Skip to content

Commit

Permalink
f2d94210e434b01a02923ec509c280c5b105fe15
Browse files Browse the repository at this point in the history
  • Loading branch information
@github-actions
github-actions[bot] committed Mar 25, 2024
1 parent 4b67547 commit e35db16
Show file tree
Hide file tree
Showing 4 changed files with 448 additions and 158 deletions.
7 changes: 3 additions & 4 deletions esm/client/0_markdown.d.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,18 +5,17 @@ export declare const CODEPOINTS: {
"\0": number;
"\v": number;
"\n": number;
"<": number;
">": number;
'"': number;
" ": number;
"@": number;
_: number;
"[": number;
"]": number;
"(": number;
")": number;
"`": number;
"~": number;
"\\": number;
"*": number;
"!": number;
"|": number;
};
export declare function parseMarkdown(text_: string): [string, MessageEntity[]];
296 changes: 221 additions & 75 deletions esm/client/0_markdown.js
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@
* Copyright (c) 2024 Roj
*/
import { UNREACHABLE } from "../1_utilities.js";
import { sortMessageEntities } from "../3_types.js";
const enc = new TextEncoder();
const dec = new TextDecoder();
export const CODEPOINTS = {
Expand All @@ -11,19 +12,18 @@ export const CODEPOINTS = {
"\0": 0,
"\v": 11,
"\n": 10,
"<": 60,
">": 62,
'"': 34,
" ": 32,
"@": 64,
"_": 95,
"[": 91,
"]": 93,
"(": 40,
")": 41,
"`": 96,
"~": 126,
"\\": 92,
"*": 42,
"!": 33,
"|": 124,
};
function isUtf8CharacterFirstCodeUnit(c) {
return (c & 0xC0) !== 0x80;
Expand Down Expand Up @@ -58,118 +58,264 @@ function getLinkUserId(url_) {
return 0;
}
}
function getLinkCustomEmojiId(url_) {
try {
const url = new URL(url_);
if (url.protocol != "tg:" || url.hostname != "emoji" || url.pathname.slice(1) != "" || url.port != "") {
return "";
}
const id_ = url.searchParams.get("id");
if (!id_) {
return "";
}
const id = BigInt(id_);
if (!id) {
return "";
}
else {
return String(id);
}
}
catch {
return "";
}
}
export function parseMarkdown(text_) {
const text = enc.encode(text_);
let resultSize = 0;
const entities = [];
const size = text.length;
let entities = [];
let utf16Offset = 0;
for (let i = 0; i < size; ++i) {
const c = text[i], next = text[i + 1];
if (c === CODEPOINTS["\\"] &&
(next === CODEPOINTS["_"] || next === CODEPOINTS["*"] || next === CODEPOINTS["`"] || next === CODEPOINTS["["])) {
++i;
const nestedEntities = [];
for (let i = 0; i < text.length; i++) {
const c = text[i];
if (c === CODEPOINTS["\\"] && text[i + 1] != null && text[i + 1] > 0 && text[i + 1] <= 126) {
i++;
utf16Offset += 1;
text[resultSize++] = text[i];
utf16Offset++;
continue;
}
if (c != null && c !== CODEPOINTS["_"] && c !== CODEPOINTS["*"] && c !== CODEPOINTS["`"] && c !== CODEPOINTS["["]) {
let reservedCharacters = enc.encode("_*[]()~`>#+-=|{}.!");
if (nestedEntities.length !== 0) {
switch (nestedEntities[nestedEntities.length - 1].type) {
case "code":
case "pre":
reservedCharacters = Uint8Array.of(CODEPOINTS["`"]);
break;
default:
break;
}
}
if (!reservedCharacters.includes(text[i])) {
if (isUtf8CharacterFirstCodeUnit(c)) {
utf16Offset += 1 + ((c >= 0xf0) ? 1 : 0);
utf16Offset += 1 + (c >= 0xf0 ? 1 : 0);
}
text[resultSize++] = text[i];
continue;
}
const beginPos = i;
let endCharacter = text[i];
let isPre = false;
if (c === CODEPOINTS["["])
endCharacter = CODEPOINTS["]"];
++i;
let language = undefined;
if (c === CODEPOINTS["`"] && text[i] === CODEPOINTS["`"] && text[i + 1] === CODEPOINTS["`"]) {
i += 2;
isPre = true;
let languageEnd = i;
while (text[languageEnd] != null && !isWhitespace(text[languageEnd]) && text[languageEnd] !== CODEPOINTS["`"]) {
