@yozora/tokenizer-autolink-extension 2.0.4 → 2.0.5

package/lib/cjs/{index.js → index.cjs}

@@ -281,21 +281,20 @@ const uniqueName = '@yozora/tokenizer-autolink-extension';
 
 class AutolinkExtensionTokenizer extends coreTokenizer.BaseInlineTokenizer {
     constructor(props = {}) {
-        var _a, _b;
         super({
-            name: (_a = props.name) !== null && _a !== void 0 ? _a : uniqueName,
-            priority: (_b = props.priority) !== null && _b !== void 0 ? _b : coreTokenizer.TokenizerPriority.LINKS,
+            name: props.name ?? uniqueName,
+            priority: props.priority ?? coreTokenizer.TokenizerPriority.LINKS,
         });
-        this.match = match;
-        this.parse = parse;
     }
+    match = match;
+    parse = parse;
 }
 
 exports.AutolinkExtensionTokenizer = AutolinkExtensionTokenizer;
 exports.AutolinkExtensionTokenizerName = uniqueName;
 exports.autolinkExtensionMatch = match;
 exports.autolinkExtensionParse = parse;
-exports["default"] = AutolinkExtensionTokenizer;
+exports.default = AutolinkExtensionTokenizer;
 exports.eatDomainSegment = eatDomainSegment;
 exports.eatExtendEmailAddress = eatExtendEmailAddress;
 exports.eatExtendedUrl = eatExtendedUrl;

package/lib/esm/{index.js → index.mjs}

@@ -277,14 +277,13 @@ const uniqueName = '@yozora/tokenizer-autolink-extension';
 
 class AutolinkExtensionTokenizer extends BaseInlineTokenizer {
     constructor(props = {}) {
-        var _a, _b;
         super({
-            name: (_a = props.name) !== null && _a !== void 0 ? _a : uniqueName,
-            priority: (_b = props.priority) !== null && _b !== void 0 ? _b : TokenizerPriority.LINKS,
+            name: props.name ?? uniqueName,
+            priority: props.priority ?? TokenizerPriority.LINKS,
         });
-        this.match = match;
-        this.parse = parse;
     }
+    match = match;
+    parse = parse;
 }
 
 export { AutolinkExtensionTokenizer, uniqueName as AutolinkExtensionTokenizerName, match as autolinkExtensionMatch, parse as autolinkExtensionParse, AutolinkExtensionTokenizer as default, eatDomainSegment, eatExtendEmailAddress, eatExtendedUrl, eatOptionalDomainFollows, eatValidDomain, eatWWWDomain };

package/lib/types/index.d.ts

@@ -58,10 +58,10 @@ declare function eatDomainSegment(nodePoints: ReadonlyArray<INodePoint>, startIn
     hasUnderscore: boolean;
 };
 
-declare type T = LinkType;
-declare type INode = Link;
+type T = LinkType;
+type INode = Link;
 declare const uniqueName = "@yozora/tokenizer-autolink-extension";
-declare type AutolinkExtensionContentType = AutolinkContentType | 'uri-www';
+type AutolinkExtensionContentType = AutolinkContentType | 'uri-www';
 interface IToken extends IPartialYastInlineToken<T> {
     /**
      * Autolink content type: absolute uri or email.
@@ -75,8 +75,8 @@ interface IDelimiter extends IYastTokenDelimiter {
      */
     contentType: AutolinkExtensionContentType;
 }
-declare type IThis = ITokenizer;
-declare type ITokenizerProps = Partial<IBaseInlineTokenizerProps>;
+type IThis = ITokenizer;
+type ITokenizerProps = Partial<IBaseInlineTokenizerProps>;
 
