-red: #FF6E90;
--ansi-bright-green: #D4F6A8;
--ansi-bright-yellow: #FFE082;
--ansi-bright-blue: #A8C7FA;
--ansi-bright-magenta: #E1ACFF;
--ansi-bright-cyan: #A8F5FF;
--ansi-bright-white: #FFFFFF;
}
@font-face {
font-family: 'C64ProMono';
src: url('../../assets/C64_Pro_Mono-STYLE.woff2') format('woff2');
font-weight: normal;
font-style: normal;
}
@font-face {
font-family: 'Pixeloid Mono';
src: url('../../assets/PixeloidMono.ttf') format('truetype');
font-weight: normal;
font-style: normal;
}
* {
margin: 0;
padding: 0;
box-sizing: border-box;
}
body {
background-color: var(--bg-editor);
#output {
flex: 1;
background: var(--bg-output);
color: var(--text-output);
padding: 20px;
overflow-y: auto;
white-space: pre-wrap;
font-family: 'Pixeloid Mono', 'Courier New', monospace;
font-size: 18px;
height: 100vh;
}
#root {
height: 100vh;
#status-bar {
height: 30px;
background: var(--bg-status-bar);
color: var(--text-status);
display: flex;
align-items: center;
padding: 0 10px;
font-size: 14px;
border-top: 3px solid var(--bg-status-border);
border-bottom: 3px solid var(--bg-status-border);
display: flex;
justify-content: space-between;
}
#status-bar .left,
#status-bar .right {
display: flex;
justify-content: center;
gap: 2rem;
}
#status-bar .multiline {
display: flex;
.dot {
padding-top: 1px;
margin-right: 4px;
}
.active {
color: var(--color-string);
}
.inactive {
color: inherit;
}
}
.syntax-error {
text-decoration: underline dotted var(--color-error);
}

View File

@ -1,70 +1,140 @@
import { EditorView, basicSetup } from 'codemirror'
import { Shrimp } from '#/index'
import { EditorView } from '@codemirror/view'
import { asciiEscapeToHtml, assertNever, log, toElement } from '#utils/utils'
import { Signal } from '#utils/signal'
import { getContent } from '#editor/plugins/persistence'
import type { HtmlEscapedString } from 'hono/utils/html'
import { connectToNose, noseSignals } from '#editor/noseClient'
import type { Value } from 'reefvm'
import { Compartment } from '@codemirror/state'
import { lineNumbers } from '@codemirror/view'
import { shrimpSetup } from '#editor/plugins/shrimpSetup'
import { shrimpTheme } from './theme'
import { createShrimpDiagnostics } from './diagnostics'
import { shrimpHighlighter } from './highlighter'
import { createShrimpCompletions } from './completions'
import { shrimpKeymap } from './keymap'
import { getContent, persistence } from './persistence'
import '#editor/editor.css'
type EditorProps = {
initialCode?: string
onChange?: (code: string) => void
extensions?: import('@codemirror/state').Extension[]
shrimp?: Shrimp
}
const lineNumbersCompartment = new Compartment()
export const Editor = ({
initialCode = '',
onChange,
extensions: customExtensions = [],
shrimp = new Shrimp(),
}: EditorProps) => {
connectToNose()
export const outputSignal = new Signal<Value | string>()
export const errorSignal = new Signal<string>()
export const multilineModeSignal = new Signal<boolean>()
export const Editor = () => {
return (
<>
<div
ref={(el: Element) => {
if (!el?.querySelector('.cm-editor'))
createEditorView(el, getContent() ?? initialCode, onChange, customExtensions, shrimp)
ref={(ref: Element) => {
if (ref?.querySelector('.cm-editor')) return
const view = new EditorView({
parent: ref,
doc: getContent(),
extensions: shrimpSetup(lineNumbersCompartment),
})
multilineModeSignal.connect((isMultiline) => {
view.dispatch({
effects: lineNumbersCompartment.reconfigure(isMultiline ? lineNumbers() : []),
})
})
requestAnimationFrame(() => view.focus())
}}
/>
<div id="status-bar">
<div className="left"></div>
<div className="right"></div>
</div>
<div id="output"></div>
<div id="error"></div>
</>
)
}
const createEditorView = (
el: Element,
initialCode: string,
onChange: ((code: string) => void) | undefined,
customExtensions: import('@codemirror/state').Extension[],
shrimp: Shrimp
) => {
const extensions = [
basicSetup,
shrimpTheme,
createShrimpDiagnostics(shrimp),
createShrimpCompletions(shrimp),
shrimpHighlighter,
shrimpKeymap,
persistence,
...customExtensions,
]
if (onChange) {
extensions.push(
EditorView.updateListener.of((update) => {
if (update.docChanged) {
onChange(update.state.doc.toString())
noseSignals.connect((message) => {
if (message.type === 'error') {
log.error(`Nose error: ${message.data}`)
errorSignal.emit(`Nose error: ${message.data}`)
} else if (message.type === 'reef-output') {
const x = outputSignal.emit(message.data)
} else if (message.type === 'connected') {
outputSignal.emit(`╞ Connected to Nose VM`)
}
})
outputSignal.connect((value) => {
const el = document.querySelector('#output')!
el.innerHTML = ''
el.innerHTML = asciiEscapeToHtml(valueToString(value))
})
errorSignal.connect((error) => {
const el = document.querySelector('#output')!
el.innerHTML = ''
el.classList.add('error')
el.innerHTML = asciiEscapeToHtml(error)
})
type StatusBarMessage = {
side: 'left' | 'right'
message: string | Promise<HtmlEscapedString>
className: string
order?: number
}
export const statusBarSignal = new Signal<StatusBarMessage>()
statusBarSignal.connect(async ({ side, message, className, order }) => {
document.querySelector(`#status-bar .${className}`)?.remove()
const sideEl = document.querySelector(`#status-bar .${side}`)!
const messageEl = (
<div data-order={order ?? 0} className={className}>
{await message}
</div>
)
}
new EditorView({
parent: el,
doc: initialCode,
extensions,
// Now go through the nodes and put it in the right spot based on order. Higher number means further right
const nodes = Array.from(sideEl.childNodes)
const index = nodes.findIndex((node) => {
if (!(node instanceof HTMLElement)) return false
return Number(node.dataset.order) > (order ?? 0)
})
// Trigger onChange with initial content
onChange?.(initialCode)
if (index === -1) {
sideEl.appendChild(toElement(messageEl))
} else {
sideEl.insertBefore(toElement(messageEl), nodes[index]!)
}
})
const valueToString = (value: Value | string): string => {
if (typeof value === 'string') {
return value
}
switch (value.type) {
case 'null':
return 'null'
case 'boolean':
return value.value ? 'true' : 'false'
case 'number':
return value.value.toString()
case 'string':
return value.value
case 'array':
return `${value.value.map(valueToString).join('\n')}`
case 'dict': {
const entries = Array.from(value.value.entries()).map(
([key, val]) => `"${key}": ${valueToString(val)}`
)
return `{${entries.join(', ')}}`
}
case 'regex':
return `/${value.value.source}/`
case 'function':
return `<function>`
case 'native':
return `<function ${value.fn.name}>`
default:
assertNever(value)
return `<unknown value type: ${(value as any).type}>`
}
}

View File

@ -1,6 +0,0 @@
import { Editor } from '#editor/editor'
import { render } from 'hono/jsx/dom'
if (typeof document !== 'undefined') {
render(<Editor initialCode={'# type some code'} />, document.getElementById('root')!)
}

View File

@ -1,11 +0,0 @@
<!DOCTYPE html>
<html>
<head>
<meta charset="UTF-8" />
<link rel="stylesheet" href="../editor.css" />
</head>
<body>
<div id="root"></div>
<script type="module" src="./frontend.tsx"></script>
</body>
</html>

View File

@ -1,14 +0,0 @@
import index from './index.html'
const server = Bun.serve({
port: 3000,
routes: {
'/': index,
},
development: {
hmr: true,
console: true,
},
})
console.log(`Editor running at ${server.url}`)

View File

@ -1,73 +0,0 @@
import { EditorView, Decoration, DecorationSet, ViewPlugin, ViewUpdate } from '@codemirror/view'
import { RangeSetBuilder } from '@codemirror/state'
import { Shrimp } from '#/index'
import { type SyntaxNode } from '#parser/node'
import { log } from '#utils/utils'
const shrimp = new Shrimp()
export const shrimpHighlighter = ViewPlugin.fromClass(
class {
decorations: DecorationSet
constructor(view: EditorView) {
this.decorations = this.highlight(view)
}
update(update: ViewUpdate) {
if (!update.docChanged) return
this.decorations = this.highlight(update.view)
}
highlight(view: EditorView): DecorationSet {
const builder = new RangeSetBuilder<Decoration>()
const code = view.state.doc.toString()
try {
const tree = shrimp.parse(code)
const decorations: { from: number; to: number; class: string }[] = []
tree.iterate({
enter: (node) => {
const cls = tokenStyles[node.type.name]
const isLeaf = node.children.length === 0
if (cls && isLeaf) {
decorations.push({ from: node.from, to: node.to, class: cls })
}
},
})
// Sort by position (required by RangeSetBuilder)
decorations.sort((a, b) => a.from - b.from)
for (const d of decorations) {
builder.add(d.from, d.to, Decoration.mark({ class: d.class }))
}
} catch (error) {
log('Parsing error in highlighter', error)
}
return builder.finish()
}
},
{
decorations: (v) => v.decorations,
}
)
// Map node types to CSS classes
const tokenStyles: Record<string, string> = {
keyword: 'tok-keyword',
String: 'tok-string',
StringFragment: 'tok-string',
CurlyString: 'tok-string',
Number: 'tok-number',
Boolean: 'tok-bool',
Null: 'tok-null',
Identifier: 'tok-identifier',
AssignableIdentifier: 'tok-variable-def',
Comment: 'tok-comment',
operator: 'tok-operator',
Regex: 'tok-regex',
}

View File

@ -1 +0,0 @@
export { Editor } from './editor'

View File

@ -1,8 +0,0 @@
import { keymap } from '@codemirror/view'
import { acceptCompletion } from '@codemirror/autocomplete'
import { indentWithTab } from '@codemirror/commands'
export const shrimpKeymap = keymap.of([
{ key: 'Tab', run: acceptCompletion },
indentWithTab,
])

59
src/editor/noseClient.ts Normal file
View File

@ -0,0 +1,59 @@
import { Signal } from '#utils/signal'
import type { Bytecode, Value } from 'reefvm'
let ws: WebSocket
type IncomingMessage =
| { type: 'connected' }
| { type: 'ping'; data: number }
| { type: 'commands'; data: number }
| {
type: 'apps'
data: {
name: string
type: 'browser' | 'server'
}[]
}
| {
type: 'session:start'
data: {
NOSE_DIR: string
cwd: string
hostname: string
mode: string
project: string
}
}
| { type: 'reef-output'; data: Value }
| { type: 'error'; data: string }
export const noseSignals = new Signal<IncomingMessage>()
export const connectToNose = (url: string = 'ws://localhost:3000/ws') => {
ws = new WebSocket(url)
ws.onopen = () => noseSignals.emit({ type: 'connected' })
ws.onmessage = (event) => {
const message = JSON.parse(event.data)
noseSignals.emit(message)
}
ws.onerror = (event) => {
console.error(`💥WebSocket error:`, event)
}
ws.onclose = () => {
console.log(`🚪 Connection closed`)
}
}
let id = 0
export const sendToNose = (code: Bytecode) => {
if (!ws) {
throw new Error('WebSocket is not connected.')
} else if (ws.readyState !== WebSocket.OPEN) {
throw new Error(`WebSocket is not open, current status is ${ws.readyState}.`)
}
id += 1
ws.send(JSON.stringify({ type: 'reef-bytecode', data: code, id }))
}

View File

@ -0,0 +1,9 @@
import { errorSignal } from '#editor/editor'
import { EditorView } from '@codemirror/view'
export const catchErrors = EditorView.exceptionSink.of((exception) => {
console.error('CodeMirror error:', exception)
errorSignal.emit(
`Editor error: ${exception instanceof Error ? exception.message : String(exception)}`
)
})

View File

@ -0,0 +1,35 @@
import { EditorView, ViewPlugin, ViewUpdate } from '@codemirror/view'
import { syntaxTree } from '@codemirror/language'
import { statusBarSignal } from '#editor/editor'
export const debugTags = ViewPlugin.fromClass(
class {
update(update: ViewUpdate) {
if (update.docChanged || update.selectionSet || update.geometryChanged) {
this.updateStatusBar(update.view)
}
}
updateStatusBar(view: EditorView) {
const pos = view.state.selection.main.head + 1
const tree = syntaxTree(view.state)
let tags: string[] = []
let node = tree.resolveInner(pos, -1)
while (node) {
tags.push(node.type.name)
node = node.parent!
if (!node) break
}
const debugText = tags.length ? tags.reverse().slice(1).join(' > ') : 'No nodes'
statusBarSignal.emit({
side: 'right',
message: debugText,
className: 'debug-tags',
order: -1,
})
}
}
)

View File

@ -0,0 +1,62 @@
import { parser } from '#parser/shrimp'
import type { Timeout } from '#utils/utils'
import { Range } from '@codemirror/state'
import {
Decoration,
EditorView,
ViewPlugin,
ViewUpdate,
type DecorationSet,
} from '@codemirror/view'
export const shrimpErrors = ViewPlugin.fromClass(
class {
timeout?: Timeout
decorations: DecorationSet = Decoration.none
constructor(view: EditorView) {
this.updateErrors(view)
}
update(update: ViewUpdate) {
if (update.docChanged) {
this.debounceUpdate(update.view)
}
}
updateErrors(view: EditorView) {
this.decorations = Decoration.none
try {
const decorations: Range<Decoration>[] = []
const tree = parser.parse(view.state.doc.toString())
tree.iterate({
enter: (node) => {
if (!node.type.isError) return
// Skip empty error nodes
if (node.from === node.to) return
const decoration = Decoration.mark({
class: 'syntax-error',
attributes: { title: 'COREY REPLACE THIS' },
}).range(node.from, node.to)
decorations.push(decoration)
},
})
this.decorations = Decoration.set(decorations)
// requestAnimationFrame(() => view.dispatch({}))
} catch (e) {
console.error('🙈 Error parsing document', e)
}
}
debounceUpdate = (view: EditorView) => {
clearTimeout(this.timeout)
this.timeout = setTimeout(() => this.updateErrors(view), 250)
}
},
{
decorations: (v) => v.decorations,
}
)

View File

@ -0,0 +1,232 @@
import {
ViewPlugin,
ViewUpdate,
EditorView,
Decoration,
type DecorationSet,
} from '@codemirror/view'
import { syntaxTree } from '@codemirror/language'
import { type SyntaxNode } from '@lezer/common'
import { WidgetType } from '@codemirror/view'
import { toElement } from '#utils/utils'
import { matchingCommands } from '#editor/commands'
import * as Terms from '#parser/shrimp.terms'
const ghostTextTheme = EditorView.theme({
'.ghost-text': {
color: '#666',
opacity: '0.6',
fontStyle: 'italic',
},
})
type Hint = { cursor: number; hintText?: string; completionText?: string }
export const inlineHints = [
ViewPlugin.fromClass(
class {
decorations: DecorationSet = Decoration.none
currentHint?: Hint
update(update: ViewUpdate) {
if (!update.docChanged && !update.selectionSet) return
this.clearHints()
let hint = this.getContext(update.view)
this.currentHint = hint
this.showHint(hint)
}
handleTab(view: EditorView) {
if (!this.currentHint?.completionText) return false
this.decorations = Decoration.none
view.dispatch({
changes: {
from: this.currentHint.cursor,
insert: this.currentHint.completionText,
},
selection: {
anchor: this.currentHint.cursor + this.currentHint.completionText.length,
},
})
this.currentHint = undefined
return true
}
clearHints() {
this.currentHint = undefined
this.decorations = Decoration.none
}
getContext(view: EditorView): Hint {
const cursor = view.state.selection.main.head
const isCursorAtEnd = cursor === view.state.doc.length
if (!isCursorAtEnd) return { cursor }
const token = this.getCommandContextToken(view, cursor)
if (!token) return { cursor }
const text = view.state.doc.sliceString(token.from, token.to)
const tokenId = token.type.id
let completionText = ''
let hintText = ''
const justSpaces = view.state.doc.sliceString(cursor - 1, cursor) === ' '
if (tokenId === Terms.CommandPartial) {
const { partialMatches } = matchingCommands(text)
const match = partialMatches[0]
if (match) {
completionText = match.command.slice(text.length) + ' '
hintText = completionText
}
} else if (
tokenId === Terms.Identifier &&
token.parent?.type.id === Terms.Arg &&
!justSpaces
) {
const { availableArgs } = this.getCommandContext(view, token)
const matchingArgs = availableArgs.filter((arg) => arg.name.startsWith(text))
const match = matchingArgs[0]
if (match) {
hintText = `${match.name.slice(text.length)}=<${match.type}>`
completionText = `${match.name.slice(text.length)}=`
}
} else if (this.containedBy(token, Terms.PartialNamedArg)) {
const { availableArgs } = this.getCommandContext(view, token)
const textWithoutEquals = text.slice(0, -1)
const matchingArgs = availableArgs.filter((arg) => arg.name == textWithoutEquals)
const match = matchingArgs[0]
if (match) {
hintText = `<${match.type}>`
completionText = 'default' in match ? `${match.default}` : ''
}
} else {
const { availableArgs } = this.getCommandContext(view, token)
const nextArg = Array.from(availableArgs)[0]
const space = justSpaces ? '' : ' '
if (nextArg) {
hintText = `${space}${nextArg.name}=<${nextArg.type}>`
if (nextArg) {
completionText = `${space}${nextArg.name}=`
}
}
}
return { completionText, hintText, cursor }
}
getCommandContextToken(view: EditorView, cursor: number) {
const tree = syntaxTree(view.state)
let node = tree.resolveInner(cursor, -1)
// If we're in a CommandCall, return the token before cursor
if (this.containedBy(node, Terms.CommandCall)) {
return tree.resolveInner(cursor, -1)
}
// If we're in Program, look backward
while (node.name === 'Program' && cursor > 0) {
cursor -= 1
node = tree.resolveInner(cursor, -1)
if (this.containedBy(node, Terms.CommandCall)) {
return tree.resolveInner(cursor, -1)
}
}
}
containedBy(node: SyntaxNode, nodeId: number): SyntaxNode | undefined {
let current: SyntaxNode | undefined = node
while (current) {
if (current.type.id === nodeId) {
return current
}
current = current.parent ?? undefined
}
}
showHint(hint: Hint) {
if (!hint.hintText) return
const widget = new GhostTextWidget(hint.hintText)
const afterCursor = 1
const decoration = Decoration.widget({ widget, side: afterCursor }).range(hint.cursor)
this.decorations = Decoration.set([decoration])
}
getCommandContext(view: EditorView, currentToken: SyntaxNode) {
let commandCallNode = currentToken.parent
while (commandCallNode?.type.name !== 'CommandCall') {
if (!commandCallNode) {
throw new Error('No CommandCall parent found, must be an error in the grammar')
}
commandCallNode = commandCallNode.parent
}
const commandToken = commandCallNode.firstChild
if (!commandToken) {
throw new Error('CommandCall has no children, must be an error in the grammar')
}
const commandText = view.state.doc.sliceString(commandToken.from, commandToken.to)
const { match: commandShape } = matchingCommands(commandText)
if (!commandShape) {
throw new Error(`No command shape found for command "${commandText}"`)
}
let availableArgs = [...commandShape.args]
// Walk through all NamedArg children
let child = commandToken.nextSibling
while (child) {
console.log('child', child.type.name, child.to - child.from)
if (child.type.id === Terms.NamedArg) {
const argName = child.firstChild // Should be the Identifier
if (argName) {
const argText = view.state.doc.sliceString(argName.from, argName.to - 1)
availableArgs = availableArgs.filter((arg) => arg.name !== argText)
}
} else if (child.type.id == Terms.Arg) {
const hasSpaceAfter = view.state.doc.sliceString(child.to, child.to + 1) === ' '
if (hasSpaceAfter) {
availableArgs.shift()
}
}
child = child.nextSibling
}
return { commandShape, availableArgs }
}
},
{
decorations: (v) => v.decorations,
eventHandlers: {
keydown(event, view) {
if (event.key === 'Tab') {
event.preventDefault()
const plugin = view.plugin(inlineHints[0]! as ViewPlugin<any>)
plugin?.handleTab(view)
}
},
},
}
),
ghostTextTheme,
]
class GhostTextWidget extends WidgetType {
constructor(private text: string) {
super()
}
toDOM() {
const el = <span className="ghost-text">{this.text}</span>
return toElement(el)
}
}

View File

@ -0,0 +1,184 @@
import { multilineModeSignal, outputSignal } from '#editor/editor'
import { printBytecodeOutput, printParserOutput, runCode } from '#editor/runCode'
import { EditorState } from '@codemirror/state'
import { keymap } from '@codemirror/view'
let multilineMode = false
const customKeymap = keymap.of([
{
key: 'Enter',
run: (view) => {
if (multilineMode) return false
const input = view.state.doc.toString()
history.push(input)
runCode(input)
view.dispatch({
changes: { from: 0, to: view.state.doc.length, insert: '' },
selection: { anchor: 0 },
})
return true
},
},
{
key: 'Shift-Enter',
run: (view) => {
if (multilineMode) {
const input = view.state.doc.toString()
runCode(input)
return true
} else {
outputSignal.emit('Press Shift+Enter to insert run the code.')
}
multilineModeSignal.emit(true)
multilineMode = true
view.dispatch({
changes: { from: view.state.doc.length, insert: '\n' },
selection: { anchor: view.state.doc.length + 1 },
})
return true
},
},
{
key: 'Tab',
preventDefault: true,
run: (view) => {
view.dispatch({
changes: { from: view.state.selection.main.from, insert: ' ' },
selection: { anchor: view.state.selection.main.from + 2 },
})
return true
},
},
{
key: 'ArrowUp',
run: (view) => {
if (multilineMode) return false
const command = history.previous()
if (command === undefined) return false
view.dispatch({
changes: { from: 0, to: view.state.doc.length, insert: command },
selection: { anchor: command.length },
})
return true
},
},
{
key: 'ArrowDown',
run: (view) => {
if (multilineMode) return false
const command = history.next()
if (command === undefined) return false
view.dispatch({
changes: { from: 0, to: view.state.doc.length, insert: command },
selection: { anchor: command.length },
})
return true
},
},
{
key: 'Mod-k 1',
preventDefault: true,
run: (view) => {
const input = view.state.doc.toString()
printParserOutput(input)
return true
},
},
{
key: 'Mod-k 2',
preventDefault: true,
run: (view) => {
const input = view.state.doc.toString()
printBytecodeOutput(input)
return true
},
},
])
let firstTime = true
const singleLineFilter = EditorState.transactionFilter.of((transaction) => {
if (multilineMode) return transaction // Allow everything in multiline mode
if (firstTime) {
firstTime = false
if (transaction.newDoc.toString().includes('\n')) {
multilineMode = true
return transaction
}
}
transaction.changes.iterChanges((fromA, toA, fromB, toB, inserted) => {
if (inserted.toString().includes('\n')) {
multilineMode = true
return
}
})
return transaction
})
export const shrimpKeymap = [customKeymap, singleLineFilter]
class History {
private commands: string[] = []
private index: number | undefined
private storageKey = 'shrimp-command-history'
constructor() {
try {
this.commands = JSON.parse(localStorage.getItem(this.storageKey) || '[]')
} catch {
console.warn('Failed to load command history from localStorage')
}
}
push(command: string) {
this.commands.push(command)
// Limit to last 50 commands
this.commands = this.commands.slice(-50)
localStorage.setItem(this.storageKey, JSON.stringify(this.commands))
this.index = undefined
}
previous(): string | undefined {
if (this.commands.length === 0) return
if (this.index === undefined) {
this.index = this.commands.length - 1
} else if (this.index > 0) {
this.index -= 1
}
return this.commands[this.index]
}
next(): string | undefined {
if (this.commands.length === 0 || this.index === undefined) return
if (this.index < this.commands.length - 1) {
this.index += 1
return this.commands[this.index]
} else {
this.index = undefined
return ''
}
}
}
const history = new History()

