MZC/KnowMathBot/frontend/node_modules/micromark/dev/lib/initialize/document.js

/**
 * @import {
 *   Construct,
 *   ContainerState,
 *   InitialConstruct,
 *   Initializer,
 *   Point,
 *   State,
 *   TokenizeContext,
 *   Tokenizer,
 *   Token
 * } from 'micromark-util-types'
 */

/**
 * @typedef {[Construct, ContainerState]} StackItem
 *   Construct and its state.
 */

import {ok as assert} from 'devlop'
import {factorySpace} from 'micromark-factory-space'
import {markdownLineEnding} from 'micromark-util-character'
import {splice} from 'micromark-util-chunked'
import {codes, constants, types} from 'micromark-util-symbol'

/** @type {InitialConstruct} */
export const document = {tokenize: initializeDocument}

/** @type {Construct} */
const containerConstruct = {tokenize: tokenizeContainer}

/**
 * @this {TokenizeContext}
 *   Self.
 * @type {Initializer}
 *   Initializer.
 */
function initializeDocument(effects) {
  const self = this
  /** @type {Array<StackItem>} */
  const stack = []
  let continued = 0
  /** @type {TokenizeContext | undefined} */
  let childFlow
  /** @type {Token | undefined} */
  let childToken
  /** @type {number} */
  let lineStartOffset

  return start

  /** @type {State} */
  function start(code) {
    // First we iterate through the open blocks, starting with the root
    // document, and descending through last children down to the last open
    // block.
    // Each block imposes a condition that the line must satisfy if the block is
    // to remain open.
    // For example, a block quote requires a `>` character.
    // A paragraph requires a non-blank line.
    // In this phase we may match all or just some of the open blocks.
    // But we cannot close unmatched blocks yet, because we may have a lazy
    // continuation line.
    if (continued < stack.length) {
      const item = stack[continued]
      self.containerState = item[1]
      assert(
        item[0].continuation,
        'expected `continuation` to be defined on container construct'
      )
      return effects.attempt(
        item[0].continuation,
        documentContinue,
        checkNewContainers
      )(code)
    }

    // Done.
    return checkNewContainers(code)
  }

  /** @type {State} */
  function documentContinue(code) {
    assert(
      self.containerState,
      'expected `containerState` to be defined after continuation'
    )

    continued++

    // Note: this field is called `_closeFlow` but it also closes containers.
    // Perhaps a good idea to rename it but it’s already used in the wild by
    // extensions.
    if (self.containerState._closeFlow) {
      self.containerState._closeFlow = undefined

      if (childFlow) {
        closeFlow()
      }

      // Note: this algorithm for moving events around is similar to the
      // algorithm when dealing with lazy lines in `writeToChild`.
      const indexBeforeExits = self.events.length
      let indexBeforeFlow = indexBeforeExits
      /** @type {Point | undefined} */
      let point

      // Find the flow chunk.
      while (indexBeforeFlow--) {
        if (
          self.events[indexBeforeFlow][0] === 'exit' &&
          self.events[indexBeforeFlow][1].type === types.chunkFlow
        ) {
          point = self.events[indexBeforeFlow][1].end
          break
        }
      }

      assert(point, 'could not find previous flow chunk')

      exitContainers(continued)

      // Fix positions.
      let index = indexBeforeExits

      while (index < self.events.length) {
        self.events[index][1].end = {...point}
        index++
      }

      // Inject the exits earlier (they’re still also at the end).
      splice(
        self.events,
        indexBeforeFlow + 1,
        0,
        self.events.slice(indexBeforeExits)
      )

      // Discard the duplicate exits.
      self.events.length = index

      return checkNewContainers(code)
    }

    return start(code)
  }

  /** @type {State} */
  function checkNewContainers(code) {
    // Next, after consuming the continuation markers for existing blocks, we
    // look for new block starts (e.g. `>` for a block quote).
    // If we encounter a new block start, we close any blocks unmatched in
    // step 1 before creating the new block as a child of the last matched
    // block.
    if (continued === stack.length) {
      // No need to `check` whether there’s a container, of `exitContainers`
      // would be moot.
      // We can instead immediately `attempt` to parse one.
      if (!childFlow) {
        return documentContinued(code)
      }

      // If we have concrete content, such as block HTML or fenced code,
      // we can’t have containers “pierce” into them, so we can immediately
      // start.
      if (childFlow.currentConstruct && childFlow.currentConstruct.concrete) {
        return flowStart(code)
      }

      // If we do have flow, it could still be a blank line,
      // but we’d be interrupting it w/ a new container if there’s a current
      // construct.
      // To do: next major: remove `_gfmTableDynamicInterruptHack` (no longer
      // needed in micromark-extension-gfm-table@1.0.6).
      self.interrupt = Boolean(
        childFlow.currentConstruct && !childFlow._gfmTableDynamicInterruptHack
      )
    }

    // Check if there is a new container.
    self.containerState = {}
    return effects.check(
      containerConstruct,
      thereIsANewContainer,
      thereIsNoNewContainer
    )(code)
  }

  /** @type {State} */
  function thereIsANewContainer(code) {
    if (childFlow) closeFlow()
    exitContainers(continued)
    return documentContinued(code)
  }

  /** @type {State} */
  function thereIsNoNewContainer(code) {
    self.parser.lazy[self.now().line] = continued !== stack.length
    lineStartOffset = self.now().offset
    return flowStart(code)
  }

  /** @type {State} */
  function documentContinued(code) {
    // Try new containers.
    self.containerState = {}
    return effects.attempt(
      containerConstruct,
      containerContinue,
      flowStart
    )(code)
  }