 /**
  * @see https://github.github.com/gfm/#autolinks-extension-

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@yozora/tokenizer-autolink-extension",
-  "version": "2.0.4",
+  "version": "2.0.5",
   "author": {
     "name": "guanghechen",
     "url": "https://github.com/guanghechen/"
@@ -11,34 +11,38 @@
     "directory": "tokenizers/autolink-extension"
   },
   "homepage": "https://github.com/yozorajs/yozora/tree/release-2.x.x/tokenizers/autolink-extension",
-  "main": "lib/cjs/index.js",
-  "module": "lib/esm/index.js",
-  "types": "lib/types/index.d.ts",
-  "source": "src/index.ts",
+  "type": "module",
+  "exports": {
+    "types": "./lib/types/index.d.ts",
+    "import": "./lib/esm/index.mjs",
+    "require": "./lib/cjs/index.cjs"
+  },
+  "source": "./src/index.ts",
+  "types": "./lib/types/index.d.ts",
+  "main": "./lib/cjs/index.cjs",
+  "module": "./lib/esm/index.mjs",
   "license": "MIT",
   "engines": {
     "node": ">= 16.0.0"
   },
   "files": [
     "lib/",
-    "!lib/**/*.js.map",
-    "!lib/**/*.d.ts.map",
+    "src/",
     "package.json",
     "CHANGELOG.md",
     "LICENSE",
     "README.md"
   ],
   "scripts": {
-    "build": "cross-env NODE_ENV=production rollup -c ../../rollup.config.js",
-    "prebuild": "rimraf lib/",
+    "build": "rimraf lib/ && cross-env NODE_ENV=production rollup -c ../../rollup.config.mjs",
     "prepublishOnly": "cross-env ROLLUP_SHOULD_SOURCEMAP=false yarn build",
-    "test": "cross-env TS_NODE_FILES=true jest --config ../../jest.config.js --rootDir ."
+    "test": "cross-env TS_NODE_FILES=true NODE_OPTIONS=--experimental-vm-modules jest --config ../../jest.config.mjs --rootDir ."
   },
   "dependencies": {
-    "@yozora/ast": "^2.0.4",
-    "@yozora/character": "^2.0.4",
-    "@yozora/core-tokenizer": "^2.0.4",
-    "@yozora/tokenizer-autolink": "^2.0.4"
+    "@yozora/ast": "^2.0.5",
+    "@yozora/character": "^2.0.5",
+    "@yozora/core-tokenizer": "^2.0.5",
+    "@yozora/tokenizer-autolink": "^2.0.5"
   },
-  "gitHead": "c980b95254394dcacba0cbb4bea251350b09397c"
+  "gitHead": "7ba3bab49fe65cf2f57082c0503af73da9356cf0"
 }

package/src/index.ts

@@ -0,0 +1,11 @@
+export * from './util/email'
+export * from './util/uri'
+export { match as autolinkExtensionMatch } from './match'
+export { parse as autolinkExtensionParse } from './parse'
+export { AutolinkExtensionTokenizer, AutolinkExtensionTokenizer as default } from './tokenizer'
+export { uniqueName as AutolinkExtensionTokenizerName } from './types'
+export type {
+  IThis as IAutolinkExtensionHookContext,
+  IToken as IAutolinkExtensionToken,
+  ITokenizerProps as IAutolinkExtensionTokenizerProps,
+} from './types'