languageEnd++;
}
if (i !== languageEnd && languageEnd < size && text[languageEnd] !== CODEPOINTS["`"]) {
language = text.slice(i, languageEnd);
i = languageEnd;
}
const current = text[i], next = text[i + 1];
if (current === CODEPOINTS["\n"] || current === CODEPOINTS["\r"]) {
if ((next === CODEPOINTS["\n"] || next === CODEPOINTS["\r"]) && current !== next) {
i += 2;
}
else {
++i;
let isEndOfAnEntity = false;
if (nestedEntities.length !== 0) {
isEndOfAnEntity = (() => {
switch (nestedEntities[nestedEntities.length - 1].type) {
case "bold":
return c === CODEPOINTS["*"];
case "italic":
return c === CODEPOINTS["_"] && text[i + 1] !== CODEPOINTS["_"];
case "code":
return c === CODEPOINTS["`"];
case "pre":
return c === CODEPOINTS["`"] && text[i + 1] === CODEPOINTS["`"] && text[i + 2] === CODEPOINTS["`"];
case "textLink":
return c === CODEPOINTS["]"];
case "underline":
return c === CODEPOINTS["_"] && text[i + 1] === CODEPOINTS["_"];
case "strikethrough":
return c === CODEPOINTS["~"];
case "spoiler":
return c === CODEPOINTS["|"] && text[i + 1] === CODEPOINTS["|"];
case "customEmoji":
return c === CODEPOINTS["]"];
default:
UNREACHABLE();
}
}
}
const entityOffset = utf16Offset;
while (i < size &&
(text[i] !== endCharacter ||
(isPre && !(text[i + 1] === CODEPOINTS["`"] && text[i + 2] === CODEPOINTS["`"])))) {
const curCh = text[i];
if (isUtf8CharacterFirstCodeUnit(curCh)) {
utf16Offset += 1 + (curCh >= 0xf0 ? 1 : 0);
}
text[resultSize++] = text[i++];
})();
}
if (i === size) {
throw new Error("Can't find end of the entity starting at byte offset " + beginPos);
}
if (entityOffset !== utf16Offset) {
const entityLength = utf16Offset - entityOffset;
if (!isEndOfAnEntity) {
let type;
let argument = new Uint8Array();
const entityByteOffset = i;
switch (c) {
case CODEPOINTS["_"]:
entities.push({ type: "italic", offset: entityOffset, length: entityLength });
if (text[i + 1] === CODEPOINTS["_"]) {
type = "underline";
i++;
}
else {
type = "italic";
}
break;
case CODEPOINTS["*"]:
entities.push({ type: "bold", offset: entityOffset, length: entityLength });
type = "bold";
break;
case CODEPOINTS["~"]:
type = "strikethrough";
break;
case CODEPOINTS["|"]:
if (text[i + 1] === CODEPOINTS["|"]) {
i++;
type = "spoiler";
}
else {
throw new Error(`Character '${String.fromCharCode(c)}' is reserved and must be escaped with the preceding '\\'`);
}
break;
case CODEPOINTS["["]:
type = "textLink";
break;
case CODEPOINTS["`"]:
if (text[i + 1] === CODEPOINTS["`"] && text[i + 2] === CODEPOINTS["`"]) {
i += 3;
type = "code";
let languageEnd = i;
while (text[languageEnd] != null && !isWhitespace(text[languageEnd]) && text[languageEnd] !== CODEPOINTS["`"]) {
languageEnd++;
}
if (i !== languageEnd && languageEnd < text.length && text[languageEnd] !== CODEPOINTS["`"]) {
type = "pre";
argument = text.slice(i, languageEnd);
i = languageEnd;
}
const current = text[i], next = text[i + 1];
if (current === CODEPOINTS["\n"] || current === CODEPOINTS["\r"]) {
if ((next === CODEPOINTS["\n"] || next === CODEPOINTS["\r"]) && current !== next) {
i += 2;
}
else {
i++;
}
}
i--;
}
else {
type = "code";
}
break;
case CODEPOINTS["!"]:
if (text[i + 1] === CODEPOINTS["["]) {
i++;
type = "customEmoji";
}
else {
throw new Error(`Character '${String.fromCharCode(text[i])}' is reserved and must be escaped with the preceding '\\'`);
}
break;
default:
throw new Error(`Character '${String.fromCharCode(text[i])}' is reserved and must be escaped with the preceding '\\'`);
}
nestedEntities.push({ type, argument, entityOffset: utf16Offset, entityByteOffset, entityBeginPos: resultSize });