View File

@ -1,6 +1,6 @@
import { ViewPlugin, ViewUpdate } from '@codemirror/view'
export const persistence = ViewPlugin.fromClass(
export const persistencePlugin = ViewPlugin.fromClass(
class {
saveTimeout?: ReturnType<typeof setTimeout>
@ -24,6 +24,6 @@ export const getContent = () => {
return localStorage.getItem('shrimp-editor-content') || ''
}
export const setContent = (data: string) => {
const setContent = (data: string) => {
localStorage.setItem('shrimp-editor-content', data)
}

View File

@ -0,0 +1,9 @@
import { parser } from '#/parser/shrimp'
import { LRLanguage, LanguageSupport } from '@codemirror/language'
import { highlighting } from '#/parser/highlight.js'
const language = LRLanguage.define({
parser: parser.configure({ props: [highlighting] }),
})
export const shrimpLanguage = new LanguageSupport(language)

View File

@ -0,0 +1,35 @@
import { history, defaultKeymap, historyKeymap } from '@codemirror/commands'
import { bracketMatching, indentOnInput } from '@codemirror/language'
import { highlightSpecialChars, drawSelection, dropCursor, keymap } from '@codemirror/view'
import { closeBrackets, autocompletion, completionKeymap } from '@codemirror/autocomplete'
import { EditorState, Compartment } from '@codemirror/state'
import { searchKeymap } from '@codemirror/search'
import { shrimpKeymap } from './keymap'
import { shrimpTheme, shrimpHighlighting } from './theme'
import { shrimpLanguage } from './shrimpLanguage'
import { shrimpErrors } from './errors'
import { persistencePlugin } from './persistence'
import { catchErrors } from './catchErrors'
export const shrimpSetup = (lineNumbersCompartment: Compartment) => {
return [
catchErrors,
shrimpKeymap,
highlightSpecialChars(),
history(),
drawSelection(),
dropCursor(),
EditorState.allowMultipleSelections.of(true),
bracketMatching(),
closeBrackets(),
autocompletion(),
indentOnInput(),
keymap.of([...defaultKeymap, ...historyKeymap, ...searchKeymap, ...completionKeymap]),
lineNumbersCompartment.of([]),
shrimpTheme,
shrimpLanguage,
shrimpHighlighting,
shrimpErrors,
persistencePlugin,
]
}

View File

@ -0,0 +1,60 @@
import { EditorView } from '@codemirror/view'
import { HighlightStyle, syntaxHighlighting } from '@codemirror/language'
import { tags } from '@lezer/highlight'
const highlightStyle = HighlightStyle.define([
{ tag: tags.keyword, color: 'var(--color-keyword)' },
{ tag: tags.name, color: 'var(--color-function)' },
{ tag: tags.string, color: 'var(--color-string)' },
{ tag: tags.number, color: 'var(--color-number)' },
{ tag: tags.bool, color: 'var(--color-bool)' },
{ tag: tags.operator, color: 'var(--color-operator)' },
{ tag: tags.paren, color: 'var(--color-paren)' },
{ tag: tags.regexp, color: 'var(--color-regex)' },
{ tag: tags.function(tags.variableName), color: 'var(--color-function-call)' },
{ tag: tags.function(tags.invalid), color: 'white' },
{
tag: tags.definition(tags.variableName),
color: 'var(--color-variable-def)',
backgroundColor: 'var(--bg-variable-def)',
padding: '1px 2px',
borderRadius: '2px',
fontWeight: '500',
},
])
export const shrimpHighlighting = syntaxHighlighting(highlightStyle)
export const shrimpTheme = EditorView.theme(
{
'&': {
color: 'var(--text-editor)',
backgroundColor: 'var(--bg-editor)',
height: '100%',
fontSize: '18px',
},
'.cm-content': {
fontFamily: '"Pixeloid Mono", "Courier New", monospace',
caretColor: 'var(--caret)',
padding: '0px',
},
'&.cm-focused .cm-cursor': {
borderLeftColor: 'var(--caret)',
},
'&.cm-focused .cm-selectionBackground, ::selection': {
backgroundColor: 'var(--bg-selection)',
},
'.cm-editor': {
border: 'none',
outline: 'none',
height: '100%',
},
'.cm-matchingBracket': {
backgroundColor: 'var(--color-bool)',
},
'.cm-nonmatchingBracket': {
backgroundColor: 'var(--color-string)',
},
},
{ dark: true }
)

38
src/editor/runCode.tsx Normal file
View File

@ -0,0 +1,38 @@
import { outputSignal, errorSignal } from '#editor/editor'
import { Compiler } from '#compiler/compiler'
import { errorMessage, log } from '#utils/utils'
import { bytecodeToString } from 'reefvm'
import { parser } from '#parser/shrimp'
import { sendToNose } from '#editor/noseClient'
import { treeToString } from '#utils/tree'
export const runCode = async (input: string) => {
try {
const compiler = new Compiler(input)
sendToNose(compiler.bytecode)
} catch (error) {
log.error(error)
errorSignal.emit(`${errorMessage(error)}`)
}
}
export const printParserOutput = (input: string) => {
try {
const cst = parser.parse(input)
const string = treeToString(cst, input)
outputSignal.emit(string)
} catch (error) {
log.error(error)
errorSignal.emit(`${errorMessage(error)}`)
}
}
export const printBytecodeOutput = (input: string) => {
try {
const compiler = new Compiler(input)
outputSignal.emit(bytecodeToString(compiler.bytecode))
} catch (error) {
log.error(error)
errorSignal.emit(`${errorMessage(error)}`)
}
}

View File

@ -1,51 +0,0 @@
import { EditorView } from '@codemirror/view'
export const shrimpTheme = EditorView.theme(
{
'&': {
color: 'var(--text-editor)',
backgroundColor: 'var(--bg-editor)',
height: '100%',
fontSize: '18px',
},
'.cm-content': {
fontFamily: '"Pixeloid Mono", "Courier New", monospace',
caretColor: 'var(--caret)',
padding: '0px',
borderTop: '1px solid var(--bg-editor)',
},
'&.cm-focused .cm-cursor': {
borderLeftColor: 'var(--caret)',
},
'.cm-selectionLayer .cm-selectionBackground': {
backgroundColor: 'var(--bg-selection)',
},
'.cm-editor': {
border: 'none',
outline: 'none',
height: '100%',
},
'.cm-matchingBracket': {
backgroundColor: 'var(--color-bool)',
},
'.cm-nonmatchingBracket': {
backgroundColor: 'var(--color-string)',
},
'.cm-activeLine': {
backgroundColor: 'rgba(255, 255, 255, 0.03)',
},
// Token highlighting
'.tok-keyword': { color: 'var(--color-keyword)' },
'.tok-string': { color: 'var(--color-string)' },
'.tok-number': { color: 'var(--color-number)' },
'.tok-bool': { color: 'var(--color-bool)' },
'.tok-null': { color: 'var(--color-number)', fontStyle: 'italic' },
'.tok-identifier': { color: 'var(--color-function)' },
'.tok-variable-def': { color: 'var(--color-variable-def)' },
'.tok-comment': { color: 'var(--ansi-bright-black)', fontStyle: 'italic' },
'.tok-operator': { color: 'var(--color-operator)' },
'.tok-regex': { color: 'var(--color-regex)' },
},
{ dark: true }
)

View File

@ -3,7 +3,7 @@ import { VM, fromValue, toValue, isValue, type Bytecode } from 'reefvm'
import { Compiler } from '#compiler/compiler'
import { parse } from '#parser/parser2'
import { Tree } from '#parser/node'
import { globals as parserGlobals, setGlobals as setParserGlobals } from '#parser/parser2'
import { globals as parserGlobals, setGlobals as setParserGlobals } from '#parser/tokenizer'
import { globals as prelude } from '#prelude'
export { Compiler } from '#compiler/compiler'
@ -53,10 +53,7 @@ export class Shrimp {
let bytecode
if (typeof code === 'string') {
const compiler = new Compiler(
code,
Object.keys(Object.assign({}, prelude, this.globals ?? {}, locals ?? {})),
)
const compiler = new Compiler(code, Object.keys(Object.assign({}, prelude, this.globals ?? {}, locals ?? {})))
bytecode = compiler.bytecode
} else {
bytecode = code
@ -69,6 +66,7 @@ export class Shrimp {
return this.vm.stack.length ? fromValue(this.vm.stack.at(-1)!, this.vm) : null
}
}
export async function runFile(path: string, globals?: Record<string, any>): Promise<any> {

View File

@ -1,6 +1,7 @@
import { parser } from '#parser/shrimp.ts'
import { parse } from '#parser/parser2'
import type { SyntaxNode } from '#parser/node'
import { isIdentStart, isIdentChar } from './tokenizer2'
import { isIdentStart, isIdentChar } from './tokenizer'
// Turns a { curly string } into strings and nodes for interpolation
export const tokenizeCurlyString = (value: string): (string | [string, SyntaxNode])[] => {
@ -38,13 +39,14 @@ export const tokenizeCurlyString = (value: string): (string | [string, SyntaxNod
const input = value.slice(start + 2, pos) // skip '$('
tokens.push([input, parse(input)])
start = pos + 1 // start after ')'
start = ++pos // skip ')'
} else {
char = value[++pos]
if (!char) break
if (!isIdentStart(char.charCodeAt(0))) break
while (char && isIdentChar(char.charCodeAt(0))) char = value[++pos]
while (char && isIdentChar(char.charCodeAt(0)))
char = value[++pos]
const input = value.slice(start + 1, pos) // skip '$'
tokens.push([input, parse(input)])

View File

@ -1,17 +1,21 @@
import { type Token, TokenType } from './tokenizer2'
import * as term from './shrimp.terms'
export type NodeType =
| 'Program'
| 'Block'
| 'FunctionCall'
| 'FunctionCallOrIdentifier'
| 'FunctionCallWithBlock'
| 'PositionalArg'
| 'NamedArg'
| 'NamedArgPrefix'
| 'FunctionDef'
| 'Params'
| 'NamedParam'
| 'Null'
| 'Boolean'
| 'Number'
@ -29,6 +33,7 @@ export type NodeType =
| 'Array'
| 'Dict'
| 'Comment'
| 'BinOp'
| 'ConditionalOp'
| 'ParenExpr'
@ -36,6 +41,7 @@ export type NodeType =
| 'CompoundAssign'
| 'DotGet'
| 'PipeExpr'
| 'IfExpr'
| 'ElseIfExpr'
| 'ElseExpr'
@ -44,12 +50,13 @@ export type NodeType =
| 'CatchExpr'
| 'FinallyExpr'
| 'Throw'
| 'Not'
| 'Eq'
| 'Modulo'
| 'Plus'
| 'Star'
| 'Slash'
| 'Import'
| 'Do'
| 'Underscore'
@ -60,13 +67,13 @@ export type NodeType =
// TODO: remove this when we switch from lezer
export const operators: Record<string, any> = {
// Logic
and: 'And',
or: 'Or',
'and': 'And',
'or': 'Or',
// Bitwise
band: 'Band',
bor: 'Bor',
bxor: 'Bxor',
'band': 'Band',
'bor': 'Bor',
'bxor': 'Bxor',
'>>>': 'Ushr',
'>>': 'Shr',
'<<': 'Shl',
@ -132,6 +139,180 @@ export class Tree {
}
}
// TODO: TEMPORARY SHIM
class SyntaxNodeType {
constructor(public nodeType: NodeType) { }
is(other: string) {
return this.nodeType === other
}
get id(): number {
switch (this.nodeType) {
case 'Program':
return term.Program
case 'Block':
return term.Block
case 'FunctionCall':
return term.FunctionCall
case 'FunctionCallOrIdentifier':
return term.FunctionCallOrIdentifier
case 'FunctionCallWithBlock':
return term.FunctionCallWithBlock
case 'PositionalArg':
return term.PositionalArg
case 'NamedArg':
return term.NamedArg
case 'FunctionDef':
return term.FunctionDef
case 'Params':
return term.Params
case 'NamedParam':
return term.NamedParam
case 'Null':
return term.Null
case 'Boolean':
return term.Boolean
case 'Number':
return term.Number
case 'String':
return term.String
case 'StringFragment':
return term.StringFragment
case 'CurlyString':
return term.CurlyString
case 'DoubleQuote':
return term.DoubleQuote
case 'EscapeSeq':
return term.EscapeSeq
case 'Interpolation':
return term.Interpolation
case 'Regex':
return term.Regex
case 'Identifier':
return term.Identifier
case 'AssignableIdentifier':
return term.AssignableIdentifier
case 'IdentifierBeforeDot':
return term.IdentifierBeforeDot
case 'Word':
return term.Word
case 'Array':
return term.Array
case 'Dict':
return term.Dict
case 'Comment':
return term.Comment
case 'BinOp':
return term.BinOp
case 'ConditionalOp':
return term.ConditionalOp
case 'ParenExpr':
return term.ParenExpr
case 'Assign':
return term.Assign
case 'CompoundAssign':
return term.CompoundAssign
case 'DotGet':
return term.DotGet
case 'PipeExpr':
return term.PipeExpr
case 'IfExpr':
return term.IfExpr
case 'ElseIfExpr':
return term.ElseIfExpr
case 'ElseExpr':
return term.ElseExpr
case 'WhileExpr':
return term.WhileExpr
case 'TryExpr':
return term.TryExpr
case 'CatchExpr':
return term.CatchExpr
case 'FinallyExpr':
return term.FinallyExpr
case 'Throw':
return term.Throw
case 'Eq':
return term.Eq
case 'Modulo':
return term.Modulo
case 'Plus':
return term.Plus
case 'Star':
return term.Star
case 'Slash':
return term.Slash
case 'Import':
return term.Import
case 'Do':
return term.Do
case 'Underscore':
return term.Underscore
case 'colon':
return term.colon
case 'keyword':
return term.keyword
}
return 0
}
get name(): string {
return this.nodeType
}
}
export class SyntaxNode {
#type: NodeType
#isError = false
@ -151,18 +332,8 @@ export class SyntaxNode {
return new SyntaxNode(TokenType[token.type] as NodeType, token.from, token.to, parent ?? null)
}
get type(): {
type: NodeType
name: NodeType
isError: boolean
is: (other: NodeType) => boolean
} {
return {
type: this.#type,
name: this.#type,
isError: this.#isError,
is: (other: NodeType) => other === this.#type,
}
get type(): SyntaxNodeType {
return new SyntaxNodeType(this.#type)
}
set type(name: NodeType) {
@ -209,7 +380,7 @@ export class SyntaxNode {
}
push(...nodes: SyntaxNode[]): SyntaxNode {
nodes.forEach((child) => (child.parent = this))
nodes.forEach(child => child.parent = this)
this.children.push(...nodes)
return this
}
@ -222,8 +393,8 @@ export class SyntaxNode {
// Operator precedence (binding power) - higher = tighter binding
export const precedence: Record<string, number> = {
// Logical
or: 10,
and: 20,
'or': 10,
'and': 20,
// Comparison
'==': 30,
@ -246,9 +417,9 @@ export const precedence: Record<string, number> = {
'-': 40,
// Bitwise AND/OR/XOR (higher precedence than addition)
band: 45,
bor: 45,
bxor: 45,
'band': 45,
'bor': 45,
'bxor': 45,
// Multiplication/Division/Modulo
'*': 50,
@ -259,6 +430,10 @@ export const precedence: Record<string, number> = {
'**': 60,
}
export const conditionals = new Set(['==', '!=', '<', '>', '<=', '>=', '??', 'and', 'or'])
export const conditionals = new Set([
'==', '!=', '<', '>', '<=', '>=', '??', 'and', 'or'
])
export const compounds = ['??=', '+=', '-=', '*=', '/=', '%=']
export const compounds = [
'??=', '+=', '-=', '*=', '/=', '%='
]

View File

@ -0,0 +1,99 @@
import { ExternalTokenizer, InputStream } from '@lezer/lr'
import * as terms from './shrimp.terms'
type Operator = { str: string; tokenName: keyof typeof terms }
const operators: Array<Operator> = [
{ str: 'and', tokenName: 'And' },
{ str: 'or', tokenName: 'Or' },
{ str: 'band', tokenName: 'Band' },
{ str: 'bor', tokenName: 'Bor' },
{ str: 'bxor', tokenName: 'Bxor' },
{ str: '>>>', tokenName: 'Ushr' }, // Must come before >>
{ str: '>>', tokenName: 'Shr' },
{ str: '<<', tokenName: 'Shl' },
{ str: '>=', tokenName: 'Gte' },
{ str: '<=', tokenName: 'Lte' },
{ str: '!=', tokenName: 'Neq' },
{ str: '==', tokenName: 'EqEq' },
// Compound assignment operators (must come before single-char operators)
{ str: '??=', tokenName: 'NullishEq' },
{ str: '+=', tokenName: 'PlusEq' },
{ str: '-=', tokenName: 'MinusEq' },
{ str: '*=', tokenName: 'StarEq' },
{ str: '/=', tokenName: 'SlashEq' },
{ str: '%=', tokenName: 'ModuloEq' },
// Nullish coalescing (must come before it could be mistaken for other tokens)
{ str: '??', tokenName: 'NullishCoalesce' },
// Single-char operators
{ str: '*', tokenName: 'Star' },
{ str: '=', tokenName: 'Eq' },
{ str: '/', tokenName: 'Slash' },
{ str: '+', tokenName: 'Plus' },
{ str: '-', tokenName: 'Minus' },
{ str: '>', tokenName: 'Gt' },
{ str: '<', tokenName: 'Lt' },
{ str: '%', tokenName: 'Modulo' },
]
export const operatorTokenizer = new ExternalTokenizer((input: InputStream) => {
for (let operator of operators) {
if (!matchesString(input, 0, operator.str)) continue
const afterOpPos = operator.str.length
const charAfterOp = input.peek(afterOpPos)
if (!isWhitespace(charAfterOp)) continue
// Accept the operator token
const token = terms[operator.tokenName]
if (token === undefined) {
throw new Error(`Unknown token name: ${operator.tokenName}`)
}
input.advance(afterOpPos)
input.acceptToken(token)
return
}
})
const isWhitespace = (ch: number): boolean => {
return matchesChar(ch, [' ', '\t', '\n'])
}
const matchesChar = (ch: number, chars: (string | number)[]): boolean => {
for (const c of chars) {
if (typeof c === 'number') {
if (ch === c) {
return true
}
} else if (ch === c.charCodeAt(0)) {
return true
}
}
return false
}
const matchesString = (input: InputStream, pos: number, str: string): boolean => {
for (let i = 0; i < str.length; i++) {
if (input.peek(pos + i) !== str.charCodeAt(i)) {
return false
}
}
return true
}
const peek = (numChars: number, input: InputStream): string => {
let result = ''
for (let i = 0; i < numChars; i++) {
const ch = input.peek(i)
if (ch === -1) {
result += 'EOF'
break
} else {
result += String.fromCharCode(ch)
}
}
return result
}

