dns-witch/dns-witch.net.eu.org
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
dns-witch.net.eu.org/node_modules/markdown-it/dist/markdown-it.js

6964 lines
292 KiB
JavaScript
Raw Normal View History

11ty + RIP DNS Witch EU.ORG service
2025-07-05 16:19:59 +00:00
/*! markdown-it 14.1.0 https://github.com/markdown-it/markdown-it @license MIT */
(function(global, factory) {
typeof exports === "object" && typeof module !== "undefined" ? module.exports = factory() : typeof define === "function" && define.amd ? define(factory) : (global = typeof globalThis !== "undefined" ? globalThis : global || self,
global.markdownit = factory());
})(this, (function() {
"use strict";
/* eslint-disable no-bitwise */ const decodeCache = {};
function getDecodeCache(exclude) {
let cache = decodeCache[exclude];
if (cache) {
return cache;
}
cache = decodeCache[exclude] = [];
for (let i = 0; i < 128; i++) {
const ch = String.fromCharCode(i);
cache.push(ch);
}
for (let i = 0; i < exclude.length; i++) {
const ch = exclude.charCodeAt(i);
cache[ch] = "%" + ("0" + ch.toString(16).toUpperCase()).slice(-2);
}
return cache;
}
// Decode percent-encoded string.
function decode$1(string, exclude) {
if (typeof exclude !== "string") {
exclude = decode$1.defaultChars;
}
const cache = getDecodeCache(exclude);
return string.replace(/(%[a-f0-9]{2})+/gi, (function(seq) {
let result = "";
for (let i = 0, l = seq.length; i < l; i += 3) {
const b1 = parseInt(seq.slice(i + 1, i + 3), 16);
if (b1 < 128) {
result += cache[b1];
continue;
}
if ((b1 & 224) === 192 && i + 3 < l) {
// 110xxxxx 10xxxxxx
const b2 = parseInt(seq.slice(i + 4, i + 6), 16);
if ((b2 & 192) === 128) {
const chr = b1 << 6 & 1984 | b2 & 63;
if (chr < 128) {
result += "\ufffd\ufffd";
} else {
result += String.fromCharCode(chr);
}
i += 3;
continue;
}
}
if ((b1 & 240) === 224 && i + 6 < l) {
// 1110xxxx 10xxxxxx 10xxxxxx
const b2 = parseInt(seq.slice(i + 4, i + 6), 16);
const b3 = parseInt(seq.slice(i + 7, i + 9), 16);
if ((b2 & 192) === 128 && (b3 & 192) === 128) {
const chr = b1 << 12 & 61440 | b2 << 6 & 4032 | b3 & 63;
if (chr < 2048 || chr >= 55296 && chr <= 57343) {
result += "\ufffd\ufffd\ufffd";
} else {
result += String.fromCharCode(chr);
}
i += 6;
continue;
}
}
if ((b1 & 248) === 240 && i + 9 < l) {
// 111110xx 10xxxxxx 10xxxxxx 10xxxxxx
const b2 = parseInt(seq.slice(i + 4, i + 6), 16);
const b3 = parseInt(seq.slice(i + 7, i + 9), 16);
const b4 = parseInt(seq.slice(i + 10, i + 12), 16);
if ((b2 & 192) === 128 && (b3 & 192) === 128 && (b4 & 192) === 128) {
let chr = b1 << 18 & 1835008 | b2 << 12 & 258048 | b3 << 6 & 4032 | b4 & 63;
if (chr < 65536 || chr > 1114111) {
result += "\ufffd\ufffd\ufffd\ufffd";
} else {
chr -= 65536;
result += String.fromCharCode(55296 + (chr >> 10), 56320 + (chr & 1023));
}
i += 9;
continue;
}
}
result += "\ufffd";
}
return result;
}));
}
decode$1.defaultChars = ";/?:@&=+$,#";
decode$1.componentChars = "";
const encodeCache = {};
// Create a lookup array where anything but characters in `chars` string
// and alphanumeric chars is percent-encoded.
function getEncodeCache(exclude) {
let cache = encodeCache[exclude];
if (cache) {
return cache;
}
cache = encodeCache[exclude] = [];
for (let i = 0; i < 128; i++) {
const ch = String.fromCharCode(i);
if (/^[0-9a-z]$/i.test(ch)) {
// always allow unencoded alphanumeric characters
cache.push(ch);
} else {
cache.push("%" + ("0" + i.toString(16).toUpperCase()).slice(-2));
}
}
for (let i = 0; i < exclude.length; i++) {
cache[exclude.charCodeAt(i)] = exclude[i];
}
return cache;
}
// Encode unsafe characters with percent-encoding, skipping already
// encoded sequences.
// - string - string to encode
// - exclude - list of characters to ignore (in addition to a-zA-Z0-9)
// - keepEscaped - don't encode '%' in a correct escape sequence (default: true)
function encode$1(string, exclude, keepEscaped) {
if (typeof exclude !== "string") {
// encode(string, keepEscaped)
keepEscaped = exclude;
exclude = encode$1.defaultChars;
}
if (typeof keepEscaped === "undefined") {
keepEscaped = true;
}
const cache = getEncodeCache(exclude);
let result = "";
for (let i = 0, l = string.length; i < l; i++) {
const code = string.charCodeAt(i);
if (keepEscaped && code === 37 /* % */ && i + 2 < l) {
if (/^[0-9a-f]{2}$/i.test(string.slice(i + 1, i + 3))) {
result += string.slice(i, i + 3);
i += 2;
continue;
}
}
if (code < 128) {
result += cache[code];
continue;
}
if (code >= 55296 && code <= 57343) {
if (code >= 55296 && code <= 56319 && i + 1 < l) {
const nextCode = string.charCodeAt(i + 1);
if (nextCode >= 56320 && nextCode <= 57343) {
result += encodeURIComponent(string[i] + string[i + 1]);
i++;
continue;
}
}
result += "%EF%BF%BD";
continue;
}
result += encodeURIComponent(string[i]);
}
return result;
}
encode$1.defaultChars = ";/?:@&=+$,-_.!~*'()#";
encode$1.componentChars = "-_.!~*'()";
function format(url) {
let result = "";
result += url.protocol || "";
result += url.slashes ? "//" : "";
result += url.auth ? url.auth + "@" : "";
if (url.hostname && url.hostname.indexOf(":") !== -1) {
// ipv6 address
result += "[" + url.hostname + "]";
} else {
result += url.hostname || "";
}
result += url.port ? ":" + url.port : "";
result += url.pathname || "";
result += url.search || "";
result += url.hash || "";
return result;
}
// Copyright Joyent, Inc. and other Node contributors.
// Permission is hereby granted, free of charge, to any person obtaining a
// copy of this software and associated documentation files (the
// "Software"), to deal in the Software without restriction, including
// without limitation the rights to use, copy, modify, merge, publish,
// distribute, sublicense, and/or sell copies of the Software, and to permit
// persons to whom the Software is furnished to do so, subject to the
// following conditions:
// The above copyright notice and this permission notice shall be included
// in all copies or substantial portions of the Software.
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
// OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
// NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
// DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
// OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
// USE OR OTHER DEALINGS IN THE SOFTWARE.
// Changes from joyent/node:
// 1. No leading slash in paths,
// e.g. in `url.parse('http://foo?bar')` pathname is ``, not `/`
// 2. Backslashes are not replaced with slashes,
// so `http:\\example.org\` is treated like a relative path
// 3. Trailing colon is treated like a part of the path,
// i.e. in `http://example.org:foo` pathname is `:foo`
// 4. Nothing is URL-encoded in the resulting object,
// (in joyent/node some chars in auth and paths are encoded)
// 5. `url.parse()` does not have `parseQueryString` argument
// 6. Removed extraneous result properties: `host`, `path`, `query`, etc.,
// which can be constructed using other parts of the url.
function Url() {
this.protocol = null;
this.slashes = null;
this.auth = null;
this.port = null;
this.hostname = null;
this.hash = null;
this.search = null;
this.pathname = null;
}
// Reference: RFC 3986, RFC 1808, RFC 2396
// define these here so at least they only have to be
// compiled once on the first module load.
const protocolPattern = /^([a-z0-9.+-]+:)/i;
const portPattern = /:[0-9]*$/;
// Special case for a simple path URL
/* eslint-disable-next-line no-useless-escape */ const simplePathPattern = /^(\/\/?(?!\/)[^\?\s]*)(\?[^\s]*)?$/;
// RFC 2396: characters reserved for delimiting URLs.
// We actually just auto-escape these.
const delims = [ "<", ">", '"', "`", " ", "\r", "\n", "\t" ];
// RFC 2396: characters not allowed for various reasons.
const unwise = [ "{", "}", "|", "\\", "^", "`" ].concat(delims);
// Allowed by RFCs, but cause of XSS attacks. Always escape these.
const autoEscape = [ "'" ].concat(unwise);
// Characters that are never ever allowed in a hostname.
// Note that any invalid chars are also handled, but these
// are the ones that are *expected* to be seen, so we fast-path
// them.
const nonHostChars = [ "%", "/", "?", ";", "#" ].concat(autoEscape);
const hostEndingChars = [ "/", "?", "#" ];
const hostnameMaxLen = 255;
const hostnamePartPattern = /^[+a-z0-9A-Z_-]{0,63}$/;
const hostnamePartStart = /^([+a-z0-9A-Z_-]{0,63})(.*)$/;
// protocols that can allow "unsafe" and "unwise" chars.
// protocols that never have a hostname.
const hostlessProtocol = {
javascript: true,
"javascript:": true
};
// protocols that always contain a // bit.
const slashedProtocol = {
http: true,
https: true,
ftp: true,
gopher: true,
file: true,
"http:": true,
"https:": true,
"ftp:": true,
"gopher:": true,
"file:": true
};
function urlParse(url, slashesDenoteHost) {
if (url && url instanceof Url) return url;
const u = new Url;
u.parse(url, slashesDenoteHost);
return u;
}
Url.prototype.parse = function(url, slashesDenoteHost) {
let lowerProto, hec, slashes;
let rest = url;
// trim before proceeding.
// This is to support parse stuff like " http://foo.com \n"
rest = rest.trim();
if (!slashesDenoteHost && url.split("#").length === 1) {
// Try fast path regexp
const simplePath = simplePathPattern.exec(rest);
if (simplePath) {
this.pathname = simplePath[1];
if (simplePath[2]) {
this.search = simplePath[2];
}
return this;
}
}
let proto = protocolPattern.exec(rest);
if (proto) {
proto = proto[0];
lowerProto = proto.toLowerCase();
this.protocol = proto;
rest = rest.substr(proto.length);
}
// figure out if it's got a host
// user@server is *always* interpreted as a hostname, and url
// resolution will treat //foo/bar as host=foo,path=bar because that's
// how the browser resolves relative URLs.
/* eslint-disable-next-line no-useless-escape */ if (slashesDenoteHost || proto || rest.match(/^\/\/[^@\/]+@[^@\/]+/)) {
slashes = rest.substr(0, 2) === "//";
if (slashes && !(proto && hostlessProtocol[proto])) {
rest = rest.substr(2);
this.slashes = true;
}
}
if (!hostlessProtocol[proto] && (slashes || proto && !slashedProtocol[proto])) {
// there's a hostname.
// the first instance of /, ?, ;, or # ends the host.
// If there is an @ in the hostname, then non-host chars *are* allowed
// to the left of the last @ sign, unless some host-ending character
// comes *before* the @-sign.
// URLs are obnoxious.
// ex:
// http://a@b@c/ => user:a@b host:c
// http://a@b?@c => user:a host:c path:/?@c
// v0.12 TODO(isaacs): This is not quite how Chrome does things.
// Review our test case against browsers more comprehensively.
// find the first instance of any hostEndingChars
let hostEnd = -1;
for (let i = 0; i < hostEndingChars.length; i++) {
hec = rest.indexOf(hostEndingChars[i]);
if (hec !== -1 && (hostEnd === -1 || hec < hostEnd)) {
hostEnd = hec;
}
}
// at this point, either we have an explicit point where the
// auth portion cannot go past, or the last @ char is the decider.
let auth, atSign;
if (hostEnd === -1) {
// atSign can be anywhere.
atSign = rest.lastIndexOf("@");
} else {
// atSign must be in auth portion.
// http://a@b/c@d => host:b auth:a path:/c@d
atSign = rest.lastIndexOf("@", hostEnd);
}
// Now we have a portion which is definitely the auth.
// Pull that off.
if (atSign !== -1) {
auth = rest.slice(0, atSign);
rest = rest.slice(atSign + 1);
this.auth = auth;
}
// the host is the remaining to the left of the first non-host char
hostEnd = -1;
for (let i = 0; i < nonHostChars.length; i++) {
hec = rest.indexOf(nonHostChars[i]);
if (hec !== -1 && (hostEnd === -1 || hec < hostEnd)) {
hostEnd = hec;
}
}
// if we still have not hit it, then the entire thing is a host.
if (hostEnd === -1) {
hostEnd = rest.length;
}
if (rest[hostEnd - 1] === ":") {
hostEnd--;
}
const host = rest.slice(0, hostEnd);
rest = rest.slice(hostEnd);
// pull out port.
this.parseHost(host);
// we've indicated that there is a hostname,
// so even if it's empty, it has to be present.
this.hostname = this.hostname || "";
// if hostname begins with [ and ends with ]
// assume that it's an IPv6 address.
const ipv6Hostname = this.hostname[0] === "[" && this.hostname[this.hostname.length - 1] === "]";
// validate a little.
if (!ipv6Hostname) {
const hostparts = this.hostname.split(/\./);
for (let i = 0, l = hostparts.length; i < l; i++) {
const part = hostparts[i];
if (!part) {
continue;
}
if (!part.match(hostnamePartPattern)) {
let newpart = "";
for (let j = 0, k = part.length; j < k; j++) {
if (part.charCodeAt(j) > 127) {
// we replace non-ASCII char with a temporary placeholder
// we need this to make sure size of hostname is not
// broken by replacing non-ASCII by nothing
newpart += "x";
} else {
newpart += part[j];
}
}
// we test again with ASCII char only
if (!newpart.match(hostnamePartPattern)) {
const validParts = hostparts.slice(0, i);
const notHost = hostparts.slice(i + 1);
const bit = part.match(hostnamePartStart);
if (bit) {
validParts.push(bit[1]);
notHost.unshift(bit[2]);
}
if (notHost.length) {
rest = notHost.join(".") + rest;
}
this.hostname = validParts.join(".");
break;
}
}
}
}
if (this.hostname.length > hostnameMaxLen) {
this.hostname = "";
}
// strip [ and ] from the hostname
// the host field still retains them, though
if (ipv6Hostname) {
this.hostname = this.hostname.substr(1, this.hostname.length - 2);
}
}
// chop off from the tail first.
const hash = rest.indexOf("#");
if (hash !== -1) {
// got a fragment string.
