OFM Comment extension (parser/Hast converter) This is not much useful, though.

Changes

11 changed files (+685/-3)

• content_parser/obsidian_markdown.ts
  • @@ -14,6 +14,8 @@ import { ofmHighlight } from "./obsidian_markdown/micromark_extension_ofm_highlight.ts";
    import { ofmImageSize } from "./obsidian_markdown/mdast_util_ofm_image_size.ts"; import { ofmWikilink } from "./obsidian_markdown/micromark_extension_ofm_wikilink.ts"; import { ofmWikilinkFromMarkdown } from "./obsidian_markdown/mdast_util_ofm_wikilink.ts"; + import { ofmComment } from "./obsidian_markdown/micromark_extension_ofm_comment.ts"; + import { ofmCommentFromMarkdown } from "./obsidian_markdown/mdast_util_ofm_comment.ts"; import { macanaMarkAssets } from "./obsidian_markdown/mdast_util_macana_mark_assets.ts"; import { macanaMarkDocumentToken } from "./obsidian_markdown/mdast_util_macana_mark_document_token.ts"; import { ofmCalloutFromMarkdown } from "./obsidian_markdown/mdast_util_ofm_callout.ts";
  • @@ -29,6 +31,7 @@ export { macanaReplaceAssetTokens } from "./obsidian_markdown/mdast_util_macana_replace_asset_tokens.ts";
    export { macanaReplaceDocumentToken } from "./obsidian_markdown/mdast_util_macana_replace_document_tokens.ts"; export { ofmWikilinkToHastHandlers } from "./obsidian_markdown/mdast_util_ofm_wikilink.ts"; export { ofmCalloutToHastHandlers } from "./obsidian_markdown/mdast_util_ofm_callout.ts"; + export { ofmCommentToHastHandlers } from "./obsidian_markdown/mdast_util_ofm_comment.ts"; export type { CalloutType } from "./obsidian_markdown/mdast_util_ofm_callout.ts"; function getFrontMatterValue(
  • @@ -101,8 +104,9 @@ "getAssetToken" | "getDocumentToken"
    >, ): Promise<Mdast.Root> { const mdast = fromMarkdown(markdown, { - extensions: [gfm(), ofmHighlight(), ofmWikilink()], + extensions: [gfm(), ofmHighlight(), ofmWikilink(), ofmComment()], mdastExtensions: [ + ofmCommentFromMarkdown(), gfmFromMarkdown(), ofmHighlightFromMarkdown(), ofmWikilinkFromMarkdown(),
• content_parser/obsidian_markdown/mdast_util_ofm_comment.test.ts (new)
  • @@ -0,0 +1,45 @@
    + // SPDX-FileCopyrightText: 2024 Shota FUJI <pockawoooh@gmail.com> + // + // SPDX-License-Identifier: Apache-2.0 + + import { assertObjectMatch } from "../../deps/deno.land/std/assert/mod.ts"; + + import { fromMarkdown } from "../../deps/esm.sh/mdast-util-from-markdown/mod.ts"; + + import { ofmComment } from "./micromark_extension_ofm_comment.ts"; + import { ofmCommentFromMarkdown } from "./mdast_util_ofm_comment.ts"; + + Deno.test("Should convert comment into Mdast", () => { + const mdast = fromMarkdown("This is an %%inline%% comment.", { + extensions: [ofmComment()], + mdastExtensions: [ofmCommentFromMarkdown()], + }); + + assertObjectMatch(mdast, { + type: "root", + children: [ + { + type: "paragraph", + children: [ + { + type: "text", + value: "This is an ", + }, + { + type: "ofmComment", + children: [ + { + type: "ofmCommentBody", + value: "inline", + }, + ], + }, + { + type: "text", + value: " comment.", + }, + ], + }, + ], + }); + });
• content_parser/obsidian_markdown/mdast_util_ofm_comment.ts (new)
  • @@ -0,0 +1,97 @@
