import {Tree, TreeFragment} from "@lezer/common" import ist from "ist" import {parser} from "../dist/index.js" import {compareTree} from "./compare-tree.js" let doc1 = ` Header --- One **two** three *four* five. > Start of quote > > 1. Nested list > > 2. More content > inside the [list][link] > > Continued item > > ~~~ > Block of code > ~~~ > > 3. And so on [link]: /ref [another]: /one And a final paragraph. *** The end. ` type ChangeSpec = {from: number, to?: number, insert?: string}[] class State { constructor(readonly doc: string, readonly tree: Tree, readonly fragments: readonly TreeFragment[]) {} static start(doc: string) { let tree = parser.parse(doc) return new State(doc, tree, TreeFragment.addTree(tree)) } update(changes: ChangeSpec, reparse = true) { let changed = [], doc = this.doc, off = 0 for (let {from, to = from, insert = ""} of changes) { doc = doc.slice(0, from) + insert + doc.slice(to) changed.push({fromA: from - off, toA: to - off, fromB: from, toB: from + insert.length}) off += insert.length - (to - from) } let fragments = TreeFragment.applyChanges(this.fragments, changed, 2) if (!reparse) return new State(doc, Tree.empty, fragments) let tree = parser.parse(doc, fragments) return new State(doc, tree, TreeFragment.addTree(tree, fragments)) } } let _state1: State | null = null, state1 = () => _state1 || (_state1 = State.start(doc1)) function overlap(a: Tree, b: Tree) { let inA = new Set(), shared = 0, sharingTo = 0 for (let cur = a.cursor(); cur.next();) if (cur.tree) inA.add(cur.tree) for (let cur = b.cursor(); cur.next();) if (cur.tree && inA.has(cur.tree) && cur.type.is("Block") && cur.from >= sharingTo) { shared += cur.to - cur.from sharingTo = cur.to } return Math.round(shared * 100 / b.length) } function testChange(change: ChangeSpec, reuse = 10) { let state = state1().update(change) compareTree(state.tree, parser.parse(state.doc)) if (reuse) ist(overlap(state.tree, state1().tree), reuse, ">") } describe("Markdown incremental parsing", () => { it("can produce the proper tree", () => { // Replace 'three' with 'bears' let state = state1().update([{from: 24, to: 29, insert: "bears"}]) compareTree(state.tree, state1().tree) }) it("reuses nodes from the previous parse", () => { // Replace 'three' with 'bears' let state = state1().update([{from: 24, to: 29, insert: "bears"}]) ist(overlap(state1().tree, state.tree), 80, ">") }) it("can reuse content for a change in a block context", () => { // Replace 'content' with 'monkeys' let state = state1().update([{from: 92, to: 99, insert: "monkeys"}]) compareTree(state.tree, state1().tree) ist(overlap(state1().tree, state.tree), 20, ">") }) it("can handle deleting a quote mark", () => testChange([{from: 82, to: 83}])) it("can handle adding to a quoted block", () => testChange([{from: 37, insert: "> "}, {from: 45, insert: "> "}])) it("can handle a change in a post-linkref paragraph", () => testChange([{from: 249, to: 251}])) it("can handle a change in a paragraph-adjacent linkrefs", () => testChange([{from: 230, to: 231}])) it("can deal with multiple changes applied separately", () => { let state = state1().update([{from: 190, to: 191}], false).update([{from: 30, insert: "hi\n\nyou"}]) compareTree(state.tree, parser.parse(state.doc)) ist(overlap(state.tree, state1().tree), 20, ">") }) it("works when a change happens directly after a block", () => testChange([{from: 150, to: 167}])) it("works when a change deletes a blank line after a paragraph", () => testChange([{from: 207, to: 213}])) it("doesn't get confused by removing paragraph-breaking markup", () => testChange([{from: 264, to: 265}])) function r(n: number) { return Math.floor(Math.random() * n) } function rStr(len: number) { let result = "", chars = "\n>x-" while (result.length < len) result += chars[r(chars.length)] return result } it("survives random changes", () => { for (let i = 