package/src/match.ts

@@ -0,0 +1,114 @@
+import { LinkType } from '@yozora/ast'
+import type { INodePoint } from '@yozora/character'
+import { AsciiCodePoint, isWhitespaceCharacter } from '@yozora/character'
+import type {
+  IMatchInlineHookCreator,
+  IResultOfProcessSingleDelimiter,
+} from '@yozora/core-tokenizer'
+import { genFindDelimiter } from '@yozora/core-tokenizer'
+import type {
+  AutolinkExtensionContentType,
+  ContentHelper,
+  IDelimiter,
+  IThis,
+  IToken,
+  T,
+} from './types'
+import { eatExtendEmailAddress } from './util/email'
+import { eatExtendedUrl, eatWWWDomain } from './util/uri'
+
+const helpers: ReadonlyArray<ContentHelper> = [
+  { contentType: 'uri', eat: eatExtendedUrl },
+  { contentType: 'uri-www', eat: eatWWWDomain },
+  { contentType: 'email', eat: eatExtendEmailAddress },
+]
+
+/**
+ * @see https://github.github.com/gfm/#autolinks-extension-
+ */
+export const match: IMatchInlineHookCreator<T, IDelimiter, IToken, IThis> = function (api) {
+  return {
+    findDelimiter: () => genFindDelimiter<IDelimiter>(_findDelimiter),
+    processSingleDelimiter,
+  }
+
+  function _findDelimiter(startIndex: number, endIndex: number): IDelimiter | null {
+    const nodePoints: ReadonlyArray<INodePoint> = api.getNodePoints()
+    const blockStartIndex: number = api.getBlockStartIndex()
+    for (let i = startIndex; i < endIndex; ++i) {
+      /**
+       * Autolinks can also be constructed without requiring the use of '<' and
+       * to '>' to delimit them, although they will be recognized under a
+       * smaller set of circumstances. All such recognized autolinks can only
+       * come at the beginning of a line, after whitespace, or any of the
+       * delimiting characters '*', '_', '~', and '('.
+       * @see https://github.github.com/gfm/#autolinks-extension-
+       */
+      {
+        let j = i
+        let flag = false
+        for (; j < endIndex; ++j) {
+          const c = nodePoints[j].codePoint
+          if (
+            isWhitespaceCharacter(c) ||
+            c === AsciiCodePoint.ASTERISK ||
+            c === AsciiCodePoint.UNDERSCORE ||
+            c === AsciiCodePoint.TILDE ||
+            c === AsciiCodePoint.OPEN_PARENTHESIS
+          ) {
+            flag = true
+            continue
+          }
+
+          if (flag || j === blockStartIndex) break
+          flag = false
+        }
+
+        if (j >= endIndex) break
+        i = j
+      }
+
+      let nextIndex: number = endIndex
+      let contentType: AutolinkExtensionContentType | null = null
+      for (const helper of helpers) {
+        const eatResult = helper.eat(nodePoints, i, endIndex)
+        nextIndex = Math.min(nextIndex, eatResult.nextIndex)
+        if (eatResult.valid) {
+          contentType = helper.contentType
+          nextIndex = eatResult.nextIndex
+          break
+        }
+      }
+
+      // Optimization: move forward to the next latest potential position.
+      if (contentType == null) {
+        i = Math.max(i, nextIndex - 1)
+        continue
+      }
+
+      if (nextIndex <= endIndex) {
+        return {
+          type: 'full',
+          startIndex: i,
+          endIndex: nextIndex,
+          contentType,
+        }
+      }
+      i = nextIndex - 1
+    }
+    return null
+  }
+
+  function processSingleDelimiter(
+    delimiter: IDelimiter,
+  ): IResultOfProcessSingleDelimiter<T, IToken> {
+    const token: IToken = {
+      nodeType: LinkType,
+      startIndex: delimiter.startIndex,
+      endIndex: delimiter.endIndex,
+      contentType: delimiter.contentType,
+      children: api.resolveFallbackTokens([], delimiter.startIndex, delimiter.endIndex),
+    }
+    return [token]
+  }
+}

package/src/parse.ts

@@ -0,0 +1,35 @@
+import type { Node } from '@yozora/ast'
+import { LinkType } from '@yozora/ast'
+import type { INodePoint } from '@yozora/character'
+import { calcStringFromNodePoints } from '@yozora/character'
+import type { IParseInlineHookCreator } from '@yozora/core-tokenizer'
+import type { INode, IThis, IToken, T } from './types'
+
+export const parse: IParseInlineHookCreator<T, IToken, INode, IThis> = function (api) {
+  return {
+    parse: tokens =>
+      tokens.map(token => {
+        const nodePoints: ReadonlyArray<INodePoint> = api.getNodePoints()
+
+        // Backslash-escapes do not work inside autolink.
+        let url = calcStringFromNodePoints(nodePoints, token.startIndex, token.endIndex)
+
+        switch (token.contentType) {
+          // Add 'mailto:' prefix to email address type autolink.
+          case 'email':
+            url = 'mailto:' + url
+            break
+          // Add 'http://' prefix to email address type autolink.
+          case 'uri-www':
+            url = 'http://' + url
+            break
+        }
+
+        const children: Node[] = api.parseInlineTokens(token.children)
+        const node: INode = api.shouldReservePosition
+          ? { type: LinkType, position: api.calcPosition(token), url, children }
+          : { type: LinkType, url, children }
+        return node
+      }),
+  }
+}

