@tinta-lang/syntax 0.0.0-stage → 0.1.0-dev.0

package/vendor/micromark-core-commonmark/lib/attention.js

@@ -0,0 +1,231 @@
+/**
+ * @import {
+ *   Code,
+ *   Construct,
+ *   Event,
+ *   Point,
+ *   Resolver,
+ *   State,
+ *   TokenizeContext,
+ *   Tokenizer,
+ *   Token
+ * } from 'micromark-util-types'
+ */
+
+import { push, splice } from 'micromark-util-chunked';
+import { classifyCharacter } from 'micromark-util-classify-character';
+import { resolveAll } from 'micromark-util-resolve-all';
+/** @type {Construct} */
+export const attention = {
+  name: 'attention',
+  resolveAll: resolveAllAttention,
+  tokenize: tokenizeAttention
+};
+
+/**
+ * Take all events and resolve attention to emphasis or strong.
+ *
+ * @type {Resolver}
+ */
+// eslint-disable-next-line complexity
+function resolveAllAttention(events, context) {
+  let index = -1;
+  /** @type {Array<Event>} */
+  let nextEvents;
+
+  // Walk through all events.
+  //
+  // Note: performance of this is fine on an mb of normal markdown, but it’s
+  // a bottleneck for malicious stuff.
+  while (++index < events.length) {
+    // Find a token that can close.
+    if (events[index][0] === 'enter' && events[index][1].type === 'attentionSequence' && events[index][1]._close) {
+      let open = index;
+
+      // Now walk back to find an opener.
+      while (open--) {
+        // Find a token that can open the closer.
+        if (events[open][0] === 'exit' && events[open][1].type === 'attentionSequence' && events[open][1]._open &&
+        // If the markers are the same:
+        context.sliceSerialize(events[open][1]).charCodeAt(0) === context.sliceSerialize(events[index][1]).charCodeAt(0)) {
+          // If the opening can close or the closing can open,
+          // and the close size *is not* a multiple of three,
+          // but the sum of the opening and closing size *is* multiple of three,
+          // then don’t match.
+          if ((events[open][1]._close || events[index][1]._open) && (events[index][1].end.offset - events[index][1].start.offset) % 3 && !((events[open][1].end.offset - events[open][1].start.offset + events[index][1].end.offset - events[index][1].start.offset) % 3)) {
+            continue;
+          }
+
+          // Number of markers to use from the sequence.
+          const use = events[open][1].end.offset - events[open][1].start.offset > 1 && events[index][1].end.offset - events[index][1].start.offset > 1 ? 2 : 1;
+          const start = {
+            ...events[open][1].end
+          };
+          const end = {
+            ...events[index][1].start
+          };
+          movePoint(start, -use);
+          movePoint(end, use);
+          const openingSequence = {
+            type: use > 1 ? "strongSequence" : "emphasisSequence",
+            start,
+            end: {
+              ...events[open][1].end
+            }
+          };
+          const closingSequence = {
+            type: use > 1 ? "strongSequence" : "emphasisSequence",
+            start: {
+              ...events[index][1].start
+            },
+            end
+          };
+          const text = {
+            type: use > 1 ? "strongText" : "emphasisText",
+            start: {
+              ...events[open][1].end
+            },
+            end: {
+              ...events[index][1].start
+            }
+          };
+          const group = {
+            type: use > 1 ? "strong" : "emphasis",
+            start: {
+              ...openingSequence.start
+            },
+            end: {
+              ...closingSequence.end
+            }
+          };
+          events[open][1].end = {
+            ...openingSequence.start
+          };
+          events[index][1].start = {
+            ...closingSequence.end
+          };