0, l = doc1.length; i < 20; i++) { let c = 1 + r(4), changes = [] for (let i = 0, rFrom = 0; i < c; i++) { let rTo = rFrom + Math.floor((l - rFrom) / (c - i)) let from = rFrom + r(rTo - rFrom - 1), to = r(2) == 1 ? from : from + r(Math.min(rTo - from, 20)) let iR = r(3), insert = iR == 0 && from != to ? "" : iR == 1 ? "\n\n" : rStr(r(5) + 1) changes.push({from, to, insert}) l += insert.length - (to - from) rFrom = to + insert.length } testChange(changes, 0) } }) it("can handle large documents", () => { let doc = doc1.repeat(50) let state = State.start(doc) let newState = state.update([{from: doc.length >> 1, insert: "a\n\nb"}]) ist(overlap(state.tree, newState.tree), 90, ">") }) it("properly re-parses a continued indented code block", () => { let state = State.start(` One paragraph to create a bit of string length here Code Block Another paragraph that is long enough to create a fragment `).update([{from: 76, insert: " "}]) compareTree(state.tree, parser.parse(state.doc)) }) it("properly re-parses a continued list", () => { let state = State.start(` One paragraph to create a bit of string length here * List More content Another paragraph that is long enough to create a fragment `).update([{from: 65, insert: " * "}]) compareTree(state.tree, parser.parse(state.doc)) }) it("can recover from incremental parses that stop in the middle of a list", () => { let doc = ` 1. I am a list item with ***some* emphasized content inside** and the parser hopefully stops parsing after me. 2. Oh no the list continues. ` let parse = parser.startParse(doc), tree parse.advance() ist(parse.parsedPos, doc.length, "<") parse.stopAt(parse.parsedPos) while (!(tree = parse.advance())) {} let state = new State(doc, tree, TreeFragment.addTree(tree)).update([]) ist(state.tree.topNode.lastChild!.from, 1) }) it("can reuse list items", () => { let start = State.start(" - List item\n".repeat(100)) let state = start.update([{from: 18, to: 19}]) ist(overlap(start.tree, state.tree), 80, ">") }) it("returns a tree starting at the first range", () => { let result = parser.parse("foo\n\nbar", [], [{from: 5, to: 8}]) ist(result.toString(), "Document(Paragraph)") ist(result.length, 3) ist(result.positions[0], 0) }) it("Allows gaps in the input", () => { let doc = ` The first X *y* X< >X paragraph. - And *a X<*>X list* ` let tree = parser.parse(doc, [], [{from: 0, to: 11}, {from: 12, to: 17}, {from: 23, to: 46}, {from: 51, to: 58}]) ist(tree.toString(), "Document(Paragraph(Emphasis(EmphasisMark,EmphasisMark)),BulletList(ListItem(ListMark,Paragraph(Emphasis(EmphasisMark,EmphasisMark)))))") ist(tree.length, doc.length) let top = tree.topNode, first = top.firstChild! ist(first.name, "Paragraph") ist(first.from, 1) ist(first.to, 34) let last = top.lastChild!.lastChild!.lastChild!, em = last.lastChild! ist(last.name, "Paragraph") ist(last.from, 39) ist(last.to, 57) ist(em.name, "Emphasis") ist(em.from, 43) ist(em.to, 57) }) it("can reuse nodes at the end of the document", () => { let doc = `* List item ~~~js function foo() { return false } ~~~ ` let tree = parser.parse(doc) let ins = 11 let doc2 = doc.slice(0, ins) + "\n* " + doc.slice(ins) let fragments = TreeFragment.applyChanges(TreeFragment.addTree(tree), [{fromA: ins, toA: ins, fromB: ins, toB: ins + 3}]) let tree2 = parser.parse(doc2, fragments) ist(tree2.topNode.lastChild!.tree, tree.topNode.lastChild!.tree) }) it("places reused nodes at the right position when there are gaps before them", () => { let doc = " {{}}\nb\n{{}}" let ast1 = parser.parse(doc, undefined, [{from: 0, to: 1}, {from: 5, to: 8}]) let frag = TreeFragment.applyChanges(TreeFragment.addTree(ast1), [{fromA: 0, toA: 0, fromB: 0, toB: 1}]) let ast2 = parser.parse(" " + doc, frag, [{from: 0, to: 2}, {from: 6, to: 9}]) ist(ast2.toString(), "Document(Paragraph)") let p = ast2.topNode.firstChild! ist(p.from, 7) ist(p.to, 8) }) })