this.hash = rest.substr(hash);
rest = rest.slice(0, hash);
}
const qm = rest.indexOf("?");
if (qm !== -1) {
this.search = rest.substr(qm);
rest = rest.slice(0, qm);
}
if (rest) {
this.pathname = rest;
}
if (slashedProtocol[lowerProto] && this.hostname && !this.pathname) {
this.pathname = "";
}
return this;
};
Url.prototype.parseHost = function(host) {
let port = portPattern.exec(host);
if (port) {
port = port[0];
if (port !== ":") {
this.port = port.substr(1);
}
host = host.substr(0, host.length - port.length);
}
if (host) {
this.hostname = host;
}
};
var mdurl = Object.freeze({
__proto__: null,
decode: decode$1,
encode: encode$1,
format: format,
parse: urlParse
});
var Any = /[\0-\uD7FF\uE000-\uFFFF]|[\uD800-\uDBFF][\uDC00-\uDFFF]|[\uD800-\uDBFF](?![\uDC00-\uDFFF])|(?:[^\uD800-\uDBFF]|^)[\uDC00-\uDFFF]/;
var Cc = /[\0-\x1F\x7F-\x9F]/;
var regex$1 = /[\xAD\u0600-\u0605\u061C\u06DD\u070F\u0890\u0891\u08E2\u180E\u200B-\u200F\u202A-\u202E\u2060-\u2064\u2066-\u206F\uFEFF\uFFF9-\uFFFB]|\uD804[\uDCBD\uDCCD]|\uD80D[\uDC30-\uDC3F]|\uD82F[\uDCA0-\uDCA3]|\uD834[\uDD73-\uDD7A]|\uDB40[\uDC01\uDC20-\uDC7F]/;
var P = /[!-#%-\*,-\/:;\?@\[-\]_\{\}\xA1\xA7\xAB\xB6\xB7\xBB\xBF\u037E\u0387\u055A-\u055F\u0589\u058A\u05BE\u05C0\u05C3\u05C6\u05F3\u05F4\u0609\u060A\u060C\u060D\u061B\u061D-\u061F\u066A-\u066D\u06D4\u0700-\u070D\u07F7-\u07F9\u0830-\u083E\u085E\u0964\u0965\u0970\u09FD\u0A76\u0AF0\u0C77\u0C84\u0DF4\u0E4F\u0E5A\u0E5B\u0F04-\u0F12\u0F14\u0F3A-\u0F3D\u0F85\u0FD0-\u0FD4\u0FD9\u0FDA\u104A-\u104F\u10FB\u1360-\u1368\u1400\u166E\u169B\u169C\u16EB-\u16ED\u1735\u1736\u17D4-\u17D6\u17D8-\u17DA\u1800-\u180A\u1944\u1945\u1A1E\u1A1F\u1AA0-\u1AA6\u1AA8-\u1AAD\u1B5A-\u1B60\u1B7D\u1B7E\u1BFC-\u1BFF\u1C3B-\u1C3F\u1C7E\u1C7F\u1CC0-\u1CC7\u1CD3\u2010-\u2027\u2030-\u2043\u2045-\u2051\u2053-\u205E\u207D\u207E\u208D\u208E\u2308-\u230B\u2329\u232A\u2768-\u2775\u27C5\u27C6\u27E6-\u27EF\u2983-\u2998\u29D8-\u29DB\u29FC\u29FD\u2CF9-\u2CFC\u2CFE\u2CFF\u2D70\u2E00-\u2E2E\u2E30-\u2E4F\u2E52-\u2E5D\u3001-\u3003\u3008-\u3011\u3014-\u301F\u3030\u303D\u30A0\u30FB\uA4FE\uA4FF\uA60D-\uA60F\uA673\uA67E\uA6F2-\uA6F7\uA874-\uA877\uA8CE\uA8CF\uA8F8-\uA8FA\uA8FC\uA92E\uA92F\uA95F\uA9C1-\uA9CD\uA9DE\uA9DF\uAA5C-\uAA5F\uAADE\uAADF\uAAF0\uAAF1\uABEB\uFD3E\uFD3F\uFE10-\uFE19\uFE30-\uFE52\uFE54-\uFE61\uFE63\uFE68\uFE6A\uFE6B\uFF01-\uFF03\uFF05-\uFF0A\uFF0C-\uFF0F\uFF1A\uFF1B\uFF1F\uFF20\uFF3B-\uFF3D\uFF3F\uFF5B\uFF5D\uFF5F-\uFF65]|\uD800[\uDD00-\uDD02\uDF9F\uDFD0]|\uD801\uDD6F|\uD802[\uDC57\uDD1F\uDD3F\uDE50-\uDE58\uDE7F\uDEF0-\uDEF6\uDF39-\uDF3F\uDF99-\uDF9C]|\uD803[\uDEAD\uDF55-\uDF59\uDF86-\uDF89]|\uD804[\uDC47-\uDC4D\uDCBB\uDCBC\uDCBE-\uDCC1\uDD40-\uDD43\uDD74\uDD75\uDDC5-\uDDC8\uDDCD\uDDDB\uDDDD-\uDDDF\uDE38-\uDE3D\uDEA9]|\uD805[\uDC4B-\uDC4F\uDC5A\uDC5B\uDC5D\uDCC6\uDDC1-\uDDD7\uDE41-\uDE43\uDE60-\uDE6C\uDEB9\uDF3C-\uDF3E]|\uD806[\uDC3B\uDD44-\uDD46\uDDE2\uDE3F-\uDE46\uDE9A-\uDE9C\uDE9E-\uDEA2\uDF00-\uDF09]|\uD807[\uDC41-\uDC45\uDC70\uDC71\uDEF7\uDEF8\uDF43-\uDF4F\uDFFF]|\uD809[\uDC70-\uDC74]|\uD80B[\uDFF1\uDFF2]|\uD81A[\uDE6E\uDE6F\uDEF5\uDF37-\uDF3B\uDF44]|\uD81B[\uDE97-\uDE9A\uDFE2]|\uD82F\uDC9F|\uD836[\uDE87-\uDE8B]|\uD83A[\uDD5E\uDD5F]/;
var regex = /[\$\+<->\^`\|~\xA2-\xA6\xA8\xA9\xAC\xAE-\xB1\xB4\xB8\xD7\xF7\u02C2-\u02C5\u02D2-\u02DF\u02E5-\u02EB\u02ED\u02EF-\u02FF\u0375\u0384\u0385\u03F6\u0482\u058D-\u058F\u0606-\u0608\u060B\u060E\u060F\u06DE\u06E9\u06FD\u06FE\u07F6\u07FE\u07FF\u0888\u09F2\u09F3\u09FA\u09FB\u0AF1\u0B70\u0BF3-\u0BFA\u0C7F\u0D4F\u0D79\u0E3F\u0F01-\u0F03\u0F13\u0F15-\u0F17\u0F1A-\u0F1F\u0F34\u0F36\u0F38\u0FBE-\u0FC5\u0FC7-\u0FCC\u0FCE\u0FCF\u0FD5-\u0FD8\u109E\u109F\u1390-\u1399\u166D\u17DB\u1940\u19DE-\u19FF\u1B61-\u1B6A\u1B74-\u1B7C\u1FBD\u1FBF-\u1FC1\u1FCD-\u1FCF\u1FDD-\u1FDF\u1FED-\u1FEF\u1FFD\u1FFE\u2044\u2052\u207A-\u207C\u208A-\u208C\u20A0-\u20C0\u2100\u2101\u2103-\u2106\u2108\u2109\u2114\u2116-\u2118\u211E-\u2123\u2125\u2127\u2129\u212E\u213A\u213B\u2140-\u2144\u214A-\u214D\u214F\u218A\u218B\u2190-\u2307\u230C-\u2328\u232B-\u2426\u2440-\u244A\u249C-\u24E9\u2500-\u2767\u2794-\u27C4\u27C7-\u27E5\u27F0-\u2982\u2999-\u29D7\u29DC-\u29FB\u29FE-\u2B73\u2B76-\u2B95\u2B97-\u2BFF\u2CE5-\u2CEA\u2E50\u2E51\u2E80-\u2E99\u2E9B-\u2EF3\u2F00-\u2FD5\u2FF0-\u2FFF\u3004\u3012\u3013\u3020\u3036\u3037\u303E\u303F\u309B\u309C\u3190\u3191\u3196-\u319F\u31C0-\u31E3\u31EF\u3200-\u321E\u322A-\u3247\u3250\u3260-\u327F\u328A-\u32B0\u32C0-\u33FF\u4DC0-\u4DFF\uA490-\uA4C6\uA700-\uA716\uA720\uA721\uA789\uA78A\uA828-\uA82B\uA836-\uA839\uAA77-\uAA79\uAB5B\uAB6A\uAB6B\uFB29\uFBB2-\uFBC2\uFD40-\uFD4F\uFDCF\uFDFC-\uFDFF\uFE62\uFE64-\uFE66\uFE69\uFF04\uFF0B\uFF1C-\uFF1E\uFF3E\uFF40\uFF5C\uFF5E\uFFE0-\uFFE6\uFFE8-\uFFEE\uFFFC\uFFFD]|\uD800[\uDD37-\uDD3F\uDD79-\uDD89\uDD8C-\uDD8E\uDD90-\uDD9C\uDDA0\uDDD0-\uDDFC]|\uD802[\uDC77\uDC78\uDEC8]|\uD805\uDF3F|\uD807[\uDFD5-\uDFF1]|\uD81A[\uDF3C-\uDF3F\uDF45]|\uD82F\uDC9C|\uD833[\uDF50-\uDFC3]|\uD834[\uDC00-\uDCF5\uDD00-\uDD26\uDD29-\uDD64\uDD6A-\uDD6C\uDD83\uDD84\uDD8C-\uDDA9\uDDAE-\uDDEA\uDE00-\uDE41\uDE45\uDF00-\uDF56]|\uD835[\uDEC1\uDEDB\uDEFB\uDF15\uDF35\uDF4F\uDF6F\uDF89\uDFA9\uDFC3]|\uD836[\uDC00-\uDDFF\uDE37-\uDE3A\uDE6D-\uDE74\uDE76-\uDE83\uDE85\uDE86]|\uD838[\uDD4F\uDEFF]|\uD83B[\uDCAC\uDCB0\uDD2E\uDEF0\uDEF1]|\uD83C[\uDC00-\uDC2B\uDC30-\uDC93\uDCA0-\uDCAE\uDCB1-\uDCBF\uDCC1-\uDCCF\uDCD1-\uDCF5\uDD0D-\uDDAD\uDDE6-\uDE02\uDE10-\uDE3B\uDE40-\uDE48\uDE50\uDE51\uDE60-\uDE65\uDF00-\uDFFF]|\uD83D[\uDC00-\uDED7\uDEDC-\uDEEC\uDEF0-\uDEFC\uDF00-\uDF76\uDF7B-\uDFD9\uDFE0-\uDFEB\uDFF0]|\uD83E[\uDC00-\uDC0B\uDC10-\uDC47\uDC50-\uDC59\uDC60-\uDC87\uDC90-\uDCAD\uDCB0\uDCB1\uDD00-\uDE53\uDE60-\uDE6D\uDE70-\uDE7C\uDE80-\uDE88\uDE90-\uDEBD\uDEBF-\uDEC5\uDECE-\uDEDB\uDEE0-\uDEE8\uDEF0-\uDEF8\uDF00-\uDF92\uDF94-\uDFCA]/;
var Z = /[ \xA0\u1680\u2000-\u200A\u2028\u2029\u202F\u205F\u3000]/;
var ucmicro = Object.freeze({
__proto__: null,
Any: Any,
Cc: Cc,
Cf: regex$1,
P: P,
S: regex,
Z: Z
});
// Generated using scripts/write-decode-map.ts
var htmlDecodeTree = new Uint16Array(
// prettier-ignore
'\u1d41<\xd5\u0131\u028a\u049d\u057b\u05d0\u0675\u06de\u07a2\u07d6\u080f\u0a4a\u0a91\u0da1\u0e6d\u0f09\u0f26\u10ca\u1228\u12e1\u1415\u149d\u14c3\u14df\u1525\0\0\0\0\0\0\u156b\u16cd\u198d\u1c12\u1ddd\u1f7e\u2060\u21b0\u228d\u23c0\u23fb\u2442\u2824\u2912\u2d08\u2e48\u2fce\u3016\u32ba\u3639\u37ac\u38fe\u3a28\u3a71\u3ae0\u3b2e\u0800EMabcfglmnoprstu\\bfms\x7f\x84\x8b\x90\x95\x98\xa6\xb3\xb9\xc8\xcflig\u803b\xc6\u40c6P\u803b&\u4026cute\u803b\xc1\u40c1reve;\u4102\u0100iyx}rc\u803b\xc2\u40c2;\u4410r;\uc000\ud835\udd04rave\u803b\xc0\u40c0pha;\u4391acr;\u4100d;\u6a53\u0100gp\x9d\xa1on;\u4104f;\uc000\ud835\udd38plyFunction;\u6061ing\u803b\xc5\u40c5\u0100cs\xbe\xc3r;\uc000\ud835\udc9cign;\u6254ilde\u803b\xc3\u40c3ml\u803b\xc4\u40c4\u0400aceforsu\xe5\xfb\xfe\u0117\u011c\u0122\u0127\u012a\u0100cr\xea\xf2kslash;\u6216\u0176\xf6\xf8;\u6ae7ed;\u6306y;\u4411\u0180crt\u0105\u010b\u0114ause;\u6235noullis;\u612ca;\u4392r;\uc000\ud835\udd05pf;\uc000\ud835\udd39eve;\u42d8c\xf2\u0113mpeq;\u624e\u0700HOacdefhilorsu\u014d\u0151\u0156\u0180\u019e\u01a2\u01b5\u01b7\u01ba\u01dc\u0215\u0273\u0278\u027ecy;\u4427PY\u803b\xa9\u40a9\u0180cpy\u015d\u0162\u017aute;\u4106\u0100;i\u0167\u0168\u62d2talDifferentialD;\u6145leys;\u612d\u0200aeio\u0189\u018e\u0194\u0198ron;\u410cdil\u803b\xc7\u40c7rc;\u4108nint;\u6230ot;\u410a\u0100dn\u01a7\u01adilla;\u40b8terDot;\u40b7\xf2\u017fi;\u43a7rcle\u0200DMPT\u01c7\u01cb\u01d1\u01d6ot;\u6299inus;\u6296lus;\u6295imes;\u6297o\u0100cs\u01e2\u01f8kwiseContourIntegral;\u6232eCurly\u0100DQ\u0203\u020foubleQuote;\u601duote;\u6019\u0200lnpu\u021e\u0228\u0247\u0255on\u0100;e\u0225\u0226\u6237;\u6a74\u0180git\u022f\u0236\u023aruent;\u6261nt;\u622fourIntegral;\u622e\u0100fr\u024c\u024e;\u6102oduct;\u6210nterClockwiseContourIntegral;\u6233oss;\u6a2fcr;\uc000\ud835\udc9ep\u0100;C\u0284\u0285\u62d3ap;\u624d\u0580DJSZacefios\u02a0\u02ac\u02b0\u02b4\u02b8\u02cb\u02d7\u02e1\u02e6\u0333\u048d\u0100;o\u0179\u02a5trahd;\u6911cy;\u4402cy;\u4405cy;\u440f\u0180grs\u02bf\u02c4\u02c7ger;\u6021r;\u61a1hv;\u6ae4\u0100ay\u02d0\u02d5ron;\u410e;\u4414l\u0100;t\u02dd\u02de\u6207a;\u4394r;\uc000\ud835\udd07\u0100af\u02eb\u0327\u0100cm\u02f0\u0322ritical\u0200ADGT\u0300\u0306\u0316\u031ccute;\u40b4o\u0174\u030b\u030d;\u42d9bleAcute;\u42ddrave;\u4060ilde;\u42dcond;\u62c4ferentialD;\u6146\u0470\u033d\0\0\0\u0342\u0354\0\u0405f;\uc000\ud835\udd3b\u0180;DE\u0348\u0349\u034d\u40a8ot;\u60dcqual;\u6250ble\u0300CDLRUV\u0363\u0372\u0382\u03cf\u03e2\u03f8ontourIntegra\xec\u0239o\u0274\u0379\0\0\u037b\xbb\u0349nArrow;\u61d3\u0100eo\u0387\u03a4ft\u0180ART\u0390\u0396\u03a1rrow;\u61d0ightArrow;\u61d4e\xe5\u02cang\u0100LR\u03ab\u03c4eft\u0100AR\u03b3\u03b9rrow;\u67f8ightArrow;\u67faightArrow;\u67f9ight\u0100AT\u03d8\u03derrow;\u61d2ee;\u62a8p\u0241\u03e9\0\0\u03efrrow;\u61d1ownArrow;\u61d5erticalBar;\u6225n\u0300ABLRTa\u0412\u042a\u0430\u045e\u047f\u037crrow\u0180;BU\u041d\u041e\u0422\u6193ar;\u6913pArrow;\u61f5reve;\u4311eft\u02d2\u043a\0\u0446\0\u0450ightVector;\u6950eeVector;\u695eector\u0100;B\u0459\u045a\u61bdar;\u6956ight\u01d4\u0467\0\u0471eeVector;\u695fector\u0100;B\u047a\u047b\u61c1ar;\u6957ee\u0100;A\u0486\u0487\u62a4rrow;\u61a7\u0100ct\u0492\u0497r;\uc000\ud835\udc9frok;\u4110\u0800NTacdfglmopqstux\u04bd\u04c0\u04c4\u04cb\u04de\u04e2\u04e7\u04ee\u04f5\u0521\u052f\u0536\u0552\u055d\u0560\u0565G;\u414aH\u803b\xd0\u40d0cute\u803b\xc9\u40c9\u0180aiy\u04d2\u04d7\u04dcron;\u411arc\u803b\xca\u40ca;\u442dot;\u4116r;\uc000\ud835\udd08rave\u803b\xc8\u40c8ement;\u6208\u0100ap\u04fa\u04fecr;\u4112ty\u0253\u0506\0\0\u0512mallSquare;\u65fberySmallSquare;\u65ab\u0100gp\u0526\u052aon;\u4118f;\uc000\ud835\udd3csilon;\u4395u\u0100ai\u053c\u0549l\u0100;T\u0542\u0543\u6a75ilde;\u6242librium;\u61cc\u0100ci\u0557\u055ar;\u6130m;\u6a73a;\u4397ml\u803b\xcb\u40cb\u0100ip\u056a\u056fsts;\u6203onentialE;\u6147\u0280cfios\u0585\u0588\u058d\u05b2\u05ccy;\u4424r;\uc000\ud835\udd09lled\u0253\u0597\0\0\u05a3mallSquare;\u65fcerySmallSquare;\u65aa\u0370\u05ba\0\u05bf\0\0\u05c4f;\uc000\ud835\udd3dAll;\u6200riertrf;\u6131c\xf2\u05cb\u0600JTabcdfgorst\u05e8\u05ec\u05ef\u05fa\u0600\u0612\u
// Generated using scripts/write-decode-map.ts
var xmlDecodeTree = new Uint16Array(
// prettier-ignore
"\u0200aglq\t\x15\x18\x1b\u026d\x0f\0\0\x12p;\u4026os;\u4027t;\u403et;\u403cuot;\u4022".split("").map((c => c.charCodeAt(0))));
// Adapted from https://github.com/mathiasbynens/he/blob/36afe179392226cf1b6ccdb16ebbb7a5a844d93a/src/he.js#L106-L134
var _a;
const decodeMap = new Map([ [ 0, 65533 ],
// C1 Unicode control character reference replacements
[ 128, 8364 ], [ 130, 8218 ], [ 131, 402 ], [ 132, 8222 ], [ 133, 8230 ], [ 134, 8224 ], [ 135, 8225 ], [ 136, 710 ], [ 137, 8240 ], [ 138, 352 ], [ 139, 8249 ], [ 140, 338 ], [ 142, 381 ], [ 145, 8216 ], [ 146, 8217 ], [ 147, 8220 ], [ 148, 8221 ], [ 149, 8226 ], [ 150, 8211 ], [ 151, 8212 ], [ 152, 732 ], [ 153, 8482 ], [ 154, 353 ], [ 155, 8250 ], [ 156, 339 ], [ 158, 382 ], [ 159, 376 ] ]);
/**
* Polyfill for `String.fromCodePoint`. It is used to create a string from a Unicode code point.
*/ const fromCodePoint$1 =
// eslint-disable-next-line @typescript-eslint/no-unnecessary-condition, node/no-unsupported-features/es-builtins
(_a = String.fromCodePoint) !== null && _a !== void 0 ? _a : function(codePoint) {
let output = "";
if (codePoint > 65535) {
codePoint -= 65536;
output += String.fromCharCode(codePoint >>> 10 & 1023 | 55296);
codePoint = 56320 | codePoint & 1023;
}
output += String.fromCharCode(codePoint);
return output;
};
/**
* Replace the given code point with a replacement character if it is a
* surrogate or is outside the valid range. Otherwise return the code
* point unchanged.
*/ function replaceCodePoint(codePoint) {
var _a;
if (codePoint >= 55296 && codePoint <= 57343 || codePoint > 1114111) {
return 65533;
}
return (_a = decodeMap.get(codePoint)) !== null && _a !== void 0 ? _a : codePoint;
}
var CharCodes;
(function(CharCodes) {
CharCodes[CharCodes["NUM"] = 35] = "NUM";
CharCodes[CharCodes["SEMI"] = 59] = "SEMI";
CharCodes[CharCodes["EQUALS"] = 61] = "EQUALS";
CharCodes[CharCodes["ZERO"] = 48] = "ZERO";
CharCodes[CharCodes["NINE"] = 57] = "NINE";
CharCodes[CharCodes["LOWER_A"] = 97] = "LOWER_A";
CharCodes[CharCodes["LOWER_F"] = 102] = "LOWER_F";
CharCodes[CharCodes["LOWER_X"] = 120] = "LOWER_X";
CharCodes[CharCodes["LOWER_Z"] = 122] = "LOWER_Z";
CharCodes[CharCodes["UPPER_A"] = 65] = "UPPER_A";
CharCodes[CharCodes["UPPER_F"] = 70] = "UPPER_F";
CharCodes[CharCodes["UPPER_Z"] = 90] = "UPPER_Z";
})(CharCodes || (CharCodes = {}));
/** Bit that needs to be set to convert an upper case ASCII character to lower case */ const TO_LOWER_BIT = 32;
var BinTrieFlags;
(function(BinTrieFlags) {
BinTrieFlags[BinTrieFlags["VALUE_LENGTH"] = 49152] = "VALUE_LENGTH";
BinTrieFlags[BinTrieFlags["BRANCH_LENGTH"] = 16256] = "BRANCH_LENGTH";
BinTrieFlags[BinTrieFlags["JUMP_TABLE"] = 127] = "JUMP_TABLE";
})(BinTrieFlags || (BinTrieFlags = {}));
function isNumber(code) {
return code >= CharCodes.ZERO && code <= CharCodes.NINE;
}
function isHexadecimalCharacter(code) {
return code >= CharCodes.UPPER_A && code <= CharCodes.UPPER_F || code >= CharCodes.LOWER_A && code <= CharCodes.LOWER_F;
}
function isAsciiAlphaNumeric(code) {
return code >= CharCodes.UPPER_A && code <= CharCodes.UPPER_Z || code >= CharCodes.LOWER_A && code <= CharCodes.LOWER_Z || isNumber(code);
}
/**
* Checks if the given character is a valid end character for an entity in an attribute.
*
* Attribute values that aren't terminated properly aren't parsed, and shouldn't lead to a parser error.
* See the example in https://html.spec.whatwg.org/multipage/parsing.html#named-character-reference-state
*/ function isEntityInAttributeInvalidEnd(code) {
return code === CharCodes.EQUALS || isAsciiAlphaNumeric(code);
}
var EntityDecoderState;
(function(EntityDecoderState) {
EntityDecoderState[EntityDecoderState["EntityStart"] = 0] = "EntityStart";
EntityDecoderState[EntityDecoderState["NumericStart"] = 1] = "NumericStart";
EntityDecoderState[EntityDecoderState["NumericDecimal"] = 2] = "NumericDecimal";
EntityDecoderState[EntityDecoderState["NumericHex"] = 3] = "NumericHex";
EntityDecoderState[EntityDecoderState["NamedEntity"] = 4] = "NamedEntity";
})(EntityDecoderState || (EntityDecoderState = {}));
var DecodingMode;
(function(DecodingMode) {
/** Entities in text nodes that can end with any character. */
DecodingMode[DecodingMode["Legacy"] = 0] = "Legacy";
/** Only allow entities terminated with a semicolon. */ DecodingMode[DecodingMode["Strict"] = 1] = "Strict";
/** Entities in attributes have limitations on ending characters. */ DecodingMode[DecodingMode["Attribute"] = 2] = "Attribute";
})(DecodingMode || (DecodingMode = {}));
/**
* Token decoder with support of writing partial entities.