package/src/tokenizer.ts

@@ -0,0 +1,36 @@
+import type {
+  IInlineTokenizer,
+  IMatchInlineHookCreator,
+  IParseInlineHookCreator,
+} from '@yozora/core-tokenizer'
+import { BaseInlineTokenizer, TokenizerPriority } from '@yozora/core-tokenizer'
+import { match } from './match'
+import { parse } from './parse'
+import type { IDelimiter, INode, IThis, IToken, ITokenizerProps, T } from './types'
+import { uniqueName } from './types'
+
+/**
+ * Lexical Analyzer for Autolink (extension).
+ *
+ * @see https://github.github.com/gfm/#autolinks-extension-
+ */
+export class AutolinkExtensionTokenizer
+  extends BaseInlineTokenizer<T, IDelimiter, IToken, INode, IThis>
+  implements IInlineTokenizer<T, IDelimiter, IToken, INode, IThis>
+{
+  /* istanbul ignore next */
+  constructor(props: ITokenizerProps = {}) {
+    super({
+      name: props.name ?? uniqueName,
+      /**
+       * Autolink has the same priority as links.
+       * @see https://github.github.com/gfm/#example-509
+       */
+      priority: props.priority ?? TokenizerPriority.LINKS,
+    })
+  }
+
+  public override readonly match: IMatchInlineHookCreator<T, IDelimiter, IToken, IThis> = match
+
+  public override readonly parse: IParseInlineHookCreator<T, IToken, INode, IThis> = parse
+}

package/src/types.ts

@@ -0,0 +1,47 @@
+import type { Link, LinkType } from '@yozora/ast'
+import type { INodePoint } from '@yozora/character'
+import type {
+  IBaseInlineTokenizerProps,
+  IPartialYastInlineToken,
+  IResultOfRequiredEater,
+  ITokenizer,
+  IYastTokenDelimiter,
+} from '@yozora/core-tokenizer'
+import type { AutolinkContentType } from '@yozora/tokenizer-autolink'
+
+export type T = LinkType
+export type INode = Link
+export const uniqueName = '@yozora/tokenizer-autolink-extension'
+
+// Content type of autolink
+export type AutolinkExtensionContentType = AutolinkContentType | 'uri-www'
+
+export interface IToken extends IPartialYastInlineToken<T> {
+  /**
+   * Autolink content type: absolute uri or email.
+   */
+  contentType: AutolinkExtensionContentType
+}
+
+export interface IDelimiter extends IYastTokenDelimiter {
+  type: 'full'
+  /**
+   * Autolink and autolink-extension content types.
+   */
+  contentType: AutolinkExtensionContentType
+}
+
+export type IThis = ITokenizer
+
+export type ITokenizerProps = Partial<IBaseInlineTokenizerProps>
+
+export type ContentEater = (
+  nodePoints: ReadonlyArray<INodePoint>,
+  startIndex: number,
+  endIndex: number,
+) => IResultOfRequiredEater
+
+export interface ContentHelper {
+  contentType: AutolinkExtensionContentType
+  eat: ContentEater
+}