View File

@ -1,17 +1,11 @@
import { CompilerError } from '#compiler/compilerError'
import { Scanner, type Token, TokenType } from './tokenizer2'
import { SyntaxNode, operators, precedence, conditionals, compounds } from './node'
import { globals } from './tokenizer'
import { parseString } from './stringParser'
const $T = TokenType
// tell the dotGet searcher about builtin globals
export const globals: string[] = []
export const setGlobals = (newGlobals: string[] | Record<string, any>) => {
globals.length = 0
globals.push(...(Array.isArray(newGlobals) ? newGlobals : Object.keys(newGlobals)))
}
export const parse = (input: string): SyntaxNode => {
const parser = new Parser()
return parser.parse(input)
@ -42,7 +36,7 @@ export class Parser {
pos = 0
inParens = 0
input = ''
scope = new Scope()
scope = new Scope
inTestExpr = false
parse(input: string): SyntaxNode {
@ -78,11 +72,14 @@ export class Parser {
// statement is a line of code
statement(): SyntaxNode | null {
if (this.is($T.Comment)) return this.comment()
if (this.is($T.Comment))
return this.comment()
while (this.is($T.Newline) || this.is($T.Semicolon)) this.next()
while (this.is($T.Newline) || this.is($T.Semicolon))
this.next()
if (this.isEOF() || this.isExprEndKeyword()) return null
if (this.isEOF() || this.isExprEndKeyword())
return null
return this.expression()
}
@ -96,38 +93,51 @@ export class Parser {
let expr
// x = value
if (
this.is($T.Identifier) &&
(this.nextIs($T.Operator, '=') || compounds.some((x) => this.nextIs($T.Operator, x)))
)
if (this.is($T.Identifier) && (
this.nextIs($T.Operator, '=') || compounds.some(x => this.nextIs($T.Operator, x))
))
expr = this.assign()
// if, while, do, etc
else if (this.is($T.Keyword)) expr = this.keywords()
else if (this.is($T.Keyword))
expr = this.keywords()
// dotget
else if (this.nextIs($T.Operator, '.')) expr = this.dotGetFunctionCall()
else if (this.nextIs($T.Operator, '.'))
expr = this.dotGetFunctionCall()
// echo hello world
else if (this.is($T.Identifier) && !this.nextIs($T.Operator) && !this.nextIsExprEnd())
expr = this.functionCall()
// bare-function-call
else if (this.is($T.Identifier) && this.nextIsExprEnd()) expr = this.functionCallOrIdentifier()
else if (this.is($T.Identifier) && this.nextIsExprEnd())
expr = this.functionCallOrIdentifier()
// everything else
else expr = this.exprWithPrecedence()
else
expr = this.exprWithPrecedence()
// check for destructuring
if (expr.type.is('Array') && this.is($T.Operator, '=')) return this.destructure(expr)
if (expr.type.is('Array') && this.is($T.Operator, '='))
return this.destructure(expr)
// check for parens function call
// ex: (ref my-func) my-arg
if (expr.type.is('ParenExpr') && !this.isExprEnd()) expr = this.functionCall(expr)
if (expr.type.is('ParenExpr') && !this.isExprEnd())
expr = this.functionCall(expr)
// if dotget is followed by binary operator, continue parsing as binary expression
if (expr.type.is('DotGet') && this.is($T.Operator) && !this.is($T.Operator, '|'))
expr = this.dotGetBinOp(expr)
// one | echo
if (allowPipe && this.isPipe()) return this.pipe(expr)
if (allowPipe && this.isPipe())
return this.pipe(expr)
// regular
else return expr
else
return expr
}
// piping | stuff | is | cool
@ -191,19 +201,23 @@ export class Parser {
// if, while, do, etc
keywords(): SyntaxNode {
if (this.is($T.Keyword, 'if')) return this.if()
if (this.is($T.Keyword, 'if'))
return this.if()
if (this.is($T.Keyword, 'while')) return this.while()
if (this.is($T.Keyword, 'while'))
return this.while()
if (this.is($T.Keyword, 'do')) return this.do()
if (this.is($T.Keyword, 'do'))
return this.do()
if (this.is($T.Keyword, 'try')) return this.try()
if (this.is($T.Keyword, 'try'))
return this.try()
if (this.is($T.Keyword, 'throw')) return this.throw()
if (this.is($T.Keyword, 'throw'))
return this.throw()
if (this.is($T.Keyword, 'not')) return this.not()
if (this.is($T.Keyword, 'import')) return this.import()
if (this.is($T.Keyword, 'import'))
return this.import()
return this.expect($T.Keyword, 'if/while/do/import') as never
}
@ -215,12 +229,15 @@ export class Parser {
// 3. binary operations
// 4. anywhere an expression can be used
value(): SyntaxNode {
if (this.is($T.OpenParen)) return this.parens()
if (this.is($T.OpenParen))
return this.parens()
if (this.is($T.OpenBracket)) return this.arrayOrDict()
if (this.is($T.OpenBracket))
return this.arrayOrDict()
// dotget
if (this.nextIs($T.Operator, '.')) return this.dotGet()
if (this.nextIs($T.Operator, '.'))
return this.dotGet()
return this.atom()
}
@ -298,7 +315,8 @@ export class Parser {
}
// probably an array
if (curr.type !== $T.Comment && curr.type !== $T.Semicolon && curr.type !== $T.Newline) break
if (curr.type !== $T.Comment && curr.type !== $T.Semicolon && curr.type !== $T.Newline)
break
curr = this.peek(peek++)
}
@ -316,7 +334,7 @@ export class Parser {
const node = new SyntaxNode(
opToken.value === '=' ? 'Assign' : 'CompoundAssign',
ident.from,
expr.to,
expr.to
)
return node.push(ident, op, expr)
@ -333,7 +351,8 @@ export class Parser {
// atoms are the basic building blocks: literals, identifiers, words
atom(): SyntaxNode {
if (this.is($T.String)) return this.string()
if (this.is($T.String))
return this.string()
if (this.isAny($T.Null, $T.Boolean, $T.Number, $T.Identifier, $T.Word, $T.Regex, $T.Underscore))
return SyntaxNode.from(this.next())
@ -374,7 +393,8 @@ export class Parser {
const keyword = this.keyword('catch')
let catchVar
if (this.is($T.Identifier)) catchVar = this.identifier()
if (this.is($T.Identifier))
catchVar = this.identifier()
const block = this.block()
@ -478,14 +498,12 @@ export class Parser {
this.scope.add(varName)
let arg
if (this.is($T.Identifier)) arg = this.identifier()
else if (this.is($T.NamedArgPrefix)) arg = this.namedParam()
if (this.is($T.Identifier))
arg = this.identifier()
else if (this.is($T.NamedArgPrefix))
arg = this.namedParam()
else
throw new CompilerError(
`Expected Identifier or NamedArgPrefix, got ${TokenType[this.current().type]}`,
this.current().from,
this.current().to,
)
throw new CompilerError(`Expected Identifier or NamedArgPrefix, got ${TokenType[this.current().type]}`, this.current().from, this.current().to)
params.push(arg)
}
@ -493,9 +511,11 @@ export class Parser {
const block = this.block(false)
let catchNode, finalNode
if (this.is($T.Keyword, 'catch')) catchNode = this.catch()
if (this.is($T.Keyword, 'catch'))
catchNode = this.catch()
if (this.is($T.Keyword, 'finally')) finalNode = this.finally()
if (this.is($T.Keyword, 'finally'))
finalNode = this.finally()
const end = this.keyword('end')
@ -507,7 +527,11 @@ export class Parser {
node.add(doNode)
const paramsNode = new SyntaxNode('Params', params[0]?.from ?? 0, params.at(-1)?.to ?? 0)
const paramsNode = new SyntaxNode(
'Params',
params[0]?.from ?? 0,
params.at(-1)?.to ?? 0
)
if (params.length) paramsNode.push(...params)
node.add(paramsNode)
@ -528,7 +552,8 @@ export class Parser {
const ident = this.input.slice(left.from, left.to)
// not in scope, just return Word
if (!this.scope.has(ident)) return this.word(left)
if (!this.scope.has(ident))
return this.word(left)
if (left.type.is('Identifier')) left.type = 'IdentifierBeforeDot'
@ -568,13 +593,16 @@ export class Parser {
const dotGet = this.dotGet()
// if followed by a binary operator (not pipe), return dotGet/Word as-is for expression parser
if (this.is($T.Operator) && !this.is($T.Operator, '|')) return dotGet
if (this.is($T.Operator) && !this.is($T.Operator, '|'))
return dotGet
// dotget not in scope, regular Word
if (dotGet.type.is('Word')) return dotGet
if (this.isExprEnd()) return this.functionCallOrIdentifier(dotGet)
else return this.functionCall(dotGet)
if (this.isExprEnd())
return this.functionCallOrIdentifier(dotGet)
else
return this.functionCall(dotGet)
}
// can be used in functions or try block
@ -726,30 +754,17 @@ export class Parser {
const val = this.value()
if (!['Null', 'Boolean', 'Number', 'String'].includes(val.type.name))
throw new CompilerError(
`Default value must be null, boolean, number, or string, got ${val.type.name}`,
val.from,
val.to,
)
throw new CompilerError(`Default value must be null, boolean, number, or string, got ${val.type.name}`, val.from, val.to)
const node = new SyntaxNode('NamedParam', prefix.from, val.to)
return node.push(prefix, val)
}
// not blah
not(): SyntaxNode {
const keyword = this.keyword('not')
const val = this.expression()
const node = new SyntaxNode('Not', keyword.from, val.to)
return node.push(keyword, val)
}
// operators like + - =
op(op?: string): SyntaxNode {
const token = op ? this.expect($T.Operator, op) : this.expect($T.Operator)
const name = operators[token.value!]
if (!name)
throw new CompilerError(`Operator not registered: ${token.value!}`, token.from, token.to)
if (!name) throw new CompilerError(`Operator not registered: ${token.value!}`, token.from, token.to)
return new SyntaxNode(name, token.from, token.to)
}
@ -796,9 +811,11 @@ export class Parser {
let last = tryBlock.at(-1)
let catchNode, finalNode
if (this.is($T.Keyword, 'catch')) catchNode = this.catch()
if (this.is($T.Keyword, 'catch'))
catchNode = this.catch()
if (this.is($T.Keyword, 'finally')) finalNode = this.finally()
if (this.is($T.Keyword, 'finally'))
finalNode = this.finally()
const end = this.keyword('end')
@ -808,9 +825,11 @@ export class Parser {
const node = new SyntaxNode('TryExpr', tryNode.from, last!.to)
node.push(tryNode, ...tryBlock)
if (catchNode) node.push(catchNode)
if (catchNode)
node.push(catchNode)
if (finalNode) node.push(finalNode)
if (finalNode)
node.push(finalNode)
return node.push(end)
}
@ -832,7 +851,8 @@ export class Parser {
while (this.is($T.Operator, '.')) {
this.next()
if (this.isAny($T.Word, $T.Identifier, $T.Number)) parts.push(this.next())
if (this.isAny($T.Word, $T.Identifier, $T.Number))
parts.push(this.next())
}
return new SyntaxNode('Word', parts[0]!.from, parts.at(-1)!.to)
@ -855,7 +875,8 @@ export class Parser {
let offset = 1
let peek = this.peek(offset)
while (peek && peek.type === $T.Newline) peek = this.peek(++offset)
while (peek && peek.type === $T.Newline)
peek = this.peek(++offset)
if (!peek || peek.type !== type) return false
if (value !== undefined && peek.value !== value) return false
@ -876,7 +897,7 @@ export class Parser {
}
isAny(...type: TokenType[]): boolean {
return type.some((x) => this.is(x))
return type.some(x => this.is(x))
}
nextIs(type: TokenType, value?: string): boolean {
@ -887,58 +908,43 @@ export class Parser {
}
nextIsAny(...type: TokenType[]): boolean {
return type.some((x) => this.nextIs(x))
return type.some(x => this.nextIs(x))
}
isExprEnd(): boolean {
return (
this.isAny($T.Colon, $T.Semicolon, $T.Newline, $T.CloseParen, $T.CloseBracket) ||
return this.isAny($T.Colon, $T.Semicolon, $T.Newline, $T.CloseParen, $T.CloseBracket) ||
this.is($T.Operator, '|') ||
this.isExprEndKeyword() ||
!this.current()
)
this.isExprEndKeyword() || !this.current()
}
nextIsExprEnd(): boolean {
// pipes act like expression end for function arg parsing
if (this.nextIs($T.Operator, '|')) return true
if (this.nextIs($T.Operator, '|'))
return true
return (
this.nextIsAny($T.Colon, $T.Semicolon, $T.Newline, $T.CloseBracket, $T.CloseParen) ||
this.nextIs($T.Keyword, 'end') ||
this.nextIs($T.Keyword, 'else') ||
this.nextIs($T.Keyword, 'catch') ||
this.nextIs($T.Keyword, 'finally') ||
return this.nextIsAny($T.Colon, $T.Semicolon, $T.Newline, $T.CloseBracket, $T.CloseParen) ||
this.nextIs($T.Keyword, 'end') || this.nextIs($T.Keyword, 'else') ||
this.nextIs($T.Keyword, 'catch') || this.nextIs($T.Keyword, 'finally') ||
!this.peek()
)
}
isExprEndKeyword(): boolean {
return (
this.is($T.Keyword, 'end') ||
this.is($T.Keyword, 'else') ||
this.is($T.Keyword, 'catch') ||
this.is($T.Keyword, 'finally')
)
return this.is($T.Keyword, 'end') || this.is($T.Keyword, 'else') ||
this.is($T.Keyword, 'catch') || this.is($T.Keyword, 'finally')
}
isPipe(): boolean {
// inside parens, only look for pipes on same line (don't look past newlines)
const canLookPastNewlines = this.inParens === 0
return (
this.is($T.Operator, '|') || (canLookPastNewlines && this.peekPastNewlines($T.Operator, '|'))
)
return this.is($T.Operator, '|') ||
(canLookPastNewlines && this.peekPastNewlines($T.Operator, '|'))
}
expect(type: TokenType, value?: string): Token | never {
if (!this.is(type, value)) {
const token = this.current()
throw new CompilerError(
`Expected ${TokenType[type]}${value ? ` "${value}"` : ''}, got ${TokenType[token?.type || 0]}${token?.value ? ` "${token.value}"` : ''} at position ${this.pos}`,
token.from,
token.to,
)
throw new CompilerError(`Expected ${TokenType[type]}${value ? ` "${value}"` : ''}, got ${TokenType[token?.type || 0]}${token?.value ? ` "${token.value}"` : ''} at position ${this.pos}`, token.from, token.to)
}
return this.next()
}
@ -958,7 +964,7 @@ function collapseDotGets(origNodes: SyntaxNode[]): SyntaxNode {
if (left.type.is('Identifier')) left.type = 'IdentifierBeforeDot'
const dot = new SyntaxNode('DotGet', left.from, right.to)
const dot = new SyntaxNode("DotGet", left.from, right.to)
dot.push(left, right)
right = dot

View File

@ -0,0 +1,129 @@
import { ContextTracker, InputStream } from '@lezer/lr'
import * as terms from './shrimp.terms'
export class Scope {
constructor(public parent: Scope | null, public vars = new Set<string>()) { }
has(name: string): boolean {
return this.vars.has(name) || (this.parent?.has(name) ?? false)
}
hash(): number {
let h = 0
for (const name of this.vars) {
for (let i = 0; i < name.length; i++) {
h = (h << 5) - h + name.charCodeAt(i)
h |= 0
}
}
if (this.parent) {
h = (h << 5) - h + this.parent.hash()
h |= 0
}
return h
}
// Static methods that return new Scopes (immutable operations)
static add(scope: Scope, ...names: string[]): Scope {
const newVars = new Set(scope.vars)
names.forEach((name) => newVars.add(name))
return new Scope(scope.parent, newVars)
}
push(): Scope {
return new Scope(this, new Set())
}
pop(): Scope {
return this.parent ?? this
}
}
// Tracker context that combines Scope with temporary pending identifiers
class TrackerContext {
constructor(public scope: Scope, public pendingIds: string[] = []) { }
}
// Extract identifier text from input stream
const readIdentifierText = (input: InputStream, start: number, end: number): string => {
let text = ''
for (let i = start; i < end; i++) {
const offset = i - input.pos
const ch = input.peek(offset)
if (ch === -1) break
text += String.fromCharCode(ch)
}
return text
}
let inParams = false
export const trackScope = new ContextTracker<TrackerContext>({
start: new TrackerContext(new Scope(null, new Set())),
shift(context, term, stack, input) {
if (term == terms.Do) inParams = true
if (term === terms.AssignableIdentifier) {
const text = readIdentifierText(input, input.pos, stack.pos)
return new TrackerContext(Scope.add(context.scope, text), context.pendingIds)
}
if (inParams && term === terms.Identifier) {
const text = readIdentifierText(input, input.pos, stack.pos)
return new TrackerContext(context.scope, [...context.pendingIds, text])
}
// Track identifiers in array destructuring: [ a b ] = ...
if (!inParams && term === terms.Identifier && isArrayDestructuring(input)) {
const text = readIdentifierText(input, input.pos, stack.pos)
return new TrackerContext(Scope.add(context.scope, text), context.pendingIds)
}
return context
},
reduce(context, term) {
if (term === terms.Params) {
inParams = false
let newScope = context.scope.push()
if (context.pendingIds.length > 0) {
newScope = Scope.add(newScope, ...context.pendingIds)
}
return new TrackerContext(newScope, [])
}
// Pop scope when exiting function
if (term === terms.FunctionDef) {
return new TrackerContext(context.scope.pop(), [])
}
return context
},
hash: (context) => context.scope.hash(),
})
// Check if we're parsing array destructuring: [ a b ] = ...
const isArrayDestructuring = (input: InputStream): boolean => {
let pos = 0
// Find closing bracket
while (pos < 200 && input.peek(pos) !== 93 /* ] */) {
if (input.peek(pos) === -1) return false // EOF
pos++
}
if (input.peek(pos) !== 93 /* ] */) return false
pos++
// Skip whitespace
while (input.peek(pos) === 32 /* space */ ||
input.peek(pos) === 9 /* tab */ ||
input.peek(pos) === 10 /* \n */) {
pos++
}
return input.peek(pos) === 61 /* = */
}

293
src/parser/shrimp.grammar Normal file
View File

@ -0,0 +1,293 @@
@external propSource highlighting from "./highlight"
@context trackScope from "./parserScopeContext"
@skip { space | Comment }
@top Program { item* }
@external tokens operatorTokenizer from "./operatorTokenizer" { Star, Slash, Plus, Minus, And, Or, Eq, EqEq, Neq, Lt, Lte, Gt, Gte, Modulo, PlusEq, MinusEq, StarEq, SlashEq, ModuloEq, Band, Bor, Bxor, Shl, Shr, Ushr, NullishCoalesce, NullishEq }
@tokens {
@precedence { Number Regex }
StringFragment { !['\\$]+ }
DoubleQuote { '"' !["]* '"' }
NamedArgPrefix { $[a-z] $[a-z0-9-]* "=" }
Number {
("-" | "+")? "0x" $[0-9a-fA-F]+ |
("-" | "+")? "0b" $[01]+ |
("-" | "+")? "0o" $[0-7]+ |
("-" | "+")? $[0-9]+ ("_"? $[0-9]+)* ('.' $[0-9]+ ("_"? $[0-9]+)*)?
}
Boolean { "true" | "false" }
semicolon { ";" }
eof { @eof }
space { " " | "\t" }
Comment { "#" ![\n]* }
leftParen { "(" }
rightParen { ")" }
colon[closedBy="end", @name="colon"] { ":" }
Underscore { "_" }
Dollar { "$" }
Regex { "//" (![/\\\n[] | "\\" ![\n] | "[" (![\n\\\]] | "\\" ![\n])* "]")+ ("//" $[gimsuy]*)? } // Stolen from the lezer JavaScript grammar
"|"[@name=operator]
}
newlineOrSemicolon { newline | semicolon }
end { @specialize[@name=keyword]<Identifier, "end"> }
while { @specialize[@name=keyword]<Identifier, "while"> }
if { @specialize[@name=keyword]<Identifier, "if"> }
else { @specialize[@name=keyword]<Identifier, "else"> }
try { @specialize[@name=keyword]<Identifier, "try"> }
catch { @specialize[@name=keyword]<Identifier, "catch"> }
finally { @specialize[@name=keyword]<Identifier, "finally"> }
throw { @specialize[@name=keyword]<Identifier, "throw"> }
import { @specialize[@name=keyword]<Identifier, "import"> }
null { @specialize[@name=Null]<Identifier, "null"> }
@external tokens tokenizer from "./tokenizer" { Identifier, AssignableIdentifier, Word, IdentifierBeforeDot, CurlyString }
@external tokens pipeStartsLineTokenizer from "./tokenizer" { newline, pipeStartsLine }
@external specialize {Identifier} specializeKeyword from "./tokenizer" { Do }
@precedence {
pipe @left,
or @left,
and @left,
nullish @left,
comparison @left,
multiplicative @left,
additive @left,
bitwise @left,
call,
functionWithNewlines
}
item {
consumeToTerminator newlineOrSemicolon |
consumeToTerminator eof |
newlineOrSemicolon // allow blank lines
}
consumeToTerminator {
PipeExpr |
WhileExpr |
FunctionCallWithBlock |
ambiguousFunctionCall |
TryExpr |
Throw |
Import |
IfExpr |
FunctionDef |
CompoundAssign |
Assign |
BinOp |
ConditionalOp |
expressionWithoutIdentifier
}
PipeExpr {
pipeOperand (!pipe (pipeStartsLine? "|") newlineOrSemicolon* pipeOperand)+
}
pipeOperand {
consumeToTerminator
}
WhileExpr {
while (ConditionalOp | expression) colon Block end
}
Block {
consumeToTerminator | newlineOrSemicolon block
}
FunctionCallWithBlock {
ambiguousFunctionCall colon Block CatchExpr? FinallyExpr? end
}
FunctionCallOrIdentifier {
DotGet | Identifier
}
ambiguousFunctionCall {
FunctionCall | FunctionCallOrIdentifier
}
FunctionCall {
(DotGet | Identifier | ParenExpr) arg+
}
arg {
PositionalArg | NamedArg
}
PositionalArg {
expression | FunctionDef | Underscore
}
NamedArg {
NamedArgPrefix (expression | FunctionDef | Underscore)
}
FunctionDef {
Do Params colon (consumeToTerminator | newlineOrSemicolon block) CatchExpr? FinallyExpr? end
}
ifTest {
ConditionalOp | expression | FunctionCall
}
IfExpr {
if ifTest colon Block ElseIfExpr* ElseExpr? end
}
ElseIfExpr {
else if ifTest colon Block
}
ElseExpr {
else colon Block
}
TryExpr {
try colon Block CatchExpr? FinallyExpr? end
}
CatchExpr {
catch Identifier colon Block
}
FinallyExpr {
finally colon Block
}
Throw {
throw (BinOp | ConditionalOp | expression)
}
// this has to be in the parse tree so the scope tracker can use it
Import {
import NamedArg* Identifier+ NamedArg*
}
ConditionalOp {
expression !comparison EqEq expression |
expression !comparison Neq expression |
expression !comparison Lt expression |
expression !comparison Lte expression |
expression !comparison Gt expression |
expression !comparison Gte expression |
(expression | ConditionalOp) !and And (expression | ConditionalOp) |
(expression | ConditionalOp) !or Or (expression | ConditionalOp) |
(expression | ConditionalOp) !nullish NullishCoalesce (expression | ConditionalOp)
}
Params {
Identifier* NamedParam*
}
NamedParam {
NamedArgPrefix (String | Number | Boolean | null)
}
Assign {
(AssignableIdentifier | Array) Eq consumeToTerminator
}
CompoundAssign {
AssignableIdentifier (PlusEq | MinusEq | StarEq | SlashEq | ModuloEq | NullishEq) consumeToTerminator
}
BinOp {
expression !multiplicative Modulo expression |
(expression | BinOp) !multiplicative Star (expression | BinOp) |
(expression | BinOp) !multiplicative Slash (expression | BinOp) |
(expression | BinOp) !additive Plus (expression | BinOp) |
(expression | BinOp) !additive Minus (expression | BinOp) |
(expression | BinOp) !bitwise Band (expression | BinOp) |
(expression | BinOp) !bitwise Bor (expression | BinOp) |
(expression | BinOp) !bitwise Bxor (expression | BinOp) |
(expression | BinOp) !bitwise Shl (expression | BinOp) |
(expression | BinOp) !bitwise Shr (expression | BinOp) |
(expression | BinOp) !bitwise Ushr (expression | BinOp)
}
ParenExpr {
leftParen newlineOrSemicolon* (
FunctionCallWithNewlines |
IfExpr |
ambiguousFunctionCall |
BinOp newlineOrSemicolon* |
expressionWithoutIdentifier |
ConditionalOp newlineOrSemicolon* |
PipeExpr |
FunctionDef
)
rightParen
}
FunctionCallWithNewlines[@name=FunctionCall] {
(DotGet | Identifier | ParenExpr) newlineOrSemicolon+ arg !functionWithNewlines (newlineOrSemicolon+ arg)* newlineOrSemicolon*
}
expression {
expressionWithoutIdentifier | DotGet | Identifier
}
@local tokens {
dot { "." }
}
@skip {} {
DotGet {
IdentifierBeforeDot dot (DotGet | Number | Identifier | ParenExpr) |
Dollar dot (DotGet | Number | Identifier | ParenExpr)
}
String {
"'" stringContent* "'" | CurlyString | DoubleQuote
}
}
stringContent {
StringFragment |
Interpolation |
EscapeSeq
}
Interpolation {
"$" FunctionCallOrIdentifier |
"$" ParenExpr
}
EscapeSeq {
"\\" ("$" | "n" | "t" | "r" | "\\" | "'")
}
Dict {
"[=]" |
"[" newlineOrSemicolon* NamedArg (newlineOrSemicolon | NamedArg)* "]"
}
Array {
"[" newlineOrSemicolon* (expression (newlineOrSemicolon | expression)*)? "]"
}
// We need expressionWithoutIdentifier to avoid conflicts in consumeToTerminator.
// Without this, when parsing "my-var" at statement level, the parser can't decide:
// - ambiguousFunctionCall → FunctionCallOrIdentifier → Identifier
// - expression → Identifier
// Both want the same Identifier token! So we use expressionWithoutIdentifier
// to remove Identifier from the second path, forcing standalone identifiers
// to go through ambiguousFunctionCall (which is what we want semantically).
// Yes, it is annoying and I gave up trying to use GLR to fix it.
expressionWithoutIdentifier {
ParenExpr | Word | String | Number | Boolean | Regex | Dict | Array | null
}
block {
(consumeToTerminator? newlineOrSemicolon)*
}