*/ class EntityDecoder {
constructor(/** The tree used to decode entities. */
decodeTree,
/**
* The function that is called when a codepoint is decoded.
*
* For multi-byte named entities, this will be called multiple times,
* with the second codepoint, and the same `consumed` value.
*
* @param codepoint The decoded codepoint.
* @param consumed The number of bytes consumed by the decoder.
*/
emitCodePoint, /** An object that is used to produce errors. */
errors) {
this.decodeTree = decodeTree;
this.emitCodePoint = emitCodePoint;
this.errors = errors;
/** The current state of the decoder. */ this.state = EntityDecoderState.EntityStart;
/** Characters that were consumed while parsing an entity. */ this.consumed = 1;
/**
* The result of the entity.
*
* Either the result index of a numeric entity, or the codepoint of a
* numeric entity.
*/ this.result = 0;
/** The current index in the decode tree. */ this.treeIndex = 0;
/** The number of characters that were consumed in excess. */ this.excess = 1;
/** The mode in which the decoder is operating. */ this.decodeMode = DecodingMode.Strict;
}
/** Resets the instance to make it reusable. */ startEntity(decodeMode) {
this.decodeMode = decodeMode;
this.state = EntityDecoderState.EntityStart;
this.result = 0;
this.treeIndex = 0;
this.excess = 1;
this.consumed = 1;
}
/**
* Write an entity to the decoder. This can be called multiple times with partial entities.
* If the entity is incomplete, the decoder will return -1.
*
* Mirrors the implementation of `getDecoder`, but with the ability to stop decoding if the
* entity is incomplete, and resume when the next string is written.
*
* @param string The string containing the entity (or a continuation of the entity).
* @param offset The offset at which the entity begins. Should be 0 if this is not the first call.
* @returns The number of characters that were consumed, or -1 if the entity is incomplete.
*/ write(str, offset) {
switch (this.state) {
case EntityDecoderState.EntityStart:
{
if (str.charCodeAt(offset) === CharCodes.NUM) {
this.state = EntityDecoderState.NumericStart;
this.consumed += 1;
return this.stateNumericStart(str, offset + 1);
}
this.state = EntityDecoderState.NamedEntity;
return this.stateNamedEntity(str, offset);
}
case EntityDecoderState.NumericStart:
{
return this.stateNumericStart(str, offset);
}
case EntityDecoderState.NumericDecimal:
{
return this.stateNumericDecimal(str, offset);
}
case EntityDecoderState.NumericHex:
{
return this.stateNumericHex(str, offset);
}
case EntityDecoderState.NamedEntity:
{
return this.stateNamedEntity(str, offset);
}
}
}
/**
* Switches between the numeric decimal and hexadecimal states.
*
* Equivalent to the `Numeric character reference state` in the HTML spec.
*
* @param str The string containing the entity (or a continuation of the entity).
* @param offset The current offset.
* @returns The number of characters that were consumed, or -1 if the entity is incomplete.
*/ stateNumericStart(str, offset) {
if (offset >= str.length) {
return -1;
}
if ((str.charCodeAt(offset) | TO_LOWER_BIT) === CharCodes.LOWER_X) {
this.state = EntityDecoderState.NumericHex;
this.consumed += 1;
return this.stateNumericHex(str, offset + 1);
}
this.state = EntityDecoderState.NumericDecimal;
return this.stateNumericDecimal(str, offset);
}
addToNumericResult(str, start, end, base) {
if (start !== end) {
const digitCount = end - start;
this.result = this.result * Math.pow(base, digitCount) + parseInt(str.substr(start, digitCount), base);
this.consumed += digitCount;
}
}
/**
* Parses a hexadecimal numeric entity.
*
* Equivalent to the `Hexademical character reference state` in the HTML spec.
*
* @param str The string containing the entity (or a continuation of the entity).
* @param offset The current offset.
* @returns The number of characters that were consumed, or -1 if the entity is incomplete.
*/ stateNumericHex(str, offset) {
const startIdx = offset;
while (offset < str.length) {
const char = str.charCodeAt(offset);
if (isNumber(char) || isHexadecimalCharacter(char)) {
offset += 1;
} else {
this.addToNumericResult(str, startIdx, offset, 16);
return this.emitNumericEntity(char, 3);
}
}
this.addToNumericResult(str, startIdx, offset, 16);
return -1;
}
/**
* Parses a decimal numeric entity.
*
* Equivalent to the `Decimal character reference state` in the HTML spec.
*
* @param str The string containing the entity (or a continuation of the entity).
* @param offset The current offset.
* @returns The number of characters that were consumed, or -1 if the entity is incomplete.
*/ stateNumericDecimal(str, offset) {
const startIdx = offset;
while (offset < str.length) {
const char = str.charCodeAt(offset);
if (isNumber(char)) {
offset += 1;
} else {
this.addToNumericResult(str, startIdx, offset, 10);
return this.emitNumericEntity(char, 2);
}
}
this.addToNumericResult(str, startIdx, offset, 10);
return -1;
}
/**
* Validate and emit a numeric entity.
*
* Implements the logic from the `Hexademical character reference start
* state` and `Numeric character reference end state` in the HTML spec.
*
* @param lastCp The last code point of the entity. Used to see if the
* entity was terminated with a semicolon.
* @param expectedLength The minimum number of characters that should be
* consumed. Used to validate that at least one digit
* was consumed.
* @returns The number of characters that were consumed.
*/ emitNumericEntity(lastCp, expectedLength) {
var _a;
// Ensure we consumed at least one digit.
if (this.consumed <= expectedLength) {
(_a = this.errors) === null || _a === void 0 ? void 0 : _a.absenceOfDigitsInNumericCharacterReference(this.consumed);
return 0;
}
// Figure out if this is a legit end of the entity
if (lastCp === CharCodes.SEMI) {
this.consumed += 1;
} else if (this.decodeMode === DecodingMode.Strict) {
return 0;
}
this.emitCodePoint(replaceCodePoint(this.result), this.consumed);
if (this.errors) {
if (lastCp !== CharCodes.SEMI) {
this.errors.missingSemicolonAfterCharacterReference();
}
this.errors.validateNumericCharacterReference(this.result);
}
return this.consumed;
}
/**
* Parses a named entity.
*
* Equivalent to the `Named character reference state` in the HTML spec.
*
* @param str The string containing the entity (or a continuation of the entity).
* @param offset The current offset.
* @returns The number of characters that were consumed, or -1 if the entity is incomplete.
*/ stateNamedEntity(str, offset) {
const {decodeTree: decodeTree} = this;
let current = decodeTree[this.treeIndex];
// The mask is the number of bytes of the value, including the current byte.
let valueLength = (current & BinTrieFlags.VALUE_LENGTH) >> 14;
for (;offset < str.length; offset++, this.excess++) {
const char = str.charCodeAt(offset);
this.treeIndex = determineBranch(decodeTree, current, this.treeIndex + Math.max(1, valueLength), char);
if (this.treeIndex < 0) {
return this.result === 0 ||
// If we are parsing an attribute
this.decodeMode === DecodingMode.Attribute && (
// We shouldn't have consumed any characters after the entity,
valueLength === 0 ||
// And there should be no invalid characters.
isEntityInAttributeInvalidEnd(char)) ? 0 : this.emitNotTerminatedNamedEntity();
}
current = decodeTree[this.treeIndex];
valueLength = (current & BinTrieFlags.VALUE_LENGTH) >> 14;
// If the branch is a value, store it and continue
if (valueLength !== 0) {
// If the entity is terminated by a semicolon, we are done.
if (char === CharCodes.SEMI) {
return this.emitNamedEntityData(this.treeIndex, valueLength, this.consumed + this.excess);
}
// If we encounter a non-terminated (legacy) entity while parsing strictly, then ignore it.
if (this.decodeMode !== DecodingMode.Strict) {
this.result = this.treeIndex;
this.consumed += this.excess;
this.excess = 0;
}
}
}
return -1;
}
/**
* Emit a named entity that was not terminated with a semicolon.
*
* @returns The number of characters consumed.
*/ emitNotTerminatedNamedEntity() {
var _a;
const {result: result, decodeTree: decodeTree} = this;
const valueLength = (decodeTree[result] & BinTrieFlags.VALUE_LENGTH) >> 14;
this.emitNamedEntityData(result, valueLength, this.consumed);
(_a = this.errors) === null || _a === void 0 ? void 0 : _a.missingSemicolonAfterCharacterReference();
return this.consumed;
}
/**
* Emit a named entity.
*
* @param result The index of the entity in the decode tree.
* @param valueLength The number of bytes in the entity.
* @param consumed The number of characters consumed.
*
* @returns The number of characters consumed.
*/ emitNamedEntityData(result, valueLength, consumed) {
const {decodeTree: decodeTree} = this;
this.emitCodePoint(valueLength === 1 ? decodeTree[result] & ~BinTrieFlags.VALUE_LENGTH : decodeTree[result + 1], consumed);
if (valueLength === 3) {
// For multi-byte values, we need to emit the second byte.
this.emitCodePoint(decodeTree[result + 2], consumed);
}
return consumed;
}
/**
* Signal to the parser that the end of the input was reached.
*
* Remaining data will be emitted and relevant errors will be produced.
*
* @returns The number of characters consumed.
*/ end() {
var _a;
switch (this.state) {
case EntityDecoderState.NamedEntity:
{
// Emit a named entity if we have one.
return this.result !== 0 && (this.decodeMode !== DecodingMode.Attribute || this.result === this.treeIndex) ? this.emitNotTerminatedNamedEntity() : 0;
}
// Otherwise, emit a numeric entity if we have one.
case EntityDecoderState.NumericDecimal:
{
return this.emitNumericEntity(0, 2);
}
case EntityDecoderState.NumericHex:
{
return this.emitNumericEntity(0, 3);
}
case EntityDecoderState.NumericStart:
{
(_a = this.errors) === null || _a === void 0 ? void 0 : _a.absenceOfDigitsInNumericCharacterReference(this.consumed);
return 0;
}
case EntityDecoderState.EntityStart:
{
// Return 0 if we have no entity.
return 0;
}
}
}
}
/**
* Creates a function that decodes entities in a string.
*
* @param decodeTree The decode tree.
* @returns A function that decodes entities in a string.
*/ function getDecoder(decodeTree) {
let ret = "";
const decoder = new EntityDecoder(decodeTree, (str => ret += fromCodePoint$1(str)));
return function decodeWithTrie(str, decodeMode) {
let lastIndex = 0;
let offset = 0;
while ((offset = str.indexOf("&", offset)) >= 0) {
ret += str.slice(lastIndex, offset);
decoder.startEntity(decodeMode);
const len = decoder.write(str,
// Skip the "&"
offset + 1);
if (len < 0) {
lastIndex = offset + decoder.end();
break;
}
lastIndex = offset + len;
// If `len` is 0, skip the current `&` and continue.
offset = len === 0 ? lastIndex + 1 : lastIndex;
}
const result = ret + str.slice(lastIndex);
// Make sure we don't keep a reference to the final string.
ret = "";
return result;
};
}
/**
* Determines the branch of the current node that is taken given the current
* character. This function is used to traverse the trie.
*
* @param decodeTree The trie.
* @param current The current node.
* @param nodeIdx The index right after the current node and its value.
* @param char The current character.
* @returns The index of the next node, or -1 if no branch is taken.
*/ function determineBranch(decodeTree, current, nodeIdx, char) {
const branchCount = (current & BinTrieFlags.BRANCH_LENGTH) >> 7;
const jumpOffset = current & BinTrieFlags.JUMP_TABLE;
// Case 1: Single branch encoded in jump offset
if (branchCount === 0) {
return jumpOffset !== 0 && char === jumpOffset ? nodeIdx : -1;
}
// Case 2: Multiple branches encoded in jump table
if (jumpOffset) {
const value = char - jumpOffset;
return value < 0 || value >= branchCount ? -1 : decodeTree[nodeIdx + value] - 1;
}
// Case 3: Multiple branches encoded in dictionary
// Binary search for the character.
let lo = nodeIdx;
let hi = lo + branchCount - 1;
while (lo <= hi) {
const mid = lo + hi >>> 1;
const midVal = decodeTree[mid];
if (midVal < char) {
lo = mid + 1;
} else if (midVal > char) {
hi = mid - 1;
} else {
return decodeTree[mid + branchCount];
}
}
return -1;
}
const htmlDecoder = getDecoder(htmlDecodeTree);
getDecoder(xmlDecodeTree);
/**
* Decodes an HTML string.
*
* @param str The string to decode.
* @param mode The decoding mode.
* @returns The decoded string.
*/ function decodeHTML(str, mode = DecodingMode.Legacy) {
return htmlDecoder(str, mode);
}
// Utilities
function _class$1(obj) {
return Object.prototype.toString.call(obj);
}
function isString$1(obj) {
return _class$1(obj) === "[object String]";
}
const _hasOwnProperty = Object.prototype.hasOwnProperty;
function has(object, key) {
return _hasOwnProperty.call(object, key);
}
// Merge objects
function assign$1(obj /* from1, from2, from3, ... */) {
const sources = Array.prototype.slice.call(arguments, 1);
sources.forEach((function(source) {
if (!source) {
return;
}
if (typeof source !== "object") {
throw new TypeError(source + "must be object");
}
Object.keys(source).forEach((function(key) {
obj[key] = source[key];
}));
}));
return obj;
}
// Remove element from array and put another array at those position.
// Useful for some operations with tokens
function arrayReplaceAt(src, pos, newElements) {
return [].concat(src.slice(0, pos), newElements, src.slice(pos + 1));
}
function isValidEntityCode(c) {
/* eslint no-bitwise:0 */
// broken sequence
if (c >= 55296 && c <= 57343) {
return false;
}
// never used
if (c >= 64976 && c <= 65007) {
return false;
}
if ((c & 65535) === 65535 || (c & 65535) === 65534) {
return false;
}
// control codes
if (c >= 0 && c <= 8) {
return false;
}
if (c === 11) {
return false;
}
if (c >= 14 && c <= 31) {
return false;
}
if (c >= 127 && c <= 159) {
return false;
}
// out of range
if (c > 1114111) {
return false;
}
return true;
}
function fromCodePoint(c) {
/* eslint no-bitwise:0 */
if (c > 65535) {
c -= 65536;
const surrogate1 = 55296 + (c >> 10);
const surrogate2 = 56320 + (c & 1023);
return String.fromCharCode(surrogate1, surrogate2);
}
return String.fromCharCode(c);
}
const UNESCAPE_MD_RE = /\\([!"#$%&'()*+,\-./:;<=>?@[\\\]^_`{|}~])/g;
const ENTITY_RE = /&([a-z#][a-z0-9]{1,31});/gi;
const UNESCAPE_ALL_RE = new RegExp(UNESCAPE_MD_RE.source + "|" + ENTITY_RE.source, "gi");
const DIGITAL_ENTITY_TEST_RE = /^#((?:x[a-f0-9]{1,8}|[0-9]{1,8}))$/i;
function replaceEntityPattern(match, name) {
if (name.charCodeAt(0) === 35 /* # */ && DIGITAL_ENTITY_TEST_RE.test(name)) {
const code = name[1].toLowerCase() === "x" ? parseInt(name.slice(2), 16) : parseInt(name.slice(1), 10);
if (isValidEntityCode(code)) {
return fromCodePoint(code);
}
return match;
}
const decoded = decodeHTML(match);
if (decoded !== match) {
return decoded;
}
return match;
}
/* function replaceEntities(str) {
if (str.indexOf('&') < 0) { return str; }
return str.replace(ENTITY_RE, replaceEntityPattern);
} */ function unescapeMd(str) {
if (str.indexOf("\\") < 0) {
return str;
}
return str.replace(UNESCAPE_MD_RE, "$1");
}
function unescapeAll(str) {
if (str.indexOf("\\") < 0 && str.indexOf("&") < 0) {
return str;
}
return str.replace(UNESCAPE_ALL_RE, (function(match, escaped, entity) {
if (escaped) {
return escaped;
}
return replaceEntityPattern(match, entity);
}));
}
const HTML_ESCAPE_TEST_RE = /[&<>"]/;
const HTML_ESCAPE_REPLACE_RE = /[&<>"]/g;
const HTML_REPLACEMENTS = {
"&": "&amp;",
"<": "&lt;",
">": "&gt;",
'"': "&quot;"
};
function replaceUnsafeChar(ch) {
return HTML_REPLACEMENTS[ch];
}
function escapeHtml(str) {
if (HTML_ESCAPE_TEST_RE.test(str)) {
return str.replace(HTML_ESCAPE_REPLACE_RE, replaceUnsafeChar);
}
return str;
}
const REGEXP_ESCAPE_RE = /[.?*+^$[\]\\(){}|-]/g;
function escapeRE$1(str) {
return str.replace(REGEXP_ESCAPE_RE, "\\$&");
}
function isSpace(code) {
switch (code) {
case 9:
case 32:
return true;
}
return false;
}
// Zs (unicode class) || [\t\f\v\r\n]
function isWhiteSpace(code) {
if (code >= 8192 && code <= 8202) {
return true;
}
switch (code) {
case 9:
// \t
case 10:
// \n
case 11:
// \v
case 12:
// \f
case 13:
// \r
case 32:
case 160:
case 5760:
case 8239:
case 8287:
case 12288:
return true;
}
return false;
}
/* eslint-disable max-len */
// Currently without astral characters support.
function isPunctChar(ch) {
return P.test(ch) || regex.test(ch);
}
// Markdown ASCII punctuation characters.
// !, ", #, $, %, &, ', (, ), *, +, ,, -, ., /, :, ;, <, =, >, ?, @, [, \, ], ^, _, `, {, |, }, or ~
// http://spec.commonmark.org/0.15/#ascii-punctuation-character
// Don't confuse with unicode punctuation !!! It lacks some chars in ascii range.
function isMdAsciiPunct(ch) {
switch (ch) {
case 33 /* ! */ :
case 34 /* " */ :
case 35 /* # */ :
case 36 /* $ */ :
case 37 /* % */ :
case 38 /* & */ :
case 39 /* ' */ :
case 40 /* ( */ :
case 41 /* ) */ :
case 42 /* * */ :
case 43 /* + */ :
case 44 /* , */ :
case 45 /* - */ :
case 46 /* . */ :
case 47 /* / */ :
case 58 /* : */ :
case 59 /* ; */ :
case 60 /* < */ :
case 61 /* = */ :
case 62 /* > */ :
case 63 /* ? */ :
case 64 /* @ */ :
case 91 /* [ */ :
case 92 /* \ */ :
case 93 /* ] */ :
case 94 /* ^ */ :
case 95 /* _ */ :
case 96 /* ` */ :
case 123 /* { */ :
case 124 /* | */ :
case 125 /* } */ :
case 126 /* ~ */ :
return true;
default:
return false;
}
}
// Hepler to unify [reference labels].
function normalizeReference(str) {
// Trim and collapse whitespace
str = str.trim().replace(/\s+/g, " ");
// In node v10 'ẞ'.toLowerCase() === 'Ṿ', which is presumed to be a bug
// fixed in v12 (couldn't find any details).