package/src/util/email.ts

@@ -0,0 +1,78 @@
+import type { INodePoint } from '@yozora/character'
+import { AsciiCodePoint, isAlphanumeric } from '@yozora/character'
+import type { IResultOfRequiredEater } from '@yozora/core-tokenizer'
+
+/**
+ * An extended email autolink will be recognised when an email address is
+ * recognised within any text node. Email addresses are recognised according to
+ * the following rules:
+ *
+ *  - One ore more characters which are alphanumeric, or '.', '-', '_', or '+'.
+ *  - An '@' symbol.
+ *  - One or more characters which are alphanumeric, or '-' or '_', separated
+ *    by periods (.). There must be at least one period. The last character must
+ *    not be one of '-' or '_'.
+ *
+ * @see https://github.github.com/gfm/#extended-email-autolink
+ */
+export function eatExtendEmailAddress(
+  nodePoints: ReadonlyArray<INodePoint>,
+  startIndex: number,
+  endIndex: number,
+): IResultOfRequiredEater {
+  let i = startIndex
+  if (i >= endIndex || !isAlphanumeric(nodePoints[i].codePoint)) {
+    return { valid: false, nextIndex: i + 1 }
+  }
+
+  for (i += 1; i < endIndex; i += 1) {
+    const c = nodePoints[i].codePoint
+    if (
+      isAlphanumeric(c) ||
+      c === AsciiCodePoint.DOT ||
+      c === AsciiCodePoint.MINUS_SIGN ||
+      c === AsciiCodePoint.UNDERSCORE ||
+      c === AsciiCodePoint.PLUS_SIGN
+    )
+      continue
+    break
+  }
+
+  // Match an '@' symbol.
+  if (
+    i === startIndex ||
+    i + 2 >= endIndex ||
+    nodePoints[i].codePoint !== AsciiCodePoint.AT_SIGN ||
+    !isAlphanumeric(nodePoints[i + 1].codePoint)
+  )
+    return { valid: false, nextIndex: i + 1 }
+
+  let countOfPeriod = 0
+  for (i += 2; i < endIndex; i += 1) {
+    const c = nodePoints[i].codePoint
+    if (c === AsciiCodePoint.DOT) {
+      countOfPeriod += 1
+      continue
+    }
+
+    if (isAlphanumeric(c) || c === AsciiCodePoint.MINUS_SIGN || c === AsciiCodePoint.UNDERSCORE)
+      continue
+    break
+  }
+
+  // '.', '-', and '_' can occur on both sides of the '@', but only '.' may
+  // occur at the end of the email address, in which case it will not be
+  // considered part of the address.
+  const lastCharacter = nodePoints[i - 1].codePoint
+  if (lastCharacter === AsciiCodePoint.MINUS_SIGN || lastCharacter === AsciiCodePoint.UNDERSCORE)
+    return { valid: false, nextIndex: i }
+
+  if (lastCharacter === AsciiCodePoint.DOT) {
+    i -= 1
+    countOfPeriod -= 1
+  }
+
+  // There must be at least one period.
+  if (countOfPeriod <= 0) return { valid: false, nextIndex: i }
+  return { valid: true, nextIndex: i }
+}

