Code mirror plugin lsp interface (#1444)

* better named dirs

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* move some stuff around

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* more logging

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* updates

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* less logging

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* add fs in

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* updates

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* file reader

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* workspace

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* start of workspace folders

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* start of workspace folders

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* cleanup workspace folders

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* cleanup logs

Signed-off-by: Jess Frazelle <github@jessfraz.com>

* updates

Signed-off-by: Jess Frazelle <github@jessfraz.com>

---------

Signed-off-by: Jess Frazelle <github@jessfraz.com>

src/editor/plugins/lsp/kcl/parser.ts

@@ -0,0 +1,168 @@
+// Extends the codemirror Parser for kcl.
+
+import {
+  Parser,
+  Input,
+  TreeFragment,
+  PartialParse,
+  Tree,
+  NodeType,
+  NodeSet,
+} from '@lezer/common'
+import { LanguageServerClient } from 'editor/plugins/lsp'
+import { posToOffset } from 'editor/plugins/lsp/util'
+import { SemanticToken } from './semantic_tokens'
+import { DocInput } from '@codemirror/language'
+import { tags, styleTags } from '@lezer/highlight'
+
+export default class KclParser extends Parser {
+  private client: LanguageServerClient
+
+  constructor(client: LanguageServerClient) {
+    super()
+    this.client = client
+  }
+
+  createParse(
+    input: Input,
+    fragments: readonly TreeFragment[],
+    ranges: readonly { from: number; to: number }[]
+  ): PartialParse {
+    let parse: PartialParse = new Context(this, input, fragments, ranges)
+    return parse
+  }
+
+  getTokenTypes(): string[] {
+    return this.client.getServerCapabilities().semanticTokensProvider!.legend
+      .tokenTypes
+  }
+
+  getSemanticTokens(): SemanticToken[] {
+    return this.client.getSemanticTokens()
+  }
+}
+
+class Context implements PartialParse {
+  private parser: KclParser
+  private input: DocInput
+  private fragments: readonly TreeFragment[]
+  private ranges: readonly { from: number; to: number }[]
+
+  private nodeTypes: { [key: string]: NodeType }
+  stoppedAt: number = 0
+
+  private semanticTokens: SemanticToken[] = []
+  private currentLine: number = 0
+  private currentColumn: number = 0
+  private nodeSet: NodeSet
+
+  constructor(
+    /// The parser configuration used.
+    parser: KclParser,
+    input: Input,
+    fragments: readonly TreeFragment[],
+    ranges: readonly { from: number; to: number }[]
+  ) {
+    this.parser = parser
+    this.input = input as DocInput
+    this.fragments = fragments
+    this.ranges = ranges
+
+    // Iterate over the semantic token types and create a node type for each.
+    this.nodeTypes = {}
+    let nodeArray: NodeType[] = []
+    this.parser.getTokenTypes().forEach((tokenType, index) => {
+      const nodeType = NodeType.define({
+        id: index,
+        name: tokenType,
+        // props: [this.styleTags],
+      })
+      this.nodeTypes[tokenType] = nodeType
+      nodeArray.push(nodeType)
+    })
+
+    this.semanticTokens = this.parser.getSemanticTokens()
+    const styles = styleTags({
+      number: tags.number,
+      variable: tags.variableName,
+      operator: tags.operator,
+      keyword: tags.keyword,
+      string: tags.string,
+      comment: tags.comment,
+      function: tags.function(tags.variableName),
+    })
+    this.nodeSet = new NodeSet(nodeArray).extend(styles)
+  }
+
+  get parsedPos(): number {
+    return 0
+  }
+
+  advance(): Tree | null {
+    if (this.semanticTokens.length === 0) {
+      return new Tree(NodeType.none, [], [], 0)
+    }
+    const tree = this.createTree(this.semanticTokens[0], 0)
+    this.stoppedAt = this.input.doc.length
+    return tree
+  }
+
+  createTree(token: SemanticToken, index: number): Tree {
+    const changedLine = token.delta_line !== 0
+    this.currentLine += token.delta_line
+    if (changedLine) {
+      this.currentColumn = 0
+    }
+    this.currentColumn += token.delta_start
+
+    // Let's get our position relative to the start of the file.
+    let currentPosition = posToOffset(this.input.doc, {
+      line: this.currentLine,
+      character: this.currentColumn,
+    })
+
+    const nodeType = this.nodeSet.types[this.nodeTypes[token.token_type].id]
+
+    if (currentPosition === undefined) {
+      // This is bad and weird.
+      return new Tree(nodeType, [], [], token.length)
+    }
+
+    if (index >= this.semanticTokens.length - 1) {
+      // We have no children.
+      return new Tree(nodeType, [], [], token.length)
+    }
+
+    const nextIndex = index + 1
+    const nextToken = this.semanticTokens[nextIndex]
+    const changedLineNext = nextToken.delta_line !== 0
+    const nextLine = this.currentLine + nextToken.delta_line
+    const nextColumn = changedLineNext
+      ? nextToken.delta_start
+      : this.currentColumn + nextToken.delta_start
+    const nextPosition = posToOffset(this.input.doc, {
+      line: nextLine,
+      character: nextColumn,
+    })
+
+    if (nextPosition === undefined) {
+      // This is bad and weird.
+      return new Tree(nodeType, [], [], token.length)
+    }
+
+    // Let's get the
+
+    return new Tree(
+      nodeType,
+      [this.createTree(nextToken, nextIndex)],
+
+      // The positions (offsets relative to the start of this tree) of the children.
+      [nextPosition - currentPosition],
+      token.length
+    )
+  }
+
+  stopAt(pos: number) {
+    this.stoppedAt = pos
+  }
+}