// So treat this one as a special case
// (remove this when node v10 is no longer supported).
if ("\u1e9e".toLowerCase() === "\u1e7e") {
str = str.replace(/\u1e9e/g, "\xdf");
}
// .toLowerCase().toUpperCase() should get rid of all differences
// between letter variants.
// Simple .toLowerCase() doesn't normalize 125 code points correctly,
// and .toUpperCase doesn't normalize 6 of them (list of exceptions:
// İ, ϴ, ẞ, Ω, K, Å - those are already uppercased, but have differently
// uppercased versions).
// Here's an example showing how it happens. Lets take greek letter omega:
// uppercase U+0398 (Θ), U+03f4 (ϴ) and lowercase U+03b8 (θ), U+03d1 (ϑ)
// Unicode entries:
// 0398;GREEK CAPITAL LETTER THETA;Lu;0;L;;;;;N;;;;03B8;
// 03B8;GREEK SMALL LETTER THETA;Ll;0;L;;;;;N;;;0398;;0398
// 03D1;GREEK THETA SYMBOL;Ll;0;L;<compat> 03B8;;;;N;GREEK SMALL LETTER SCRIPT THETA;;0398;;0398
// 03F4;GREEK CAPITAL THETA SYMBOL;Lu;0;L;<compat> 0398;;;;N;;;;03B8;
// Case-insensitive comparison should treat all of them as equivalent.
// But .toLowerCase() doesn't change ϑ (it's already lowercase),
// and .toUpperCase() doesn't change ϴ (already uppercase).
// Applying first lower then upper case normalizes any character:
// '\u0398\u03f4\u03b8\u03d1'.toLowerCase().toUpperCase() === '\u0398\u0398\u0398\u0398'
// Note: this is equivalent to unicode case folding; unicode normalization
// is a different step that is not required here.
// Final result should be uppercased, because it's later stored in an object
// (this avoid a conflict with Object.prototype members,
// most notably, `__proto__`)
return str.toLowerCase().toUpperCase();
}
// Re-export libraries commonly used in both markdown-it and its plugins,
// so plugins won't have to depend on them explicitly, which reduces their
// bundled size (e.g. a browser build).
const lib = {
mdurl: mdurl,
ucmicro: ucmicro
};
var utils = Object.freeze({
__proto__: null,
arrayReplaceAt: arrayReplaceAt,
assign: assign$1,
escapeHtml: escapeHtml,
escapeRE: escapeRE$1,
fromCodePoint: fromCodePoint,
has: has,
isMdAsciiPunct: isMdAsciiPunct,
isPunctChar: isPunctChar,
isSpace: isSpace,
isString: isString$1,
isValidEntityCode: isValidEntityCode,
isWhiteSpace: isWhiteSpace,
lib: lib,
normalizeReference: normalizeReference,
unescapeAll: unescapeAll,
unescapeMd: unescapeMd
});
// Parse link label
// this function assumes that first character ("[") already matches;
// returns the end of the label
function parseLinkLabel(state, start, disableNested) {
let level, found, marker, prevPos;
const max = state.posMax;
const oldPos = state.pos;
state.pos = start + 1;
level = 1;
while (state.pos < max) {
marker = state.src.charCodeAt(state.pos);
if (marker === 93 /* ] */) {
level--;
if (level === 0) {
found = true;
break;
}
}
prevPos = state.pos;
state.md.inline.skipToken(state);
if (marker === 91 /* [ */) {
if (prevPos === state.pos - 1) {
// increase level if we find text `[`, which is not a part of any token
level++;
} else if (disableNested) {
state.pos = oldPos;
return -1;
}
}
}
let labelEnd = -1;
if (found) {
labelEnd = state.pos;
}
// restore old state
state.pos = oldPos;
return labelEnd;
}
// Parse link destination
function parseLinkDestination(str, start, max) {
let code;
let pos = start;
const result = {
ok: false,
pos: 0,
str: ""
};
if (str.charCodeAt(pos) === 60 /* < */) {
pos++;
while (pos < max) {
code = str.charCodeAt(pos);
if (code === 10 /* \n */) {
return result;
}
if (code === 60 /* < */) {
return result;
}
if (code === 62 /* > */) {
result.pos = pos + 1;
result.str = unescapeAll(str.slice(start + 1, pos));
result.ok = true;
return result;
}
if (code === 92 /* \ */ && pos + 1 < max) {
pos += 2;
continue;
}
pos++;
}
// no closing '>'
return result;
}
// this should be ... } else { ... branch
let level = 0;
while (pos < max) {
code = str.charCodeAt(pos);
if (code === 32) {
break;
}
// ascii control characters
if (code < 32 || code === 127) {
break;
}
if (code === 92 /* \ */ && pos + 1 < max) {
if (str.charCodeAt(pos + 1) === 32) {
break;
}
pos += 2;
continue;
}
if (code === 40 /* ( */) {
level++;
if (level > 32) {
return result;
}
}
if (code === 41 /* ) */) {
if (level === 0) {
break;
}
level--;
}
pos++;
}
if (start === pos) {
return result;
}
if (level !== 0) {
return result;
}
result.str = unescapeAll(str.slice(start, pos));
result.pos = pos;
result.ok = true;
return result;
}
// Parse link title
// Parse link title within `str` in [start, max] range,
// or continue previous parsing if `prev_state` is defined (equal to result of last execution).
function parseLinkTitle(str, start, max, prev_state) {
let code;
let pos = start;
const state = {
// if `true`, this is a valid link title
ok: false,
// if `true`, this link can be continued on the next line
can_continue: false,
// if `ok`, it's the position of the first character after the closing marker
pos: 0,
// if `ok`, it's the unescaped title
str: "",
// expected closing marker character code
marker: 0
};
if (prev_state) {
// this is a continuation of a previous parseLinkTitle call on the next line,
// used in reference links only
state.str = prev_state.str;
state.marker = prev_state.marker;
} else {
if (pos >= max) {
return state;
}
let marker = str.charCodeAt(pos);
if (marker !== 34 /* " */ && marker !== 39 /* ' */ && marker !== 40 /* ( */) {
return state;
}
start++;
pos++;
// if opening marker is "(", switch it to closing marker ")"
if (marker === 40) {
marker = 41;
}
state.marker = marker;
}
while (pos < max) {
code = str.charCodeAt(pos);
if (code === state.marker) {
state.pos = pos + 1;
state.str += unescapeAll(str.slice(start, pos));
state.ok = true;
return state;
} else if (code === 40 /* ( */ && state.marker === 41 /* ) */) {
return state;
} else if (code === 92 /* \ */ && pos + 1 < max) {
pos++;
}
pos++;
}
// no closing marker found, but this link title may continue on the next line (for references)
state.can_continue = true;
state.str += unescapeAll(str.slice(start, pos));
return state;
}
// Just a shortcut for bulk export
var helpers = Object.freeze({
__proto__: null,
parseLinkDestination: parseLinkDestination,
parseLinkLabel: parseLinkLabel,
parseLinkTitle: parseLinkTitle
});
/**
* class Renderer
*
* Generates HTML from parsed token stream. Each instance has independent
* copy of rules. Those can be rewritten with ease. Also, you can add new
* rules if you create plugin and adds new token types.
**/ const default_rules = {};
default_rules.code_inline = function(tokens, idx, options, env, slf) {
const token = tokens[idx];
return "<code" + slf.renderAttrs(token) + ">" + escapeHtml(token.content) + "</code>";
};
default_rules.code_block = function(tokens, idx, options, env, slf) {
const token = tokens[idx];
return "<pre" + slf.renderAttrs(token) + "><code>" + escapeHtml(tokens[idx].content) + "</code></pre>\n";
};
default_rules.fence = function(tokens, idx, options, env, slf) {
const token = tokens[idx];
const info = token.info ? unescapeAll(token.info).trim() : "";
let langName = "";
let langAttrs = "";
if (info) {
const arr = info.split(/(\s+)/g);
langName = arr[0];
langAttrs = arr.slice(2).join("");
}
let highlighted;
if (options.highlight) {
highlighted = options.highlight(token.content, langName, langAttrs) || escapeHtml(token.content);
} else {
highlighted = escapeHtml(token.content);
}
if (highlighted.indexOf("<pre") === 0) {
return highlighted + "\n";
}
// If language exists, inject class gently, without modifying original token.
// May be, one day we will add .deepClone() for token and simplify this part, but
// now we prefer to keep things local.
if (info) {
const i = token.attrIndex("class");
const tmpAttrs = token.attrs ? token.attrs.slice() : [];
if (i < 0) {
tmpAttrs.push([ "class", options.langPrefix + langName ]);
} else {
tmpAttrs[i] = tmpAttrs[i].slice();
tmpAttrs[i][1] += " " + options.langPrefix + langName;
}
// Fake token just to render attributes
const tmpToken = {
attrs: tmpAttrs
};
return `<pre><code${slf.renderAttrs(tmpToken)}>${highlighted}</code></pre>\n`;
}
return `<pre><code${slf.renderAttrs(token)}>${highlighted}</code></pre>\n`;
};
default_rules.image = function(tokens, idx, options, env, slf) {
const token = tokens[idx];
// "alt" attr MUST be set, even if empty. Because it's mandatory and
// should be placed on proper position for tests.
// Replace content with actual value
token.attrs[token.attrIndex("alt")][1] = slf.renderInlineAsText(token.children, options, env);
return slf.renderToken(tokens, idx, options);
};
default_rules.hardbreak = function(tokens, idx, options /*, env */) {
return options.xhtmlOut ? "<br />\n" : "<br>\n";
};
default_rules.softbreak = function(tokens, idx, options /*, env */) {
return options.breaks ? options.xhtmlOut ? "<br />\n" : "<br>\n" : "\n";
};
default_rules.text = function(tokens, idx /*, options, env */) {
return escapeHtml(tokens[idx].content);
};
default_rules.html_block = function(tokens, idx /*, options, env */) {
return tokens[idx].content;
};
default_rules.html_inline = function(tokens, idx /*, options, env */) {
return tokens[idx].content;
};
/**
* new Renderer()
*
* Creates new [[Renderer]] instance and fill [[Renderer#rules]] with defaults.
**/ function Renderer() {
/**
* Renderer#rules -> Object
*
* Contains render rules for tokens. Can be updated and extended.
*
* ##### Example
*
* ```javascript
* var md = require('markdown-it')();
*
* md.renderer.rules.strong_open = function () { return '<b>'; };
* md.renderer.rules.strong_close = function () { return '</b>'; };
*
* var result = md.renderInline(...);
* ```
*
* Each rule is called as independent static function with fixed signature:
*
* ```javascript
* function my_token_render(tokens, idx, options, env, renderer) {
* // ...
* return renderedHTML;
* }
* ```
*
* See [source code](https://github.com/markdown-it/markdown-it/blob/master/lib/renderer.mjs)
* for more details and examples.
**/
this.rules = assign$1({}, default_rules);
}
/**
* Renderer.renderAttrs(token) -> String
*
* Render token attributes to string.
**/ Renderer.prototype.renderAttrs = function renderAttrs(token) {
let i, l, result;
if (!token.attrs) {
return "";
}
result = "";
for (i = 0, l = token.attrs.length; i < l; i++) {
result += " " + escapeHtml(token.attrs[i][0]) + '="' + escapeHtml(token.attrs[i][1]) + '"';
}
return result;
};
/**
* Renderer.renderToken(tokens, idx, options) -> String
* - tokens (Array): list of tokens
* - idx (Numbed): token index to render
* - options (Object): params of parser instance
*
* Default token renderer. Can be overriden by custom function
* in [[Renderer#rules]].
**/ Renderer.prototype.renderToken = function renderToken(tokens, idx, options) {
const token = tokens[idx];
let result = "";
// Tight list paragraphs
if (token.hidden) {
return "";
}
// Insert a newline between hidden paragraph and subsequent opening
// block-level tag.
// For example, here we should insert a newline before blockquote:
// - a
// >
if (token.block && token.nesting !== -1 && idx && tokens[idx - 1].hidden) {
result += "\n";
}
// Add token name, e.g. `<img`
result += (token.nesting === -1 ? "</" : "<") + token.tag;
// Encode attributes, e.g. `<img src="foo"`
result += this.renderAttrs(token);
// Add a slash for self-closing tags, e.g. `<img src="foo" /`
if (token.nesting === 0 && options.xhtmlOut) {
result += " /";
}
// Check if we need to add a newline after this tag
let needLf = false;
if (token.block) {
needLf = true;
if (token.nesting === 1) {
if (idx + 1 < tokens.length) {
const nextToken = tokens[idx + 1];
if (nextToken.type === "inline" || nextToken.hidden) {
// Block-level tag containing an inline tag.
needLf = false;
} else if (nextToken.nesting === -1 && nextToken.tag === token.tag) {
// Opening tag + closing tag of the same type. E.g. `<li></li>`.
needLf = false;
}
}
}
}
result += needLf ? ">\n" : ">";
return result;
};
/**
* Renderer.renderInline(tokens, options, env) -> String
* - tokens (Array): list on block tokens to render
* - options (Object): params of parser instance
* - env (Object): additional data from parsed input (references, for example)
*
* The same as [[Renderer.render]], but for single token of `inline` type.
**/ Renderer.prototype.renderInline = function(tokens, options, env) {
let result = "";
const rules = this.rules;
for (let i = 0, len = tokens.length; i < len; i++) {
const type = tokens[i].type;
if (typeof rules[type] !== "undefined") {
result += rules[type](tokens, i, options, env, this);
} else {
result += this.renderToken(tokens, i, options);
}
}
return result;
};
/** internal
* Renderer.renderInlineAsText(tokens, options, env) -> String
* - tokens (Array): list on block tokens to render
* - options (Object): params of parser instance
* - env (Object): additional data from parsed input (references, for example)
*
* Special kludge for image `alt` attributes to conform CommonMark spec.
* Don't try to use it! Spec requires to show `alt` content with stripped markup,
* instead of simple escaping.
**/ Renderer.prototype.renderInlineAsText = function(tokens, options, env) {
let result = "";
for (let i = 0, len = tokens.length; i < len; i++) {
switch (tokens[i].type) {
case "text":
result += tokens[i].content;
break;
case "image":
result += this.renderInlineAsText(tokens[i].children, options, env);
break;
case "html_inline":
case "html_block":
result += tokens[i].content;
break;
case "softbreak":
case "hardbreak":
result += "\n";
break;
// all other tokens are skipped
}
}
return result;
};
/**
* Renderer.render(tokens, options, env) -> String
* - tokens (Array): list on block tokens to render
* - options (Object): params of parser instance
* - env (Object): additional data from parsed input (references, for example)
*
* Takes token stream and generates HTML. Probably, you will never need to call
* this method directly.
**/ Renderer.prototype.render = function(tokens, options, env) {
let result = "";
const rules = this.rules;
for (let i = 0, len = tokens.length; i < len; i++) {
const type = tokens[i].type;
if (type === "inline") {
result += this.renderInline(tokens[i].children, options, env);
} else if (typeof rules[type] !== "undefined") {
result += rules[type](tokens, i, options, env, this);
} else {
result += this.renderToken(tokens, i, options, env);
}
}
return result;
};
/**
* class Ruler
*
* Helper class, used by [[MarkdownIt#core]], [[MarkdownIt#block]] and
* [[MarkdownIt#inline]] to manage sequences of functions (rules):
*
* - keep rules in defined order
* - assign the name to each rule
* - enable/disable rules
* - add/replace rules
* - allow assign rules to additional named chains (in the same)
* - cacheing lists of active rules
*
* You will not need use this class directly until write plugins. For simple
* rules control use [[MarkdownIt.disable]], [[MarkdownIt.enable]] and
* [[MarkdownIt.use]].
**/
/**
* new Ruler()
**/ function Ruler() {
// List of added rules. Each element is:
// {
// name: XXX,
// enabled: Boolean,
// fn: Function(),
// alt: [ name2, name3 ]
// }
this.__rules__ = [];
// Cached rule chains.
// First level - chain name, '' for default.
// Second level - diginal anchor for fast filtering by charcodes.
this.__cache__ = null;
}
// Helper methods, should not be used directly
// Find rule index by name
Ruler.prototype.__find__ = function(name) {
for (let i = 0; i < this.__rules__.length; i++) {
if (this.__rules__[i].name === name) {
return i;
}
}
return -1;
};
// Build rules lookup cache
Ruler.prototype.__compile__ = function() {
const self = this;
const chains = [ "" ];
// collect unique names
self.__rules__.forEach((function(rule) {
if (!rule.enabled) {
return;
}
rule.alt.forEach((function(altName) {
if (chains.indexOf(altName) < 0) {
chains.push(altName);
}
}));
}));
self.__cache__ = {};
chains.forEach((function(chain) {
self.__cache__[chain] = [];
self.__rules__.forEach((function(rule) {
if (!rule.enabled) {
return;
}
if (chain && rule.alt.indexOf(chain) < 0) {
return;
}
self.__cache__[chain].push(rule.fn);
}));
}));
};
/**
* Ruler.at(name, fn [, options])
* - name (String): rule name to replace.
* - fn (Function): new rule function.
* - options (Object): new rule options (not mandatory).
*
* Replace rule by name with new function & options. Throws error if name not
* found.
*
* ##### Options:
*
* - __alt__ - array with names of "alternate" chains.
*
* ##### Example
*
* Replace existing typographer replacement rule with new one:
*
* ```javascript
* var md = require('markdown-it')();
*
* md.core.ruler.at('replacements', function replace(state) {
* //...
* });
* ```
**/ Ruler.prototype.at = function(name, fn, options) {
const index = this.__find__(name);
const opt = options || {};
if (index === -1) {
throw new Error("Parser rule not found: " + name);
}
this.__rules__[index].fn = fn;
this.__rules__[index].alt = opt.alt || [];
this.__cache__ = null;
};
/**
* Ruler.before(beforeName, ruleName, fn [, options])
* - beforeName (String): new rule will be added before this one.
* - ruleName (String): name of added rule.
* - fn (Function): rule function.
* - options (Object): rule options (not mandatory).
*
* Add new rule to chain before one with given name. See also
* [[Ruler.after]], [[Ruler.push]].
*
* ##### Options:
*
* - __alt__ - array with names of "alternate" chains.
*
* ##### Example
*
* ```javascript
* var md = require('markdown-it')();
*
* md.block.ruler.before('paragraph', 'my_rule', function replace(state) {
* //...
* });
* ```
**/ Ruler.prototype.before = function(beforeName, ruleName, fn, options) {
const index = this.__find__(beforeName);
const opt = options || {};
if (index === -1) {
throw new Error("Parser rule not found: " + beforeName);
}
this.__rules__.splice(index, 0, {
name: ruleName,
enabled: true,
fn: fn,
alt: opt.alt || []
});
this.__cache__ = null;
};
/**
* Ruler.after(afterName, ruleName, fn [, options])
* - afterName (String): new rule will be added after this one.
* - ruleName (String): name of added rule.
* - fn (Function): rule function.
* - options (Object): rule options (not mandatory).
*
* Add new rule to chain after one with given name. See also
* [[Ruler.before]], [[Ruler.push]].
*
* ##### Options:
*
* - __alt__ - array with names of "alternate" chains.
*
* ##### Example
*
* ```javascript
* var md = require('markdown-it')();
*
* md.inline.ruler.after('text', 'my_rule', function replace(state) {
* //...