4
src/parser/shrimp.grammar.d.ts vendored Normal file
View File

@ -0,0 +1,4 @@
declare module '*.grammar' {
const content: string
export default content
}

View File

@ -0,0 +1,81 @@
// This file was generated by lezer-generator. You probably shouldn't edit it.
export const
Star = 1,
Slash = 2,
Plus = 3,
Minus = 4,
And = 5,
Or = 6,
Eq = 7,
EqEq = 8,
Neq = 9,
Lt = 10,
Lte = 11,
Gt = 12,
Gte = 13,
Modulo = 14,
PlusEq = 15,
MinusEq = 16,
StarEq = 17,
SlashEq = 18,
ModuloEq = 19,
Band = 20,
Bor = 21,
Bxor = 22,
Shl = 23,
Shr = 24,
Ushr = 25,
NullishCoalesce = 26,
NullishEq = 27,
Identifier = 28,
AssignableIdentifier = 29,
Word = 30,
IdentifierBeforeDot = 31,
CurlyString = 32,
newline = 101,
pipeStartsLine = 102,
Do = 33,
Comment = 34,
Program = 35,
PipeExpr = 36,
WhileExpr = 38,
keyword = 84,
ConditionalOp = 40,
ParenExpr = 41,
FunctionCallWithNewlines = 42,
DotGet = 43,
Number = 44,
Dollar = 45,
PositionalArg = 46,
FunctionDef = 47,
Params = 48,
NamedParam = 49,
NamedArgPrefix = 50,
String = 51,
StringFragment = 52,
Interpolation = 53,
FunctionCallOrIdentifier = 54,
EscapeSeq = 55,
DoubleQuote = 56,
Boolean = 57,
Null = 58,
colon = 59,
CatchExpr = 60,
Block = 62,
FinallyExpr = 63,
Underscore = 66,
NamedArg = 67,
IfExpr = 68,
FunctionCall = 70,
ElseIfExpr = 71,
ElseExpr = 73,
BinOp = 74,
Regex = 75,
Dict = 76,
Array = 77,
FunctionCallWithBlock = 78,
TryExpr = 79,
Throw = 81,
Import = 83,
CompoundAssign = 85,
Assign = 86

27
src/parser/shrimp.ts Normal file
View File

@ -0,0 +1,27 @@
// This file was generated by lezer-generator. You probably shouldn't edit it.
import {LRParser, LocalTokenGroup} from "@lezer/lr"
import {operatorTokenizer} from "./operatorTokenizer"
import {tokenizer, pipeStartsLineTokenizer, specializeKeyword} from "./tokenizer"
import {trackScope} from "./parserScopeContext"
import {highlighting} from "./highlight"
const spec_Identifier = {__proto__:null,while:78, null:116, catch:122, finally:128, end:130, if:138, else:144, try:160, throw:164, import:168}
export const parser = LRParser.deserialize({
version: 14,
states: "?[QYQ!SOOOOQ!Q'#Ek'#EkO!sO!bO'#DXO%kQ!TO'#DdO&UOSO'#DaOOQ!R'#Da'#DaO)SQ!TO'#EnOOQ!Q'#E{'#E{O)pQRO'#DxO+xQ!TO'#EjO,fQ!SO'#DVOOQ!R'#Dz'#DzO/WQ!SO'#D{OOQ!R'#En'#EnO/_Q!TO'#EnO1cQ!TO'#EmO2qQ!TO'#EjO3OQRO'#ETOOQ!Q'#Ej'#EjO3gQ!SO'#EjO3nQrO'#EiOOQ!Q'#Ei'#EiOOQ!Q'#EV'#EVQYQ!SOOO4PQbO'#D]O4[QbO'#DrO5YQbO'#DSO6WQQO'#D}O5YQbO'#EPO6]QbO'#ERO6eObO,59sOOQ!Q'#D['#D[O6vQbO'#DqOOQ!Q'#Eq'#EqOOQ!Q'#E_'#E_O7QQ!SO,5:`OOQ!R'#Em'#EmO8QQbO'#DcO8`QWO'#DeOOOO'#Es'#EsOOOO'#E['#E[O8tOSO,59{OOQ!R,59{,59{O5YQbO,5:dO5YQbO,5:dO5YQbO,5:dO5YQbO,5:dO5YQbO,59pO5YQbO,59pO5YQbO,59pO5YQbO,59pOOQ!Q'#EX'#EXO,fQ!SO,59qO9SQ!TO'#DdO9^Q!TO'#EnO9hQsO,59qO9uQQO,59qO9zQrO,59qO:VQrO,59qO:eQsO,59qO;TQsO,59qO;[QrO'#DQO;dQ!SO,5:gO;kQrO,5:fOOQ!R,5:g,5:gO;yQ!SO,5:gO<WQbO,5:pO<WQbO,5:oOYQ!SO,5:hO=kQ!SO,59lOOQ!Q,5;T,5;TOYQ!SO'#EWO>]QQO'#EWOOQ!Q-E8T-E8TOOQ!Q'#EY'#EYO>bQbO'#D^O>mQbO'#D_OOQO'#EZ'#EZO>eQQO'#D^O?RQQO,59wO?WQcO'#EmO@TQRO'#EzOAQQRO'#EzOOQO'#Ez'#EzOAXQQO,5:^OA^QRO,59nOAeQRO,59nOYQ!SO,5:iOAsQ!TO,5:kOCXQ!TO,5:kOC{Q!TO,5:kODYQ!SO,5:mOOQ!Q'#Ec'#EcO6]QbO,5:mOOQ!R1G/_1G/_OOQ!Q,5:],5:]OOQ!Q-E8]-E8]OOOO'#Dd'#DdOOOO,59},59}OOOO,5:P,5:POOOO-E8Y-E8YOOQ!R1G/g1G/gOOQ!R1G0O1G0OOF_Q!TO1G0OOFiQ!TO1G0OOG}Q!TO1G0OOHXQ!TO1G0OOHfQ!TO1G0OOOQ!R1G/[1G/[OI}Q!TO1G/[OJUQ!TO1G/[OJ]Q!TO1G/[OKbQ!TO1G/[OJdQ!TO1G/[OOQ!Q-E8V-E8VOKxQsO1G/]OLVQQO1G/]OL[QrO1G/]OLgQrO1G/]OLuQsO1G/]OL|QsO1G/]OMTQ!SO,59rOM_QrO1G/]OOQ!R1G/]1G/]OMjQrO1G0QOOQ!R1G0R1G0ROMxQ!SO1G0ROOQp'#Ea'#EaOMjQrO1G0QOOQ!R1G0Q1G0QOOQ!Q'#Eb'#EbOMxQ!SO1G0RONVQ!SO1G0[ONwQ!SO1G0ZO! iQ!SO'#DlO! }Q!SO'#DlO!!_QbO1G0SOOQ!Q-E8U-E8UOYQ!SO,5:rOOQ!Q,5:r,5:rOYQ!SO,5:rOOQ!Q-E8W-E8WO!!jQQO,59xOOQO,59y,59yOOQO-E8X-E8XOYQ!SO1G/cOYQ!SO1G/xOYQ!SO1G/YO!!rQbO1G0TO!!}Q!SO1G0XO!#rQ!SO1G0XOOQ!Q-E8a-E8aO!#yQrO7+$wOOQ!R7+$w7+$wO!$UQrO1G/^O!$aQrO7+%lOOQ!R7+%l7+%lO!$oQ!SO7+%mOOQ!R7+%m7+%mOOQp-E8_-E8_OOQ!Q-E8`-E8`OOQ!Q'#E]'#E]O!$|QrO'#E]O!%[Q!SO'#EyOOQ`,5:W,5:WO!%lQbO'#DjO!%qQQO'#DmOOQ!Q7+%n7+%nO!%vQbO7+%nO!%{QbO7+%nOOQ!Q1G0^1G0^OYQ!SO1G0^O!&TQ!SO7+$}O!&fQ!SO7+$}O!&sQbO7+%dO!&{QbO7+$tOOQ!Q7+%o7+%oO!'QQbO7+%oO!'VQbO7+%oO!'_Q!SO7+%sOOQ!R<<Hc<<HcO!(SQ!SO7+$xO!(aQrO7+$xOOQ!R<<IW<<IWOOQ!R<<IX<<IXOOQ!Q,5:w,5:wOOQ!Q-E8Z-E8ZO!(lQQO,5:UOYQ!SO,5:XOOQ!Q<<IY<<IYO!(qQbO<<IYOOQ!Q7+%x7+%xOOQ!Q<<Hi<<HiO!(vQbO<<HiO!({QbO<<HiO!)TQbO<<HiOOQ`'#E`'#E`O!)`QbO<<IOO!)hQbO'#DwOOQ!Q<<IO<<IOO!)pQbO<<IOOOQ!Q<<H`<<H`OOQ!Q<<IZ<<IZO!)uQbO<<IZOOQp,5:x,5:xO!)zQ!SO<<HdOOQp-E8[-E8[OYQ!SO1G/pOOQ`1G/s1G/sOOQ!QAN>tAN>tOOQ!QAN>TAN>TO!*XQbOAN>TO!*^QbOAN>TOOQ`-E8^-E8^OOQ!QAN>jAN>jO!*fQbOAN>jO4[QbO,5:aOYQ!SO,5:cOOQ!QAN>uAN>uPMTQ!SO'#EXOOQ`7+%[7+%[OOQ!QG23oG23oO!*kQbOG23oP!)kQbO'#DuOOQ!QG24UG24UO!*pQQO1G/{OOQ`1G/}1G/}OOQ!QLD)ZLD)ZOYQ!SO7+%gOOQ`<<IR<<IRO!*uObO,59sO!+WO!bO'#DX",
stateData: "!+`~O#[OSrOS~OlROmaOn]OoQOpTOqhOwjO|]O}QO!YTO!Z]O![]O!giO!m]O!rkO!tlO!vmO#XPO#`PO#cYO#fSO#qZO#r[O~O#dnO~OltOn]OoQOpTOqhO|]O}QO!SpO!YTO!Z]O![]O!doO!m]O#cYO#fSO#qZO#r[OP#aXQ#aXR#aXS#aXT#aXU#aXW#aXX#aXY#aXZ#aX[#aX]#aX^#aXd#aXe#aXf#aXg#aXh#aXi#aXj#aXu!WX!]!WX#Y!WX#p!WX~O#X!WX#`!WX#t!WX!_!WX!b!WX!c!WX!j!WX~P!xO!UwO#fzO#huO#ivO~OltOn]OoQOpTOqhO|]O}QO!SpO!YTO!Z]O![]O!doO!m]O#cYO#fSO#qZO#r[OP#bXQ#bXR#bXS#bXT#bXU#bXW#bXX#bXY#bXZ#bX[#bX]#bX^#bXd#bXe#bXf#bXg#bXh#bXi#bXj#bXu#bX#Y#bX#p#bX~O#X#bX#`#bX#t#bX!]#bX!_#bX!b#bX!c#bX!j#bX~P&dOP|OQ|OR}OS}OT!QOU!ROW!POX!POY!POZ!PO[!PO]!PO^{Od!OOe!OOf!OOg!OOh!OOi!OOj!SO~OP|OQ|OR}OS}Od!OOe!OOf!OOg!OOh!OOi!OOu#^X#Y#^X~O#X#^X#`#^X#t#^X!_#^X!b#^X!c#^X#p#^X!j#^X~P+QOl!VOmaOn]OoQOpTOqhOwjO|]O}QO!YTO!Z]O![]O!giO!m]O!rkO!tlO!vmO#XPO#`PO#cYO#fSO#qZO#r[O~OltOn]OoQOpTO|]O}QO!SpO!YTO!Z]O![]O!m]O#XPO#`PO#cYO#fSO#qZO#r[O~O#s!bO~P.POV!dO#X#bX#`#bX#t#bX!_#bX!b#bX!c#bX!j#bX~P'iOP#aXQ#aXR#aXS#aXT#aXU#aXW#aXX#aXY#aXZ#aX[#aX]#aX^#aXd#aXe#aXf#aXg#aXh#aXi#aXj#aXu#^X#Y#^X~O#X#^X#`#^X#t#^X!_#^X!b#^X!c#^X#p#^X!j#^X~P/{Ou#^X#X#^X#Y#^X#`#^X#t#^X!_#^X!b#^X!c#^X#p#^X!j#^X~OT!QOU!ROj!SO~P2POV!dO_!eO`!eOa!eOb!eOc!eOk!eO~O!]!fO~P2POu!iO#XPO#Y!jO#`PO#t!hO~Ol!lO!S!nO!]!QP~Ol!rOn]OoQOpTO|]O}QO!YTO!Z]O![]O!m]O#cYO#fSO#qZO#r[O~OltOn]OoQOpTO|]O}QO!YTO!Z]O![]O!m]O#cYO#fSO#qZO#r[O~O!]!yO~Ol!lO!SpO~Ol#QOoQO|#QO}QO#cYO~OqhO!d#RO~P5YOqhO!SpO!doOu!ha!]!ha#X!ha#Y!ha#`!ha#t!ha#p!ha!_!ha!b!ha!c!ha!j!ha~P5YOl#TOo&PO}&PO#cYO~O#f#VO#h#VO#i#VO#j#VO#k#VO#l#VO~O!UwO#f#XO#huO#ivO~O#XPO#`PO~P!xO#XPO#`PO~P&dO#XPO#`PO#p#oO~P+QO#p#oO~O#p#oOu#^X#Y#^X~O!]!fO#p#oOu#^X#Y#^X~O#p#oO~P/{OT!QOU!ROj!SO#XPO#`POu#^X#Y#^X~O#p#oO~P:lOu!iO#Y!jO~O#s#qO~P.PO!SpO#XPO#`PO#s#uO~O#XPO#`PO#s#qO~P5YOlROmaOn]OoQOpTOqhOwjO|]O}QO!YTO!Z]O![]O!giO!m]O!rkO!tlO!vmO#cYO#fSO#qZO#r[O~Ou!iO#Y!jO#Xta#`ta#tta#pta!_ta!bta!cta!jta~Ou$QO~Ol!lO!S!nO!]!QX~OpTO|$TO!YTO!Z$TO![$TO#fSO~O!]$VO~OqhO!SpO!doOT#aXU#aXW#aXX#aXY#aXZ#aX[#aX]#aXj#aX!]#aX~P5YOT!QOU!ROj!SO!]#nX~OT!QOU!ROW!POX!POY!POZ!PO[!PO]!POj!SO~O!]#nX~P@cO!]$WO~O!]$XO~P@cOT!QOU!ROj!SO!]$XO~Ou!sa#X!sa#Y!sa#`!sa#t!sa!_!sa!b!sa!c!sa#p!sa!j!sa~P)pOu!sa#X!sa#Y!sa#`!sa#t!sa!_!sa!b!sa!c!sa#p!sa!j!sa~OP|OQ|OR}OS}Od!OOe!OOf!OOg!OOh!OOi!OO~PBgOT!QOU!ROj!SO~PBgOl!lO!SpOu!ua#X!ua#Y!ua#`!ua#t!ua!_!ua!b!ua!c!ua#p!ua!j!ua~O^{OR!liS!lid!lie!lif!lig!lih!lii!liu!li#X!li#Y!li#`!li#t!li#p!li!_!li!b!li!c!li!j!li~OP!liQ!li~PEQOP|OQ|O~PEQOP|OQ|Od!lie!lif!lig!lih!lii!liu!li#X!li#Y!li#`!li#t!li#p!li!_!li!b!li!c!li!j!li~OR!liS!li~PFsOR}OS}O^{O~PFsOR}OS}O~PFsOW!POX!POY!POZ!PO[!PO]!POTxijxiuxi#Xxi#Yxi#`xi#txi#pxi!]xi!_xi!bxi!cxi!jxi~OU!RO~PHpOU!RO~PISOUxi~PHpOT!QOU!ROjxiuxi#Xxi#Yxi#`xi#txi#pxi!]xi!_xi!bxi!cxi!jxi~OW!POX!POY!POZ!PO[!PO]!PO~PJdO#XPO#`PO#p$_O~P+QO#p$_O~O#p$_Ou#^X#Y#^X~O!]!fO#p$_Ou#^X#Y#^X~O#p$_O~P/{O#p$_O~P:lOqhO!doO~P.PO#XPO#`PO#p$_O~O!SpO#XPO#`PO#s$bO~O#XPO#`PO#s$dO~P5YOu!iO#Y!jO#X!xi#`!xi#t!xi!_!xi!b!xi!c!xi#p!xi!j!xi~Ou!iO#Y!jO#X!wi#`!wi#t!wi!_!wi!b!wi!c!wi#p!wi!j!wi~Ou!iO#Y!jO!_!`X!b!`X!c!`X!j!`X~O!_#mP!b#mP!c#mP!j#mP~PYO!_$kO!b$lO!c$mO~O!S!nO!]!Qa~O!_$kO!b$lO!c$vO~O!SpOu!ui#X!ui#Y!ui#`!ui#t!ui!_!ui!b!ui!c!ui#p!ui!j!ui~Ol!lO~P!!}O#XPO#`PO#p$zO~O#XPO#`PO#pzi~O!SpO#XPO#`PO#s$}O~O#XPO#`PO#s%OO~P5YOu!iO#XPO#Y!jO#`PO~O!_#mX!b#mX!c#mX!j#mX~PYOl%RO~O!]%SO~O!c%TO~O!b$lO!c%TO~Ou!iO!_$kO!b$lO!c%WO#Y!jO~O!_#mP!b#mP!c#mP~PYO!c%_O!j%^O~O!c%aO~O!c%bO~O!b$lO!c%bO~O!SpOu!uq#X!uq#Y!uq#`!uq#t!uq!_!uq!b!uq!c!uq#p!uq!j!uq~OqhO!doO#pzq~P.PO#XPO#`PO#pzq~O!]%gO~O!c%iO~O!c%jO~O!b$lO!c%jO~O!_$kO!b$lO!c%jO~O!c%nO!j%^O~O!]%qO!g%pO~O!c%nO~O!c%rO~OqhO!doO#pzy~P.PO!c%uO~O!b$lO!c%uO~O!c%xO~O!c%{O~O!]%|O~Ol#QOo&PO|#QO}&PO#cYO~O#d&OO~O|!m~",
goto: "<[#pPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP#qP$_P$w%x'['bPP(v)S*P*SP*YP+d+h+dPPPP,TP,a,yPPP-a#qP.R.oP.s.yP/s0z$_$_P$_P$_P$_$_2T2Z2g3c3q3{4R4Y4`4j4p4z5UPPPPP5d5h6dP7v9oPP:|P;^PPPPP;b;h;nxbOg!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|Q!ZYR#i!U}bOYg!U!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|x`Og!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|Q!^YS!si%pQ!xjQ!|lQ#`!RQ#b!QQ#e!SR#l!U|UOgi!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%p%q%|!W]RU[jlps{|}!O!P!Q!R!S!V!W!`!c!r#m#r#w$c${%e%sS!WY!US#Qn&OR#UuQ!YYR#h!UxROg!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|!WtRU[jlps{|}!O!P!Q!R!S!V!W!`!c!r#m#r#w$c${%e%sS!VY!US!ri%pS#Qn&OR#TueqRUs!V!W!r#m${%e%sxbOg!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|doRUs!V!W!r#m${%e%sQ!ZYQ#RpR#i!UR!qhX!oh!m!p$S#Y]ORUY[gijlps{|}!O!P!Q!R!S!U!V!W!`!c!d!e!f!i!r!y#m#r#w#{$O$Q$V$W$X$c$i$q$s${%S%e%g%p%q%s%|R$T!nTwSy|VOYg!U!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|R#UuQ$o#|Q$x$YQ%Y$rR%l%ZQ#|!fQ$Y!yQ$t$WQ$u$XQ%h%SQ%t%gQ%z%qR%}%|Q$n#|Q$w$YQ%U$oQ%X$rQ%c$xS%k%Y%ZR%v%ldqRUs!V!W!r#m${%e%sQ!a[[#Om!}#P$Z$[$yQ#p!`X#s!a#p#t$a|VOYg!U!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|T!ui%pT%[$t%]Q%`$tR%o%]xXOg!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|Q!XYQ!{lQ#Y|Q#]}Q#_!OR#g!U#Z]ORUY[gijlps{|}!O!P!Q!R!S!U!V!W!`!c!d!e!f!i!r!y#m#r#w#{$O$Q$V$W$X$c$i$q$s${%S%e%g%p%q%s%|![]RU[ijlps{|}!O!P!Q!R!S!V!W!`!c!r#m#r#w$c${%e%p%s}^OYg!U!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|QgOR!kg^!gd!_#x#y#z$h$rR#}!gQ!UYQ!`[d#f!U!`#m#n$O$^$q${%e%sS#m!V!WS#n!X!^Q$O!iS$^#g#lQ$q$QQ${$`R%e$|Q!mhQ!}mU$R!m!}$[R$[#PQ!phQ$S!mT$U!p$SQySR#WyS$i#{$sR%Q$iQ$|$`R%f$|YsRU!V!W!rR#SsQ%]$tR%m%]Q#t!aQ$a#pT$e#t$aQ#w!cQ$c#rT$f#w$cQ#PmQ$Z!}U$]#P$Z$yR$y$[TfOgSdOgS!_Y!UQ#x!dQ#y!e`#z!f!y$W$X%S%g%q%|Q$P!iU$h#{$i$sS$p$O$QQ$r$VR%V$qSeOg|!TY[!U!V!W!X!^!`!i#g#l#m#n$O$Q$^$`$q${$|%e%sQ!hdW#s!a#p#t$aW#v!c#r#w$c`#{!f!y$W$X%S%g%q%|U$g#{$i$sQ$s$VR%P$h|WOYg!U!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|doRUs!V!W!r#m${%e%sQ!c[S!ti%pQ!wjQ!zlQ#RpQ#Y{Q#Z|Q#[}Q#^!OQ#`!PQ#a!QQ#c!RQ#d!SQ#r!`X#v!c#r#w$cx_Og!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|![tRU[ijlps{|}!O!P!Q!R!S!V!W!`!c!r#m#r#w$c${%e%p%sQ!]YR#k!U[rRUs!V!W!rQ$`#mV%d${%e%sTxSyQ$j#{R%Z$sQ!viR%y%pxcOg!d!e!f!i!y#{$O$Q$V$W$X$i$q$s%S%g%q%|Q![YR#j!U",
nodeNames: "⚠ Star Slash Plus Minus And Or Eq EqEq Neq Lt Lte Gt Gte Modulo PlusEq MinusEq StarEq SlashEq ModuloEq Band Bor Bxor Shl Shr Ushr NullishCoalesce NullishEq Identifier AssignableIdentifier Word IdentifierBeforeDot CurlyString Do Comment Program PipeExpr operator WhileExpr keyword ConditionalOp ParenExpr FunctionCall DotGet Number Dollar PositionalArg FunctionDef Params NamedParam NamedArgPrefix String StringFragment Interpolation FunctionCallOrIdentifier EscapeSeq DoubleQuote Boolean Null colon CatchExpr keyword Block FinallyExpr keyword keyword Underscore NamedArg IfExpr keyword FunctionCall ElseIfExpr keyword ElseExpr BinOp Regex Dict Array FunctionCallWithBlock TryExpr keyword Throw keyword Import keyword CompoundAssign Assign",
maxTerm: 128,
context: trackScope,
nodeProps: [
["closedBy", 59,"end"]
],
propSources: [highlighting],
skippedNodes: [0,34],
repeatNodeCount: 13,
tokenData: "Lp~R}OX$OXY$mYp$Opq$mqr$Ors%Wst'^tu(uuw$Owx(|xy)Ryz)lz{$O{|*V|}$O}!O*V!O!P$O!P!Q3r!Q!R*w!R![-l![!]<_!]!^<x!^!}$O!}#O=c#O#P?X#P#Q?^#Q#R$O#R#S?w#S#T$O#T#Y@b#Y#ZA|#Z#b@b#b#cGj#c#f@b#f#gHm#g#h@b#h#iIp#i#o@b#o#p$O#p#qLQ#q;'S$O;'S;=`$g<%l~$O~O$O~~LkS$TU!USOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OS$jP;=`<%l$O^$tU!US#[YOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OU%]Z!USOr%Wrs&Ost%Wtu&iuw%Wwx&ix#O%W#O#P&i#P;'S%W;'S;=`'W<%lO%WU&VU!YQ!USOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OQ&lTOr&irs&{s;'S&i;'S;=`'Q<%lO&iQ'QO!YQQ'TP;=`<%l&iU'ZP;=`<%l%W^'eZrY!USOY'^YZ$OZt'^tu(Wuw'^wx(Wx#O'^#O#P(W#P;'S'^;'S;=`(o<%lO'^Y(]SrYOY(WZ;'S(W;'S;=`(i<%lO(WY(lP;=`<%l(W^(rP;=`<%l'^^(|O#h[}Q~)RO#f~U)YU!US#cQOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OU)sU!US#pQOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OU*[X!USOt$Ouw$Ox!Q$O!Q!R*w!R![-l![#O$O#P;'S$O;'S;=`$g<%lO$OU+Ob!US|QOt$Ouw$Ox!O$O!O!P,W!P!Q$O!Q![-l![#O$O#P#R$O#R#S.i#S#U$O#U#V/W#V#c$O#c#d0l#d#l$O#l#m1z#m;'S$O;'S;=`$g<%lO$OU,]W!USOt$Ouw$Ox!Q$O!Q![,u![#O$O#P;'S$O;'S;=`$g<%lO$OU,|Y!US|QOt$Ouw$Ox!Q$O!Q![,u![#O$O#P#R$O#R#S,W#S;'S$O;'S;=`$g<%lO$OU-s[!US|QOt$Ouw$Ox!O$O!O!P,W!P!Q$O!Q![-l![#O$O#P#R$O#R#S.i#S;'S$O;'S;=`$g<%lO$OU.nW!USOt$Ouw$Ox!Q$O!Q![-l![#O$O#P;'S$O;'S;=`$g<%lO$OU/]X!USOt$Ouw$Ox!Q$O!Q!R/x!R!S/x!S#O$O#P;'S$O;'S;=`$g<%lO$OU0PX!US|QOt$Ouw$Ox!Q$O!Q!R/x!R!S/x!S#O$O#P;'S$O;'S;=`$g<%lO$OU0qW!USOt$Ouw$Ox!Q$O!Q!Y1Z!Y#O$O#P;'S$O;'S;=`$g<%lO$OU1bW!US|QOt$Ouw$Ox!Q$O!Q!Y1Z!Y#O$O#P;'S$O;'S;=`$g<%lO$OU2P[!USOt$Ouw$Ox!Q$O!Q![2u![!c$O!c!i2u!i#O$O#P#T$O#T#Z2u#Z;'S$O;'S;=`$g<%lO$OU2|[!US|QOt$Ouw$Ox!Q$O!Q![2u![!c$O!c!i2u!i#O$O#P#T$O#T#Z2u#Z;'S$O;'S;=`$g<%lO$OU3wW!USOt$Ouw$Ox!P$O!P!Q4a!Q#O$O#P;'S$O;'S;=`$g<%lO$OU4f^!USOY5bYZ$OZt5btu6euw5bwx6ex!P5b!P!Q$O!Q!}5b!}#O;W#O#P8s#P;'S5b;'S;=`<X<%lO5bU5i^!US!mQOY5bYZ$OZt5btu6euw5bwx6ex!P5b!P!Q9Y!Q!}5b!}#O;W#O#P8s#P;'S5b;'S;=`<X<%lO5bQ6jX!mQOY6eZ!P6e!P!Q7V!Q!}6e!}#O7t#O#P8s#P;'S6e;'S;=`9S<%lO6eQ7YP!P!Q7]Q7bU!mQ#Z#[7]#]#^7]#a#b7]#g#h7]#i#j7]#m#n7]Q7wVOY7tZ#O7t#O#P8^#P#Q6e#Q;'S7t;'S;=`8m<%lO7tQ8aSOY7tZ;'S7t;'S;=`8m<%lO7tQ8pP;=`<%l7tQ8vSOY6eZ;'S6e;'S;=`9S<%lO6eQ9VP;=`<%l6eU9_W!USOt$Ouw$Ox!P$O!P!Q9w!Q#O$O#P;'S$O;'S;=`$g<%lO$OU:Ob!US!mQOt$Ouw$Ox#O$O#P#Z$O#Z#[9w#[#]$O#]#^9w#^#a$O#a#b9w#b#g$O#g#h9w#h#i$O#i#j9w#j#m$O#m#n9w#n;'S$O;'S;=`$g<%lO$OU;][!USOY;WYZ$OZt;Wtu7tuw;Wwx7tx#O;W#O#P8^#P#Q5b#Q;'S;W;'S;=`<R<%lO;WU<UP;=`<%l;WU<[P;=`<%l5bU<fU!US!]QOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OU=PU!US#`QOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OU=jW#rQ!USOt$Ouw$Ox!_$O!_!`>S!`#O$O#P;'S$O;'S;=`$g<%lO$OU>XV!USOt$Ouw$Ox#O$O#P#Q>n#Q;'S$O;'S;=`$g<%lO$OU>uU#qQ!USOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$O~?^O#i~U?eU#sQ!USOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OU@OU!US!dQOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OU@g^!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#o@b#o;'S$O;'S;=`$g<%lO$OUAjU!SQ!USOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$OUBR_!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#UCQ#U#o@b#o;'S$O;'S;=`$g<%lO$OUCV`!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#`@b#`#aDX#a#o@b#o;'S$O;'S;=`$g<%lO$OUD^`!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#g@b#g#hE`#h#o@b#o;'S$O;'S;=`$g<%lO$OUEe`!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#X@b#X#YFg#Y#o@b#o;'S$O;'S;=`$g<%lO$OUFn^!ZQ!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#o@b#o;'S$O;'S;=`$g<%lO$O^Gq^#jW!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#o@b#o;'S$O;'S;=`$g<%lO$O^Ht^#lW!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#o@b#o;'S$O;'S;=`$g<%lO$O^Iw`#kW!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#f@b#f#gJy#g#o@b#o;'S$O;'S;=`$g<%lO$OUKO`!USOt$Ouw$Ox}$O}!O@b!O!Q$O!Q![@b![!_$O!_!`Ac!`#O$O#P#T$O#T#i@b#i#jE`#j#o@b#o;'S$O;'S;=`$g<%lO$OULXUuQ!USOt$Ouw$Ox#O$O#P;'S$O;'S;=`$g<%lO$O~LpO#t~",
tokenizers: [operatorTokenizer, 1, 2, 3, tokenizer, pipeStartsLineTokenizer, new LocalTokenGroup("[~RP!O!PU~ZO#d~~", 11)],
topRules: {"Program":[0,35]},
specialized: [{term: 28, get: (value: any, stack: any) => (specializeKeyword(value, stack) << 1), external: specializeKeyword},{term: 28, get: (value: keyof typeof spec_Identifier) => spec_Identifier[value] || -1}],
tokenPrec: 2589
})