+          nextEvents = [];
+
+          // If there are more markers in the opening, add them before.
+          if (events[open][1].end.offset - events[open][1].start.offset) {
+            nextEvents = push(nextEvents, [['enter', events[open][1], context], ['exit', events[open][1], context]]);
+          }
+
+          // Opening.
+          nextEvents = push(nextEvents, [['enter', group, context], ['enter', openingSequence, context], ['exit', openingSequence, context], ['enter', text, context]]);
+
+          // Always populated by defaults.
+
+          // Between.
+          nextEvents = push(nextEvents, resolveAll(context.parser.constructs.insideSpan.null, events.slice(open + 1, index), context));
+
+          // Closing.
+          nextEvents = push(nextEvents, [['exit', text, context], ['enter', closingSequence, context], ['exit', closingSequence, context], ['exit', group, context]]);
+
+          /** @type {number} */
+          let offset = 0;
+
+          // If there are more markers in the closing, add them after.
+          if (events[index][1].end.offset - events[index][1].start.offset) {
+            offset = 2;
+            nextEvents = push(nextEvents, [['enter', events[index][1], context], ['exit', events[index][1], context]]);
+          }
+          splice(events, open - 1, index - open + 3, nextEvents);
+          index = open + nextEvents.length - offset - 2;
+          break;
+        }
+      }
+    }
+  }
+
+  // Remove remaining sequences.
+  index = -1;
+  while (++index < events.length) {
+    if (events[index][1].type === 'attentionSequence') {
+      events[index][1].type = 'data';
+    }
+  }
+  return events;
+}
+
+/**
+ * @this {TokenizeContext}
+ *   Context.
+ * @type {Tokenizer}
+ */
+function tokenizeAttention(effects, ok) {
+  const attentionMarkers = this.parser.constructs.attentionMarkers.null;
+  const previous = this.previous;
+  const before = classifyCharacter(previous);
+
+  /** @type {NonNullable<Code>} */
+  let marker;
+  return start;
+
+  /**
+   * Before a sequence.
+   *
+   * ```markdown
+   * > | **
+   *     ^
+   * ```
+   *
+   * @type {State}
+   */
+  function start(code) {
+    marker = code;
+    effects.enter('attentionSequence');
+    return inside(code);
+  }
+
+  /**
+   * In a sequence.
+   *
+   * ```markdown
+   * > | **
+   *     ^^
+   * ```
+   *
+   * @type {State}
+   */
+  function inside(code) {
+    if (code === marker) {
+      effects.consume(code);
+      return inside;
+    }
+    const token = effects.exit('attentionSequence');
+
+    // To do: next major: move this to resolver, just like `markdown-rs`.
+    const after = classifyCharacter(code);
+
+    // Always populated by defaults.
+
+    // Note: `*` and `_` are in `attentionMarkers` through core.
+    // They are excluded here as the loosening is meant for markers registered
+    // by *other* constructs (such as GFM strikethrough’s `~`).
+    const open = !after || after === 2 && before || attentionMarkers.includes(code) && code !== 42 && code !== 95;
+    const close = !before || before === 2 && after || attentionMarkers.includes(previous) && previous !== 42 && previous !== 95;
+    token._open = Boolean(marker === 42 ? open : open && (before || !close));
+    token._close = Boolean(marker === 42 ? close : close && (after || !open));
+    return ok(code);
+  }
+}
+
+/**
+ * Move a point a bit.
+ *
+ * Note: `move` only works inside lines! It’s not possible to move past other
+ * chunks (replacement characters, tabs, or line endings).
+ *
+ * @param {Point} point
+ *   Point.
+ * @param {number} offset
+ *   Amount to move.
+ * @returns {undefined}
+ *   Nothing.
+ */
+function movePoint(point, offset) {
+  point.column += offset;
+  point.offset += offset;
+  point._bufferIndex += offset;
+}