}
else {
let { type, argument } = nestedEntities[nestedEntities.length - 1];
let userId = 0;
let customEmojiId = "";
let skipEntity = utf16Offset === nestedEntities.at(-1).entityOffset;
switch (type) {
case "bold":
case "italic":
case "code":
case "strikethrough":
break;
case "underline":
case "spoiler":
i++;
break;
case "pre":
i += 2;
break;
case CODEPOINTS["["]: {
case "textLink": {
let url;
if (text[i + 1] !== CODEPOINTS["("]) {
url = text.slice(beginPos + 1, i);
url = text.slice(nestedEntities.at(-1).entityBeginPos, resultSize);
}
else {
i += 2;
const urlBeginPos = i;
const url_ = [];
while (i < size && text[i] !== CODEPOINTS[")"]) {
while (i < text.length && text[i] !== CODEPOINTS[")"]) {
if (text[i] === CODEPOINTS["\\"] && text[i + 1] > 0 && text[i + 1] <= 126) {
url_.push(text[i + 1]);
i += 2;
continue;
}
url_.push(text[i++]);
}
url = Uint8Array.from(url_);
if (text[i] !== CODEPOINTS[")"]) {
throw new Error("Can't find end of a URL at byte offset " + urlBeginPos);
}
}
const userId = getLinkUserId(dec.decode(url));
if (userId) {
entities.push({ type: "textMention", offset: entityOffset, length: entityLength, userId });
}
else {
userId = getLinkUserId(dec.decode(url));
if (!userId) {
const url_ = getUrl(dec.decode(url));
if (url_) {
entities.push({ type: "textLink", offset: entityOffset, length: entityLength, url: url_ });
if (!url_) {
skipEntity = true;
}
else {
argument = url_;
}
}
break;
}
case CODEPOINTS["`"]:
if (isPre) {
entities.push({ type: "pre", offset: entityOffset, length: entityLength, language: language ? dec.decode(language) : "" });
case "customEmoji": {
if (text[i + 1] !== CODEPOINTS["("]) {
throw new Error("Custom emoji entity must contain a tg://emoji URL");
}
else {
entities.push({ type: "code", offset: entityOffset, length: entityLength });
i += 2;
const url_ = [];
const urlBeginPos = i;
while (i < text.length && text[i] !== CODEPOINTS[")"]) {
if (text[i] === CODEPOINTS["\\"] && text[i + 1] > 0 && text[i + 1] <= 126) {
url_.push(text[i + 1]);
i += 2;
continue;
}
url_.push(text[i++]);
}
const url = Uint8Array.from(url_);
if (text[i] !== CODEPOINTS[")"]) {
throw new Error("Can't find end of a custom emoji URL at byte offset " + urlBeginPos);
}
customEmojiId = getLinkCustomEmojiId(dec.decode(url));
break;
}
default:
UNREACHABLE();
}
if (!skipEntity) {
const entityOffset = nestedEntities.at(-1).entityOffset;
const entityLength = utf16Offset - entityOffset;
if (userId) {
entities.push({ type: "textMention", offset: entityOffset, length: entityLength, userId });
}
else if (customEmojiId) {
entities.push({ type: "customEmoji", offset: entityOffset, length: entityLength, customEmojiId });
}
else if (type == "textLink") {
entities.push({ type, offset: entityOffset, length: entityLength, url: typeof argument === "string" ? argument : dec.decode(argument) });
}
else if (type == "pre") {
entities.push({ type, offset: entityOffset, length: entityLength, language: typeof argument === "string" ? argument : dec.decode(argument) });
}
else if (type != "customEmoji") {
entities.push({ type, offset: entityOffset, length: entityLength });
}
}
nestedEntities.pop();
}
if (isPre) {
i += 2;
}
}
if (nestedEntities.length !== 0) {
const last = nestedEntities[nestedEntities.length - 1];
throw new Error(`Can't find end of ${last.type} entity at byte offset ${last.entityByteOffset}`);
}
entities = sortMessageEntities(entities);
return [dec.decode(text.slice(0, resultSize)), entities];
}
Loading

0 comments on commit e35db16

Please sign in to comment.