View File

@ -39,18 +39,11 @@ export const parseString = (input: string, from: number, to: number, parser: any
* Parse single-quoted string: 'hello $name\n'
* Supports: interpolation ($var, $(expr)), escape sequences (\n, \$, etc)
*/
const parseSingleQuoteString = (
stringNode: SyntaxNode,
input: string,
from: number,
to: number,
parser: any,
) => {
const parseSingleQuoteString = (stringNode: SyntaxNode, input: string, from: number, to: number, parser: any) => {
let pos = from + 1 // Skip opening '
let fragmentStart = pos
while (pos < to - 1) {
// -1 to skip closing '
while (pos < to - 1) { // -1 to skip closing '
const char = input[pos]
// Escape sequence
@ -122,13 +115,7 @@ const parseSingleQuoteString = (
* Supports: interpolation ($var, $(expr)), nested braces
* Does NOT support: escape sequences (raw content)
*/
const parseCurlyString = (
stringNode: SyntaxNode,
input: string,
from: number,
to: number,
parser: any,
) => {
const parseCurlyString = (stringNode: SyntaxNode, input: string, from: number, to: number, parser: any) => {
let pos = from + 1 // Skip opening {
let fragmentStart = from // Include the opening { in the fragment
let depth = 1
@ -201,11 +188,7 @@ const parseCurlyString = (
* Returns the parsed expression node and the position after the closing )
* pos is position of the opening ( in the full input string
*/
const parseInterpolationExpr = (
input: string,
pos: number,
parser: any,
): { node: SyntaxNode; endPos: number } => {
const parseInterpolationExpr = (input: string, pos: number, parser: any): { node: SyntaxNode, endPos: number } => {
// Find matching closing paren
let depth = 1
let start = pos

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('null', () => {
test('parses null', () => {
expect('null').toMatchTree(`Null null`)

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('bitwise operators - grammar', () => {
test('parses band (bitwise AND)', () => {
expect('5 band 3').toMatchTree(`

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('if/else if/else', () => {
test('parses single line if', () => {
expect(`if y == 1: 'cool' end`).toMatchTree(`
@ -195,6 +197,7 @@ describe('if/else if/else', () => {
`)
})
test('parses function calls in else-if tests', () => {
expect(`if false: true else if var? 'abc': true end`).toMatchTree(`
IfExpr
@ -285,6 +288,7 @@ describe('while', () => {
keyword end`)
})
test('compound expression', () => {
expect(`while a > 0 and b < 100 and c < 1000: true end`).toMatchTree(`
WhileExpr
@ -342,6 +346,7 @@ describe('while', () => {
keyword end`)
})
test('multiline compound expression', () => {
expect(`
while a > 0 and b < 100 and c < 1000:

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('Array destructuring', () => {
test('parses array pattern with two variables', () => {
expect('[ a b ] = [ 1 2 3 4]').toMatchTree(`

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('try/catch/finally/throw', () => {
test('parses try with catch', () => {
expect(`try:

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('single line function blocks', () => {
test('work with no args', () => {
expect(`trap: echo bye bye end`).toMatchTree(`
@ -14,7 +16,8 @@ describe('single line function blocks', () => {
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
test('work with one arg', () => {
@ -32,7 +35,8 @@ describe('single line function blocks', () => {
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
test('work with named args', () => {
@ -52,9 +56,11 @@ describe('single line function blocks', () => {
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
test('work with dot-get', () => {
expect(`signals = [=]; signals.trap 'EXIT': echo bye bye end`).toMatchTree(`
Assign
@ -77,7 +83,8 @@ describe('single line function blocks', () => {
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
})
@ -99,7 +106,8 @@ end
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
test('work with one arg', () => {
@ -120,7 +128,8 @@ end`).toMatchTree(`
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
test('work with named args', () => {
@ -146,9 +155,11 @@ end`).toMatchTree(`
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
test('work with dot-get', () => {
expect(`
signals = [=]
@ -175,7 +186,8 @@ end`).toMatchTree(`
Identifier bye
PositionalArg
Identifier bye
keyword end`)
keyword end`
)
})
})
@ -252,7 +264,8 @@ end`).toMatchTree(`
p:
h1 class=bright style='font-family: helvetica' Heya
h2 man that is (b wild)!
end`).toMatchTree(`
end`)
.toMatchTree(`
FunctionCallWithBlock
FunctionCallOrIdentifier
Identifier p

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('calling functions', () => {
test('call with no args', () => {
expect('tail').toMatchTree(`
@ -92,6 +94,7 @@ describe('calling functions', () => {
`)
})
test('command with arg that is also a command', () => {
expect('tail tail').toMatchTree(`
FunctionCall

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('import', () => {
test('parses single import', () => {
expect(`import str`).toMatchTree(`

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('number literals', () => {
test('binary numbers', () => {
expect('0b110').toMatchTree(`

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('multiline', () => {
test('parses multiline strings', () => {
expect(`'first'\n'second'`).toMatchTree(`

View File

@ -1,4 +1,7 @@
import { expect, describe, test } from 'bun:test'
import { parser } from '../shrimp'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('pipe expressions', () => {
test('simple pipe expression', () => {

View File

@ -1,5 +1,7 @@
import { expect, describe, test } from 'bun:test'
import '../shrimp.grammar' // Importing this so changes cause it to retest!
describe('string interpolation', () => {
test('string with variable interpolation', () => {
expect("'hello $name'").toMatchTree(`
@ -161,7 +163,7 @@ describe('curly strings', () => {
})
describe('double quoted strings', () => {
test('work', () => {
test("work", () => {
expect(`"hello world"`).toMatchTree(`
String
DoubleQuote "hello world"`)

View File

@ -15,7 +15,10 @@ describe('numbers', () => {
test('non-numbers', () => {
expect(`1st`).toMatchToken('Word', '1st')
expect(`1_`).toMatchToken('Word', '1_')
expect(`100.`).toMatchTokens({ type: 'Number', value: '100' }, { type: 'Operator', value: '.' })
expect(`100.`).toMatchTokens(
{ type: 'Number', value: '100' },
{ type: 'Operator', value: '.' },
)
})
test('simple numbers', () => {
@ -127,7 +130,10 @@ describe('identifiers', () => {
expect('dog#pound').toMatchToken('Word', 'dog#pound')
expect('http://website.com').toMatchToken('Word', 'http://website.com')
expect('school$cool').toMatchToken('Identifier', 'school$cool')
expect('EXIT:').toMatchTokens({ type: 'Word', value: 'EXIT' }, { type: 'Colon' })
expect('EXIT:').toMatchTokens(
{ type: 'Word', value: 'EXIT' },
{ type: 'Colon' },
)
expect(`if y == 1: 'cool' end`).toMatchTokens(
{ type: 'Keyword', value: 'if' },
{ type: 'Identifier', value: 'y' },
@ -208,24 +214,18 @@ describe('curly strings', () => {
expect(`{
one
two
three }`).toMatchToken(
'String',
`{
three }`).toMatchToken('String', `{
one
two
three }`,
)
three }`)
})
test('can contain other curlies', () => {
expect(`{ { one }
two
{ three } }`).toMatchToken(
'String',
`{ { one }
{ three } }`).toMatchToken('String', `{ { one }
two
{ three } }`,
)
{ three } }`)
})
test('empty curly string', () => {
@ -304,7 +304,6 @@ describe('keywords', () => {
expect(`catch`).toMatchToken('Keyword', 'catch')
expect(`finally`).toMatchToken('Keyword', 'finally')
expect(`throw`).toMatchToken('Keyword', 'throw')
expect(`not`).toMatchToken('Keyword', 'not')
})
})
@ -408,12 +407,12 @@ f
]`).toMatchTokens(
{ type: 'OpenBracket' },
{ type: 'Identifier', value: 'a' },
{ type: 'Identifier', value: 'b' },
{ type: 'Identifier', value: 'c' },
{ type: 'Identifier', value: 'd' },
{ type: 'Identifier', value: 'e' },
{ type: 'Identifier', value: 'f' },
{ type: 'Identifier', value: "a" },
{ type: 'Identifier', value: "b" },
{ type: 'Identifier', value: "c" },
{ type: 'Identifier', value: "d" },
{ type: 'Identifier', value: "e" },
{ type: 'Identifier', value: "f" },
{ type: 'CloseBracket' },
)
})
@ -506,6 +505,7 @@ f
{ type: 'Identifier', value: 'y' },
)
expect(`if (var? 'abc'): y`).toMatchTokens(
{ type: 'Keyword', value: 'if' },
{ type: 'OpenParen' },
@ -551,25 +551,25 @@ end`).toMatchTokens(
test('dot operator beginning word with slash', () => {
expect(`(basename ./cool)`).toMatchTokens(
{ type: 'OpenParen' },
{ type: 'Identifier', value: 'basename' },
{ type: 'Word', value: './cool' },
{ type: 'CloseParen' },
{ 'type': 'OpenParen' },
{ 'type': 'Identifier', 'value': 'basename' },
{ 'type': 'Word', 'value': './cool' },
{ 'type': 'CloseParen' }
)
})
test('dot word after identifier with space', () => {
expect(`expand-path .git`).toMatchTokens(
{ type: 'Identifier', value: 'expand-path' },
{ type: 'Word', value: '.git' },
{ 'type': 'Identifier', 'value': 'expand-path' },
{ 'type': 'Word', 'value': '.git' },
)
})
test('dot operator after identifier without space', () => {
expect(`config.path`).toMatchTokens(
{ type: 'Identifier', value: 'config' },
{ type: 'Operator', value: '.' },
{ type: 'Identifier', value: 'path' },
{ 'type': 'Identifier', 'value': 'config' },
{ 'type': 'Operator', 'value': '.' },
{ 'type': 'Identifier', 'value': 'path' },
)
})
})
@ -648,7 +648,11 @@ describe('empty and whitespace input', () => {
})
test('only newlines', () => {
expect('\n\n\n').toMatchTokens({ type: 'Newline' }, { type: 'Newline' }, { type: 'Newline' })
expect('\n\n\n').toMatchTokens(
{ type: 'Newline' },
{ type: 'Newline' },
{ type: 'Newline' },
)
})
})
@ -660,14 +664,14 @@ describe('named args', () => {
)
})
test('can have spaces', () => {
test("can have spaces", () => {
expect(`named= arg`).toMatchTokens(
{ type: 'NamedArgPrefix', value: 'named=' },
{ type: 'Identifier', value: 'arg' },
)
})
test('can include numbers', () => {
test("can include numbers", () => {
expect(`named123= arg`).toMatchTokens(
{ type: 'NamedArgPrefix', value: 'named123=' },
{ type: 'Identifier', value: 'arg' },