* });
* ```
**/ Ruler.prototype.after = function(afterName, ruleName, fn, options) {
const index = this.__find__(afterName);
const opt = options || {};
if (index === -1) {
throw new Error("Parser rule not found: " + afterName);
}
this.__rules__.splice(index + 1, 0, {
name: ruleName,
enabled: true,
fn: fn,
alt: opt.alt || []
});
this.__cache__ = null;
};
/**
* Ruler.push(ruleName, fn [, options])
* - ruleName (String): name of added rule.
* - fn (Function): rule function.
* - options (Object): rule options (not mandatory).
*
* Push new rule to the end of chain. See also
* [[Ruler.before]], [[Ruler.after]].
*
* ##### Options:
*
* - __alt__ - array with names of "alternate" chains.
*
* ##### Example
*
* ```javascript
* var md = require('markdown-it')();
*
* md.core.ruler.push('my_rule', function replace(state) {
* //...
* });
* ```
**/ Ruler.prototype.push = function(ruleName, fn, options) {
const opt = options || {};
this.__rules__.push({
name: ruleName,
enabled: true,
fn: fn,
alt: opt.alt || []
});
this.__cache__ = null;
};
/**
* Ruler.enable(list [, ignoreInvalid]) -> Array
* - list (String|Array): list of rule names to enable.
* - ignoreInvalid (Boolean): set `true` to ignore errors when rule not found.
*
* Enable rules with given names. If any rule name not found - throw Error.
* Errors can be disabled by second param.
*
* Returns list of found rule names (if no exception happened).
*
* See also [[Ruler.disable]], [[Ruler.enableOnly]].
**/ Ruler.prototype.enable = function(list, ignoreInvalid) {
if (!Array.isArray(list)) {
list = [ list ];
}
const result = [];
// Search by name and enable
list.forEach((function(name) {
const idx = this.__find__(name);
if (idx < 0) {
if (ignoreInvalid) {
return;
}
throw new Error("Rules manager: invalid rule name " + name);
}
this.__rules__[idx].enabled = true;
result.push(name);
}), this);
this.__cache__ = null;
return result;
};
/**
* Ruler.enableOnly(list [, ignoreInvalid])
* - list (String|Array): list of rule names to enable (whitelist).
* - ignoreInvalid (Boolean): set `true` to ignore errors when rule not found.
*
* Enable rules with given names, and disable everything else. If any rule name
* not found - throw Error. Errors can be disabled by second param.
*
* See also [[Ruler.disable]], [[Ruler.enable]].
**/ Ruler.prototype.enableOnly = function(list, ignoreInvalid) {
if (!Array.isArray(list)) {
list = [ list ];
}
this.__rules__.forEach((function(rule) {
rule.enabled = false;
}));
this.enable(list, ignoreInvalid);
};
/**
* Ruler.disable(list [, ignoreInvalid]) -> Array
* - list (String|Array): list of rule names to disable.
* - ignoreInvalid (Boolean): set `true` to ignore errors when rule not found.
*
* Disable rules with given names. If any rule name not found - throw Error.
* Errors can be disabled by second param.
*
* Returns list of found rule names (if no exception happened).
*
* See also [[Ruler.enable]], [[Ruler.enableOnly]].
**/ Ruler.prototype.disable = function(list, ignoreInvalid) {
if (!Array.isArray(list)) {
list = [ list ];
}
const result = [];
// Search by name and disable
list.forEach((function(name) {
const idx = this.__find__(name);
if (idx < 0) {
if (ignoreInvalid) {
return;
}
throw new Error("Rules manager: invalid rule name " + name);
}
this.__rules__[idx].enabled = false;
result.push(name);
}), this);
this.__cache__ = null;
return result;
};
/**
* Ruler.getRules(chainName) -> Array
*
* Return array of active functions (rules) for given chain name. It analyzes
* rules configuration, compiles caches if not exists and returns result.
*
* Default chain name is `''` (empty string). It can't be skipped. That's
* done intentionally, to keep signature monomorphic for high speed.
**/ Ruler.prototype.getRules = function(chainName) {
if (this.__cache__ === null) {
this.__compile__();
}
// Chain can be empty, if rules disabled. But we still have to return Array.
return this.__cache__[chainName] || [];
};
// Token class
/**
* class Token
**/
/**
* new Token(type, tag, nesting)
*
* Create new token and fill passed properties.
**/ function Token(type, tag, nesting) {
/**
* Token#type -> String
*
* Type of the token (string, e.g. "paragraph_open")
**/
this.type = type;
/**
* Token#tag -> String
*
* html tag name, e.g. "p"
**/ this.tag = tag;
/**
* Token#attrs -> Array
*
* Html attributes. Format: `[ [ name1, value1 ], [ name2, value2 ] ]`
**/ this.attrs = null;
/**
* Token#map -> Array
*
* Source map info. Format: `[ line_begin, line_end ]`
**/ this.map = null;
/**
* Token#nesting -> Number
*
* Level change (number in {-1, 0, 1} set), where:
*
* - `1` means the tag is opening
* - `0` means the tag is self-closing
* - `-1` means the tag is closing
**/ this.nesting = nesting;
/**
* Token#level -> Number
*
* nesting level, the same as `state.level`
**/ this.level = 0;
/**
* Token#children -> Array
*
* An array of child nodes (inline and img tokens)
**/ this.children = null;
/**
* Token#content -> String
*
* In a case of self-closing tag (code, html, fence, etc.),
* it has contents of this tag.
**/ this.content = "";
/**
* Token#markup -> String
*
* '*' or '_' for emphasis, fence string for fence, etc.
**/ this.markup = "";
/**
* Token#info -> String
*
* Additional information:
*
* - Info string for "fence" tokens
* - The value "auto" for autolink "link_open" and "link_close" tokens
* - The string value of the item marker for ordered-list "list_item_open" tokens
**/ this.info = "";
/**
* Token#meta -> Object
*
* A place for plugins to store an arbitrary data
**/ this.meta = null;
/**
* Token#block -> Boolean
*
* True for block-level tokens, false for inline tokens.
* Used in renderer to calculate line breaks
**/ this.block = false;
/**
* Token#hidden -> Boolean
*
* If it's true, ignore this element when rendering. Used for tight lists
* to hide paragraphs.
**/ this.hidden = false;
}
/**
* Token.attrIndex(name) -> Number
*
* Search attribute index by name.
**/ Token.prototype.attrIndex = function attrIndex(name) {
if (!this.attrs) {
return -1;
}
const attrs = this.attrs;
for (let i = 0, len = attrs.length; i < len; i++) {
if (attrs[i][0] === name) {
return i;
}
}
return -1;
};
/**
* Token.attrPush(attrData)
*
* Add `[ name, value ]` attribute to list. Init attrs if necessary
**/ Token.prototype.attrPush = function attrPush(attrData) {
if (this.attrs) {
this.attrs.push(attrData);
} else {
this.attrs = [ attrData ];
}
};
/**
* Token.attrSet(name, value)
*
* Set `name` attribute to `value`. Override old value if exists.
**/ Token.prototype.attrSet = function attrSet(name, value) {
const idx = this.attrIndex(name);
const attrData = [ name, value ];
if (idx < 0) {
this.attrPush(attrData);
} else {
this.attrs[idx] = attrData;
}
};
/**
* Token.attrGet(name)
*
* Get the value of attribute `name`, or null if it does not exist.
**/ Token.prototype.attrGet = function attrGet(name) {
const idx = this.attrIndex(name);
let value = null;
if (idx >= 0) {
value = this.attrs[idx][1];
}
return value;
};
/**
* Token.attrJoin(name, value)
*
* Join value to existing attribute via space. Or create new attribute if not
* exists. Useful to operate with token classes.
**/ Token.prototype.attrJoin = function attrJoin(name, value) {
const idx = this.attrIndex(name);
if (idx < 0) {
this.attrPush([ name, value ]);
} else {
this.attrs[idx][1] = this.attrs[idx][1] + " " + value;
}
};
// Core state object
function StateCore(src, md, env) {
this.src = src;
this.env = env;
this.tokens = [];
this.inlineMode = false;
this.md = md;
// link to parser instance
}
// re-export Token class to use in core rules
StateCore.prototype.Token = Token;
// Normalize input string
// https://spec.commonmark.org/0.29/#line-ending
const NEWLINES_RE = /\r\n?|\n/g;
const NULL_RE = /\0/g;
function normalize(state) {
let str;
// Normalize newlines
str = state.src.replace(NEWLINES_RE, "\n");
// Replace NULL characters
str = str.replace(NULL_RE, "\ufffd");
state.src = str;
}
function block(state) {
let token;
if (state.inlineMode) {
token = new state.Token("inline", "", 0);
token.content = state.src;
token.map = [ 0, 1 ];
token.children = [];
state.tokens.push(token);
} else {
state.md.block.parse(state.src, state.md, state.env, state.tokens);
}
}
function inline(state) {
const tokens = state.tokens;
// Parse inlines
for (let i = 0, l = tokens.length; i < l; i++) {
const tok = tokens[i];
if (tok.type === "inline") {
state.md.inline.parse(tok.content, state.md, state.env, tok.children);
}
}
}
// Replace link-like texts with link nodes.
// Currently restricted by `md.validateLink()` to http/https/ftp
function isLinkOpen$1(str) {
return /^<a[>\s]/i.test(str);
}
function isLinkClose$1(str) {
return /^<\/a\s*>/i.test(str);
}
function linkify$1(state) {
const blockTokens = state.tokens;
if (!state.md.options.linkify) {
return;
}
for (let j = 0, l = blockTokens.length; j < l; j++) {
if (blockTokens[j].type !== "inline" || !state.md.linkify.pretest(blockTokens[j].content)) {
continue;
}
let tokens = blockTokens[j].children;
let htmlLinkLevel = 0;
// We scan from the end, to keep position when new tags added.
// Use reversed logic in links start/end match
for (let i = tokens.length - 1; i >= 0; i--) {
const currentToken = tokens[i];
// Skip content of markdown links
if (currentToken.type === "link_close") {
i--;
while (tokens[i].level !== currentToken.level && tokens[i].type !== "link_open") {
i--;
}
continue;
}
// Skip content of html tag links
if (currentToken.type === "html_inline") {
if (isLinkOpen$1(currentToken.content) && htmlLinkLevel > 0) {
htmlLinkLevel--;
}
if (isLinkClose$1(currentToken.content)) {
htmlLinkLevel++;
}
}
if (htmlLinkLevel > 0) {
continue;
}
if (currentToken.type === "text" && state.md.linkify.test(currentToken.content)) {
const text = currentToken.content;
let links = state.md.linkify.match(text);
// Now split string to nodes
const nodes = [];
let level = currentToken.level;
let lastPos = 0;
// forbid escape sequence at the start of the string,
// this avoids http\://example.com/ from being linkified as
// http:<a href="//example.com/">//example.com/</a>
if (links.length > 0 && links[0].index === 0 && i > 0 && tokens[i - 1].type === "text_special") {
links = links.slice(1);
}
for (let ln = 0; ln < links.length; ln++) {
const url = links[ln].url;
const fullUrl = state.md.normalizeLink(url);
if (!state.md.validateLink(fullUrl)) {
continue;
}
let urlText = links[ln].text;
// Linkifier might send raw hostnames like "example.com", where url
// starts with domain name. So we prepend http:// in those cases,
// and remove it afterwards.
if (!links[ln].schema) {
urlText = state.md.normalizeLinkText("http://" + urlText).replace(/^http:\/\//, "");
} else if (links[ln].schema === "mailto:" && !/^mailto:/i.test(urlText)) {
urlText = state.md.normalizeLinkText("mailto:" + urlText).replace(/^mailto:/, "");
} else {
urlText = state.md.normalizeLinkText(urlText);
}
const pos = links[ln].index;
if (pos > lastPos) {
const token = new state.Token("text", "", 0);
token.content = text.slice(lastPos, pos);
token.level = level;
nodes.push(token);
}
const token_o = new state.Token("link_open", "a", 1);
token_o.attrs = [ [ "href", fullUrl ] ];
token_o.level = level++;
token_o.markup = "linkify";
token_o.info = "auto";
nodes.push(token_o);
const token_t = new state.Token("text", "", 0);
token_t.content = urlText;
token_t.level = level;
nodes.push(token_t);
const token_c = new state.Token("link_close", "a", -1);
token_c.level = --level;
token_c.markup = "linkify";
token_c.info = "auto";
nodes.push(token_c);
lastPos = links[ln].lastIndex;
}
if (lastPos < text.length) {
const token = new state.Token("text", "", 0);
token.content = text.slice(lastPos);
token.level = level;
nodes.push(token);
}
// replace current node
blockTokens[j].children = tokens = arrayReplaceAt(tokens, i, nodes);
}
}
}
}
// Simple typographic replacements
// (c) (C) → ©
// (tm) (TM) → ™
// (r) (R) → ®
// +- → ±
// ... → … (also ?.... → ?.., !.... → !..)
// ???????? → ???, !!!!! → !!!, `,,` → `,`
// -- → &ndash;, --- → &mdash;
// TODO:
// - fractionals 1/2, 1/4, 3/4 -> ½, ¼, ¾
// - multiplications 2 x 4 -> 2 × 4
const RARE_RE = /\+-|\.\.|\?\?\?\?|!!!!|,,|--/;
// Workaround for phantomjs - need regex without /g flag,
// or root check will fail every second time
const SCOPED_ABBR_TEST_RE = /\((c|tm|r)\)/i;
const SCOPED_ABBR_RE = /\((c|tm|r)\)/gi;
const SCOPED_ABBR = {
c: "\xa9",
r: "\xae",
tm: "\u2122"
};
function replaceFn(match, name) {
return SCOPED_ABBR[name.toLowerCase()];
}
function replace_scoped(inlineTokens) {
let inside_autolink = 0;
for (let i = inlineTokens.length - 1; i >= 0; i--) {
const token = inlineTokens[i];
if (token.type === "text" && !inside_autolink) {
token.content = token.content.replace(SCOPED_ABBR_RE, replaceFn);
}
if (token.type === "link_open" && token.info === "auto") {
inside_autolink--;
}
if (token.type === "link_close" && token.info === "auto") {
inside_autolink++;
}
}
}
function replace_rare(inlineTokens) {
let inside_autolink = 0;
for (let i = inlineTokens.length - 1; i >= 0; i--) {
const token = inlineTokens[i];
if (token.type === "text" && !inside_autolink) {
if (RARE_RE.test(token.content)) {
token.content = token.content.replace(/\+-/g, "\xb1").replace(/\.{2,}/g, "\u2026").replace(/([?!])\u2026/g, "$1..").replace(/([?!]){4,}/g, "$1$1$1").replace(/,{2,}/g, ",").replace(/(^|[^-])---(?=[^-]|$)/gm, "$1\u2014").replace(/(^|\s)--(?=\s|$)/gm, "$1\u2013").replace(/(^|[^-\s])--(?=[^-\s]|$)/gm, "$1\u2013");
}
}
if (token.type === "link_open" && token.info === "auto") {
inside_autolink--;
}
if (token.type === "link_close" && token.info === "auto") {
inside_autolink++;
}
}
}
function replace(state) {
let blkIdx;
if (!state.md.options.typographer) {
return;
}
for (blkIdx = state.tokens.length - 1; blkIdx >= 0; blkIdx--) {
if (state.tokens[blkIdx].type !== "inline") {
continue;
}
if (SCOPED_ABBR_TEST_RE.test(state.tokens[blkIdx].content)) {
replace_scoped(state.tokens[blkIdx].children);
}
if (RARE_RE.test(state.tokens[blkIdx].content)) {
replace_rare(state.tokens[blkIdx].children);
}
}
}
// Convert straight quotation marks to typographic ones
const QUOTE_TEST_RE = /['"]/;
const QUOTE_RE = /['"]/g;
const APOSTROPHE = "\u2019";
/* ’ */ function replaceAt(str, index, ch) {
return str.slice(0, index) + ch + str.slice(index + 1);
}
function process_inlines(tokens, state) {
let j;
const stack = [];
for (let i = 0; i < tokens.length; i++) {
const token = tokens[i];
const thisLevel = tokens[i].level;
for (j = stack.length - 1; j >= 0; j--) {
if (stack[j].level <= thisLevel) {
break;
}
}
stack.length = j + 1;
if (token.type !== "text") {
continue;
}
let text = token.content;
let pos = 0;
let max = text.length;
/* eslint no-labels:0,block-scoped-var:0 */ OUTER: while (pos < max) {
QUOTE_RE.lastIndex = pos;
const t = QUOTE_RE.exec(text);
if (!t) {
break;
}
let canOpen = true;
let canClose = true;
pos = t.index + 1;
const isSingle = t[0] === "'";
// Find previous character,
// default to space if it's the beginning of the line
let lastChar = 32;
if (t.index - 1 >= 0) {
lastChar = text.charCodeAt(t.index - 1);
} else {
for (j = i - 1; j >= 0; j--) {
if (tokens[j].type === "softbreak" || tokens[j].type === "hardbreak") break;
// lastChar defaults to 0x20
if (!tokens[j].content) continue;
// should skip all tokens except 'text', 'html_inline' or 'code_inline'
lastChar = tokens[j].content.charCodeAt(tokens[j].content.length - 1);
break;
}
}
// Find next character,
// default to space if it's the end of the line
let nextChar = 32;
if (pos < max) {
nextChar = text.charCodeAt(pos);
} else {
for (j = i + 1; j < tokens.length; j++) {
if (tokens[j].type === "softbreak" || tokens[j].type === "hardbreak") break;
// nextChar defaults to 0x20
if (!tokens[j].content) continue;
// should skip all tokens except 'text', 'html_inline' or 'code_inline'
nextChar = tokens[j].content.charCodeAt(0);
break;
}
}
const isLastPunctChar = isMdAsciiPunct(lastChar) || isPunctChar(String.fromCharCode(lastChar));
const isNextPunctChar = isMdAsciiPunct(nextChar) || isPunctChar(String.fromCharCode(nextChar));
const isLastWhiteSpace = isWhiteSpace(lastChar);
const isNextWhiteSpace = isWhiteSpace(nextChar);
if (isNextWhiteSpace) {
canOpen = false;
} else if (isNextPunctChar) {
if (!(isLastWhiteSpace || isLastPunctChar)) {
canOpen = false;
}
}
if (isLastWhiteSpace) {
canClose = false;
} else if (isLastPunctChar) {
if (!(isNextWhiteSpace || isNextPunctChar)) {
canClose = false;
}
}
if (nextChar === 34 /* " */ && t[0] === '"') {
if (lastChar >= 48 /* 0 */ && lastChar <= 57 /* 9 */) {
// special case: 1"" - count first quote as an inch
canClose = canOpen = false;
}
}
if (canOpen && canClose) {
// Replace quotes in the middle of punctuation sequence, but not
// in the middle of the words, i.e.:
// 1. foo " bar " baz - not replaced
// 2. foo-"-bar-"-baz - replaced
// 3. foo"bar"baz - not replaced
canOpen = isLastPunctChar;
canClose = isNextPunctChar;
}
if (!canOpen && !canClose) {
// middle of word
if (isSingle) {
token.content = replaceAt(token.content, t.index, APOSTROPHE);
}
continue;
}
if (canClose) {
// this could be a closing quote, rewind the stack to get a match
for (j = stack.length - 1; j >= 0; j--) {
let item = stack[j];
if (stack[j].level < thisLevel) {
break;
}
if (item.single === isSingle && stack[j].level === thisLevel) {
item = stack[j];
let openQuote;
let closeQuote;
if (isSingle) {
openQuote = state.md.options.quotes[2];
closeQuote = state.md.options.quotes[3];
} else {
openQuote = state.md.options.quotes[0];
closeQuote = state.md.options.quotes[1];
}
// replace token.content *before* tokens[item.token].content,
// because, if they are pointing at the same token, replaceAt
// could mess up indices when quote length != 1
token.content = replaceAt(token.content, t.index, closeQuote);
tokens[item.token].content = replaceAt(tokens[item.token].content, item.pos, openQuote);
pos += closeQuote.length - 1;
if (item.token === i) {
pos += openQuote.length - 1;
}
text = token.content;
max = text.length;
stack.length = j;
continue OUTER;
}
}
}
if (canOpen) {
stack.push({
token: i,
pos: t.index,
single: isSingle,
level: thisLevel
});
} else if (canClose && isSingle) {
token.content = replaceAt(token.content, t.index, APOSTROPHE);
}
}
}
}
function smartquotes(state) {
/* eslint max-depth:0 */
if (!state.md.options.typographer) {
return;
}
for (let blkIdx = state.tokens.length - 1; blkIdx >= 0; blkIdx--) {
if (state.tokens[blkIdx].type !== "inline" || !QUOTE_TEST_RE.test(state.tokens[blkIdx].content)) {
continue;
}
process_inlines(state.tokens[blkIdx].children, state);
}
}
// Join raw text tokens with the rest of the text
// This is set as a separate rule to provide an opportunity for plugins
// to run text replacements after text join, but before escape join.