package/vendor/micromark-core-commonmark/lib/autolink.js

@@ -0,0 +1,233 @@
+/**
+ * @import {
+ *   Construct,
+ *   State,
+ *   TokenizeContext,
+ *   Tokenizer
+ * } from 'micromark-util-types'
+ */
+
+import { asciiAlphanumeric, asciiAlpha, asciiAtext, asciiControl } from 'micromark-util-character';
+/** @type {Construct} */
+export const autolink = {
+  name: 'autolink',
+  tokenize: tokenizeAutolink
+};
+
+/**
+ * @this {TokenizeContext}
+ *   Context.
+ * @type {Tokenizer}
+ */
+function tokenizeAutolink(effects, ok, nok) {
+  let size = 0;
+  return start;
+
+  /**
+   * Start of an autolink.
+   *
+   * ```markdown
+   * > | a<https://example.com>b
+   *      ^
+   * > | a<user@example.com>b
+   *      ^
+   * ```
+   *
+   * @type {State}
+   */
+  function start(code) {
+    effects.enter("autolink");
+    effects.enter("autolinkMarker");
+    effects.consume(code);
+    effects.exit("autolinkMarker");
+    effects.enter("autolinkProtocol");
+    return open;
+  }
+
+  /**
+   * After `<`, at protocol or atext.
+   *
+   * ```markdown
+   * > | a<https://example.com>b
+   *       ^
+   * > | a<user@example.com>b
+   *       ^
+   * ```
+   *
+   * @type {State}
+   */
+  function open(code) {
+    if (asciiAlpha(code)) {
+      effects.consume(code);
+      return schemeOrEmailAtext;
+    }
+    if (code === 64) {
+      return nok(code);
+    }
+    return emailAtext(code);
+  }
+
+  /**
+   * At second byte of protocol or atext.
+   *
+   * ```markdown
+   * > | a<https://example.com>b
+   *        ^
+   * > | a<user@example.com>b
+   *        ^
+   * ```
+   *
+   * @type {State}
+   */
+  function schemeOrEmailAtext(code) {
+    // ASCII alphanumeric and `+`, `-`, and `.`.
+    if (code === 43 || code === 45 || code === 46 || asciiAlphanumeric(code)) {
+      // Count the previous alphabetical from `open` too.
+      size = 1;
+      return schemeInsideOrEmailAtext(code);
+    }
+    return emailAtext(code);
+  }
+
+  /**
+   * In ambiguous protocol or atext.
+   *
+   * ```markdown
+   * > | a<https://example.com>b
+   *        ^
+   * > | a<user@example.com>b
+   *        ^
+   * ```
+   *
+   * @type {State}
+   */
+  function schemeInsideOrEmailAtext(code) {
+    if (code === 58) {
+      effects.consume(code);
+      size = 0;
+      return urlInside;
+    }
+
+    // ASCII alphanumeric and `+`, `-`, and `.`.
+    if ((code === 43 || code === 45 || code === 46 || asciiAlphanumeric(code)) && size++ < 32) {
+      effects.consume(code);
+      return schemeInsideOrEmailAtext;
+    }
+    size = 0;
+    return emailAtext(code);
+  }
+
+  /**
+   * After protocol, in URL.
+   *
+   * ```markdown
+   * > | a<https://example.com>b
+   *             ^
+   * ```
+   *
+   * @type {State}
+   */
+  function urlInside(code) {
+    if (code === 62) {
+      effects.exit("autolinkProtocol");
+      effects.enter("autolinkMarker");
+      effects.consume(code);
+      effects.exit("autolinkMarker");
+      effects.exit("autolink");
+      return ok;
+    }
+
+    // ASCII control, space, or `<`.
+    if (code === null || code === 32 || code === 60 || asciiControl(code)) {
+      return nok(code);
+    }
+    effects.consume(code);
+    return urlInside;
+  }
+
+  /**
+   * In email atext.
+   *
+   * ```markdown
+   * > | a<user.name@example.com>b
+   *              ^
+   * ```
+   *
+   * @type {State}
+   */
+  function emailAtext(code) {
+    if (code === 64) {
+      effects.consume(code);
+      return emailAtSignOrDot;
+    }
+    if (asciiAtext(code)) {
+      effects.consume(code);
+      return emailAtext;
+    }
+    return nok(code);
+  }
+
+  /**
+   * In label, after at-sign or dot.
+   *
+   * ```markdown
+   * > | a<user.name@example.com>b
+   *                 ^       ^
+   * ```
+   *
+   * @type {State}
+   */
+  function emailAtSignOrDot(code) {
+    return asciiAlphanumeric(code) ? emailLabel(code) : nok(code);
+  }
+
+  /**
+   * In label, where `.` and `>` are allowed.
+   *
+   * ```markdown
+   * > | a<user.name@example.com>b
+   *                   ^
+   * ```
+   *
+   * @type {State}
+   */
+  function emailLabel(code) {
+    if (code === 46) {
+      effects.consume(code);
+      size = 0;
+      return emailAtSignOrDot;
+    }
+    if (code === 62) {
+      // Exit, then change the token type.
+      effects.exit("autolinkProtocol").type = "autolinkEmail";
+      effects.enter("autolinkMarker");
+      effects.consume(code);
+      effects.exit("autolinkMarker");
+      effects.exit("autolink");
+      return ok;
+    }
+    return emailValue(code);
+  }
+
+  /**
+   * In label, where `.` and `>` are *not* allowed.
+   *
+   * Though, this is also used in `emailLabel` to parse other values.
+   *
+   * ```markdown
+   * > | a<user.name@ex-ample.com>b
+   *                    ^
+   * ```
+   *
+   * @type {State}
+   */
+  function emailValue(code) {
+    // ASCII alphanumeric or `-`.
+    if ((code === 45 || asciiAlphanumeric(code)) && size++ < 63) {
+      const next = code === 45 ? emailValue : emailLabel;
+      effects.consume(code);
+      return next;
+    }
+    return nok(code);
+  }
+}