389
src/parser/tokenizer.ts Normal file
View File

@ -0,0 +1,389 @@
import { ExternalTokenizer, InputStream, Stack } from '@lezer/lr'
import { Identifier, AssignableIdentifier, Word, IdentifierBeforeDot, Do, CurlyString, DotGet, newline, pipeStartsLine } from './shrimp.terms'
// doobie doobie do (we need the `do` keyword to know when we're defining params)
export function specializeKeyword(ident: string) {
return ident === 'do' ? Do : -1
}
// tell the dotGet searcher about builtin globals
export const globals: string[] = []
export const setGlobals = (newGlobals: string[] | Record<string, any>) => {
globals.length = 0
globals.push(...(Array.isArray(newGlobals) ? newGlobals : Object.keys(newGlobals)))
}
// The only chars that can't be words are whitespace, apostrophes, closing parens, and EOF.
export const tokenizer = new ExternalTokenizer(
(input: InputStream, stack: Stack) => {
const ch = getFullCodePoint(input, 0)
// Handle curly strings
if (ch === 123 /* { */) return consumeCurlyString(input, stack)
if (!isWordChar(ch)) return
// Don't consume things that start with digits - let Number token handle it
if (isDigit(ch)) return
// Don't consume things that start with - or + followed by a digit (negative/positive numbers)
if ((ch === 45 /* - */ || ch === 43) /* + */ && isDigit(input.peek(1))) return
const isValidStart = isIdentStart(ch)
const canBeWord = stack.canShift(Word)
// Consume all word characters, tracking if it remains a valid identifier
const { pos, isValidIdentifier, stoppedAtDot } = consumeWordToken(
input,
isValidStart,
canBeWord
)
// Check if we should emit IdentifierBeforeDot for property access
if (stoppedAtDot) {
const dotGetToken = checkForDotGet(input, stack, pos)
if (dotGetToken) {
input.advance(pos)
input.acceptToken(dotGetToken)
} else {
// Not in scope - continue consuming the dot as part of the word
const afterDot = consumeRestOfWord(input, pos + 1, canBeWord)
input.advance(afterDot)
input.acceptToken(Word)
}
return
}
// Advance past the token we consumed
input.advance(pos)
// Choose which token to emit
if (isValidIdentifier) {
const token = chooseIdentifierToken(input, stack)
input.acceptToken(token)
} else {
input.acceptToken(Word)
}
},
{ contextual: true }
)
// Build identifier text from input stream, handling surrogate pairs for emoji
const buildIdentifierText = (input: InputStream, length: number): string => {
let text = ''
for (let i = 0; i < length; i++) {
const charCode = input.peek(i)
if (charCode === -1) break
// Handle surrogate pairs for emoji (UTF-16 encoding)
if (charCode >= 0xd800 && charCode <= 0xdbff && i + 1 < length) {
const low = input.peek(i + 1)
if (low >= 0xdc00 && low <= 0xdfff) {
text += String.fromCharCode(charCode, low)
i++ // Skip the low surrogate
continue
}
}
text += String.fromCharCode(charCode)
}
return text
}
// Consume word characters, tracking if it remains a valid identifier
// Returns the position after consuming, whether it's a valid identifier, and if we stopped at a dot
const consumeWordToken = (
input: InputStream,
isValidStart: boolean,
canBeWord: boolean
): { pos: number; isValidIdentifier: boolean; stoppedAtDot: boolean } => {
let pos = getCharSize(getFullCodePoint(input, 0))
let isValidIdentifier = isValidStart
let stoppedAtDot = false
while (true) {
const ch = getFullCodePoint(input, pos)
// Stop at dot if we have a valid identifier (might be property access)
if (ch === 46 /* . */ && isValidIdentifier) {
stoppedAtDot = true
break
}
// Stop if we hit a non-word character
if (!isWordChar(ch)) break
// Context-aware termination: semicolon/colon can end a word if followed by whitespace
// This allows `hello; 2` to parse correctly while `hello;world` stays as one word
if (canBeWord && (ch === 59 /* ; */ || ch === 58) /* : */) {
const nextCh = getFullCodePoint(input, pos + 1)
if (!isWordChar(nextCh)) break
}
// Track identifier validity: must be lowercase, digit, dash, or emoji/unicode
if (!isIdentChar(ch)) {
if (!canBeWord) break
isValidIdentifier = false
}
pos += getCharSize(ch)
}
return { pos, isValidIdentifier, stoppedAtDot }
}
// Consume the rest of a word after we've decided not to treat a dot as DotGet
// Used when we have "file.txt" - we already consumed "file", now consume ".txt"
const consumeRestOfWord = (input: InputStream, startPos: number, canBeWord: boolean): number => {
let pos = startPos
while (true) {
const ch = getFullCodePoint(input, pos)
// Stop if we hit a non-word character
if (!isWordChar(ch)) break
// Context-aware termination for semicolon/colon
if (canBeWord && (ch === 59 /* ; */ || ch === 58) /* : */) {
const nextCh = getFullCodePoint(input, pos + 1)
if (!isWordChar(nextCh)) break
}
pos += getCharSize(ch)
}
return pos
}
// Consumes { curly strings } and tracks braces so you can { have { braces { inside { braces } } }
const consumeCurlyString = (input: InputStream, stack: Stack) => {
if (!stack.canShift(CurlyString)) return
let depth = 0
let pos = 0
while (true) {
const ch = input.peek(pos)
if (ch < 0) return // EOF - invalid
if (ch === 123) depth++ // {
else if (ch === 125) { // }
depth--
if (depth === 0) {
pos++ // consume final }
break
}
}
pos++
}
input.acceptToken(CurlyString, pos)
}
// Check if this identifier is in scope (for property access detection)
// Returns IdentifierBeforeDot token if in scope, null otherwise
const checkForDotGet = (input: InputStream, stack: Stack, pos: number): number | null => {
const identifierText = buildIdentifierText(input, pos)
const context = stack.context as { scope: { has(name: string): boolean } } | undefined
// Check if identifier is in scope (lexical scope or globals)
const inScope = context?.scope.has(identifierText) || globals.includes(identifierText)
// property access
if (inScope) return IdentifierBeforeDot
// Not in scope - check if we're inside a DotGet chain
// Inside the @skip {} block where DotGet is defined, Word cannot be shifted
// but Identifier can be. This tells us we're at the RHS of a DotGet.
const canShiftIdentifier = stack.canShift(Identifier)
const canShiftWord = stack.canShift(Word)
const inDotGetChain = canShiftIdentifier && !canShiftWord
// continue if we're inside a DotGet
return inDotGetChain ? IdentifierBeforeDot : null
}
// Decide between AssignableIdentifier and Identifier using grammar state + peek-ahead
const chooseIdentifierToken = (input: InputStream, stack: Stack): number => {
const canAssignable = stack.canShift(AssignableIdentifier)
const canRegular = stack.canShift(Identifier)
// Only one option is valid - use it
if (canAssignable && !canRegular) return AssignableIdentifier
if (canRegular && !canAssignable) return Identifier
// Both possible (ambiguous context) - peek ahead for '=' to disambiguate
// This happens at statement start where both `x = 5` (assign) and `echo x` (call) are valid
let peekPos = 0
while (true) {
const ch = getFullCodePoint(input, peekPos)
if (isWhiteSpace(ch)) {
peekPos += getCharSize(ch)
} else {
break
}
}
const nextCh = getFullCodePoint(input, peekPos)
const nextCh2 = getFullCodePoint(input, peekPos + 1)
const nextCh3 = getFullCodePoint(input, peekPos + 2)
// Check for ??= (three-character compound operator)
if (nextCh === 63 /* ? */ && nextCh2 === 63 /* ? */ && nextCh3 === 61 /* = */) {
const charAfterOp = getFullCodePoint(input, peekPos + 3)
if (isWhiteSpace(charAfterOp) || charAfterOp === -1 /* EOF */) {
return AssignableIdentifier
}
}
// Check for compound assignment operators: +=, -=, *=, /=, %=
if (
[43 /* + */, 45 /* - */, 42 /* * */, 47 /* / */, 37 /* % */].includes(nextCh) &&
nextCh2 === 61 /* = */
) {
// Found compound operator, check if it's followed by whitespace
const charAfterOp = getFullCodePoint(input, peekPos + 2)
if (isWhiteSpace(charAfterOp) || charAfterOp === -1 /* EOF */) {
return AssignableIdentifier
}
}
if (nextCh === 61 /* = */) {
// Found '=', but check if it's followed by whitespace
// If '=' is followed by non-whitespace (like '=cool*'), it won't be tokenized as Eq
// In that case, this should be Identifier (for function call), not AssignableIdentifier
const charAfterEquals = getFullCodePoint(input, peekPos + 1)
if (isWhiteSpace(charAfterEquals) || charAfterEquals === -1 /* EOF */) {
return AssignableIdentifier
}
}
return Identifier
}
// Character classification helpers
export const isIdentStart = (ch: number): boolean => {
return isLowercaseLetter(ch) || isEmojiOrUnicode(ch)
}
export const isIdentChar = (ch: number): boolean => {
return isLowercaseLetter(ch) || isDigit(ch) || ch === 45 /* - */ || ch === 63 /* ? */ || isEmojiOrUnicode(ch)
}
const isWhiteSpace = (ch: number): boolean => {
return ch === 32 /* space */ || ch === 9 /* tab */ || ch === 13 /* \r */
}
const isWordChar = (ch: number): boolean => {
return (
!isWhiteSpace(ch) &&
ch !== 10 /* \n */ &&
ch !== 41 /* ) */ &&
ch !== 93 /* ] */ &&
ch !== -1 /* EOF */
)
}
const isLowercaseLetter = (ch: number): boolean => {
return ch >= 97 && ch <= 122 // a-z
}
const isDigit = (ch: number): boolean => {
return ch >= 48 && ch <= 57 // 0-9
}
const getFullCodePoint = (input: InputStream, pos: number): number => {
const ch = input.peek(pos)
// Check if this is a high surrogate (0xD800-0xDBFF)
if (ch >= 0xd800 && ch <= 0xdbff) {
const low = input.peek(pos + 1)
// Check if next is low surrogate (0xDC00-0xDFFF)
if (low >= 0xdc00 && low <= 0xdfff) {
// Combine surrogate pair into full code point
return 0x10000 + ((ch & 0x3ff) << 10) + (low & 0x3ff)
}
}
return ch
}
const isEmojiOrUnicode = (ch: number): boolean => {
return (
// Basic Emoticons
(ch >= 0x1f600 && ch <= 0x1f64f) ||
// Miscellaneous Symbols and Pictographs
(ch >= 0x1f300 && ch <= 0x1f5ff) ||
// Transport and Map Symbols
(ch >= 0x1f680 && ch <= 0x1f6ff) ||
// Regional Indicator Symbols (flags)
(ch >= 0x1f1e6 && ch <= 0x1f1ff) ||
// Miscellaneous Symbols (hearts, stars, weather)
(ch >= 0x2600 && ch <= 0x26ff) ||
// Dingbats (scissors, pencils, etc)
(ch >= 0x2700 && ch <= 0x27bf) ||
// Supplemental Symbols and Pictographs (newer emojis)
(ch >= 0x1f900 && ch <= 0x1f9ff) ||
// Symbols and Pictographs Extended-A (newest emojis)
(ch >= 0x1fa70 && ch <= 0x1faff) ||
// Various Asian Characters with emoji presentation
(ch >= 0x1f018 && ch <= 0x1f270) ||
// Variation Selectors (for emoji presentation)
(ch >= 0xfe00 && ch <= 0xfe0f) ||
// Additional miscellaneous items
(ch >= 0x238c && ch <= 0x2454) ||
// Combining Diacritical Marks for Symbols
(ch >= 0x20d0 && ch <= 0x20ff) ||
// Latin-1 Supplement (includes ², ³, ¹ and other special chars)
(ch >= 0x00a0 && ch <= 0x00ff) ||
// Greek and Coptic (U+0370-U+03FF)
(ch >= 0x0370 && ch <= 0x03ff) ||
// Mathematical Alphanumeric Symbols (U+1D400-U+1D7FF)
(ch >= 0x1d400 && ch <= 0x1d7ff) ||
// Mathematical Operators (U+2200-U+22FF)
(ch >= 0x2200 && ch <= 0x22ff) ||
// Superscripts and Subscripts (U+2070-U+209F)
(ch >= 0x2070 && ch <= 0x209f) ||
// Arrows (U+2190-U+21FF)
(ch >= 0x2190 && ch <= 0x21ff) ||
// Hiragana (U+3040-U+309F)
(ch >= 0x3040 && ch <= 0x309f) ||
// Katakana (U+30A0-U+30FF)
(ch >= 0x30a0 && ch <= 0x30ff) ||
// CJK Unified Ideographs (U+4E00-U+9FFF)
(ch >= 0x4e00 && ch <= 0x9fff)
)
}
const getCharSize = (ch: number) => (ch > 0xffff ? 2 : 1) // emoji takes 2 UTF-16 code units
export const pipeStartsLineTokenizer = new ExternalTokenizer((input: InputStream, stack: Stack) => {
const ch = input.peek(0)
if (ch !== 10 /* \n */) return
// ignore whitespace
let offset = 1
let lastNewlineOffset = 0
while (true) {
const ch = input.peek(offset)
if (ch === 10 /* \n */) {
lastNewlineOffset = offset
offset++
} else if (isWhiteSpace(ch)) {
offset++
} else {
break
}
}
// look for pipe after skipping empty lines
if (input.peek(offset) === 124 /* | */) {
input.advance(lastNewlineOffset + 1)
input.acceptToken(pipeStartsLine)
} else {
input.advance(1)
input.acceptToken(newline)
}
})

View File

@ -1,10 +1,10 @@
import { isDebug } from '#utils/utils'
const DEBUG = process.env.DEBUG || false
export type Token = {
type: TokenType
value?: string
from: number
to: number
value?: string,
from: number,
to: number,
}
export enum TokenType {
@ -36,16 +36,10 @@ export enum TokenType {
const valueTokens = new Set([
TokenType.Comment,
TokenType.Keyword,
TokenType.Operator,
TokenType.Identifier,
TokenType.Word,
TokenType.NamedArgPrefix,
TokenType.Boolean,
TokenType.Number,
TokenType.String,
TokenType.Regex,
TokenType.Underscore,
TokenType.Keyword, TokenType.Operator,
TokenType.Identifier, TokenType.Word, TokenType.NamedArgPrefix,
TokenType.Boolean, TokenType.Number, TokenType.String, TokenType.Regex,
TokenType.Underscore
])
const operators = new Set([
@ -110,12 +104,11 @@ const keywords = new Set([
'catch',
'finally',
'throw',
'not',
])
// helper
function c(strings: TemplateStringsArray, ...values: any[]) {
return strings.reduce((result, str, i) => result + str + (values[i] ?? ''), '').charCodeAt(0)
return strings.reduce((result, str, i) => result + str + (values[i] ?? ""), "").charCodeAt(0)
}
function s(c: number): string {
@ -161,19 +154,13 @@ export class Scanner {
to ??= this.pos - getCharSize(this.char)
if (to < from) to = from
this.tokens.push(
Object.assign(
{},
{
this.tokens.push(Object.assign({}, {
type,
from,
to,
},
valueTokens.has(type) ? { value: this.input.slice(from, to) } : {}
)
)
}, valueTokens.has(type) ? { value: this.input.slice(from, to) } : {}))
if (isDebug()) {
if (DEBUG) {
const tok = this.tokens.at(-1)
console.log(`≫ PUSH(${from},${to})`, TokenType[tok?.type || 0], '—', tok?.value)
}
@ -250,7 +237,8 @@ export class Scanner {
}
if (char === c`\n`) {
if (this.inParen === 0 && this.inBracket === 0) this.pushChar(TokenType.Newline)
if (this.inParen === 0 && this.inBracket === 0)
this.pushChar(TokenType.Newline)
this.next()
continue
}
@ -277,20 +265,16 @@ export class Scanner {
switch (this.char) {
case c`(`:
this.inParen++
this.pushChar(TokenType.OpenParen)
break
this.pushChar(TokenType.OpenParen); break
case c`)`:
this.inParen--
this.pushChar(TokenType.CloseParen)
break
this.pushChar(TokenType.CloseParen); break
case c`[`:
this.inBracket++
this.pushChar(TokenType.OpenBracket)
break
this.pushChar(TokenType.OpenBracket); break
case c`]`:
this.inBracket--
this.pushChar(TokenType.CloseBracket)
break
this.pushChar(TokenType.CloseBracket); break
}
this.next()
}
@ -354,14 +338,29 @@ export class Scanner {
const word = this.input.slice(this.start, this.pos - getCharSize(this.char))
// classify the token based on what we read
if (word === '_') this.push(TokenType.Underscore)
else if (word === 'null') this.push(TokenType.Null)
else if (word === 'true' || word === 'false') this.push(TokenType.Boolean)
else if (isKeyword(word)) this.push(TokenType.Keyword)
else if (isOperator(word)) this.push(TokenType.Operator)
else if (isIdentifer(word)) this.push(TokenType.Identifier)
else if (word.endsWith('=')) this.push(TokenType.NamedArgPrefix)
else this.push(TokenType.Word)
if (word === '_')
this.push(TokenType.Underscore)
else if (word === 'null')
this.push(TokenType.Null)
else if (word === 'true' || word === 'false')
this.push(TokenType.Boolean)
else if (isKeyword(word))
this.push(TokenType.Keyword)
else if (isOperator(word))
this.push(TokenType.Operator)
else if (isIdentifer(word))
this.push(TokenType.Identifier)
else if (word.endsWith('='))
this.push(TokenType.NamedArgPrefix)
else
this.push(TokenType.Word)
}
readNumber() {
@ -394,7 +393,8 @@ export class Scanner {
this.next() // skip /
// read regex flags
while (this.char > 0 && isIdentStart(this.char)) this.next()
while (this.char > 0 && isIdentStart(this.char))
this.next()
// validate regex
const to = this.pos - getCharSize(this.char)
@ -421,29 +421,30 @@ export class Scanner {
}
canBeDotGet(lastToken?: Token): boolean {
return (
!this.prevIsWhitespace &&
!!lastToken &&
return !this.prevIsWhitespace && !!lastToken &&
(lastToken.type === TokenType.Identifier ||
lastToken.type === TokenType.Number ||
lastToken.type === TokenType.CloseParen ||
lastToken.type === TokenType.CloseBracket)
)
}
}
const isNumber = (word: string): boolean => {
// regular number
if (/^[+-]?\d+(_?\d+)*(\.(\d+(_?\d+)*))?$/.test(word)) return true
if (/^[+-]?\d+(_?\d+)*(\.(\d+(_?\d+)*))?$/.test(word))
return true
// binary
if (/^[+-]?0b[01]+(_?[01]+)*(\.[01](_?[01]*))?$/.test(word)) return true
if (/^[+-]?0b[01]+(_?[01]+)*(\.[01](_?[01]*))?$/.test(word))
return true
// octal
if (/^[+-]?0o[0-7]+(_?[0-7]+)*(\.[0-7](_?[0-7]*))?$/.test(word)) return true
if (/^[+-]?0o[0-7]+(_?[0-7]+)*(\.[0-7](_?[0-7]*))?$/.test(word))
return true
// hex
if (/^[+-]?0x[0-9a-f]+([0-9a-f]_?[0-9a-f]+)*(\.([0-9a-f]_?[0-9a-f]*))?$/i.test(word)) return true
if (/^[+-]?0x[0-9a-f]+([0-9a-f]_?[0-9a-f]+)*(\.([0-9a-f]_?[0-9a-f]*))?$/i.test(word))
return true
return false
}
@ -459,26 +460,26 @@ const isIdentifer = (s: string): boolean => {
chars.push(out)
}
if (chars.length === 1) return isIdentStart(chars[0]!)
else if (chars.length === 2) return isIdentStart(chars[0]!) && isIdentEnd(chars[1]!)
if (chars.length === 1)
return isIdentStart(chars[0]!)
else if (chars.length === 2)
return isIdentStart(chars[0]!) && isIdentEnd(chars[1]!)
else
return (
isIdentStart(chars[0]!) &&
return isIdentStart(chars[0]!) &&
chars.slice(1, chars.length - 1).every(isIdentChar) &&
isIdentEnd(chars.at(-1)!)
)
}
const isStringDelim = (ch: number): boolean => {
return ch === c`'` || ch === c`"`
}
export const isIdentStart = (char: number | string): boolean => {
const isIdentStart = (char: number | string): boolean => {
let ch = typeof char === 'string' ? char.charCodeAt(0) : char
return isLowercaseLetter(ch) || isEmojiOrUnicode(ch) || ch === 36 /* $ */
}
export const isIdentChar = (char: number | string): boolean => {
const isIdentChar = (char: number | string): boolean => {
let ch = typeof char === 'string' ? char.charCodeAt(0) : char
return isIdentStart(ch) || isDigit(ch) || ch === 45 /* - */ || ch === 63 /* ? */
}
@ -496,14 +497,9 @@ const isDigit = (ch: number): boolean => {
}
const isWhitespace = (ch: number): boolean => {
return (
ch === 32 /* space */ ||
ch === 9 /* tab */ ||
ch === 13 /* \r */ ||
ch === 10 /* \n */ ||
ch === -1 ||
ch === 0
) /* EOF */
return ch === 32 /* space */ || ch === 9 /* tab */ ||
ch === 13 /* \r */ || ch === 10 /* \n */ ||
ch === -1 || ch === 0 /* EOF */
}
const isWordChar = (ch: number): boolean => {
@ -530,7 +526,8 @@ const isBracket = (char: number): boolean => {
return char === c`(` || char === c`)` || char === c`[` || char === c`]`
}
const getCharSize = (ch: number) => (ch > 0xffff ? 2 : 1) // emoji takes 2 UTF-16 code units
const getCharSize = (ch: number) =>
(ch > 0xffff ? 2 : 1) // emoji takes 2 UTF-16 code units
const getFullCodePoint = (input: string, pos: number): number => {
const ch = input[pos]?.charCodeAt(0) || 0

View File

@ -1,12 +0,0 @@
export const date = {
now: () => Date.now(),
year: (time: number) => new Date(time).getFullYear(),
month: (time: number) => new Date(time).getMonth(),
date: (time: number) => new Date(time).getDate(),
hour: (time: number) => new Date(time).getHours(),
minute: (time: number) => new Date(time).getMinutes(),
second: (time: number) => new Date(time).getSeconds(),
ms: (time: number) => new Date(time).getMilliseconds(),
new: (year: number, month: number, day: number, hour = 0, minute = 0, second = 0, ms = 0) =>
new Date(year, month, day, hour, minute, second, ms).getTime(),
}

View File

@ -3,11 +3,9 @@ import { type Value, toString } from 'reefvm'
export const dict = {
keys: (dict: Record<string, any>) => Object.keys(dict),
values: (dict: Record<string, any>) => Object.values(dict),
entries: (dict: Record<string, any>) =>
Object.entries(dict).map(([k, v]) => ({ key: k, value: v })),
entries: (dict: Record<string, any>) => Object.entries(dict).map(([k, v]) => ({ key: k, value: v })),
'has?': (dict: Record<string, any>, key: string) => key in dict,
get: (dict: Record<string, any>, key: string, defaultValue: any = null) =>
dict[key] ?? defaultValue,
get: (dict: Record<string, any>, key: string, defaultValue: any = null) => dict[key] ?? defaultValue,
set: (dict: Value, key: Value, value: Value) => {
const map = dict.value as Map<string, Value>
map.set(toString(key), value)

View File

@ -1,27 +1,17 @@
import { join, resolve, basename, dirname, extname } from 'path'
import {
readdirSync,
mkdirSync,
rmdirSync,
readFileSync,
writeFileSync,
appendFileSync,
rmSync,
copyFileSync,
statSync,
lstatSync,
chmodSync,
symlinkSync,
readlinkSync,
watch,
} from 'fs'
readdirSync, mkdirSync, rmdirSync,
readFileSync, writeFileSync, appendFileSync,
rmSync, copyFileSync,
statSync, lstatSync, chmodSync, symlinkSync, readlinkSync,
watch
} from "fs"
export const fs = {
// Directory operations
ls: (path: string) => readdirSync(path),
mkdir: (path: string) => mkdirSync(path, { recursive: true }),
rmdir: (path: string) =>
rmdirSync(path === '/' || path === '' ? '/tmp/*' : path, { recursive: true }),
rmdir: (path: string) => rmdirSync(path === '/' || path === '' ? '/tmp/*' : path, { recursive: true }),
pwd: () => process.cwd(),
cd: (path: string) => process.chdir(path),
@ -68,50 +58,39 @@ export const fs = {
} catch {
return {}
}
},
'exists?': (path: string) => {
try {
statSync(path)
return true
} catch {
}
catch {
return false
}
},
'file?': (path: string) => {
try {
return statSync(path).isFile()
} catch {
return false
}
try { return statSync(path).isFile() }
catch { return false }
},
'dir?': (path: string) => {
try {
return statSync(path).isDirectory()
} catch {
return false
}
try { return statSync(path).isDirectory() }
catch { return false }
},
'symlink?': (path: string) => {
try {
return lstatSync(path).isSymbolicLink()
} catch {
return false
}
try { return lstatSync(path).isSymbolicLink() }
catch { return false }
},
'exec?': (path: string) => {
try {
const stats = statSync(path)
return !!(stats.mode & 0o111)
} catch {
return false
}
catch { return false }
},
size: (path: string) => {
try {
return statSync(path).size
} catch {
return 0
}
try { return statSync(path).size }
catch { return 0 }
},
// Permissions
@ -135,11 +114,14 @@ export const fs = {
return readdirSync(dir)
.filter((f) => f.endsWith(ext))
.map((f) => join(dir, f))
},
watch: (path: string, callback: Function) =>
watch(path, (event, filename) => callback(event, filename)),
}
; (fs as any).cat = fs.read
; (fs as any).mv = fs.move
; (fs as any).cp = fs.copy