// For example, `\:)` shouldn't be replaced with an emoji.
function text_join(state) {
let curr, last;
const blockTokens = state.tokens;
const l = blockTokens.length;
for (let j = 0; j < l; j++) {
if (blockTokens[j].type !== "inline") continue;
const tokens = blockTokens[j].children;
const max = tokens.length;
for (curr = 0; curr < max; curr++) {
if (tokens[curr].type === "text_special") {
tokens[curr].type = "text";
}
}
for (curr = last = 0; curr < max; curr++) {
if (tokens[curr].type === "text" && curr + 1 < max && tokens[curr + 1].type === "text") {
// collapse two adjacent text nodes
tokens[curr + 1].content = tokens[curr].content + tokens[curr + 1].content;
} else {
if (curr !== last) {
tokens[last] = tokens[curr];
}
last++;
}
}
if (curr !== last) {
tokens.length = last;
}
}
}
/** internal
* class Core
*
* Top-level rules executor. Glues block/inline parsers and does intermediate
* transformations.
**/ const _rules$2 = [ [ "normalize", normalize ], [ "block", block ], [ "inline", inline ], [ "linkify", linkify$1 ], [ "replacements", replace ], [ "smartquotes", smartquotes ],
// `text_join` finds `text_special` tokens (for escape sequences)
// and joins them with the rest of the text
[ "text_join", text_join ] ];
/**
* new Core()
**/ function Core() {
/**
* Core#ruler -> Ruler
*
* [[Ruler]] instance. Keep configuration of core rules.
**/
this.ruler = new Ruler;
for (let i = 0; i < _rules$2.length; i++) {
this.ruler.push(_rules$2[i][0], _rules$2[i][1]);
}
}
/**
* Core.process(state)
*
* Executes core chain rules.
**/ Core.prototype.process = function(state) {
const rules = this.ruler.getRules("");
for (let i = 0, l = rules.length; i < l; i++) {
rules[i](state);
}
};
Core.prototype.State = StateCore;
// Parser state class
function StateBlock(src, md, env, tokens) {
this.src = src;
// link to parser instance
this.md = md;
this.env = env;
// Internal state vartiables
this.tokens = tokens;
this.bMarks = [];
// line begin offsets for fast jumps
this.eMarks = [];
// line end offsets for fast jumps
this.tShift = [];
// offsets of the first non-space characters (tabs not expanded)
this.sCount = [];
// indents for each line (tabs expanded)
// An amount of virtual spaces (tabs expanded) between beginning
// of each line (bMarks) and real beginning of that line.
// It exists only as a hack because blockquotes override bMarks
// losing information in the process.
// It's used only when expanding tabs, you can think about it as
// an initial tab length, e.g. bsCount=21 applied to string `\t123`
// means first tab should be expanded to 4-21%4 === 3 spaces.
this.bsCount = [];
// block parser variables
// required block content indent (for example, if we are
// inside a list, it would be positioned after list marker)
this.blkIndent = 0;
this.line = 0;
// line index in src
this.lineMax = 0;
// lines count
this.tight = false;
// loose/tight mode for lists
this.ddIndent = -1;
// indent of the current dd block (-1 if there isn't any)
this.listIndent = -1;
// indent of the current list block (-1 if there isn't any)
// can be 'blockquote', 'list', 'root', 'paragraph' or 'reference'
// used in lists to determine if they interrupt a paragraph
this.parentType = "root";
this.level = 0;
// Create caches
// Generate markers.
const s = this.src;
for (let start = 0, pos = 0, indent = 0, offset = 0, len = s.length, indent_found = false; pos < len; pos++) {
const ch = s.charCodeAt(pos);
if (!indent_found) {
if (isSpace(ch)) {
indent++;
if (ch === 9) {
offset += 4 - offset % 4;
} else {
offset++;
}
continue;
} else {
indent_found = true;
}
}
if (ch === 10 || pos === len - 1) {
if (ch !== 10) {
pos++;
}
this.bMarks.push(start);
this.eMarks.push(pos);
this.tShift.push(indent);
this.sCount.push(offset);
this.bsCount.push(0);
indent_found = false;
indent = 0;
offset = 0;
start = pos + 1;
}
}
// Push fake entry to simplify cache bounds checks
this.bMarks.push(s.length);
this.eMarks.push(s.length);
this.tShift.push(0);
this.sCount.push(0);
this.bsCount.push(0);
this.lineMax = this.bMarks.length - 1;
// don't count last fake line
}
// Push new token to "stream".
StateBlock.prototype.push = function(type, tag, nesting) {
const token = new Token(type, tag, nesting);
token.block = true;
if (nesting < 0) this.level--;
// closing tag
token.level = this.level;
if (nesting > 0) this.level++;
// opening tag
this.tokens.push(token);
return token;
};
StateBlock.prototype.isEmpty = function isEmpty(line) {
return this.bMarks[line] + this.tShift[line] >= this.eMarks[line];
};
StateBlock.prototype.skipEmptyLines = function skipEmptyLines(from) {
for (let max = this.lineMax; from < max; from++) {
if (this.bMarks[from] + this.tShift[from] < this.eMarks[from]) {
break;
}
}
return from;
};
// Skip spaces from given position.
StateBlock.prototype.skipSpaces = function skipSpaces(pos) {
for (let max = this.src.length; pos < max; pos++) {
const ch = this.src.charCodeAt(pos);
if (!isSpace(ch)) {
break;
}
}
return pos;
};
// Skip spaces from given position in reverse.
StateBlock.prototype.skipSpacesBack = function skipSpacesBack(pos, min) {
if (pos <= min) {
return pos;
}
while (pos > min) {
if (!isSpace(this.src.charCodeAt(--pos))) {
return pos + 1;
}
}
return pos;
};
// Skip char codes from given position
StateBlock.prototype.skipChars = function skipChars(pos, code) {
for (let max = this.src.length; pos < max; pos++) {
if (this.src.charCodeAt(pos) !== code) {
break;
}
}
return pos;
};
// Skip char codes reverse from given position - 1
StateBlock.prototype.skipCharsBack = function skipCharsBack(pos, code, min) {
if (pos <= min) {
return pos;
}
while (pos > min) {
if (code !== this.src.charCodeAt(--pos)) {
return pos + 1;
}
}
return pos;
};
// cut lines range from source.
StateBlock.prototype.getLines = function getLines(begin, end, indent, keepLastLF) {
if (begin >= end) {
return "";
}
const queue = new Array(end - begin);
for (let i = 0, line = begin; line < end; line++, i++) {
let lineIndent = 0;
const lineStart = this.bMarks[line];
let first = lineStart;
let last;
if (line + 1 < end || keepLastLF) {
// No need for bounds check because we have fake entry on tail.
last = this.eMarks[line] + 1;
} else {
last = this.eMarks[line];
}
while (first < last && lineIndent < indent) {
const ch = this.src.charCodeAt(first);
if (isSpace(ch)) {
if (ch === 9) {
lineIndent += 4 - (lineIndent + this.bsCount[line]) % 4;
} else {
lineIndent++;
}
} else if (first - lineStart < this.tShift[line]) {
// patched tShift masked characters to look like spaces (blockquotes, list markers)
lineIndent++;
} else {
break;
}
first++;
}
if (lineIndent > indent) {
// partially expanding tabs in code blocks, e.g '\t\tfoobar'
// with indent=2 becomes ' \tfoobar'
queue[i] = new Array(lineIndent - indent + 1).join(" ") + this.src.slice(first, last);
} else {
queue[i] = this.src.slice(first, last);
}
}
return queue.join("");
};
// re-export Token class to use in block rules
StateBlock.prototype.Token = Token;
// GFM table, https://github.github.com/gfm/#tables-extension-
// Limit the amount of empty autocompleted cells in a table,
// see https://github.com/markdown-it/markdown-it/issues/1000,
// Both pulldown-cmark and commonmark-hs limit the number of cells this way to ~200k.
// We set it to 65k, which can expand user input by a factor of x370
// (256x256 square is 1.8kB expanded into 650kB).
const MAX_AUTOCOMPLETED_CELLS = 65536;
function getLine(state, line) {
const pos = state.bMarks[line] + state.tShift[line];
const max = state.eMarks[line];
return state.src.slice(pos, max);
}
function escapedSplit(str) {
const result = [];
const max = str.length;
let pos = 0;
let ch = str.charCodeAt(pos);
let isEscaped = false;
let lastPos = 0;
let current = "";
while (pos < max) {
if (ch === 124 /* | */) {
if (!isEscaped) {
// pipe separating cells, '|'
result.push(current + str.substring(lastPos, pos));
current = "";
lastPos = pos + 1;
} else {
// escaped pipe, '\|'
current += str.substring(lastPos, pos - 1);
lastPos = pos;
}
}
isEscaped = ch === 92 /* \ */;
pos++;
ch = str.charCodeAt(pos);
}
result.push(current + str.substring(lastPos));
return result;
}
function table(state, startLine, endLine, silent) {
// should have at least two lines
if (startLine + 2 > endLine) {
return false;
}
let nextLine = startLine + 1;
if (state.sCount[nextLine] < state.blkIndent) {
return false;
}
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[nextLine] - state.blkIndent >= 4) {
return false;
}
// first character of the second line should be '|', '-', ':',
// and no other characters are allowed but spaces;
// basically, this is the equivalent of /^[-:|][-:|\s]*$/ regexp
let pos = state.bMarks[nextLine] + state.tShift[nextLine];
if (pos >= state.eMarks[nextLine]) {
return false;
}
const firstCh = state.src.charCodeAt(pos++);
if (firstCh !== 124 /* | */ && firstCh !== 45 /* - */ && firstCh !== 58 /* : */) {
return false;
}
if (pos >= state.eMarks[nextLine]) {
return false;
}
const secondCh = state.src.charCodeAt(pos++);
if (secondCh !== 124 /* | */ && secondCh !== 45 /* - */ && secondCh !== 58 /* : */ && !isSpace(secondCh)) {
return false;
}
// if first character is '-', then second character must not be a space
// (due to parsing ambiguity with list)
if (firstCh === 45 /* - */ && isSpace(secondCh)) {
return false;
}
while (pos < state.eMarks[nextLine]) {
const ch = state.src.charCodeAt(pos);
if (ch !== 124 /* | */ && ch !== 45 /* - */ && ch !== 58 /* : */ && !isSpace(ch)) {
return false;
}
pos++;
}
let lineText = getLine(state, startLine + 1);
let columns = lineText.split("|");
const aligns = [];
for (let i = 0; i < columns.length; i++) {
const t = columns[i].trim();
if (!t) {
// allow empty columns before and after table, but not in between columns;
// e.g. allow ` |---| `, disallow ` ---||--- `
if (i === 0 || i === columns.length - 1) {
continue;
} else {
return false;
}
}
if (!/^:?-+:?$/.test(t)) {
return false;
}
if (t.charCodeAt(t.length - 1) === 58 /* : */) {
aligns.push(t.charCodeAt(0) === 58 /* : */ ? "center" : "right");
} else if (t.charCodeAt(0) === 58 /* : */) {
aligns.push("left");
} else {
aligns.push("");
}
}
lineText = getLine(state, startLine).trim();
if (lineText.indexOf("|") === -1) {
return false;
}
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
columns = escapedSplit(lineText);
if (columns.length && columns[0] === "") columns.shift();
if (columns.length && columns[columns.length - 1] === "") columns.pop();
// header row will define an amount of columns in the entire table,
// and align row should be exactly the same (the rest of the rows can differ)
const columnCount = columns.length;
if (columnCount === 0 || columnCount !== aligns.length) {
return false;
}
if (silent) {
return true;
}
const oldParentType = state.parentType;
state.parentType = "table";
// use 'blockquote' lists for termination because it's
// the most similar to tables
const terminatorRules = state.md.block.ruler.getRules("blockquote");
const token_to = state.push("table_open", "table", 1);
const tableLines = [ startLine, 0 ];
token_to.map = tableLines;
const token_tho = state.push("thead_open", "thead", 1);
token_tho.map = [ startLine, startLine + 1 ];
const token_htro = state.push("tr_open", "tr", 1);
token_htro.map = [ startLine, startLine + 1 ];
for (let i = 0; i < columns.length; i++) {
const token_ho = state.push("th_open", "th", 1);
if (aligns[i]) {
token_ho.attrs = [ [ "style", "text-align:" + aligns[i] ] ];
}
const token_il = state.push("inline", "", 0);
token_il.content = columns[i].trim();
token_il.children = [];
state.push("th_close", "th", -1);
}
state.push("tr_close", "tr", -1);
state.push("thead_close", "thead", -1);
let tbodyLines;
let autocompletedCells = 0;
for (nextLine = startLine + 2; nextLine < endLine; nextLine++) {
if (state.sCount[nextLine] < state.blkIndent) {
break;
}
let terminate = false;
for (let i = 0, l = terminatorRules.length; i < l; i++) {
if (terminatorRules[i](state, nextLine, endLine, true)) {
terminate = true;
break;
}
}
if (terminate) {
break;
}
lineText = getLine(state, nextLine).trim();
if (!lineText) {
break;
}
if (state.sCount[nextLine] - state.blkIndent >= 4) {
break;
}
columns = escapedSplit(lineText);
if (columns.length && columns[0] === "") columns.shift();
if (columns.length && columns[columns.length - 1] === "") columns.pop();
// note: autocomplete count can be negative if user specifies more columns than header,
// but that does not affect intended use (which is limiting expansion)
autocompletedCells += columnCount - columns.length;
if (autocompletedCells > MAX_AUTOCOMPLETED_CELLS) {
break;
}
if (nextLine === startLine + 2) {
const token_tbo = state.push("tbody_open", "tbody", 1);
token_tbo.map = tbodyLines = [ startLine + 2, 0 ];
}
const token_tro = state.push("tr_open", "tr", 1);
token_tro.map = [ nextLine, nextLine + 1 ];
for (let i = 0; i < columnCount; i++) {
const token_tdo = state.push("td_open", "td", 1);
if (aligns[i]) {
token_tdo.attrs = [ [ "style", "text-align:" + aligns[i] ] ];
}
const token_il = state.push("inline", "", 0);
token_il.content = columns[i] ? columns[i].trim() : "";
token_il.children = [];
state.push("td_close", "td", -1);
}
state.push("tr_close", "tr", -1);
}
if (tbodyLines) {
state.push("tbody_close", "tbody", -1);
tbodyLines[1] = nextLine;
}
state.push("table_close", "table", -1);
tableLines[1] = nextLine;
state.parentType = oldParentType;
state.line = nextLine;
return true;
}
// Code block (4 spaces padded)
function code(state, startLine, endLine /*, silent */) {
if (state.sCount[startLine] - state.blkIndent < 4) {
return false;
}
let nextLine = startLine + 1;
let last = nextLine;
while (nextLine < endLine) {
if (state.isEmpty(nextLine)) {
nextLine++;
continue;
}
if (state.sCount[nextLine] - state.blkIndent >= 4) {
nextLine++;
last = nextLine;
continue;
}
break;
}
state.line = last;
const token = state.push("code_block", "code", 0);
token.content = state.getLines(startLine, last, 4 + state.blkIndent, false) + "\n";
token.map = [ startLine, state.line ];
return true;
}
// fences (``` lang, ~~~ lang)
function fence(state, startLine, endLine, silent) {
let pos = state.bMarks[startLine] + state.tShift[startLine];
let max = state.eMarks[startLine];
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
if (pos + 3 > max) {
return false;
}
const marker = state.src.charCodeAt(pos);
if (marker !== 126 /* ~ */ && marker !== 96 /* ` */) {
return false;
}
// scan marker length
let mem = pos;
pos = state.skipChars(pos, marker);
let len = pos - mem;
if (len < 3) {
return false;
}
const markup = state.src.slice(mem, pos);
const params = state.src.slice(pos, max);
if (marker === 96 /* ` */) {
if (params.indexOf(String.fromCharCode(marker)) >= 0) {
return false;
}
}
// Since start is found, we can report success here in validation mode
if (silent) {
return true;
}
// search end of block
let nextLine = startLine;
let haveEndMarker = false;
for (;;) {
nextLine++;
if (nextLine >= endLine) {
// unclosed block should be autoclosed by end of document.
// also block seems to be autoclosed by end of parent
break;
}
pos = mem = state.bMarks[nextLine] + state.tShift[nextLine];
max = state.eMarks[nextLine];
if (pos < max && state.sCount[nextLine] < state.blkIndent) {
// non-empty line with negative indent should stop the list:
// - ```
// test
break;
}
if (state.src.charCodeAt(pos) !== marker) {
continue;
}
if (state.sCount[nextLine] - state.blkIndent >= 4) {
// closing fence should be indented less than 4 spaces
continue;
}
pos = state.skipChars(pos, marker);
// closing code fence must be at least as long as the opening one
if (pos - mem < len) {
continue;
}
// make sure tail has spaces only
pos = state.skipSpaces(pos);
if (pos < max) {
continue;
}
haveEndMarker = true;
// found!
break;
}
// If a fence has heading spaces, they should be removed from its inner block
len = state.sCount[startLine];
state.line = nextLine + (haveEndMarker ? 1 : 0);
const token = state.push("fence", "code", 0);
token.info = params;
token.content = state.getLines(startLine + 1, nextLine, len, true);
token.markup = markup;
token.map = [ startLine, state.line ];
return true;
}
// Block quotes
function blockquote(state, startLine, endLine, silent) {
let pos = state.bMarks[startLine] + state.tShift[startLine];
let max = state.eMarks[startLine];
const oldLineMax = state.lineMax;
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
// check the block quote marker
if (state.src.charCodeAt(pos) !== 62 /* > */) {
return false;
}
// we know that it's going to be a valid blockquote,
// so no point trying to find the end of it in silent mode
if (silent) {
return true;
}
const oldBMarks = [];
const oldBSCount = [];
const oldSCount = [];
const oldTShift = [];
const terminatorRules = state.md.block.ruler.getRules("blockquote");
const oldParentType = state.parentType;
state.parentType = "blockquote";
let lastLineEmpty = false;
let nextLine;
// Search the end of the block
// Block ends with either:
// 1. an empty line outside:
// ```
// > test
// ```
// 2. an empty line inside:
// ```
// >
// test
// ```
// 3. another tag:
// ```
// > test
// - - -
// ```
for (nextLine = startLine; nextLine < endLine; nextLine++) {
// check if it's outdented, i.e. it's inside list item and indented
// less than said list item:
// ```
// 1. anything
// > current blockquote
// 2. checking this line
// ```
const isOutdented = state.sCount[nextLine] < state.blkIndent;
pos = state.bMarks[nextLine] + state.tShift[nextLine];
max = state.eMarks[nextLine];
if (pos >= max) {
// Case 1: line is not inside the blockquote, and this line is empty.
break;
}
if (state.src.charCodeAt(pos++) === 62 /* > */ && !isOutdented) {
// This line is inside the blockquote.