package/src/util/uri.ts

@@ -0,0 +1,235 @@
+import type { INodePoint } from '@yozora/character'
+import {
+  AsciiCodePoint,
+  isAlphanumeric,
+  isPunctuationCharacter,
+  isWhitespaceCharacter,
+} from '@yozora/character'
+import type { IResultOfOptionalEater, IResultOfRequiredEater } from '@yozora/core-tokenizer'
+import { eatAutolinkSchema } from '@yozora/tokenizer-autolink'
+
+/**
+ * An extended url autolink will be recognised when one of the schemes 'http://',
+ * or 'https://', followed by a valid domain, then zero or more non-space non-<
+ * characters according to extended autolink path validation.
+ *
+ * @see https://github.github.com/gfm/#extended-url-autolink
+ */
+export function eatExtendedUrl(
+  nodePoints: ReadonlyArray<INodePoint>,
+  startIndex: number,
+  endIndex: number,
+): IResultOfRequiredEater {
+  const schema = eatAutolinkSchema(nodePoints, startIndex, endIndex)
+  const { nextIndex } = schema
+
+  if (
+    !schema.valid ||
+    nextIndex + 3 >= endIndex ||
+    nodePoints[nextIndex].codePoint !== AsciiCodePoint.COLON ||
+    nodePoints[nextIndex + 1].codePoint !== AsciiCodePoint.SLASH ||
+    nodePoints[nextIndex + 2].codePoint !== AsciiCodePoint.SLASH
+  )
+    return { valid: false, nextIndex: nextIndex + 1 }
+
+  const result = eatValidDomain(nodePoints, nextIndex + 3, endIndex)
+  result.nextIndex = eatOptionalDomainFollows(nodePoints, result.nextIndex, endIndex)
+  return result
+}
+
+/**
+ * An extended www autolink will be recognised when the text 'www.' is found
+ * followed by a valid domain
+ */
+export function eatWWWDomain(
+  nodePoints: ReadonlyArray<INodePoint>,
+  startIndex: number,
+  endIndex: number,
+): IResultOfRequiredEater {
+  const segment = eatDomainSegment(nodePoints, startIndex, endIndex)
+  const nextIndex = segment.nextIndex
+
+  if (
+    !segment.valid ||
+    nextIndex >= endIndex ||
+    nodePoints[nextIndex].codePoint !== AsciiCodePoint.DOT ||
+    nextIndex - startIndex !== 3
+  )
+    return { valid: false, nextIndex }
+
+  for (let i = startIndex; i < nextIndex; ++i) {
+    const c = nodePoints[i].codePoint
+    if (c !== AsciiCodePoint.LOWERCASE_W && c !== AsciiCodePoint.UPPERCASE_W)
+      return { valid: false, nextIndex }
+  }
+
+  const result = eatValidDomain(nodePoints, nextIndex + 1, endIndex)
+  result.nextIndex = eatOptionalDomainFollows(nodePoints, result.nextIndex, endIndex)
+  return result
+}
+
+/**
+ * Try to eat an optional domain follows.
+ *
+ * After a valid domain, zero or more non-space non-'<' characters may follow.
+ *
+ * @param nodePoints
+ * @param startIndex
+ * @param endIndex
+ * @see https://github.github.com/gfm/#extended-autolink-path-validation
+ */
+export function eatOptionalDomainFollows(
+  nodePoints: ReadonlyArray<INodePoint>,
+  startIndex: number,
+  endIndex: number,
+): IResultOfOptionalEater {
+  let nextIndex = startIndex
+  for (; nextIndex < endIndex; ++nextIndex) {
+    const c = nodePoints[nextIndex].codePoint
+    if (isWhitespaceCharacter(c) || c === AsciiCodePoint.OPEN_ANGLE) break
+  }
+
+  // Trailing punctuation (specifically, '?', '!', '.', ',', ':', '*', '_', and '~')
+  // will not be considered part of the autolink, though they may be included in
+  // the interior of the link
+  for (nextIndex -= 1; nextIndex >= startIndex; nextIndex -= 1) {
+    const c = nodePoints[nextIndex].codePoint
+    if (
+      isPunctuationCharacter(c) ||
+      c === AsciiCodePoint.QUESTION_MARK ||
+      c === AsciiCodePoint.EXCLAMATION_MARK ||
+      c === AsciiCodePoint.DOT ||
+      c === AsciiCodePoint.COMMA ||
+      c === AsciiCodePoint.COLON ||
+      c === AsciiCodePoint.ASTERISK ||
+      c === AsciiCodePoint.UNDERSCORE ||
+      c === AsciiCodePoint.TILDE
+    )
+      continue
+    break
+  }
+
+  /**
+   * When an autolink ends in ')', we scan the entire autolink for the total
+   * number of parentheses. If there is a greater number of closing parentheses
+   * than opening ones, we don’t consider the unmatched trailing parentheses