package/vendor/micromark-core-commonmark/lib/blank-line.js

@@ -0,0 +1,61 @@
+/**
+ * @import {
+ *   Construct,
+ *   State,
+ *   TokenizeContext,
+ *   Tokenizer
+ * } from 'micromark-util-types'
+ */
+
+import { factorySpace } from 'micromark-factory-space';
+import { markdownLineEnding, markdownSpace } from 'micromark-util-character';
+/** @type {Construct} */
+export const blankLine = {
+  partial: true,
+  tokenize: tokenizeBlankLine
+};
+
+/**
+ * @this {TokenizeContext}
+ *   Context.
+ * @type {Tokenizer}
+ */
+function tokenizeBlankLine(effects, ok, nok) {
+  return start;
+
+  /**
+   * Start of blank line.
+   *
+   * > 👉 **Note**: `␠` represents a space character.
+   *
+   * ```markdown
+   * > | ␠␠␊
+   *     ^
+   * > | ␊
+   *     ^
+   * ```
+   *
+   * @type {State}
+   */
+  function start(code) {
+    return markdownSpace(code) ? factorySpace(effects, after, "linePrefix")(code) : after(code);
+  }
+
+  /**
+   * At eof/eol, after optional whitespace.
+   *
+   * > 👉 **Note**: `␠` represents a space character.
+   *
+   * ```markdown
+   * > | ␠␠␊
+   *       ^
+   * > | ␊
+   *     ^
+   * ```
+   *
+   * @type {State}
+   */
+  function after(code) {
+    return code === null || markdownLineEnding(code) ? ok(code) : nok(code);
+  }
+}

package/vendor/micromark-core-commonmark/lib/block-quote.js

@@ -0,0 +1,143 @@
+/**
+ * @import {
+ *   Construct,
+ *   Exiter,
+ *   State,
+ *   TokenizeContext,
+ *   Tokenizer
+ * } from 'micromark-util-types'
+ */
+
+import { factorySpace } from 'micromark-factory-space';
+import { markdownSpace } from 'micromark-util-character';
+/** @type {Construct} */
+export const blockQuote = {
+  continuation: {
+    tokenize: tokenizeBlockQuoteContinuation
+  },
+  exit,
+  name: 'blockQuote',
+  tokenize: tokenizeBlockQuoteStart
+};
+
+/**
+ * @this {TokenizeContext}
+ *   Context.
+ * @type {Tokenizer}
+ */
+function tokenizeBlockQuoteStart(effects, ok, nok) {
+  const self = this;
+  return start;
+
+  /**
+   * Start of block quote.
+   *
+   * ```markdown
+   * > | > a
+   *     ^
+   * ```
+   *
+   * @type {State}
+   */
+  function start(code) {
+    if (code === 62) {
+      const state = self.containerState;
+      if (!state.open) {
+        effects.enter("blockQuote", {
+          _container: true
+        });
+        state.open = true;
+      }
+      effects.enter("blockQuotePrefix");
+      effects.enter("blockQuoteMarker");
+      effects.consume(code);
+      effects.exit("blockQuoteMarker");
+      return after;
+    }
+    return nok(code);
+  }
+
+  /**
+   * After `>`, before optional whitespace.
+   *
+   * ```markdown
+   * > | > a
+   *      ^
+   * ```
+   *
+   * @type {State}
+   */
+  function after(code) {
+    if (markdownSpace(code)) {
+      effects.enter("blockQuotePrefixWhitespace");
+      effects.consume(code);
+      effects.exit("blockQuotePrefixWhitespace");
+      effects.exit("blockQuotePrefix");
+      return ok;
+    }
+    effects.exit("blockQuotePrefix");
+    return ok(code);
+  }
+}
+
+/**
+ * Start of block quote continuation.
+ *
+ * ```markdown
+ *   | > a
+ * > | > b
+ *     ^
+ * ```
+ *
+ * @this {TokenizeContext}
+ *   Context.
+ * @type {Tokenizer}
+ */
+function tokenizeBlockQuoteContinuation(effects, ok, nok) {
+  const self = this;
+  return contStart;
+
+  /**
+   * Start of block quote continuation.
+   *
+   * Also used to parse the first block quote opening.
+   *
+   * ```markdown
+   *   | > a
+   * > | > b
+   *     ^
+   * ```
+   *
+   * @type {State}
+   */
+  function contStart(code) {
+    if (markdownSpace(code)) {
+      // Always populated by defaults.
+
+      return factorySpace(effects, contBefore, "linePrefix", self.parser.constructs.disable.null.includes('codeIndented') ? undefined : 4)(code);
+    }
+    return contBefore(code);
+  }
+
+  /**
+   * At `>`, after optional whitespace.
+   *
+   * Also used to parse the first block quote opening.
+   *
+   * ```markdown
+   *   | > a
+   * > | > b
+   *     ^
+   * ```
+   *
+   * @type {State}
+   */
+  function contBefore(code) {
+    return effects.attempt(blockQuote, ok, nok)(code);
+  }
+}
+
+/** @type {Exiter} */
+function exit(effects) {
+  effects.exit("blockQuote");
+}