// set offset past spaces and ">"
let initial = state.sCount[nextLine] + 1;
let spaceAfterMarker;
let adjustTab;
// skip one optional space after '>'
if (state.src.charCodeAt(pos) === 32 /* space */) {
// ' > test '
// ^ -- position start of line here:
pos++;
initial++;
adjustTab = false;
spaceAfterMarker = true;
} else if (state.src.charCodeAt(pos) === 9 /* tab */) {
spaceAfterMarker = true;
if ((state.bsCount[nextLine] + initial) % 4 === 3) {
// ' >\t test '
// ^ -- position start of line here (tab has width===1)
pos++;
initial++;
adjustTab = false;
} else {
// ' >\t test '
// ^ -- position start of line here + shift bsCount slightly
// to make extra space appear
adjustTab = true;
}
} else {
spaceAfterMarker = false;
}
let offset = initial;
oldBMarks.push(state.bMarks[nextLine]);
state.bMarks[nextLine] = pos;
while (pos < max) {
const ch = state.src.charCodeAt(pos);
if (isSpace(ch)) {
if (ch === 9) {
offset += 4 - (offset + state.bsCount[nextLine] + (adjustTab ? 1 : 0)) % 4;
} else {
offset++;
}
} else {
break;
}
pos++;
}
lastLineEmpty = pos >= max;
oldBSCount.push(state.bsCount[nextLine]);
state.bsCount[nextLine] = state.sCount[nextLine] + 1 + (spaceAfterMarker ? 1 : 0);
oldSCount.push(state.sCount[nextLine]);
state.sCount[nextLine] = offset - initial;
oldTShift.push(state.tShift[nextLine]);
state.tShift[nextLine] = pos - state.bMarks[nextLine];
continue;
}
// Case 2: line is not inside the blockquote, and the last line was empty.
if (lastLineEmpty) {
break;
}
// Case 3: another tag found.
let terminate = false;
for (let i = 0, l = terminatorRules.length; i < l; i++) {
if (terminatorRules[i](state, nextLine, endLine, true)) {
terminate = true;
break;
}
}
if (terminate) {
// Quirk to enforce "hard termination mode" for paragraphs;
// normally if you call `tokenize(state, startLine, nextLine)`,
// paragraphs will look below nextLine for paragraph continuation,
// but if blockquote is terminated by another tag, they shouldn't
state.lineMax = nextLine;
if (state.blkIndent !== 0) {
// state.blkIndent was non-zero, we now set it to zero,
// so we need to re-calculate all offsets to appear as
// if indent wasn't changed
oldBMarks.push(state.bMarks[nextLine]);
oldBSCount.push(state.bsCount[nextLine]);
oldTShift.push(state.tShift[nextLine]);
oldSCount.push(state.sCount[nextLine]);
state.sCount[nextLine] -= state.blkIndent;
}
break;
}
oldBMarks.push(state.bMarks[nextLine]);
oldBSCount.push(state.bsCount[nextLine]);
oldTShift.push(state.tShift[nextLine]);
oldSCount.push(state.sCount[nextLine]);
// A negative indentation means that this is a paragraph continuation
state.sCount[nextLine] = -1;
}
const oldIndent = state.blkIndent;
state.blkIndent = 0;
const token_o = state.push("blockquote_open", "blockquote", 1);
token_o.markup = ">";
const lines = [ startLine, 0 ];
token_o.map = lines;
state.md.block.tokenize(state, startLine, nextLine);
const token_c = state.push("blockquote_close", "blockquote", -1);
token_c.markup = ">";
state.lineMax = oldLineMax;
state.parentType = oldParentType;
lines[1] = state.line;
// Restore original tShift; this might not be necessary since the parser
// has already been here, but just to make sure we can do that.
for (let i = 0; i < oldTShift.length; i++) {
state.bMarks[i + startLine] = oldBMarks[i];
state.tShift[i + startLine] = oldTShift[i];
state.sCount[i + startLine] = oldSCount[i];
state.bsCount[i + startLine] = oldBSCount[i];
}
state.blkIndent = oldIndent;
return true;
}
// Horizontal rule
function hr(state, startLine, endLine, silent) {
const max = state.eMarks[startLine];
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
let pos = state.bMarks[startLine] + state.tShift[startLine];
const marker = state.src.charCodeAt(pos++);
// Check hr marker
if (marker !== 42 /* * */ && marker !== 45 /* - */ && marker !== 95 /* _ */) {
return false;
}
// markers can be mixed with spaces, but there should be at least 3 of them
let cnt = 1;
while (pos < max) {
const ch = state.src.charCodeAt(pos++);
if (ch !== marker && !isSpace(ch)) {
return false;
}
if (ch === marker) {
cnt++;
}
}
if (cnt < 3) {
return false;
}
if (silent) {
return true;
}
state.line = startLine + 1;
const token = state.push("hr", "hr", 0);
token.map = [ startLine, state.line ];
token.markup = Array(cnt + 1).join(String.fromCharCode(marker));
return true;
}
// Lists
// Search `[-+*][\n ]`, returns next pos after marker on success
// or -1 on fail.
function skipBulletListMarker(state, startLine) {
const max = state.eMarks[startLine];
let pos = state.bMarks[startLine] + state.tShift[startLine];
const marker = state.src.charCodeAt(pos++);
// Check bullet
if (marker !== 42 /* * */ && marker !== 45 /* - */ && marker !== 43 /* + */) {
return -1;
}
if (pos < max) {
const ch = state.src.charCodeAt(pos);
if (!isSpace(ch)) {
// " -test " - is not a list item
return -1;
}
}
return pos;
}
// Search `\d+[.)][\n ]`, returns next pos after marker on success
// or -1 on fail.
function skipOrderedListMarker(state, startLine) {
const start = state.bMarks[startLine] + state.tShift[startLine];
const max = state.eMarks[startLine];
let pos = start;
// List marker should have at least 2 chars (digit + dot)
if (pos + 1 >= max) {
return -1;
}
let ch = state.src.charCodeAt(pos++);
if (ch < 48 /* 0 */ || ch > 57 /* 9 */) {
return -1;
}
for (;;) {
// EOL -> fail
if (pos >= max) {
return -1;
}
ch = state.src.charCodeAt(pos++);
if (ch >= 48 /* 0 */ && ch <= 57 /* 9 */) {
// List marker should have no more than 9 digits
// (prevents integer overflow in browsers)
if (pos - start >= 10) {
return -1;
}
continue;
}
// found valid marker
if (ch === 41 /* ) */ || ch === 46 /* . */) {
break;
}
return -1;
}
if (pos < max) {
ch = state.src.charCodeAt(pos);
if (!isSpace(ch)) {
// " 1.test " - is not a list item
return -1;
}
}
return pos;
}
function markTightParagraphs(state, idx) {
const level = state.level + 2;
for (let i = idx + 2, l = state.tokens.length - 2; i < l; i++) {
if (state.tokens[i].level === level && state.tokens[i].type === "paragraph_open") {
state.tokens[i + 2].hidden = true;
state.tokens[i].hidden = true;
i += 2;
}
}
}
function list(state, startLine, endLine, silent) {
let max, pos, start, token;
let nextLine = startLine;
let tight = true;
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[nextLine] - state.blkIndent >= 4) {
return false;
}
// Special case:
// - item 1
// - item 2
// - item 3
// - item 4
// - this one is a paragraph continuation
if (state.listIndent >= 0 && state.sCount[nextLine] - state.listIndent >= 4 && state.sCount[nextLine] < state.blkIndent) {
return false;
}
let isTerminatingParagraph = false;
// limit conditions when list can interrupt
// a paragraph (validation mode only)
if (silent && state.parentType === "paragraph") {
// Next list item should still terminate previous list item;
// This code can fail if plugins use blkIndent as well as lists,
// but I hope the spec gets fixed long before that happens.
if (state.sCount[nextLine] >= state.blkIndent) {
isTerminatingParagraph = true;
}
}
// Detect list type and position after marker
let isOrdered;
let markerValue;
let posAfterMarker;
if ((posAfterMarker = skipOrderedListMarker(state, nextLine)) >= 0) {
isOrdered = true;
start = state.bMarks[nextLine] + state.tShift[nextLine];
markerValue = Number(state.src.slice(start, posAfterMarker - 1));
// If we're starting a new ordered list right after
// a paragraph, it should start with 1.
if (isTerminatingParagraph && markerValue !== 1) return false;
} else if ((posAfterMarker = skipBulletListMarker(state, nextLine)) >= 0) {
isOrdered = false;
} else {
return false;
}
// If we're starting a new unordered list right after
// a paragraph, first line should not be empty.
if (isTerminatingParagraph) {
if (state.skipSpaces(posAfterMarker) >= state.eMarks[nextLine]) return false;
}
// For validation mode we can terminate immediately
if (silent) {
return true;
}
// We should terminate list on style change. Remember first one to compare.
const markerCharCode = state.src.charCodeAt(posAfterMarker - 1);
// Start list
const listTokIdx = state.tokens.length;
if (isOrdered) {
token = state.push("ordered_list_open", "ol", 1);
if (markerValue !== 1) {
token.attrs = [ [ "start", markerValue ] ];
}
} else {
token = state.push("bullet_list_open", "ul", 1);
}
const listLines = [ nextLine, 0 ];
token.map = listLines;
token.markup = String.fromCharCode(markerCharCode);
// Iterate list items
let prevEmptyEnd = false;
const terminatorRules = state.md.block.ruler.getRules("list");
const oldParentType = state.parentType;
state.parentType = "list";
while (nextLine < endLine) {
pos = posAfterMarker;
max = state.eMarks[nextLine];
const initial = state.sCount[nextLine] + posAfterMarker - (state.bMarks[nextLine] + state.tShift[nextLine]);
let offset = initial;
while (pos < max) {
const ch = state.src.charCodeAt(pos);
if (ch === 9) {
offset += 4 - (offset + state.bsCount[nextLine]) % 4;
} else if (ch === 32) {
offset++;
} else {
break;
}
pos++;
}
const contentStart = pos;
let indentAfterMarker;
if (contentStart >= max) {
// trimming space in "- \n 3" case, indent is 1 here
indentAfterMarker = 1;
} else {
indentAfterMarker = offset - initial;
}
// If we have more than 4 spaces, the indent is 1
// (the rest is just indented code block)
if (indentAfterMarker > 4) {
indentAfterMarker = 1;
}
// " - test"
// ^^^^^ - calculating total length of this thing
const indent = initial + indentAfterMarker;
// Run subparser & write tokens
token = state.push("list_item_open", "li", 1);
token.markup = String.fromCharCode(markerCharCode);
const itemLines = [ nextLine, 0 ];
token.map = itemLines;
if (isOrdered) {
token.info = state.src.slice(start, posAfterMarker - 1);
}
// change current state, then restore it after parser subcall
const oldTight = state.tight;
const oldTShift = state.tShift[nextLine];
const oldSCount = state.sCount[nextLine];
// - example list
// ^ listIndent position will be here
// ^ blkIndent position will be here
const oldListIndent = state.listIndent;
state.listIndent = state.blkIndent;
state.blkIndent = indent;
state.tight = true;
state.tShift[nextLine] = contentStart - state.bMarks[nextLine];
state.sCount[nextLine] = offset;
if (contentStart >= max && state.isEmpty(nextLine + 1)) {
// workaround for this case
// (list item is empty, list terminates before "foo"):
// ~~~~~~~~
// -
// foo
// ~~~~~~~~
state.line = Math.min(state.line + 2, endLine);
} else {
state.md.block.tokenize(state, nextLine, endLine, true);
}
// If any of list item is tight, mark list as tight
if (!state.tight || prevEmptyEnd) {
tight = false;
}
// Item become loose if finish with empty line,
// but we should filter last element, because it means list finish
prevEmptyEnd = state.line - nextLine > 1 && state.isEmpty(state.line - 1);
state.blkIndent = state.listIndent;
state.listIndent = oldListIndent;
state.tShift[nextLine] = oldTShift;
state.sCount[nextLine] = oldSCount;
state.tight = oldTight;
token = state.push("list_item_close", "li", -1);
token.markup = String.fromCharCode(markerCharCode);
nextLine = state.line;
itemLines[1] = nextLine;
if (nextLine >= endLine) {
break;
}
// Try to check if list is terminated or continued.
if (state.sCount[nextLine] < state.blkIndent) {
break;
}
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[nextLine] - state.blkIndent >= 4) {
break;
}
// fail if terminating block found
let terminate = false;
for (let i = 0, l = terminatorRules.length; i < l; i++) {
if (terminatorRules[i](state, nextLine, endLine, true)) {
terminate = true;
break;
}
}
if (terminate) {
break;
}
// fail if list has another type
if (isOrdered) {
posAfterMarker = skipOrderedListMarker(state, nextLine);
if (posAfterMarker < 0) {
break;
}
start = state.bMarks[nextLine] + state.tShift[nextLine];
} else {
posAfterMarker = skipBulletListMarker(state, nextLine);
if (posAfterMarker < 0) {
break;
}
}
if (markerCharCode !== state.src.charCodeAt(posAfterMarker - 1)) {
break;
}
}
// Finalize list
if (isOrdered) {
token = state.push("ordered_list_close", "ol", -1);
} else {
token = state.push("bullet_list_close", "ul", -1);
}
token.markup = String.fromCharCode(markerCharCode);
listLines[1] = nextLine;
state.line = nextLine;
state.parentType = oldParentType;
// mark paragraphs tight if needed
if (tight) {
markTightParagraphs(state, listTokIdx);
}
return true;
}
function reference(state, startLine, _endLine, silent) {
let pos = state.bMarks[startLine] + state.tShift[startLine];
let max = state.eMarks[startLine];
let nextLine = startLine + 1;
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
if (state.src.charCodeAt(pos) !== 91 /* [ */) {
return false;
}
function getNextLine(nextLine) {
const endLine = state.lineMax;
if (nextLine >= endLine || state.isEmpty(nextLine)) {
// empty line or end of input
return null;
}
let isContinuation = false;
// this would be a code block normally, but after paragraph
// it's considered a lazy continuation regardless of what's there
if (state.sCount[nextLine] - state.blkIndent > 3) {
isContinuation = true;
}
// quirk for blockquotes, this line should already be checked by that rule
if (state.sCount[nextLine] < 0) {
isContinuation = true;
}
if (!isContinuation) {
const terminatorRules = state.md.block.ruler.getRules("reference");
const oldParentType = state.parentType;
state.parentType = "reference";
// Some tags can terminate paragraph without empty line.
let terminate = false;
for (let i = 0, l = terminatorRules.length; i < l; i++) {
if (terminatorRules[i](state, nextLine, endLine, true)) {
terminate = true;
break;
}
}
state.parentType = oldParentType;
if (terminate) {
// terminated by another block
return null;
}
}
const pos = state.bMarks[nextLine] + state.tShift[nextLine];
const max = state.eMarks[nextLine];
// max + 1 explicitly includes the newline
return state.src.slice(pos, max + 1);
}
let str = state.src.slice(pos, max + 1);
max = str.length;
let labelEnd = -1;
for (pos = 1; pos < max; pos++) {
const ch = str.charCodeAt(pos);
if (ch === 91 /* [ */) {
return false;
} else if (ch === 93 /* ] */) {
labelEnd = pos;
break;
} else if (ch === 10 /* \n */) {
const lineContent = getNextLine(nextLine);
if (lineContent !== null) {
str += lineContent;
max = str.length;
nextLine++;
}
} else if (ch === 92 /* \ */) {
pos++;
if (pos < max && str.charCodeAt(pos) === 10) {
const lineContent = getNextLine(nextLine);
if (lineContent !== null) {
str += lineContent;
max = str.length;
nextLine++;
}
}
}
}
if (labelEnd < 0 || str.charCodeAt(labelEnd + 1) !== 58 /* : */) {
return false;
}
// [label]: destination 'title'
// ^^^ skip optional whitespace here
for (pos = labelEnd + 2; pos < max; pos++) {
const ch = str.charCodeAt(pos);
if (ch === 10) {
const lineContent = getNextLine(nextLine);
if (lineContent !== null) {
str += lineContent;
max = str.length;
nextLine++;
}
} else if (isSpace(ch)) ; else {
break;
}
}
// [label]: destination 'title'
// ^^^^^^^^^^^ parse this
const destRes = state.md.helpers.parseLinkDestination(str, pos, max);
if (!destRes.ok) {
return false;
}
const href = state.md.normalizeLink(destRes.str);
if (!state.md.validateLink(href)) {
return false;
}
pos = destRes.pos;
// save cursor state, we could require to rollback later
const destEndPos = pos;
const destEndLineNo = nextLine;
// [label]: destination 'title'
// ^^^ skipping those spaces
const start = pos;
for (;pos < max; pos++) {
const ch = str.charCodeAt(pos);
if (ch === 10) {
const lineContent = getNextLine(nextLine);
if (lineContent !== null) {
str += lineContent;
max = str.length;
nextLine++;
}
} else if (isSpace(ch)) ; else {
break;
}
}
// [label]: destination 'title'
// ^^^^^^^ parse this
let titleRes = state.md.helpers.parseLinkTitle(str, pos, max);
while (titleRes.can_continue) {
const lineContent = getNextLine(nextLine);
if (lineContent === null) break;
str += lineContent;
pos = max;
max = str.length;
nextLine++;
titleRes = state.md.helpers.parseLinkTitle(str, pos, max, titleRes);
}
let title;
if (pos < max && start !== pos && titleRes.ok) {
title = titleRes.str;
pos = titleRes.pos;
} else {
title = "";
pos = destEndPos;
nextLine = destEndLineNo;
}
// skip trailing spaces until the rest of the line
while (pos < max) {
const ch = str.charCodeAt(pos);
if (!isSpace(ch)) {
break;
}
pos++;
}
if (pos < max && str.charCodeAt(pos) !== 10) {
if (title) {
// garbage at the end of the line after title,
// but it could still be a valid reference if we roll back
title = "";
pos = destEndPos;
nextLine = destEndLineNo;
while (pos < max) {
const ch = str.charCodeAt(pos);
if (!isSpace(ch)) {
break;
}
pos++;
}
}
}
if (pos < max && str.charCodeAt(pos) !== 10) {
// garbage at the end of the line
return false;
}
const label = normalizeReference(str.slice(1, labelEnd));
if (!label) {
// CommonMark 0.20 disallows empty labels
return false;
}
// Reference can not terminate anything. This check is for safety only.