    + // SPDX-FileCopyrightText: 2024 Shota FUJI <pockawoooh@gmail.com> + // + // SPDX-License-Identifier: Apache-2.0 + + import type * as Mdast from "../../deps/esm.sh/mdast/types.ts"; + import type { Extension } from "../../deps/esm.sh/mdast-util-from-markdown/mod.ts"; + import type { State } from "../../deps/esm.sh/mdast-util-to-hast/mod.ts"; + import type * as Hast from "../../deps/esm.sh/hast/types.ts"; + + export interface OfmComment extends Mdast.Node { + type: "ofmComment"; + children: [OfmCommentBody]; + } + + export interface OfmCommentBody extends Mdast.Node { + type: "ofmCommentBody"; + value: string; + } + + export function ofmCommentFromMarkdown(): Extension { + return { + enter: { + ofmInlineComment(token) { + this.enter({ + // @ts-expect-error: unist-related libraries heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + type: "ofmComment", + children: [], + }, token); + }, + ofmBlockComment(token) { + this.enter({ + // @ts-expect-error: unist-related libraries heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + type: "ofmComment", + children: [], + }, token); + }, + ofmCommentBody(token) { + this.enter({ + // @ts-expect-error: unist-related libraries heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + type: "ofmCommentBody", + children: [], + value: this.sliceSerialize(token), + }, token); + }, + }, + exit: { + ofmInlineComment(token) { + this.exit(token); + }, + ofmBlockComment(token) { + this.exit(token); + }, + ofmCommentBody(token) { + this.exit(token); + }, + }, + }; + } + + export interface OfmCommentToHastHandlersOptions { + preserveAsHtmlComment?: boolean; + } + + export function ofmCommentToHastHandlers( + { preserveAsHtmlComment = false }: OfmCommentToHastHandlersOptions = {}, + ) { + if (!preserveAsHtmlComment) { + return { + ofmComment(): Hast.Nodes { + return { + type: "comment", + value: "", + }; + }, + }; + } + + return { + ofmComment(_state: State, node: OfmComment): Hast.Nodes { + const [body] = node.children; + if (!body) { + return { + type: "comment", + value: "", + }; + } + + return { + type: "comment", + value: body.value, + }; + }, + }; + }
• content_parser/obsidian_markdown/micromark_extension_ofm_comment.test.ts (new)
  • @@ -0,0 +1,128 @@
    + // SPDX-FileCopyrightText: 2024 Shota FUJI <pockawoooh@gmail.com> + // + // SPDX-License-Identifier: Apache-2.0 + + import { assertEquals } from "../../deps/deno.land/std/assert/mod.ts"; + + import { micromark } from "../../deps/esm.sh/micromark/mod.ts"; + + import { + ofmComment, + ofmCommentHtml, + type OfmCommentHtmlOptions, + } from "./micromark_extension_ofm_comment.ts"; + + function f(markdown: string, opts?: OfmCommentHtmlOptions): string { + return micromark(markdown, { + extensions: [ofmComment()], + htmlExtensions: [ofmCommentHtml(opts)], + }); + } + + Deno.test("Should parse inline comment", () => { + assertEquals( + f("This is an %%inline%% comment."), + `<p>This is an comment.</p>`, + ); + }); + + Deno.test("Should not treat non-terminated inline comment", () => { + assertEquals( + f("This is an %%inline comment."), + `<p>This is an %%inline comment.</p>`, + ); + }); + + Deno.test("Should abort if it finds line-endings inside inline comment", () => { + assertEquals( + f("This is an %%inline\n comment.%%"), + `<p>This is an %%inline\ncomment.%%</p>`, + ); + }); + + Deno.test("Should keep as HTML comment", () => { + assertEquals( + f("This is an %%inline%% comment.", { + preserveAsHtmlComment: true, + }), + `<p>This is an <!--inline--> comment.</p>`, + ); + }); + + Deno.test("Should ignore non-sequential percent sign", () => { + assertEquals( + f("This is an %%%inline%% comment.", { + preserveAsHtmlComment: true, + }), + `<p>This is an <!--%inline--> comment.</p>`, + ); + }); + + Deno.test("Should handle escape", () => { + assertEquals( + f("This is an \\%%inline%% comment%%.", { + preserveAsHtmlComment: true, + }), + `<p>This is an %%inline<!-- comment-->.</p>`, + ); + }); + + Deno.test("Should parse block comment", () => { + const code = ` + %% + This is a block comment. + + Block comments can span multiple lines. + %% + `.trim(); + assertEquals( + f(code, { preserveAsHtmlComment: true }), + `<!