  /** @type {State} */
  function containerContinue(code) {
    assert(
      self.currentConstruct,
      'expected `currentConstruct` to be defined on tokenizer'
    )
    assert(
      self.containerState,
      'expected `containerState` to be defined on tokenizer'
    )
    continued++
    stack.push([self.currentConstruct, self.containerState])
    // Try another.
    return documentContinued(code)
  }

  /** @type {State} */
  function flowStart(code) {
    if (code === codes.eof) {
      if (childFlow) closeFlow()
      exitContainers(0)
      effects.consume(code)
      return
    }

    childFlow = childFlow || self.parser.flow(self.now())
    effects.enter(types.chunkFlow, {
      _tokenizer: childFlow,
      contentType: constants.contentTypeFlow,
      previous: childToken
    })

    return flowContinue(code)
  }

  /** @type {State} */
  function flowContinue(code) {
    if (code === codes.eof) {
      writeToChild(effects.exit(types.chunkFlow), true)
      exitContainers(0)
      effects.consume(code)
      return
    }

    if (markdownLineEnding(code)) {
      effects.consume(code)
      writeToChild(effects.exit(types.chunkFlow))
      // Get ready for the next line.
      continued = 0
      self.interrupt = undefined
      return start
    }

    effects.consume(code)
    return flowContinue
  }

  /**
   * @param {Token} token
   *   Token.
   * @param {boolean | undefined} [endOfFile]
   *   Whether the token is at the end of the file (default: `false`).
   * @returns {undefined}
   *   Nothing.
   */
  function writeToChild(token, endOfFile) {
    assert(childFlow, 'expected `childFlow` to be defined when continuing')
    const stream = self.sliceStream(token)
    if (endOfFile) stream.push(null)
    token.previous = childToken
    if (childToken) childToken.next = token
    childToken = token
    childFlow.defineSkip(token.start)
    childFlow.write(stream)

    // Alright, so we just added a lazy line:
    //
    // ```markdown
    // > a
    // b.
    //
    // Or:
    //
    // > ~~~c
    // d
    //
    // Or:
    //
    // > | e |
    // f
    // ```
    //
    // The construct in the second example (fenced code) does not accept lazy
    // lines, so it marked itself as done at the end of its first line, and
    // then the content construct parses `d`.
    // Most constructs in markdown match on the first line: if the first line
    // forms a construct, a non-lazy line can’t “unmake” it.
    //
    // The construct in the third example is potentially a GFM table, and
    // those are *weird*.
    // It *could* be a table, from the first line, if the following line
    // matches a condition.
    // In this case, that second line is lazy, which “unmakes” the first line
    // and turns the whole into one content block.
    //
    // We’ve now parsed the non-lazy and the lazy line, and can figure out
    // whether the lazy line started a new flow block.
    // If it did, we exit the current containers between the two flow blocks.
    if (self.parser.lazy[token.start.line]) {
      let index = childFlow.events.length

      while (index--) {
        if (
          // The token starts before the line ending…
          childFlow.events[index][1].start.offset < lineStartOffset &&
          // …and either is not ended yet…
          (!childFlow.events[index][1].end ||
            // …or ends after it.
            childFlow.events[index][1].end.offset > lineStartOffset)
        ) {
          // Exit: there’s still something open, which means it’s a lazy line
          // part of something.
          return
        }
      }

      // Note: this algorithm for moving events around is similar to the
      // algorithm when closing flow in `documentContinue`.
      const indexBeforeExits = self.events.length
      let indexBeforeFlow = indexBeforeExits
      /** @type {boolean | undefined} */
      let seen
      /** @type {Point | undefined} */
      let point

      // Find the previous chunk (the one before the lazy line).
      while (indexBeforeFlow--) {
        if (
          self.events[indexBeforeFlow][0] === 'exit' &&
          self.events[indexBeforeFlow][1].type === types.chunkFlow
        ) {
          if (seen) {
            point = self.events[indexBeforeFlow][1].end
            break
          }

          seen = true
        }
      }

      assert(point, 'could not find previous flow chunk')

      exitContainers(continued)

      // Fix positions.
      index = indexBeforeExits

      while (index < self.events.length) {
        self.events[index][1].end = {...point}
        index++
      }

      // Inject the exits earlier (they’re still also at the end).
      splice(
        self.events,
        indexBeforeFlow + 1,
        0,
        self.events.slice(indexBeforeExits)
      )

      // Discard the duplicate exits.
      self.events.length = index
    }
  }

  /**
   * @param {number} size
   *   Size.
   * @returns {undefined}
   *   Nothing.
   */
  function exitContainers(size) {
    let index = stack.length

    // Exit open containers.
    while (index-- > size) {
      const entry = stack[index]
      self.containerState = entry[1]
      assert(
        entry[0].exit,
        'expected `exit` to be defined on container construct'
      )
      entry[0].exit.call(self, effects)
    }

    stack.length = size
  }

  function closeFlow() {
    assert(
      self.containerState,
      'expected `containerState` to be defined when closing flow'
    )
    assert(childFlow, 'expected `childFlow` to be defined when closing it')
    childFlow.write([codes.eof])
    childToken = undefined
    childFlow = undefined
    self.containerState._closeFlow = undefined
  }
}

/**
 * @this {TokenizeContext}
 *   Context.
 * @type {Tokenizer}
 *   Tokenizer.
 */
function tokenizeContainer(effects, ok, nok) {
  // Always populated by defaults.
  assert(
    this.parser.constructs.disable.null,
    'expected `disable.null` to be populated'
  )
  return factorySpace(
    effects,
    effects.attempt(this.parser.constructs.document, ok, nok),
    types.linePrefix,
    this.parser.constructs.disable.null.includes('codeIndented')
      ? undefined
      : constants.tabSize
  )
}