Escape markdown sequences (#2208)

* escape inline markdown character

* fix typo

* improve document around custom markdown plugin and add escape sequence utils

* recover inline escape sequences on edit

* remove escape sequences from plain text body

* use `s` for strike-through instead of del

* escape block markdown sequences

* fix remove escape sequence was not removing all slashes from plain text

* recover block sequences on edit
This commit is contained in:
Ajay Bura
2025-02-21 19:19:24 +11:00
committed by GitHub
parent b63868bbb5
commit 7456c152b7
19 changed files with 764 additions and 476 deletions
+1
View File
@@ -0,0 +1 @@
export * from './parser';
+40
View File
@@ -0,0 +1,40 @@
import {
BoldRule,
CodeRule,
EscapeRule,
ItalicRule1,
ItalicRule2,
LinkRule,
SpoilerRule,
StrikeRule,
UnderlineRule,
} from './rules';
import { runInlineRule, runInlineRules } from './runner';
import { InlineMDParser } from './type';
const LeveledRules = [
BoldRule,
ItalicRule1,
UnderlineRule,
ItalicRule2,
StrikeRule,
SpoilerRule,
LinkRule,
EscapeRule,
];
/**
* Parses inline markdown text into HTML using defined rules.
*
* @param text - The markdown text to be parsed.
* @returns The parsed HTML or the original text if no markdown was found.
*/
export const parseInlineMD: InlineMDParser = (text) => {
if (text === '') return text;
let result: string | undefined;
if (!result) result = runInlineRule(text, CodeRule, parseInlineMD);
if (!result) result = runInlineRules(text, LeveledRules, parseInlineMD);
return result ?? text;
};
+123
View File
@@ -0,0 +1,123 @@
import { InlineMDRule } from './type';
const MIN_ANY = '(.+?)';
const URL_NEG_LB = '(?<!(https?|ftp|mailto|magnet):\\/\\/\\S*)';
const ESC_NEG_LB = '(?<!\\\\)';
const BOLD_MD_1 = '**';
const BOLD_PREFIX_1 = `${ESC_NEG_LB}\\*{2}`;
const BOLD_NEG_LA_1 = '(?!\\*)';
const BOLD_REG_1 = new RegExp(
`${URL_NEG_LB}${BOLD_PREFIX_1}${MIN_ANY}${BOLD_PREFIX_1}${BOLD_NEG_LA_1}`
);
export const BoldRule: InlineMDRule = {
match: (text) => text.match(BOLD_REG_1),
html: (parse, match) => {
const [, , g2] = match;
return `<strong data-md="${BOLD_MD_1}">${parse(g2)}</strong>`;
},
};
const ITALIC_MD_1 = '*';
const ITALIC_PREFIX_1 = `${ESC_NEG_LB}\\*`;
const ITALIC_NEG_LA_1 = '(?!\\*)';
const ITALIC_REG_1 = new RegExp(
`${URL_NEG_LB}${ITALIC_PREFIX_1}${MIN_ANY}${ITALIC_PREFIX_1}${ITALIC_NEG_LA_1}`
);
export const ItalicRule1: InlineMDRule = {
match: (text) => text.match(ITALIC_REG_1),
html: (parse, match) => {
const [, , g2] = match;
return `<i data-md="${ITALIC_MD_1}">${parse(g2)}</i>`;
},
};
const ITALIC_MD_2 = '_';
const ITALIC_PREFIX_2 = `${ESC_NEG_LB}_`;
const ITALIC_NEG_LA_2 = '(?!_)';
const ITALIC_REG_2 = new RegExp(
`${URL_NEG_LB}${ITALIC_PREFIX_2}${MIN_ANY}${ITALIC_PREFIX_2}${ITALIC_NEG_LA_2}`
);
export const ItalicRule2: InlineMDRule = {
match: (text) => text.match(ITALIC_REG_2),
html: (parse, match) => {
const [, , g2] = match;
return `<i data-md="${ITALIC_MD_2}">${parse(g2)}</i>`;
},
};
const UNDERLINE_MD_1 = '__';
const UNDERLINE_PREFIX_1 = `${ESC_NEG_LB}_{2}`;
const UNDERLINE_NEG_LA_1 = '(?!_)';
const UNDERLINE_REG_1 = new RegExp(
`${URL_NEG_LB}${UNDERLINE_PREFIX_1}${MIN_ANY}${UNDERLINE_PREFIX_1}${UNDERLINE_NEG_LA_1}`
);
export const UnderlineRule: InlineMDRule = {
match: (text) => text.match(UNDERLINE_REG_1),
html: (parse, match) => {
const [, , g2] = match;
return `<u data-md="${UNDERLINE_MD_1}">${parse(g2)}</u>`;
},
};
const STRIKE_MD_1 = '~~';
const STRIKE_PREFIX_1 = `${ESC_NEG_LB}~{2}`;
const STRIKE_NEG_LA_1 = '(?!~)';
const STRIKE_REG_1 = new RegExp(
`${URL_NEG_LB}${STRIKE_PREFIX_1}${MIN_ANY}${STRIKE_PREFIX_1}${STRIKE_NEG_LA_1}`
);
export const StrikeRule: InlineMDRule = {
match: (text) => text.match(STRIKE_REG_1),
html: (parse, match) => {
const [, , g2] = match;
return `<s data-md="${STRIKE_MD_1}">${parse(g2)}</s>`;
},
};
const CODE_MD_1 = '`';
const CODE_PREFIX_1 = `${ESC_NEG_LB}\``;
const CODE_NEG_LA_1 = '(?!`)';
const CODE_REG_1 = new RegExp(`${URL_NEG_LB}${CODE_PREFIX_1}(.+?)${CODE_PREFIX_1}${CODE_NEG_LA_1}`);