-- + This is a block comment. + + Block comments can span multiple lines. + -->`, + ); + }); + + Deno.test("Should not parse contents inside block comment", () => { + const code = ` + %% + **This is a block comment**. + + Block comments can span multiple lines. + %% + `.trim(); + assertEquals( + f(code, { preserveAsHtmlComment: true }), + `<!-- + **This is a block comment**. + + Block comments can span multiple lines. + -->`, + ); + }); + + Deno.test("Should parse block comment in a document", () => { + const code = ` + This is a paragraph. + + %% + This is a block comment. + + Block comments can span multiple lines. + %% + + This is a paragraph too. + `.trim(); + assertEquals( + f(code, { preserveAsHtmlComment: true }), + `<p>This is a paragraph.</p> + <!-- + This is a block comment. + + Block comments can span multiple lines. + --> + <p>This is a paragraph too.</p>`, + ); + });
• content_parser/obsidian_markdown/micromark_extension_ofm_comment.ts (new)
  • @@ -0,0 +1,332 @@
    + // SPDX-FileCopyrightText: 2024 Shota FUJI <pockawoooh@gmail.com> + // + // SPDX-License-Identifier: Apache-2.0 + + import type { + CompileContext, + Construct, + Extension, + HtmlExtension, + State, + Token, + } from "../../deps/esm.sh/micromark-util-types/types.ts"; + import { codes, types } from "../../deps/esm.sh/micromark-util-symbol/mod.ts"; + import { markdownLineEnding } from "../../deps/esm.sh/micromark-util-character/mod.ts"; + + const enum TokenTypeMap { + inline = "ofmInlineComment", + block = "ofmBlockComment", + chunk = "ofmCommentBodyChunk", + body = "ofmCommentBody", + } + + export interface OfmCommentHtmlOptions { + /** + * Whether to preserve comments as HTML comment (<!-- this -->) + */ + preserveAsHtmlComment?: boolean; + } + + export function ofmCommentHtml( + { preserveAsHtmlComment = false }: OfmCommentHtmlOptions = {}, + ): HtmlExtension { + if (!preserveAsHtmlComment) { + return {}; + } + + return { + enter: { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + [TokenTypeMap.block](this: CompileContext) { + this.raw("<!--"); + }, + [TokenTypeMap.inline](this: CompileContext) { + this.raw("<!--"); + }, + [TokenTypeMap.chunk](this: CompileContext, token: Token) { + // Prevent outputting broken HTML + this.raw(this.sliceSerialize(token).replaceAll("-->", "")); + }, + }, + exit: { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + [TokenTypeMap.block](this: CompileContext) { + this.raw("-->"); + }, + [TokenTypeMap.inline](this: CompileContext) { + this.raw("-->"); + }, + }, + }; + } + + export function ofmComment(): Extension { + return { + text: { + [codes.percentSign]: { + name: "ofm-inline-comment", + tokenize(effects, ok, nok) { + /** + * ```markdown + * > | %%body%% + * ^^ + * ``` + */ + const start: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.inline); + + return effects.attempt(startMarker, beforeBody, nok)(code); + }; + + const beforeBody: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.body); + + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.chunk); + + return body(code); + }; + + const body: State = (code) => { + if (code === codes.eof || markdownLineEnding(code)) { + return nok(code); + } + + if (code === codes.percentSign) { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.exit(TokenTypeMap.chunk); + + return effects.attempt(endMarker, beforeEnd, unbondedPercentSign)( + code, + ); + } + + effects.consume(code); + return body; + }; + + const unbondedPercentSign: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.chunk); + effects.consume(code); + + return