View File

@ -2,15 +2,10 @@
import { join, resolve } from 'path'
import {
type Value,
type VM,
toValue,
extractParamInfo,
isWrapped,
getOriginalFunction,
type Value, type VM, toValue,
extractParamInfo, isWrapped, getOriginalFunction,
} from 'reefvm'
import { date } from './date'
import { dict } from './dict'
import { fs } from './fs'
import { json } from './json'
@ -18,11 +13,8 @@ import { load } from './load'
import { list } from './list'
import { math } from './math'
import { str } from './str'
import { types } from './types'
import { runningInBrowser } from '#utils/utils'
export const globals: Record<string, any> = {
date,
export const globals = {
dict,
fs,
json,
@ -32,19 +24,7 @@ export const globals: Record<string, any> = {
str,
// shrimp runtime info
$: runningInBrowser
? {
args: [],
argv: [],
env: {},
pid: 0,
cwd: '',
script: {
name: '',
path: '.',
},
}
: {
$: {
args: Bun.argv.slice(3),
argv: Bun.argv.slice(1),
env: process.env,
@ -52,18 +32,17 @@ export const globals: Record<string, any> = {
cwd: process.env.PWD,
script: {
name: Bun.argv[2] || '(shrimp)',
path: resolve(join('.', Bun.argv[2] ?? '')),
path: resolve(join('.', Bun.argv[2] ?? ''))
},
},
// hello
echo: (...args: any[]) => {
console.log(
...args.map((a) => {
console.log(...args.map(a => {
const v = toValue(a)
return ['array', 'dict'].includes(v.type) ? formatValue(v, true) : v.value
})
)
}))
return toValue(null)
},
@ -82,10 +61,11 @@ export const globals: Record<string, any> = {
},
ref: (fn: Function) => fn,
import: function (this: VM, atNamed: Record<any, string | string[]> = {}, ...idents: string[]) {
const onlyArray = Array.isArray(atNamed.only) ? atNamed.only : [atNamed.only].filter((a) => a)
const onlyArray = Array.isArray(atNamed.only) ? atNamed.only : [atNamed.only].filter(a => a)
const only = new Set(onlyArray)
const wantsOnly = only.size > 0
for (const ident of idents) {
const module = this.get(ident)
@ -104,9 +84,20 @@ export const globals: Record<string, any> = {
exit: (num: number) => process.exit(num ?? 0),
// type predicates
'string?': (v: any) => toValue(v).type === 'string',
'number?': (v: any) => toValue(v).type === 'number',
'boolean?': (v: any) => toValue(v).type === 'boolean',
'array?': (v: any) => toValue(v).type === 'array',
'dict?': (v: any) => toValue(v).type === 'dict',
'function?': (v: any) => {
const t = toValue(v).type
return t === 'function' || t === 'native'
},
'null?': (v: any) => toValue(v).type === 'null',
'some?': (v: any) => toValue(v).type !== 'null',
// boolean/logic
not: (v: any) => !v,
bnot: (n: number) => ~(n | 0),
// utilities
@ -118,13 +109,9 @@ export const globals: Record<string, any> = {
length: (v: any) => {
const value = toValue(v)
switch (value.type) {
case 'string':
case 'array':
return value.value.length
case 'dict':
return value.value.size
default:
throw new Error(`length: expected string, array, or dict, got ${value.type}`)
case 'string': case 'array': return value.value.length
case 'dict': return value.value.size
default: throw new Error(`length: expected string, array, or dict, got ${value.type}`)
}
},
at: (collection: any, index: number | string) => {
@ -132,9 +119,7 @@ export const globals: Record<string, any> = {
if (value.type === 'string' || value.type === 'array') {
const idx = typeof index === 'number' ? index : parseInt(index as string)
if (idx < 0 || idx >= value.value.length) {
throw new Error(
`at: index ${idx} out of bounds for ${value.type} of length ${value.value.length}`
)
throw new Error(`at: index ${idx} out of bounds for ${value.type} of length ${value.value.length}`)
}
return value.value[idx]
} else if (value.type === 'dict') {
@ -161,8 +146,7 @@ export const globals: Record<string, any> = {
'empty?': (v: any) => {
const value = toValue(v)
switch (value.type) {
case 'string':
case 'array':
case 'string': case 'array':
return value.value.length === 0
case 'dict':
return value.value.size === 0
@ -176,6 +160,7 @@ export const globals: Record<string, any> = {
for (const value of list) await cb(value)
return list
},
}
export const colors = {
@ -188,7 +173,7 @@ export const colors = {
red: '\x1b[31m',
blue: '\x1b[34m',
magenta: '\x1b[35m',
pink: '\x1b[38;2;255;105;180m',
pink: '\x1b[38;2;255;105;180m'
}
export function formatValue(value: Value, inner = false): string {
@ -202,15 +187,15 @@ export function formatValue(value: Value, inner = false): string {
case 'null':
return `${colors.dim}null${colors.reset}`
case 'array': {
const items = value.value.map((x) => formatValue(x, true)).join(' ')
const items = value.value.map(x => formatValue(x, true)).join(' ')
return `${colors.blue}[${colors.reset}${items}${colors.blue}]${colors.reset}`
}
case 'dict': {
const entries = Array.from(value.value.entries())
.reverse()
const entries = Array.from(value.value.entries()).reverse()
.map(([k, v]) => `${k.trim()}${colors.blue}=${colors.reset}${formatValue(v, true)}`)
.join(' ')
if (entries.length === 0) return `${colors.blue}[=]${colors.reset}`
if (entries.length === 0)
return `${colors.blue}[=]${colors.reset}`
return `${colors.blue}[${colors.reset}${entries}${colors.blue}]${colors.reset}`
}
case 'function': {
@ -228,6 +213,3 @@ export function formatValue(value: Value, inner = false): string {
return String(value)
}
}
// add types functions to top-level namespace
for (const [key, value] of Object.entries(types)) globals[key] = value

View File

@ -2,5 +2,6 @@ export const json = {
encode: (s: any) => JSON.stringify(s),
decode: (s: string) => JSON.parse(s),
}
; (json as any).parse = json.decode
; (json as any).stringify = json.encode

View File

@ -46,7 +46,7 @@ export const list = {
},
'all?': async (list: any[], cb: Function) => {
for (const value of list) {
if (!(await cb(value))) return false
if (!await cb(value)) return false
}
return true
},
@ -131,7 +131,7 @@ export const list = {
}
return [truthy, falsy]
},
compact: (list: any[]) => list.filter((x) => x != null),
compact: (list: any[]) => list.filter(x => x != null),
'group-by': async (list: any[], cb: Function) => {
const groups: Record<string, any[]> = {}
for (const value of list) {
@ -143,6 +143,7 @@ export const list = {
},
}
// raw functions deal directly in Value types, meaning we can modify collection
// careful - they MUST return a Value!
; (list.splice as any).raw = true

View File

@ -20,7 +20,8 @@ export const load = async function (this: VM, path: string): Promise<Record<stri
await this.continue()
const module: Record<string, Value> = {}
for (const [name, value] of this.scope.locals.entries()) module[name] = value
for (const [name, value] of this.scope.locals.entries())
module[name] = value
this.scope = scope
this.pc = pc

View File

@ -16,10 +16,7 @@ export const math = {
if (n < 0) throw new Error(`sqrt: cannot take square root of negative number ${n}`)
return Math.sqrt(n)
},
random: (min = 0, max = 1) => {
if (min === 0 && max === 1) return Math.random()
return Math.floor(Math.random() * (max - min + 1)) + min
},
random: () => Math.random(),
clamp: (n: number, min: number, max: number) => {
if (min > max) throw new Error(`clamp: min (${min}) must be less than or equal to max (${max})`)
return Math.min(Math.max(n, min), max)

View File

@ -17,33 +17,17 @@ export const str = {
'last-index-of': (str: string, search: string) => String(str ?? '').lastIndexOf(search),
// transformations
replace: (str: string, search: string, replacement: string) =>
String(str ?? '').replace(search, replacement),
'replace-all': (str: string, search: string, replacement: string) =>
String(str ?? '').replaceAll(search, replacement),
slice: (str: string, start: number, end?: number | null) =>
String(str ?? '').slice(start, end ?? undefined),
substring: (str: string, start: number, end?: number | null) =>
String(str ?? '').substring(start, end ?? undefined),
replace: (str: string, search: string, replacement: string) => String(str ?? '').replace(search, replacement),
'replace-all': (str: string, search: string, replacement: string) => String(str ?? '').replaceAll(search, replacement),
slice: (str: string, start: number, end?: number | null) => String(str ?? '').slice(start, end ?? undefined),
substring: (str: string, start: number, end?: number | null) => String(str ?? '').substring(start, end ?? undefined),
repeat: (str: string, count: number) => {
if (count < 0) throw new Error(`repeat: count must be non-negative, got ${count}`)
if (!Number.isInteger(count)) throw new Error(`repeat: count must be an integer, got ${count}`)
return String(str ?? '').repeat(count)
},
'pad-start': (str: string, length: number, pad: string = ' ') =>
String(str ?? '').padStart(length, pad),
'pad-end': (str: string, length: number, pad: string = ' ') =>
String(str ?? '').padEnd(length, pad),
capitalize: (str: string) => {
const s = String(str ?? '')
return s.charAt(0).toUpperCase() + s.slice(1).toLowerCase()
},
titlecase: (s: string) => {
return String(s ?? '')
.split(' ')
.map(str.capitalize)
.join(' ')
},
'pad-start': (str: string, length: number, pad: string = ' ') => String(str ?? '').padStart(length, pad),
'pad-end': (str: string, length: number, pad: string = ' ') => String(str ?? '').padEnd(length, pad),
lines: (str: string) => String(str ?? '').split('\n'),
chars: (str: string) => String(str ?? '').split(''),

View File

@ -1,170 +0,0 @@
import { expect, describe, test } from 'bun:test'
describe('date', () => {
test('date.now returns current timestamp', () => {
expect(`date.now | number?`).toEvaluateTo(true)
expect(`(date.now) > 1577836800000`).toEvaluateTo(true)
})
test('date.new creates timestamp from components', () => {
expect(`
t = date.new 2024 0 1 12 0 0 500
[
(date.year t)
(date.month t)
(date.date t)
(date.hour t)
(date.minute t)
(date.second t)
(date.ms t)
]
`).toEvaluateTo([2024, 0, 1, 12, 0, 0, 500])
})
test('date.new with minimal arguments', () => {
expect(`
t = date.new 2024 5 15
[
(date.year t)
(date.month t)
(date.date t)
(date.hour t)
(date.minute t)
(date.second t)
(date.ms t)
]
`).toEvaluateTo([2024, 5, 15, 0, 0, 0, 0])
})
test('date.year extracts year', () => {
expect(`
t = date.new 2024 0 1
date.year t
`).toEvaluateTo(2024)
expect(`
t = date.new 1999 11 31
date.year t
`).toEvaluateTo(1999)
})
test('date.month extracts month (0-indexed)', () => {
// January = 0, December = 11
expect(`
jan = date.new 2024 0 1
dec = date.new 2024 11 31
[(date.month jan) (date.month dec)]
`).toEvaluateTo([0, 11])
})
test('date.date extracts day of month', () => {
expect(`
t = date.new 2024 5 15
date.date t
`).toEvaluateTo(15)
expect(`
date.new 2024 0 1 | date.date
`).toEvaluateTo(1)
})
test('date.hour extracts hour', () => {
expect(`
t = date.new 2024 0 1 14 30 45
date.hour t
`).toEvaluateTo(14)
expect(`
t = date.new 2024 0 1 0 0 0
date.hour t
`).toEvaluateTo(0)
})
test('date.minute extracts minute', () => {
expect(`
t = date.new 2024 0 1 14 30 45
date.minute t
`).toEvaluateTo(30)
})
test('date.second extracts second', () => {
expect(`
t = date.new 2024 0 1 14 30 45
date.second t
`).toEvaluateTo(45)
})
test('date.ms extracts milliseconds', () => {
expect(`
t = date.new 2024 0 1 14 30 45 250
date.ms t
`).toEvaluateTo(250)
})
test('round-trip: create and extract components', () => {
expect(`
t = date.new 2024 6 4 15 30 45 123
year = date.year t
month = date.month t
day = date.date t
hour = date.hour t
min = date.minute t
sec = date.second t
ms = date.ms t
[year month day hour min sec ms]
`).toEvaluateTo([2024, 6, 4, 15, 30, 45, 123])
})
test('edge cases - midnight', () => {
expect(`
t = date.new 2024 0 1 0 0 0 0
[
(date.hour t)
(date.minute t)
(date.second t)
(date.ms t)
]
`).toEvaluateTo([0, 0, 0, 0])
})
test('edge cases - end of day', () => {
expect(`
t = date.new 2024 0 1 23 59 59 999
[
(date.hour t)
(date.minute t)
(date.second t)
(date.ms t)
]
`).toEvaluateTo([23, 59, 59, 999])
})
test('edge cases - leap year', () => {
expect(`
t = date.new 2024 1 29
[
(date.year t)
(date.month t)
(date.date t)
]
`).toEvaluateTo([2024, 1, 29])
})
test('combining date functions with arithmetic', () => {
expect(`
t = date.new 2024 5 15 10 30 0
next-hour = date.new 2024 5 15 11 30 0
(date.hour next-hour) - (date.hour t)
`).toEvaluateTo(1)
})
test('using date.now in calculations', () => {
// Check that date.now is in the past compared to a future timestamp
expect(`
now = (date.now)
future = date.new 2030 0 1
future > now
`).toEvaluateTo(true)
})
})

View File

@ -314,16 +314,14 @@ describe('fs - other', () => {
writeFileSync(file, 'initial')
let called = false
const watcher = fs.watch(file, () => {
called = true
})
const watcher = fs.watch(file, () => { called = true })
// Trigger change
await new Promise((resolve) => setTimeout(resolve, 100))
await new Promise(resolve => setTimeout(resolve, 100))
writeFileSync(file, 'updated')
// Wait for watcher
await new Promise((resolve) => setTimeout(resolve, 500))
await new Promise(resolve => setTimeout(resolve, 500))
expect(called).toBe(true)
watcher.close?.()

View File

@ -5,22 +5,14 @@ describe('json', () => {
expect(`json.decode '[1,2,3]'`).toEvaluateTo([1, 2, 3])
expect(`json.decode '"heya"'`).toEvaluateTo('heya')
expect(`json.decode '[true, false, null]'`).toEvaluateTo([true, false, null])
expect(`json.decode '{"a": true, "b": false, "c": "yeah"}'`).toEvaluateTo({
a: true,
b: false,
c: 'yeah',
})
expect(`json.decode '{"a": true, "b": false, "c": "yeah"}'`).toEvaluateTo({ a: true, b: false, c: "yeah" })
})
test('json.encode', () => {
expect(`json.encode [1 2 3]`).toEvaluateTo('[1,2,3]')
expect(`json.encode 'heya'`).toEvaluateTo('"heya"')
expect(`json.encode [true false null]`).toEvaluateTo('[true,false,null]')
expect(`json.encode [a=true b=false c='yeah'] | json.decode`).toEvaluateTo({
a: true,
b: false,
c: 'yeah',
})
expect(`json.encode [a=true b=false c='yeah'] | json.decode`).toEvaluateTo({ a: true, b: false, c: "yeah" })
})
test('edge cases - empty structures', () => {
@ -59,31 +51,27 @@ describe('json', () => {
})
test('nested structures - arrays', () => {
expect(`json.decode '[[1,2],[3,4],[5,6]]'`).toEvaluateTo([
[1, 2],
[3, 4],
[5, 6],
])
expect(`json.decode '[[1,2],[3,4],[5,6]]'`).toEvaluateTo([[1, 2], [3, 4], [5, 6]])
expect(`json.decode '[1,[2,[3,[4]]]]'`).toEvaluateTo([1, [2, [3, [4]]]])
})
test('nested structures - objects', () => {
expect(`json.decode '{"user":{"name":"Alice","age":30}}'`).toEvaluateTo({
user: { name: 'Alice', age: 30 },
user: { name: 'Alice', age: 30 }
})
expect(`json.decode '{"a":{"b":{"c":"deep"}}}'`).toEvaluateTo({
a: { b: { c: 'deep' } },
a: { b: { c: 'deep' } }
})
})
test('nested structures - mixed arrays and objects', () => {
expect(`json.decode '[{"id":1,"tags":["a","b"]},{"id":2,"tags":["c"]}]'`).toEvaluateTo([
{ id: 1, tags: ['a', 'b'] },
{ id: 2, tags: ['c'] },
{ id: 2, tags: ['c'] }
])
expect(`json.decode '{"items":[1,2,3],"meta":{"count":3}}'`).toEvaluateTo({
items: [1, 2, 3],
meta: { count: 3 },
meta: { count: 3 }
})
})

View File

@ -16,18 +16,6 @@ describe('string operations', () => {
await expect(`str.trim '\\n\\thello\\t\\n'`).toEvaluateTo('hello')
})
test('capitalize makes first char uppercase', async () => {
await expect(`str.capitalize 'hello'`).toEvaluateTo('Hello')
await expect(`str.capitalize 'HELLO'`).toEvaluateTo('Hello')
await expect(`str.capitalize 'hello world'`).toEvaluateTo('Hello world')
})
test('titlecase capitalizes each word', async () => {
await expect(`str.titlecase 'hello world'`).toEvaluateTo('Hello World')
await expect(`str.titlecase 'HELLO WORLD'`).toEvaluateTo('Hello World')
await expect(`str.titlecase 'the quick brown fox'`).toEvaluateTo('The Quick Brown Fox')
})
test('split divides string by separator', async () => {
await expect(`str.split 'a,b,c' ','`).toEvaluateTo(['a', 'b', 'c'])
await expect(`str.split 'hello' ''`).toEvaluateTo(['h', 'e', 'l', 'l', 'o'])
@ -117,17 +105,6 @@ describe('boolean logic', () => {
await expect(`not 42`).toEvaluateTo(false)
await expect(`not null`).toEvaluateTo(true)
})
test('not works with function calls', async () => {
await expect(`equals = do x y: x == y end; not equals 5 5`).toEvaluateTo(false)
await expect(`equals = do x y: x == y end; not equals 5 10`).toEvaluateTo(true)
})
test('not works with binary operations and comparisons', async () => {
await expect(`not 5 > 10`).toEvaluateTo(true)
await expect(`not 10 > 5`).toEvaluateTo(false)
await expect(`not true and false`).toEvaluateTo(true)
})
})
describe('utilities', () => {
@ -277,10 +254,7 @@ describe('collections', () => {
})
test('list.zip combines two arrays', async () => {
await expect(`list.zip [1 2] [3 4]`).toEvaluateTo([
[1, 3],
[2, 4],
])
await expect(`list.zip [1 2] [3 4]`).toEvaluateTo([[1, 3], [2, 4]])
})
test('list.first returns first element', async () => {
@ -450,10 +424,7 @@ describe('collections', () => {
await expect(`
gt-two = do x: x > 2 end
list.partition [1 2 3 4 5] gt-two
`).toEvaluateTo([
[3, 4, 5],
[1, 2],
])
`).toEvaluateTo([[3, 4, 5], [1, 2]])
})
test('list.compact removes null values', async () => {

