remark-lint-no-unicode refactoring

packages/remark-lint-no-unicode/fixtures/002.txt

@@ -9,8 +9,8 @@ abcd
 
 ---
 
-1:6-1:7: Forbidden character "d". https://symbl.cc/en/0064/
-3:4-3:5: Forbidden character "d". https://symbl.cc/en/0064/
-4:4-4:5: Forbidden character "d". https://symbl.cc/en/0064/
+1:6-1:7: Forbidden character "d" (see https://symbl.cc/en/0064/).
+3:4-3:5: Forbidden character "d" (see https://symbl.cc/en/0064/).
+4:4-4:5: Forbidden character "d" (see https://symbl.cc/en/0064/).
 
 ---

packages/remark-lint-no-unicode/lib/main.ts

@@ -1,91 +1,92 @@
-import {type Root, type RootContent} from "mdast"
+import {isCodeNode, isInlineCodeNode, isLiteralNode, isParentNode} from "@onlyoffice/mdast-util-is-node"
+import {type Node} from "mdast"
 import {lintRule} from "unified-lint-rule"
-import {type VFile} from "vfile"
+import {type Position} from "unist"
+import {type MessageOptions, type VFile} from "vfile"
 
-const replaceHint = new Map([
-  ["00A0", "Use a regular space instead of no-break space. https://symbl.cc/en/00A0/"],
-])
+const hint: Record<string, string> = {
+  "00A0": `Use a regular space instead of no-break space (see ${url("00A0")}).`,
+}
 
 export const remarkLintNoUnicode: any =
   lintRule("@onlyoffice:no-unicode", rule)
 
-function rule(t: Root, f: VFile, o: string[]): void {
-  if (o === null || o === undefined || o.length === 0) {
-    f.fail("No allowed characters provided")
+function rule(t: Node, f: VFile, o: string[]): void {
+  if (isCodeNode(t) || isInlineCodeNode(t)) {
+    return
   }
-  for (const c of t.children) {
-    unicodeCheck(c, f, o)
+  if (isParentNode(t)) {
+    for (const c of t.children) {
+      rule(c, f, o)
+    }
   }
-}
-
-function unicodeCheck(t: RootContent, f: VFile, o: string[]): void {
-  if (t.type === "inlineCode" || t.type === "code") {
+  if (!isLiteralNode(t)) {
     return
   }
-  if ("value" in t) {
-    const uc = getUnicodeCharacters(t.value)
-    for (const [i, u] of uc.entries()) {
-      if (!o.includes(u)) {
-        const m = replaceHint.get(u) ||
-          `Forbidden character "${String.fromCodePoint(Number.parseInt(u, 16))}". https://symbl.cc/en/${u}/`
-        if (t.position !== undefined) {
-          const line = t.position.start.line + uc.slice(0, i + 1).filter((e) => e === "000A").length
-          let {column} = t.position.start
-          for (const c of uc.slice(0, i + 1)) {
-            column += 1
-            if (c === "000A") {
-              column = 1
-            }
-          }
-          f.message(m, {
-            start: {
-              line,
-              column: column - 1,
-            },
-            end: {
-              line,
-              column,
-            },
-          })
-        } else {
-          f.message(m, {
-            ancestors: [t],
-          })
-        }
-      }
+  for (const [i, v] of [...t.value].entries()) {
+    const u = unicode(v)
+    if (o.includes(u)) {
+      continue
     }
-  }
-  if ("children" in t) {
-    for (const c of t.children) {
-      unicodeCheck(c, f, o)
+    const m = message(v, u)
+    const p: MessageOptions = {
+      ancestors: [t],
+    }
+    if (t.position) {
+      p.place = position(t.position, t.value, i)
     }
+    f.message(m, p)
   }
 }
 
-function getUnicodeCharacters(s: string): string[] {
-  if (s === null || s === undefined || s.length === 0) {
-    return []
+function message(v: string, u: string): string {
+  let m = hint[u]
+  if (!m) {
+    m = `Forbidden character "${v}" (see ${url(u)}).`
   }
-  const a = []
-  for (const c of s) {
-    const cp = c.codePointAt(0)
-    if (cp !== undefined) {
-      const u = cp.toString(16).toUpperCase()
-      switch (u.length) {
-      case 1:
-        a.push(`000${u}`)
-        break
-      case 2:
-        a.push(`00${u}`)
-        break
-      case 3:
-        a.push(`0${u}`)
-        break
-      default:
-        a.push(u)
-        break
-      }
+  return m
+}
+
+function position(p: Position, a: string, i: number): Position {
+  let {line, column} = p.start
+  for (let j = 0; j <= i; j += 1) {
+    column += 1
+    if (a[j] === "\n") {
+      column = 1
+      line += 1
     }
   }
-  return a
+  const n: Position = {
+    start: {
+      line,
+      column: column - 1,
+    },
+    end: {
+      line,
+      column,
+    },
+  }
+  return n
+}
+
+function unicode(s: string): string {
+  const p = s.codePointAt(0)
+  if (p === undefined) {
+    return ""
+  }
+  const u = p.toString(16).toUpperCase()
+  switch (u.length) {
+  case 1:
+    return `000${u}`
+  case 2:
+    return `00${u}`
+  case 3:
+    return `0${u}`
+  default:
+    return u
+  }
+}
+
+function url(u: string): string {
+  return `https://symbl.cc/en/${u}/`
 }

packages/remark-lint-no-unicode/package.json

@@ -13,7 +13,9 @@
     "test": "pnpm test:types && pnpm test:unit"
   },
   "dependencies": {
+    "@onlyoffice/mdast-util-is-node": "workspace:^",
     "@types/mdast": "4.0.3",
+    "@types/unist": "3.0.3",
     "unified-lint-rule": "3.0.0",
     "vfile": "6.0.1"
   },