body; + }; + + const beforeEnd: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.exit(TokenTypeMap.inline); + + return ok(code); + }; + + return start; + }, + }, + }, + flow: { + [codes.percentSign]: { + name: "ofm-block-comment", + concrete: true, + tokenize(effects, ok, nok) { + /** + * ```markdown + * > | %% + * ^^ + * > | body + * > | %% + * ``` + */ + const start: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.block); + + return effects.attempt(startMarker, beforeBody, nok)(code); + }; + + const beforeBody: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.body); + + return beforeChunk(code); + }; + + const beforeChunk: State = (code) => { + if (markdownLineEnding(code)) { + effects.enter(types.lineEnding); + effects.consume(code); + effects.exit(types.lineEnding); + + return beforeChunk; + } + + if (code === codes.percentSign) { + return effects.attempt(endMarker, beforeEnd, unbondedPercentSign)( + code, + ); + } + + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.chunk); + + return chunk(code); + }; + + /** + * ```markdown + * > | %% + * > | body + * ^^^^ + * > | %% + * ``` + */ + const chunk: State = (code) => { + if (markdownLineEnding(code)) { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.exit(TokenTypeMap.chunk); + effects.enter(types.lineEnding); + effects.consume(code); + effects.exit(types.lineEnding); + + return beforeChunk; + } + + if (code === codes.percentSign) { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.exit(TokenTypeMap.chunk); + + return effects.attempt(endMarker, beforeEnd, unbondedPercentSign)( + code, + ); + } + + effects.consume(code); + + return chunk; + }; + + const unbondedPercentSign: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.enter(TokenTypeMap.chunk); + + effects.consume(code); + + return chunk; + }; + + const beforeEnd: State = (code) => { + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.exit(TokenTypeMap.block); + + return ok(code); + }; + + return start; + }, + }, + }, + }; + } + + const startMarker: Construct = { + partial: true, + tokenize(effects, ok, nok) { + const { previous } = this; + + /** + * ```markdown + * > | %% + * ^ + * ``` + */ + const first: State = (code) => { + if (code !== codes.percentSign || previous === codes.backslash) { + return nok(code); + } + + effects.consume(code); + + return second; + }; + + /** + * ```markdown + * > | %% + * ^ + * ``` + */ + const second: State = (code) => { + if (code !== codes.percentSign) { + return nok(code); + } + + effects.consume(code); + + return ok(code); + }; + + return first; + }, + }; + + const endMarker: Construct = { + partial: true, + tokenize(effects, ok, nok) { + const { previous } = this; + + /** + * ```markdown + * > | %% + * ^ + * ``` + */ + const first: State = (code) => { + if (code !== codes.percentSign || previous === codes.backslash) { + return nok(code); + } + + // @ts-expect-error: micromark heavily relies on ambient module declarations, + // which Deno does not support. APIs also don't accept type parameters. + effects.exit(TokenTypeMap.body); + + effects.consume(code); + + return second; + }; + + /** + * ```markdown + * > | %% + * ^ + * ``` + */ + const second: State = (code) => { + if (code !== codes.percentSign) { + return nok(code); + } + + effects.consume(code); + + return ok(code); + }; + + return first; + }, + };
• deno.lock
  • @@ -281,6 +281,8 @@ "https://esm.sh/v135/micromark-factory-space@2.0.0/denonext/micromark-factory-space.mjs": "1ac7c90dec53f7f634767c5470c2dcf204f4df99ec318a27832786153d5c8110",