export const CodeRule: InlineMDRule = {
match: (text) => text.match(CODE_REG_1),
html: (parse, match) => {
const [, , g2] = match;
return `<code data-md="${CODE_MD_1}">${g2}</code>`;
},
};
const SPOILER_MD_1 = '||';
const SPOILER_PREFIX_1 = `${ESC_NEG_LB}\\|{2}`;
const SPOILER_NEG_LA_1 = '(?!\\|)';
const SPOILER_REG_1 = new RegExp(
`${URL_NEG_LB}${SPOILER_PREFIX_1}${MIN_ANY}${SPOILER_PREFIX_1}${SPOILER_NEG_LA_1}`
);
export const SpoilerRule: InlineMDRule = {
match: (text) => text.match(SPOILER_REG_1),
html: (parse, match) => {
const [, , g2] = match;
return `<span data-md="${SPOILER_MD_1}" data-mx-spoiler>${parse(g2)}</span>`;
},
};
const LINK_ALT = `\\[${MIN_ANY}\\]`;
const LINK_URL = `\\((https?:\\/\\/.+?)\\)`;
const LINK_REG_1 = new RegExp(`${LINK_ALT}${LINK_URL}`);
export const LinkRule: InlineMDRule = {
match: (text) => text.match(LINK_REG_1),
html: (parse, match) => {
const [, g1, g2] = match;
return `<a data-md href="${g2}">${parse(g1)}</a>`;
},
};
export const INLINE_SEQUENCE_SET = '[*_~`|]';
const ESC_SEQ_1 = `\\\\(${INLINE_SEQUENCE_SET})`;
const ESC_REG_1 = new RegExp(`${URL_NEG_LB}${ESC_SEQ_1}`);
export const EscapeRule: InlineMDRule = {
match: (text) => text.match(ESC_REG_1),
html: (parse, match) => {
const [, , g2] = match;
return g2;
},
};
+62
View File
@@ -0,0 +1,62 @@
import { MatchResult, replaceMatch } from '../internal';
import { InlineMDParser, InlineMDRule } from './type';
/**
* Runs a single markdown rule on the provided text.
*
* @param text - The text to parse.
* @param rule - The markdown rule to run.
* @param parse - A function that run the parser on remaining parts.
* @returns The text with the markdown rule applied or `undefined` if no match is found.
*/
export const runInlineRule = (
text: string,
rule: InlineMDRule,
parse: InlineMDParser
): string | undefined => {
const matchResult = rule.match(text);
if (matchResult) {
const content = rule.html(parse, matchResult);
return replaceMatch(text, matchResult, content, (txt) => [parse(txt)]).join('');
}
return undefined;
};
/**
* Runs multiple rules at the same time to better handle nested rules.
* Rules will be run in the order they appear.
*
* @param text - The text to parse.
* @param rules - The markdown rules to run.
* @param parse - A function that run the parser on remaining parts.
* @returns The text with the markdown rules applied or `undefined` if no match is found.
*/
export const runInlineRules = (
text: string,
rules: InlineMDRule[],
parse: InlineMDParser
): string | undefined => {
const matchResults = rules.map((rule) => rule.match(text));
let targetRule: InlineMDRule | undefined;
let targetResult: MatchResult | undefined;
for (let i = 0; i < matchResults.length; i += 1) {
const currentResult = matchResults[i];
if (currentResult && typeof currentResult.index === 'number') {
if (
!targetResult ||
(typeof targetResult?.index === 'number' && currentResult.index < targetResult.index)
) {
targetResult = currentResult;
targetRule = rules[i];
}
}
}
if (targetRule && targetResult) {
const content = targetRule.html(parse, targetResult);
return replaceMatch(text, targetResult, content, (txt) => [parse(txt)]).join('');
}
return undefined;
};
+26
View File
@@ -0,0 +1,26 @@
import { MatchResult, MatchRule } from '../internal';
/**
* Type for a function that parses inline markdown into HTML.
*
* @param text - The markdown text to be parsed.
* @returns The parsed HTML.
*/
export type InlineMDParser = (text: string) => string;
/**
* Type for a function that converts a match to output.
*
* @param parse - The inline markdown parser function.
* @param match - The match result.
* @returns The output string after processing the match.
*/
export type InlineMatchConverter = (parse: InlineMDParser, match: MatchResult) => string;
/**
* Type representing a markdown rule that includes a matching pattern and HTML conversion.
*/
export type InlineMDRule = {
match: MatchRule; // A function that matches a specific markdown pattern.
html: InlineMatchConverter; // A function that converts the match to HTML.
};