View File

@ -1,143 +0,0 @@
import { expect, describe, test } from 'bun:test'
describe('type predicates', () => {
test('boolean? checks if value is boolean', async () => {
await expect(`boolean? true`).toEvaluateTo(true)
await expect(`boolean? false`).toEvaluateTo(true)
await expect(`boolean? 42`).toEvaluateTo(false)
await expect(`boolean? 'hello'`).toEvaluateTo(false)
await expect(`boolean? null`).toEvaluateTo(false)
await expect(`boolean? [1 2 3]`).toEvaluateTo(false)
})
test('number? checks if value is number', async () => {
await expect(`number? 42`).toEvaluateTo(true)
await expect(`number? 3.14`).toEvaluateTo(true)
await expect(`number? 0`).toEvaluateTo(true)
await expect(`number? -5`).toEvaluateTo(true)
await expect(`number? 'hello'`).toEvaluateTo(false)
await expect(`number? true`).toEvaluateTo(false)
await expect(`number? null`).toEvaluateTo(false)
})
test('string? checks if value is string', async () => {
await expect(`string? 'hello'`).toEvaluateTo(true)
await expect(`string? ''`).toEvaluateTo(true)
await expect(`string? world`).toEvaluateTo(true)
await expect(`string? 42`).toEvaluateTo(false)
await expect(`string? true`).toEvaluateTo(false)
await expect(`string? null`).toEvaluateTo(false)
await expect(`string? [1 2 3]`).toEvaluateTo(false)
})
test('array? checks if value is array', async () => {
await expect(`array? [1 2 3]`).toEvaluateTo(true)
await expect(`array? []`).toEvaluateTo(true)
await expect(`array? ['a' 'b']`).toEvaluateTo(true)
await expect(`array? [a=1 b=2]`).toEvaluateTo(false)
await expect(`array? 42`).toEvaluateTo(false)
await expect(`array? 'hello'`).toEvaluateTo(false)
await expect(`array? null`).toEvaluateTo(false)
})
test('list? is alias for array?', async () => {
await expect(`list? [1 2 3]`).toEvaluateTo(true)
await expect(`list? []`).toEvaluateTo(true)
await expect(`list? [a=1 b=2]`).toEvaluateTo(false)
})
test('dict? checks if value is dict', async () => {
await expect(`dict? [a=1 b=2]`).toEvaluateTo(true)
await expect(`dict? [=]`).toEvaluateTo(true)
await expect(`dict? [1 2 3]`).toEvaluateTo(false)
await expect(`dict? []`).toEvaluateTo(false)
await expect(`dict? 42`).toEvaluateTo(false)
await expect(`dict? 'hello'`).toEvaluateTo(false)
})
test('function? checks if value is function', async () => {
await expect(`
my-fn = do x: x * 2 end
function? my-fn
`).toEvaluateTo(true)
await expect(`function? inc`).toEvaluateTo(true)
await expect(`function? list.map`).toEvaluateTo(true)
await expect(`function? 42`).toEvaluateTo(false)
await expect(`function? 'hello'`).toEvaluateTo(false)
await expect(`function? [1 2 3]`).toEvaluateTo(false)
})
test('null? checks if value is null', async () => {
await expect(`null? null`).toEvaluateTo(true)
await expect(`null? 0`).toEvaluateTo(false)
await expect(`null? false`).toEvaluateTo(false)
await expect(`null? ''`).toEvaluateTo(false)
await expect(`null? []`).toEvaluateTo(false)
})
})
describe('type coercion', () => {
test('boolean coerces to boolean', async () => {
await expect(`boolean true`).toEvaluateTo(true)
await expect(`boolean false`).toEvaluateTo(false)
await expect(`boolean 1`).toEvaluateTo(true)
await expect(`boolean 0`).toEvaluateTo(false)
await expect(`boolean 'hello'`).toEvaluateTo(true)
await expect(`boolean ''`).toEvaluateTo(false)
await expect(`boolean null`).toEvaluateTo(false)
await expect(`boolean [1 2 3]`).toEvaluateTo(true)
})
test('number coerces to number', async () => {
await expect(`number 42`).toEvaluateTo(42)
await expect(`number '42'`).toEvaluateTo(42)
await expect(`number '3.14'`).toEvaluateTo(3.14)
await expect(`number true`).toEvaluateTo(1)
await expect(`number false`).toEvaluateTo(0)
})
test('string coerces to string', async () => {
await expect(`string 'hello'`).toEvaluateTo('hello')
await expect(`string 42`).toEvaluateTo('42')
await expect(`string true`).toEvaluateTo('true')
await expect(`string false`).toEvaluateTo('false')
await expect(`string null`).toEvaluateTo('null')
})
})
describe('type predicates in conditionals', () => {
test('using type predicates in if statements', async () => {
await expect(`
x = 42
if (number? x):
'is-num'
else:
'not-num'
end
`).toEvaluateTo('is-num')
})
test('filtering by type', async () => {
await expect(`
items = [1 'hello' 2 'world' 3]
list.filter items number?
`).toEvaluateTo([1, 2, 3])
})
test('filtering strings', async () => {
await expect(`
items = [1 'hello' 2 'world' 3]
list.filter items string?
`).toEvaluateTo(['hello', 'world'])
})
test('checking for functions', async () => {
await expect(`
double = do x: x * 2 end
not-fn = 42
is-fn = function? double
is-not-fn = function? not-fn
is-fn and (not is-not-fn)
`).toEvaluateTo(true)
})
})

View File

@ -1,21 +0,0 @@
import { toValue } from 'reefvm'
export const types = {
'boolean?': (v: any) => toValue(v).type === 'boolean',
boolean: (v: any) => Boolean(v),
'number?': (v: any) => toValue(v).type === 'number',
number: (v: any) => Number(v),
'string?': (v: any) => toValue(v).type === 'string',
string: (v: any) => String(v),
'array?': (v: any) => toValue(v).type === 'array',
'list?': (v: any) => toValue(v).type === 'array',
'dict?': (v: any) => toValue(v).type === 'dict',
'function?': (v: any) => ['function', 'native'].includes(toValue(v).type),
'null?': (v: any) => toValue(v).type === 'null',
}

10
src/server/app.tsx Normal file
View File

@ -0,0 +1,10 @@
import { Editor } from '#/editor/editor'
import { render } from 'hono/jsx/dom'
import './index.css'
const App = () => {
return <Editor />
}
const root = document.getElementById('root')!
render(<App />, root)

84
src/server/index.css Normal file
View File

@ -0,0 +1,84 @@
:root {
/* Background colors */
--bg-editor: #011627;
--bg-output: #40318D;
--bg-status-bar: #1E2A4A;
--bg-status-border: #0E1A3A;
--bg-selection: #1D3B53;
--bg-variable-def: #1E2A4A;
/* Text colors */
--text-editor: #D6DEEB;
--text-output: #7C70DA;
--text-status: #B3A9FF55;
--caret: #80A4C2;
/* Syntax highlighting colors */
--color-keyword: #C792EA;
--color-function: #82AAFF;
--color-string: #C3E88D;
--color-number: #F78C6C;
--color-bool: #FF5370;
--color-operator: #89DDFF;
--color-paren: #676E95;
--color-function-call: #FF9CAC;
--color-variable-def: #FFCB6B;
--color-error: #FF6E6E;
--color-regex: #E1ACFF;
/* ANSI terminal colors */
--ansi-black: #011627;
--ansi-red: #FF5370;
--ansi-green: #C3E88D;
--ansi-yellow: #FFCB6B;
--ansi-blue: #82AAFF;
--ansi-magenta: #C792EA;
--ansi-cyan: #89DDFF;
--ansi-white: #D6DEEB;
/* ANSI bright colors (slightly more vibrant) */
--ansi-bright-black: #676E95;
--ansi-bright-red: #FF6E90;
--ansi-bright-green: #D4F6A8;
--ansi-bright-yellow: #FFE082;
--ansi-bright-blue: #A8C7FA;
--ansi-bright-magenta: #E1ACFF;
--ansi-bright-cyan: #A8F5FF;
--ansi-bright-white: #FFFFFF;
}
@font-face {
font-family: 'C64ProMono';
src: url('../../assets/C64_Pro_Mono-STYLE.woff2') format('woff2');
font-weight: normal;
font-style: normal;
}
@font-face {
font-family: 'Pixeloid Mono';
src: url('../../assets/PixeloidMono.ttf') format('truetype');
font-weight: normal;
font-style: normal;
}
* {
margin: 0;
padding: 0;
box-sizing: border-box;
}
body {
background: var(--bg-output);
color: var(--text-output);
font-family: 'Pixeloid Mono', 'Courier New', monospace;
font-size: 18px;
height: 100vh;
overflow: hidden;
}
#root {
height: 100vh;
background: var(--bg-output);
display: flex;
flex-direction: column;
}

12
src/server/index.html Normal file
View File

@ -0,0 +1,12 @@
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>Shrimp</title>
</head>
<body>
<div id="root"></div>
<script type="module" src="./app.tsx"></script>
</body>
</html>

29
src/server/server.tsx Normal file
View File

@ -0,0 +1,29 @@
import index from './index.html'
const server = Bun.serve({
port: process.env.PORT ? Number(process.env.PORT) : 3001,
routes: {
'/*': index,
'/api/hello': {
async GET(req) {
return Response.json({
message: 'Hello, world!',
method: 'GET',
})
},
async PUT(req) {
return Response.json({
message: 'Hello, world!',
method: 'PUT',
})
},
},
},
development: process.env.NODE_ENV !== 'production' && {
hmr: true,
console: true,
},
})
console.log(`🚀 Server running at ${server.url}`)

View File

@ -2,13 +2,36 @@ import { expect } from 'bun:test'
import { diffLines } from 'diff'
import color from 'kleur'
import { Scanner, TokenType, type Token } from '#parser/tokenizer2'
import { parse, setGlobals } from '#parser/parser2'
import { Tree } from '#parser/node'
import { parser } from '#parser/shrimp'
import { setGlobals } from '#parser/tokenizer'
import { parse } from '#parser/parser2'
import { globals as prelude } from '#prelude'
import { $ } from 'bun'
import { assert, errorMessage } from '#utils/utils'
import { Compiler } from '#compiler/compiler'
import { run, VM } from 'reefvm'
import { treeToString2, VMResultToValue } from '#utils/tree'
import { treeToString2, treeToString, VMResultToValue } from '#utils/tree'
const regenerateParser = async () => {
let generate = true
try {
const grammarStat = await Bun.file('./src/parser/shrimp.grammar').stat()
const tokenizerStat = await Bun.file('./src/parser/tokenizer.ts').stat()
const parserStat = await Bun.file('./src/parser/shrimp.ts').stat()
if (grammarStat.mtime <= parserStat.mtime && tokenizerStat.mtime <= parserStat.mtime) {
generate = false
}
} catch (e) {
console.error('Error checking or regenerating parser:', e)
} finally {
if (generate) {
await $`bun generate-parser`
}
}
}
await regenerateParser()
// Type declaration for TypeScript
declare module 'bun:test' {
@ -20,7 +43,7 @@ declare module 'bun:test' {
toFailEvaluation(): Promise<T>
toBeToken(expected: string): T
toMatchToken(typeOrValue: string, value?: string): T
toMatchTokens(...tokens: { type: string; value?: string }[]): T
toMatchTokens(...tokens: { type: string, value?: string }[]): T
}
}
@ -50,8 +73,7 @@ expect.extend({
assert(typeof received === 'string', 'toFailParse can only be used with string values')
try {
const node = parse(received)
const tree = new Tree(node)
const tree = parser.parse(received)
let hasErrors = false
tree.iterate({
enter(n) {
@ -68,7 +90,7 @@ expect.extend({
pass: true,
}
} else {
const actual = treeToString2(node, received)
const actual = treeToString(tree, received)
return {
message: () => `Expected input to fail parsing, but it parsed successfully:\n${actual}`,
pass: false,
@ -146,7 +168,7 @@ expect.extend({
return {
message: () => `Expected token type to be ${expected}, but got ${TokenType[value.type]}`,
pass: value.type === target,
pass: value.type === target
}
} catch (error) {
return {
@ -166,8 +188,7 @@ expect.extend({
if (!token) {
return {
message: () =>
`Expected token to be ${expectedValue.replaceAll('\n', '\\n')}, got ${token}`,
message: () => `Expected token to be ${expectedValue.replaceAll('\n', '\\n')}, got ${token}`,
pass: false,
}
}
@ -175,14 +196,13 @@ expect.extend({
if (expectedType && TokenType[expectedType as keyof typeof TokenType] !== token.type) {
return {
message: () => `Expected token to be ${expectedType}, but got ${TokenType[token.type]}`,
pass: false,
pass: false
}
}
return {
message: () =>
`Expected token to be ${expectedValue.replaceAll('\n', '\\n')}, but got ${token.value}`,
pass: token.value === expectedValue,
message: () => `Expected token to be ${expectedValue.replaceAll('\n', '\\n')}, but got ${token.value}`,
pass: token.value === expectedValue
}
} catch (error) {
return {
@ -191,11 +211,11 @@ expect.extend({
}
}
},
toMatchTokens(received: unknown, ...tokens: { type: string; value?: string }[]) {
toMatchTokens(received: unknown, ...tokens: { type: string, value?: string }[]) {
assert(typeof received === 'string', 'toMatchTokens can only be used with string values')
try {
const result = tokenize(received).map((t) => toHumanToken(t))
const result = tokenize(received).map(t => toHumanToken(t))
if (result.length === 0 && tokens.length > 0) {
return {
@ -209,7 +229,7 @@ expect.extend({
return {
message: () => `Tokens don't match: \n\n${diff(actual, expected)}`,
pass: expected == actual,
pass: expected == actual
}
} catch (error) {
return {
@ -217,18 +237,18 @@ expect.extend({
pass: false,
}
}
},
}
})
const tokenize = (code: string): Token[] => {
const scanner = new Scanner()
const scanner = new Scanner
return scanner.tokenize(code)
}
const toHumanToken = (tok: Token): { type: string; value?: string } => {
const toHumanToken = (tok: Token): { type: string, value?: string } => {
return {
type: TokenType[tok.type],
value: tok.value,
value: tok.value
}
}
@ -243,7 +263,7 @@ const trimWhitespace = (str: string): string => {
if (!line.startsWith(leadingWhitespace)) {
let foundWhitespace = line.match(/^(\s*)/)?.[1] || ''
throw new Error(
`Line has inconsistent leading whitespace: "${line}"(found "${foundWhitespace}", expected "${leadingWhitespace}")`,
`Line has inconsistent leading whitespace: "${line}"(found "${foundWhitespace}", expected "${leadingWhitespace}")`
)
}
return line.slice(leadingWhitespace.length)
@ -259,7 +279,7 @@ const diff = (a: string, b: string): string => {
if (expected !== actual) {
const changes = diffLines(actual, expected)
for (const part of changes) {
const sign = part.added ? '+' : part.removed ? '-' : ' '
const sign = part.added ? "+" : part.removed ? "-" : " "
let line = sign + part.value
if (part.added) {
line = color.green(line)
@ -267,7 +287,7 @@ const diff = (a: string, b: string): string => {
line = color.red(line)
}
lines.push(line.endsWith('\n') || line.endsWith('\n\u001b[39m') ? line : line + '\n')
lines.push(line.endsWith("\n") || line.endsWith("\n\u001b[39m") ? line : line + "\n")
}
}

68
src/utils/signal.ts Normal file
View File

@ -0,0 +1,68 @@
/**
* How to use a Signal:
*
* Create a signal with primitives:
* const nameSignal = new Signal<string>()
* const countSignal = new Signal<number>()
*
* Create a signal with objects:
* const chatSignal = new Signal<{ username: string, message: string }>()
*
* Create a signal with no data (void):
* const clickSignal = new Signal<void>()
* const clickSignal2 = new Signal() // Defaults to void
*
* Connect to the signal:
* const disconnect = chatSignal.connect((data) => {
* const {username, message} = data;
* console.log(`${username} said "${message}"`);
* })
*
* Emit a signal:
* nameSignal.emit("Alice")
* countSignal.emit(42)
* chatSignal.emit({ username: "Chad", message: "Hey everyone, how's it going?" });
* clickSignal.emit() // No argument for void signals
*
* Forward a signal:
* const relaySignal = new Signal<{ username: string, message: string }>()
* const disconnectRelay = chatSignal.connect(relaySignal)
* // Now, when chatSignal emits, relaySignal will also emit the same data
*
* Disconnect a single listener:
* disconnect(); // The disconnect function is returned when you connect to a signal
*
* Disconnect all listeners:
* chatSignal.disconnect()
*/
export class Signal<T = void> {
private listeners: Array<(data: T) => void> = []
connect(listenerOrSignal: Signal<T> | ((data: T) => void)) {
let listener: (data: T) => void
// If it is a signal, forward the data to the signal
if (listenerOrSignal instanceof Signal) {
listener = (data: T) => listenerOrSignal.emit(data)
} else {
listener = listenerOrSignal
}
this.listeners.push(listener)
return () => {
this.listeners = this.listeners.filter((l) => l !== listener)
}
}
emit(data: T) {
for (const listener of this.listeners) {
listener(data)
}
}
disconnect() {
this.listeners = []
}
}

View File

@ -136,12 +136,3 @@ export const asciiEscapeToHtml = (str: string): HtmlEscapedString => {
return result as HtmlEscapedString
}
export const isDebug = (): boolean => {
if (typeof process !== 'undefined' && process.env) {
return !!process.env.DEBUG
}
return false
}
export const runningInBrowser = typeof window !== 'undefined'

View File

@ -5,7 +5,10 @@
"name": "Run Extension",
"type": "extensionHost",
"request": "launch",
"args": ["--extensionDevelopmentPath=${workspaceFolder}", "--profile=Shrimp Dev"],
"args": [
"--extensionDevelopmentPath=${workspaceFolder}",
"--profile=Shrimp Dev"
],
"outFiles": [
"${workspaceFolder}/client/dist/**/*.js",
"${workspaceFolder}/server/dist/**/*.js"

View File

@ -22,7 +22,7 @@ export function activate(context: vscode.ExtensionContext) {
'shrimpLanguageServer',
'Shrimp Language Server',
serverOptions,
clientOptions,
clientOptions
)
client.start()
@ -46,7 +46,7 @@ export function activate(context: vscode.ExtensionContext) {
language: 'text',
})
await vscode.window.showTextDocument(doc, { preview: false })
}),
})
)
// Command: Show Bytecode
@ -67,7 +67,7 @@ export function activate(context: vscode.ExtensionContext) {
language: 'text',
})
await vscode.window.showTextDocument(doc, { preview: false })
}),
})
)
// Command: Run File
@ -93,7 +93,7 @@ export function activate(context: vscode.ExtensionContext) {
const terminal = vscode.window.createTerminal('Shrimp')
terminal.show()
terminal.sendText(`${binaryPath} "${filePath}"`)
}),
})
)
}

View File

@ -113,5 +113,5 @@ console.log(`✓ Generated ${names.length} prelude names to server/src/metadata/
console.log(
`✓ Generated completions for ${
Object.keys(moduleMetadata).length
} modules to server/src/metadata/prelude-completions.ts`,
} modules to server/src/metadata/prelude-completions.ts`
)

View File

@ -10,7 +10,7 @@ import { analyzeCompletionContext } from './contextAnalyzer'
*/
export const provideCompletions = (
document: TextDocument,
position: { line: number; character: number },
position: { line: number; character: number }
): CompletionItem[] => {
const context = analyzeCompletionContext(document, position)

View File

@ -14,7 +14,7 @@ export type CompletionContext =
*/
export const analyzeCompletionContext = (
document: TextDocument,
position: { line: number; character: number },
position: { line: number; character: number }
): CompletionContext => {
const offset = document.offsetAt(position)
const text = document.getText()

File diff suppressed because it is too large Load Diff

View File

@ -2,44 +2,39 @@
// Do not edit manually - run 'bun run generate-prelude-metadata' to regenerate
export const PRELUDE_NAMES = [
'$',
'array?',
'at',
'bnot',
'boolean',
'boolean?',
'date',
'dec',
'describe',
'dict',
'dict?',
'each',
'echo',
'empty?',
'exit',
'fs',
'function?',
'identity',
'import',
'inc',
'inspect',
'json',
'length',
'list',
'list?',
'load',
'math',
'not',
'null?',
'number',
'number?',
'range',
'ref',
'some?',
'str',
'string',
'string?',
'type',
'var',
'var?',
"$",
"array?",
"at",
"bnot",
"boolean?",
"dec",
"describe",
"dict",
"dict?",
"each",
"echo",
"empty?",
"exit",
"fs",
"function?",
"identity",
"import",
"inc",
"inspect",
"json",
"length",
"list",
"load",
"math",
"not",
"null?",
"number?",
"range",
"ref",
"some?",
"str",
"string?",
"type",
"var",
"var?"
] as const

View File

@ -41,7 +41,7 @@ export function buildSemanticTokens(document: TextDocument, tree: Tree): number[
function emitNamedArgPrefix(
node: SyntaxNode,
document: TextDocument,
builder: SemanticTokensBuilder,
builder: SemanticTokensBuilder
) {
const text = document.getText({
start: document.positionAt(node.from),
@ -57,7 +57,7 @@ function emitNamedArgPrefix(
start.character,
nameLength,
TOKEN_TYPES.indexOf(SemanticTokenTypes.property),
0,
0
)
// Emit token for the "=" part
@ -66,7 +66,7 @@ function emitNamedArgPrefix(
start.character + nameLength,
1, // Just the = character
TOKEN_TYPES.indexOf(SemanticTokenTypes.operator),
0,
0
)
}
@ -75,7 +75,7 @@ function walkTree(
node: SyntaxNode,
document: TextDocument,
builder: SemanticTokensBuilder,
scopeTracker: EditorScopeAnalyzer,
scopeTracker: EditorScopeAnalyzer
) {
// Special handling for NamedArgPrefix to split "name=" into two tokens
if (node.type.id === Terms.NamedArgPrefix) {
@ -102,7 +102,7 @@ type TokenInfo = { type: number; modifiers: number } | undefined
function getTokenType(
node: SyntaxNode,
document: TextDocument,
scopeTracker: EditorScopeAnalyzer,
scopeTracker: EditorScopeAnalyzer
): TokenInfo {
const nodeTypeId = node.type.id
const parentTypeId = node.parent?.type.id

View File

@ -17,6 +17,7 @@ import {
CompletionItemKind,
TextDocumentChangeEvent,
} from 'vscode-languageserver/node'
import { setGlobals } from '../../../src/parser/tokenizer'
import { globals } from '../../../src/prelude'
// Initialize parser with prelude globals so it knows dict/list/str are in scope
@ -124,7 +125,7 @@ function handleCompletion(params: any) {
if (contextCompletions.length > 0) {
console.log(
`✅ Returning ${contextCompletions.length} completions:`,
contextCompletions.map((c) => c.label).join(', '),
contextCompletions.map((c) => c.label).join(', ')
)
return contextCompletions
}

View File

@ -1,8 +1,4 @@
import {
SignatureHelp,
SignatureInformation,
ParameterInformation,
} from 'vscode-languageserver/node'
import { SignatureHelp, SignatureInformation, ParameterInformation } from 'vscode-languageserver/node'
import { TextDocument } from 'vscode-languageserver-textdocument'
import { Tree, SyntaxNode } from '@lezer/common'
import { parser } from '../../../src/parser/shrimp'
@ -10,7 +6,7 @@ import { completions } from './metadata/prelude-completions'
export const provideSignatureHelp = (
document: TextDocument,
position: { line: number; character: number },
position: { line: number; character: number }
): SignatureHelp | undefined => {
const text = document.getText()
const tree = parser.parse(text)
@ -104,6 +100,6 @@ const lookupFunctionParams = (funcName: string): string[] | undefined => {
const buildSignature = (funcName: string, params: string[]): SignatureInformation => {
const label = `${funcName}(${params.join(', ')})`
const parameters: ParameterInformation[] = params.map((p) => ({ label: p }))
const parameters: ParameterInformation[] = params.map(p => ({ label: p }))
return { label, parameters }
}

View File

@ -0,0 +1,41 @@
import { parser } from '../../src/parser/shrimp'
import { setGlobals } from '../../src/parser/tokenizer'
import { PRELUDE_NAMES } from '../server/src/prelude-names'
// Set globals for DotGet detection
setGlobals(PRELUDE_NAMES as unknown as string[])
// Test cases - does incomplete DotGet parse correctly?
const testCases = [
'dict.',
'dict.g',
'dict.get',
'$.',
'$.e',
'$.env',
]
for (const code of testCases) {
console.log(`\nTesting: "${code}"`)
const tree = parser.parse(code)
const cursor = tree.cursor()
// Print the parse tree
const printTree = (depth = 0) => {
const indent = ' '.repeat(depth)
console.log(`${indent}${cursor.name} [${cursor.from}-${cursor.to}]`)
if (cursor.firstChild()) {
do {
printTree(depth + 1)
} while (cursor.nextSibling())
cursor.parent()
}
}
printTree()
// Check at cursor position (end of string)
const node = tree.resolveInner(code.length, -1)
console.log(`Node at end: ${node.name} (type: ${node.type.id})`)
}