    "https://esm.sh/v135/micromark-factory-title@2.0.0/denonext/micromark-factory-title.mjs": "1b202816c09c57894c8e254962ffbd8ad559439b0eb90ff0eca743f0dbfff470", "https://esm.sh/v135/micromark-factory-whitespace@2.0.0/denonext/micromark-factory-whitespace.mjs": "f85efacaec053453a9445b4147d2fb7ce02c1d083f0b0da9a730a112ff934d9e", "https://esm.sh/v135/micromark-util-character@2.0.1/denonext/micromark-util-character.mjs": "18b451d148e1ccc3a9b18e5c4061d44a0485e8ec65ad805d20b2950a51c7213b", + "https://esm.sh/v135/micromark-util-character@2.1.0": "dab3a3101833f22c1e8c7f90eef3d03d9a6658b72204a004abaec47d9f238465", + "https://esm.sh/v135/micromark-util-character@2.1.0/denonext/micromark-util-character.mjs": "9d6a0ba08282a3a306b38d255e77d4636ac1447de61008f7213f725c07f6b029", "https://esm.sh/v135/micromark-util-chunked@2.0.0": "6137fb50257da8a18125ac21ac885bd56f79e6ead241b0b004b0416c7906eaa2", "https://esm.sh/v135/micromark-util-chunked@2.0.0/denonext/micromark-util-chunked.mjs": "531cf323ba53649fdc30cd39ebba54253dfd847a4b23f806058ecc6cf67bca69", "https://esm.sh/v135/micromark-util-classify-character@2.0.0": "c92c93be5f8370abbd73783460c6e183f2b31fe74085175389e1c86d96aeef9b",
• deps/esm.sh/micromark-util-character/mod.ts (new)
  • @@ -0,0 +1,5 @@
    + // SPDX-FileCopyrightText: 2024 Shota FUJI <pockawoooh@gmail.com> + // + // SPDX-License-Identifier: Apache-2.0 + + export * from "https://esm.sh/v135/micromark-util-character@2.1.0";
• docs/en/Internals/Tests/Obsidian Flavored Markdown/Comment extension.md (new)
  • @@ -0,0 +1,66 @@
    + ## Basics + + ### Inline + + ```markdown + %%This is a comment%% + ``` + + %%This is a comment%% + + ```markdown + Inline %%comment can't include + line-endings%% + ``` + + Inline %%comment can't include + line-endings%% + + The latter one is buggy inside Obsidian: renders this as "Inline (rest commented-out)" in edit view but `Inline %%comment can't include line-endings%%` in reading view. + ### Block + + ```markdown + %%This + is + a + comment, + too + %% + ``` + + %%This + is + a + comment, + too + %% + + ```markdown + %% + **bold** + %% + ``` + + %% + **bold** + %% + + Obsidian renders "bold" in bold in edit view. + + ## Escapes + + ```markdown + \%%This isn't a comment%% but this is%% + ``` + + \%%This isn't a comment%% but this is%% + + ## Edge cases + + ### Triple percent sign + + ```markdown + %%%This is a comment%% + ``` + + %%%This is a comment%%
• docs/en/Internals/Tests/Obsidian Flavored Markdown/What's this?.md
  • @@ -6,4 +6,5 @@
    Most of the unit tests are picked up from this folder. - [[Highlight extension]] - - [[Callout extension]] + - [[Callout extension]] + - [[Comment extension]]
• docs/en/Roadmap.md
  • @@ -36,7 +36,7 @@ - [ ] List from another file
    - [ ] Search results - [x] ==Highlight== - [x] Callouts - - [ ] Comments %% You can check this item once I'm no longer visible %% + - [x] Comments %% You can check this item once I'm no longer visible %% - [ ] Strip Raw HTML (only `<title>` is troublesome, but align behavior to Obsidian's) - [ ] `<script>` <script>console.log("This tag should be eliminated: escaping is not suffice")</script> - [ ] `<title>` <title>Foo</title>
• page_builder/default_theme/components/html.tsx
  • @@ -20,6 +20,7 @@ import {
    type CalloutType, type ObsidianMarkdownDocument, ofmCalloutToHastHandlers, + ofmCommentToHastHandlers, ofmWikilinkToHastHandlers, } from "../../../content_parser/obsidian_markdown.ts"; import type { JSONCanvasDocument } from "../../../content_parser/json_canvas.ts";
  • @@ -196,6 +197,7 @@ children: [],
    }; }, }), + ...ofmCommentToHastHandlers(), ...ofmWikilinkToHastHandlers, ...syntaxHighlightingHandlers(), },