+   * part of the autolink, in order to facilitate including an autolink inside
+   * a parenthesis.
+   * @see https://github.github.com/gfm/#example-624
+   * @see https://github.github.com/gfm/#example-625
+   */
+  if (
+    nextIndex >= startIndex &&
+    nextIndex + 1 < endIndex &&
+    nodePoints[nextIndex + 1].codePoint === AsciiCodePoint.CLOSE_PARENTHESIS
+  ) {
+    let countOfOpenParenthesis = 0
+    for (let i = startIndex; i < nextIndex; ++i) {
+      const c = nodePoints[i].codePoint
+      switch (c) {
+        case AsciiCodePoint.OPEN_PARENTHESIS:
+          countOfOpenParenthesis += 1
+          break
+        case AsciiCodePoint.CLOSE_PARENTHESIS:
+          countOfOpenParenthesis -= 1
+          break
+      }
+    }
+
+    if (countOfOpenParenthesis > 0) {
+      nextIndex += 2
+      countOfOpenParenthesis -= 1
+      for (; nextIndex < endIndex && countOfOpenParenthesis > 0; ) {
+        const c = nodePoints[nextIndex].codePoint
+        if (c !== AsciiCodePoint.CLOSE_PARENTHESIS) break
+        countOfOpenParenthesis -= 1
+        nextIndex += 1
+      }
+      nextIndex -= 1
+    }
+  }
+
+  /**
+   * If an autolink ends in a semicolon (;), we check to see if it appears to
+   * resemble an entity reference; if the preceding text is & followed by one
+   * or more alphanumeric characters. If so, it is excluded from the autolink.
+   * @see https://github.github.com/gfm/#example-626
+   */
+  if (
+    nextIndex + 1 < endIndex &&
+    nodePoints[nextIndex + 1].codePoint === AsciiCodePoint.SEMICOLON
+  ) {
+    let i = nextIndex
+    for (; i >= startIndex; --i) {
+      const c = nodePoints[i].codePoint
+      if (!isAlphanumeric(c)) break
+    }
+    if (i >= startIndex && nodePoints[i].codePoint === AsciiCodePoint.AMPERSAND) nextIndex = i - 1
+  }
+
+  return nextIndex + 1
+}
+
+/**
+ * A valid domain consists of segments of alphanumeric characters,
+ * underscores (_) and hyphens (-) separated by periods (.).
+ *
+ * @see https://github.github.com/gfm/#valid-domain
+ */
+export function eatValidDomain(
+  nodePoints: ReadonlyArray<INodePoint>,
+  startIndex: number,
+  endIndex: number,
+): IResultOfRequiredEater {
+  const segment = eatDomainSegment(nodePoints, startIndex, endIndex)!
+  if (!segment.valid || segment.nextIndex >= endIndex) {
+    return { valid: false, nextIndex: segment.nextIndex }
+  }
+
+  let nextIndex = segment.nextIndex,
+    countOfPeriod = 0
+  let countOfUnderscoreOfLastTwoSegment = segment.hasUnderscore ? 2 : 0
+  for (; nextIndex < endIndex; ) {
+    if (nodePoints[nextIndex].codePoint !== AsciiCodePoint.DOT) break
+
+    const segment = eatDomainSegment(nodePoints, nextIndex + 1, endIndex)
+    if (!segment.valid) break
+
+    nextIndex = segment.nextIndex
+    countOfPeriod += 1
+    countOfUnderscoreOfLastTwoSegment >>>= 1
+    countOfUnderscoreOfLastTwoSegment |= segment.hasUnderscore ? 2 : 0
+  }
+
+  // There must be at least one period, and no underscores may be present in the
+  // last two segments of the domain.
+  if (countOfPeriod <= 0 && countOfUnderscoreOfLastTwoSegment === 0) {
+    return { valid: false, nextIndex }
+  }
+  return { valid: true, nextIndex }
+}
+
+/**
+ * A valid domain segment consists of alphanumeric characters,
+ * underscores (_) and hyphens (-).
+ * @see https://github.github.com/gfm/#valid-domain
+ */
+export function eatDomainSegment(
+  nodePoints: ReadonlyArray<INodePoint>,
+  startIndex: number,
+  endIndex: number,
+): IResultOfRequiredEater & { hasUnderscore: boolean } {
+  let i = startIndex,
+    hasUnderscore = false
+  for (; i < endIndex; ++i) {
+    const c = nodePoints[i].codePoint
+    if (c === AsciiCodePoint.UNDERSCORE) {
+      hasUnderscore = true
+      continue
+    }
+    if (!isAlphanumeric(c) && c !== AsciiCodePoint.MINUS_SIGN) break
+  }
+
+  if (i > startIndex) return { valid: true, nextIndex: i, hasUnderscore }
+  return { valid: false, nextIndex: i, hasUnderscore }
+}