/* istanbul ignore if */ if (silent) {
return true;
}
if (typeof state.env.references === "undefined") {
state.env.references = {};
}
if (typeof state.env.references[label] === "undefined") {
state.env.references[label] = {
title: title,
href: href
};
}
state.line = nextLine;
return true;
}
// List of valid html blocks names, according to commonmark spec
// https://spec.commonmark.org/0.30/#html-blocks
var block_names = [ "address", "article", "aside", "base", "basefont", "blockquote", "body", "caption", "center", "col", "colgroup", "dd", "details", "dialog", "dir", "div", "dl", "dt", "fieldset", "figcaption", "figure", "footer", "form", "frame", "frameset", "h1", "h2", "h3", "h4", "h5", "h6", "head", "header", "hr", "html", "iframe", "legend", "li", "link", "main", "menu", "menuitem", "nav", "noframes", "ol", "optgroup", "option", "p", "param", "search", "section", "summary", "table", "tbody", "td", "tfoot", "th", "thead", "title", "tr", "track", "ul" ];
// Regexps to match html elements
const attr_name = "[a-zA-Z_:][a-zA-Z0-9:._-]*";
const unquoted = "[^\"'=<>`\\x00-\\x20]+";
const single_quoted = "'[^']*'";
const double_quoted = '"[^"]*"';
const attr_value = "(?:" + unquoted + "|" + single_quoted + "|" + double_quoted + ")";
const attribute = "(?:\\s+" + attr_name + "(?:\\s*=\\s*" + attr_value + ")?)";
const open_tag = "<[A-Za-z][A-Za-z0-9\\-]*" + attribute + "*\\s*\\/?>";
const close_tag = "<\\/[A-Za-z][A-Za-z0-9\\-]*\\s*>";
const comment = "\x3c!---?>|\x3c!--(?:[^-]|-[^-]|--[^>])*--\x3e";
const processing = "<[?][\\s\\S]*?[?]>";
const declaration = "<![A-Za-z][^>]*>";
const cdata = "<!\\[CDATA\\[[\\s\\S]*?\\]\\]>";
const HTML_TAG_RE = new RegExp("^(?:" + open_tag + "|" + close_tag + "|" + comment + "|" + processing + "|" + declaration + "|" + cdata + ")");
const HTML_OPEN_CLOSE_TAG_RE = new RegExp("^(?:" + open_tag + "|" + close_tag + ")");
// HTML block
// An array of opening and corresponding closing sequences for html tags,
// last argument defines whether it can terminate a paragraph or not
const HTML_SEQUENCES = [ [ /^<(script|pre|style|textarea)(?=(\s|>|$))/i, /<\/(script|pre|style|textarea)>/i, true ], [ /^<!--/, /-->/, true ], [ /^<\?/, /\?>/, true ], [ /^<![A-Z]/, />/, true ], [ /^<!\[CDATA\[/, /\]\]>/, true ], [ new RegExp("^</?(" + block_names.join("|") + ")(?=(\\s|/?>|$))", "i"), /^$/, true ], [ new RegExp(HTML_OPEN_CLOSE_TAG_RE.source + "\\s*$"), /^$/, false ] ];
function html_block(state, startLine, endLine, silent) {
let pos = state.bMarks[startLine] + state.tShift[startLine];
let max = state.eMarks[startLine];
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
if (!state.md.options.html) {
return false;
}
if (state.src.charCodeAt(pos) !== 60 /* < */) {
return false;
}
let lineText = state.src.slice(pos, max);
let i = 0;
for (;i < HTML_SEQUENCES.length; i++) {
if (HTML_SEQUENCES[i][0].test(lineText)) {
break;
}
}
if (i === HTML_SEQUENCES.length) {
return false;
}
if (silent) {
// true if this sequence can be a terminator, false otherwise
return HTML_SEQUENCES[i][2];
}
let nextLine = startLine + 1;
// If we are here - we detected HTML block.
// Let's roll down till block end.
if (!HTML_SEQUENCES[i][1].test(lineText)) {
for (;nextLine < endLine; nextLine++) {
if (state.sCount[nextLine] < state.blkIndent) {
break;
}
pos = state.bMarks[nextLine] + state.tShift[nextLine];
max = state.eMarks[nextLine];
lineText = state.src.slice(pos, max);
if (HTML_SEQUENCES[i][1].test(lineText)) {
if (lineText.length !== 0) {
nextLine++;
}
break;
}
}
}
state.line = nextLine;
const token = state.push("html_block", "", 0);
token.map = [ startLine, nextLine ];
token.content = state.getLines(startLine, nextLine, state.blkIndent, true);
return true;
}
// heading (#, ##, ...)
function heading(state, startLine, endLine, silent) {
let pos = state.bMarks[startLine] + state.tShift[startLine];
let max = state.eMarks[startLine];
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
let ch = state.src.charCodeAt(pos);
if (ch !== 35 /* # */ || pos >= max) {
return false;
}
// count heading level
let level = 1;
ch = state.src.charCodeAt(++pos);
while (ch === 35 /* # */ && pos < max && level <= 6) {
level++;
ch = state.src.charCodeAt(++pos);
}
if (level > 6 || pos < max && !isSpace(ch)) {
return false;
}
if (silent) {
return true;
}
// Let's cut tails like ' ### ' from the end of string
max = state.skipSpacesBack(max, pos);
const tmp = state.skipCharsBack(max, 35, pos);
// #
if (tmp > pos && isSpace(state.src.charCodeAt(tmp - 1))) {
max = tmp;
}
state.line = startLine + 1;
const token_o = state.push("heading_open", "h" + String(level), 1);
token_o.markup = "########".slice(0, level);
token_o.map = [ startLine, state.line ];
const token_i = state.push("inline", "", 0);
token_i.content = state.src.slice(pos, max).trim();
token_i.map = [ startLine, state.line ];
token_i.children = [];
const token_c = state.push("heading_close", "h" + String(level), -1);
token_c.markup = "########".slice(0, level);
return true;
}
// lheading (---, ===)
function lheading(state, startLine, endLine /*, silent */) {
const terminatorRules = state.md.block.ruler.getRules("paragraph");
// if it's indented more than 3 spaces, it should be a code block
if (state.sCount[startLine] - state.blkIndent >= 4) {
return false;
}
const oldParentType = state.parentType;
state.parentType = "paragraph";
// use paragraph to match terminatorRules
// jump line-by-line until empty one or EOF
let level = 0;
let marker;
let nextLine = startLine + 1;
for (;nextLine < endLine && !state.isEmpty(nextLine); nextLine++) {
// this would be a code block normally, but after paragraph
// it's considered a lazy continuation regardless of what's there
if (state.sCount[nextLine] - state.blkIndent > 3) {
continue;
}
// Check for underline in setext header
if (state.sCount[nextLine] >= state.blkIndent) {
let pos = state.bMarks[nextLine] + state.tShift[nextLine];
const max = state.eMarks[nextLine];
if (pos < max) {
marker = state.src.charCodeAt(pos);
if (marker === 45 /* - */ || marker === 61 /* = */) {
pos = state.skipChars(pos, marker);
pos = state.skipSpaces(pos);
if (pos >= max) {
level = marker === 61 /* = */ ? 1 : 2;
break;
}
}
}
}
// quirk for blockquotes, this line should already be checked by that rule
if (state.sCount[nextLine] < 0) {
continue;
}
// Some tags can terminate paragraph without empty line.
let terminate = false;
for (let i = 0, l = terminatorRules.length; i < l; i++) {
if (terminatorRules[i](state, nextLine, endLine, true)) {
terminate = true;
break;
}
}
if (terminate) {
break;
}
}
if (!level) {
// Didn't find valid underline
return false;
}
const content = state.getLines(startLine, nextLine, state.blkIndent, false).trim();
state.line = nextLine + 1;
const token_o = state.push("heading_open", "h" + String(level), 1);
token_o.markup = String.fromCharCode(marker);
token_o.map = [ startLine, state.line ];
const token_i = state.push("inline", "", 0);
token_i.content = content;
token_i.map = [ startLine, state.line - 1 ];
token_i.children = [];
const token_c = state.push("heading_close", "h" + String(level), -1);
token_c.markup = String.fromCharCode(marker);
state.parentType = oldParentType;
return true;
}
// Paragraph
function paragraph(state, startLine, endLine) {
const terminatorRules = state.md.block.ruler.getRules("paragraph");
const oldParentType = state.parentType;
let nextLine = startLine + 1;
state.parentType = "paragraph";
// jump line-by-line until empty one or EOF
for (;nextLine < endLine && !state.isEmpty(nextLine); nextLine++) {
// this would be a code block normally, but after paragraph
// it's considered a lazy continuation regardless of what's there
if (state.sCount[nextLine] - state.blkIndent > 3) {
continue;
}
// quirk for blockquotes, this line should already be checked by that rule
if (state.sCount[nextLine] < 0) {
continue;
}
// Some tags can terminate paragraph without empty line.
let terminate = false;
for (let i = 0, l = terminatorRules.length; i < l; i++) {
if (terminatorRules[i](state, nextLine, endLine, true)) {
terminate = true;
break;
}
}
if (terminate) {
break;
}
}
const content = state.getLines(startLine, nextLine, state.blkIndent, false).trim();
state.line = nextLine;
const token_o = state.push("paragraph_open", "p", 1);
token_o.map = [ startLine, state.line ];
const token_i = state.push("inline", "", 0);
token_i.content = content;
token_i.map = [ startLine, state.line ];
token_i.children = [];
state.push("paragraph_close", "p", -1);
state.parentType = oldParentType;
return true;
}
/** internal
* class ParserBlock
*
* Block-level tokenizer.
**/ const _rules$1 = [
// First 2 params - rule name & source. Secondary array - list of rules,
// which can be terminated by this one.
[ "table", table, [ "paragraph", "reference" ] ], [ "code", code ], [ "fence", fence, [ "paragraph", "reference", "blockquote", "list" ] ], [ "blockquote", blockquote, [ "paragraph", "reference", "blockquote", "list" ] ], [ "hr", hr, [ "paragraph", "reference", "blockquote", "list" ] ], [ "list", list, [ "paragraph", "reference", "blockquote" ] ], [ "reference", reference ], [ "html_block", html_block, [ "paragraph", "reference", "blockquote" ] ], [ "heading", heading, [ "paragraph", "reference", "blockquote" ] ], [ "lheading", lheading ], [ "paragraph", paragraph ] ];
/**
* new ParserBlock()
**/ function ParserBlock() {
/**
* ParserBlock#ruler -> Ruler
*
* [[Ruler]] instance. Keep configuration of block rules.
**/
this.ruler = new Ruler;
for (let i = 0; i < _rules$1.length; i++) {
this.ruler.push(_rules$1[i][0], _rules$1[i][1], {
alt: (_rules$1[i][2] || []).slice()
});
}
}
// Generate tokens for input range
ParserBlock.prototype.tokenize = function(state, startLine, endLine) {
const rules = this.ruler.getRules("");
const len = rules.length;
const maxNesting = state.md.options.maxNesting;
let line = startLine;
let hasEmptyLines = false;
while (line < endLine) {
state.line = line = state.skipEmptyLines(line);
if (line >= endLine) {
break;
}
// Termination condition for nested calls.
// Nested calls currently used for blockquotes & lists
if (state.sCount[line] < state.blkIndent) {
break;
}
// If nesting level exceeded - skip tail to the end. That's not ordinary
// situation and we should not care about content.
if (state.level >= maxNesting) {
state.line = endLine;
break;
}
// Try all possible rules.
// On success, rule should:
// - update `state.line`
// - update `state.tokens`
// - return true
const prevLine = state.line;
let ok = false;
for (let i = 0; i < len; i++) {
ok = rules[i](state, line, endLine, false);
if (ok) {
if (prevLine >= state.line) {
throw new Error("block rule didn't increment state.line");
}
break;
}
}
// this can only happen if user disables paragraph rule
if (!ok) throw new Error("none of the block rules matched");
// set state.tight if we had an empty line before current tag
// i.e. latest empty line should not count
state.tight = !hasEmptyLines;
// paragraph might "eat" one newline after it in nested lists
if (state.isEmpty(state.line - 1)) {
hasEmptyLines = true;
}
line = state.line;
if (line < endLine && state.isEmpty(line)) {
hasEmptyLines = true;
line++;
state.line = line;
}
}
};
/**
* ParserBlock.parse(str, md, env, outTokens)
*
* Process input string and push block tokens into `outTokens`
**/ ParserBlock.prototype.parse = function(src, md, env, outTokens) {
if (!src) {
return;
}
const state = new this.State(src, md, env, outTokens);
this.tokenize(state, state.line, state.lineMax);
};
ParserBlock.prototype.State = StateBlock;
// Inline parser state
function StateInline(src, md, env, outTokens) {
this.src = src;
this.env = env;
this.md = md;
this.tokens = outTokens;
this.tokens_meta = Array(outTokens.length);
this.pos = 0;
this.posMax = this.src.length;
this.level = 0;
this.pending = "";
this.pendingLevel = 0;
// Stores { start: end } pairs. Useful for backtrack
// optimization of pairs parse (emphasis, strikes).
this.cache = {};
// List of emphasis-like delimiters for current tag
this.delimiters = [];
// Stack of delimiter lists for upper level tags
this._prev_delimiters = [];
// backtick length => last seen position
this.backticks = {};
this.backticksScanned = false;
// Counter used to disable inline linkify-it execution
// inside <a> and markdown links
this.linkLevel = 0;
}
// Flush pending text
StateInline.prototype.pushPending = function() {
const token = new Token("text", "", 0);
token.content = this.pending;
token.level = this.pendingLevel;
this.tokens.push(token);
this.pending = "";
return token;
};
// Push new token to "stream".
// If pending text exists - flush it as text token
StateInline.prototype.push = function(type, tag, nesting) {
if (this.pending) {
this.pushPending();
}
const token = new Token(type, tag, nesting);
let token_meta = null;
if (nesting < 0) {
// closing tag
this.level--;
this.delimiters = this._prev_delimiters.pop();
}
token.level = this.level;
if (nesting > 0) {
// opening tag
this.level++;
this._prev_delimiters.push(this.delimiters);
this.delimiters = [];
token_meta = {
delimiters: this.delimiters
};
}
this.pendingLevel = this.level;
this.tokens.push(token);
this.tokens_meta.push(token_meta);
return token;
};
// Scan a sequence of emphasis-like markers, and determine whether
// it can start an emphasis sequence or end an emphasis sequence.
// - start - position to scan from (it should point at a valid marker);
// - canSplitWord - determine if these markers can be found inside a word
StateInline.prototype.scanDelims = function(start, canSplitWord) {
const max = this.posMax;
const marker = this.src.charCodeAt(start);
// treat beginning of the line as a whitespace
const lastChar = start > 0 ? this.src.charCodeAt(start - 1) : 32;
let pos = start;
while (pos < max && this.src.charCodeAt(pos) === marker) {
pos++;
}
const count = pos - start;
// treat end of the line as a whitespace
const nextChar = pos < max ? this.src.charCodeAt(pos) : 32;
const isLastPunctChar = isMdAsciiPunct(lastChar) || isPunctChar(String.fromCharCode(lastChar));
const isNextPunctChar = isMdAsciiPunct(nextChar) || isPunctChar(String.fromCharCode(nextChar));
const isLastWhiteSpace = isWhiteSpace(lastChar);
const isNextWhiteSpace = isWhiteSpace(nextChar);
const left_flanking = !isNextWhiteSpace && (!isNextPunctChar || isLastWhiteSpace || isLastPunctChar);
const right_flanking = !isLastWhiteSpace && (!isLastPunctChar || isNextWhiteSpace || isNextPunctChar);
const can_open = left_flanking && (canSplitWord || !right_flanking || isLastPunctChar);
const can_close = right_flanking && (canSplitWord || !left_flanking || isNextPunctChar);
return {
can_open: can_open,
can_close: can_close,
length: count
};
};
// re-export Token class to use in block rules
StateInline.prototype.Token = Token;
// Skip text characters for text token, place those to pending buffer
// and increment current pos
// Rule to skip pure text
// '{}$%@~+=:' reserved for extentions
// !, ", #, $, %, &, ', (, ), *, +, ,, -, ., /, :, ;, <, =, >, ?, @, [, \, ], ^, _, `, {, |, }, or ~
// !!!! Don't confuse with "Markdown ASCII Punctuation" chars
// http://spec.commonmark.org/0.15/#ascii-punctuation-character
function isTerminatorChar(ch) {
switch (ch) {
case 10 /* \n */ :
case 33 /* ! */ :
case 35 /* # */ :
case 36 /* $ */ :
case 37 /* % */ :
case 38 /* & */ :
case 42 /* * */ :
case 43 /* + */ :
case 45 /* - */ :
case 58 /* : */ :
case 60 /* < */ :
case 61 /* = */ :
case 62 /* > */ :
case 64 /* @ */ :
case 91 /* [ */ :
case 92 /* \ */ :
case 93 /* ] */ :
case 94 /* ^ */ :
case 95 /* _ */ :
case 96 /* ` */ :
case 123 /* { */ :
case 125 /* } */ :
case 126 /* ~ */ :
return true;
default:
return false;
}
}
function text(state, silent) {
let pos = state.pos;
while (pos < state.posMax && !isTerminatorChar(state.src.charCodeAt(pos))) {
pos++;
}
if (pos === state.pos) {
return false;
}
if (!silent) {
state.pending += state.src.slice(state.pos, pos);
}
state.pos = pos;
return true;
}
// Alternative implementation, for memory.
// It costs 10% of performance, but allows extend terminators list, if place it
// to `ParserInline` property. Probably, will switch to it sometime, such
// flexibility required.
/*
var TERMINATOR_RE = /[\n!#$%&*+\-:<=>@[\\\]^_`{}~]/;
module.exports = function text(state, silent) {
var pos = state.pos,
idx = state.src.slice(pos).search(TERMINATOR_RE);
// first char is terminator -> empty text
if (idx === 0) { return false; }
// no terminator -> text till end of string
if (idx < 0) {
if (!silent) { state.pending += state.src.slice(pos); }
state.pos = state.src.length;
return true;
}
if (!silent) { state.pending += state.src.slice(pos, pos + idx); }
state.pos += idx;
return true;
}; */
// Process links like https://example.org/
// RFC3986: scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
const SCHEME_RE = /(?:^|[^a-z0-9.+-])([a-z][a-z0-9.+-]*)$/i;
function linkify(state, silent) {
if (!state.md.options.linkify) return false;
if (state.linkLevel > 0) return false;
const pos = state.pos;
const max = state.posMax;
if (pos + 3 > max) return false;
if (state.src.charCodeAt(pos) !== 58 /* : */) return false;
if (state.src.charCodeAt(pos + 1) !== 47 /* / */) return false;
if (state.src.charCodeAt(pos + 2) !== 47 /* / */) return false;
const match = state.pending.match(SCHEME_RE);
if (!match) return false;
const proto = match[1];
const link = state.md.linkify.matchAtStart(state.src.slice(pos - proto.length));
if (!link) return false;
let url = link.url;
// invalid link, but still detected by linkify somehow;
// need to check to prevent infinite loop below
if (url.length <= proto.length) return false;
// disallow '*' at the end of the link (conflicts with emphasis)
url = url.replace(/\*+$/, "");
const fullUrl = state.md.normalizeLink(url);
if (!state.md.validateLink(fullUrl)) return false;
if (!silent) {
state.pending = state.pending.slice(0, -proto.length);
const token_o = state.push("link_open", "a", 1);
token_o.attrs = [ [ "href", fullUrl ] ];
token_o.markup = "linkify";
token_o.info = "auto";
const token_t = state.push("text", "", 0);
token_t.content = state.md.normalizeLinkText(url);
const token_c = state.push("link_close", "a", -1);
token_c.markup = "linkify";
token_c.info = "auto";
}
state.pos += url.length - proto.length;
return true;
}
// Proceess '\n'
function newline(state, silent) {
let pos = state.pos;
if (state.src.charCodeAt(pos) !== 10 /* \n */) {
return false;
}
const pmax = state.pending.length - 1;
const max = state.posMax;
// ' \n' -> hardbreak
// Lookup in pending chars is bad practice! Don't copy to other rules!
// Pending string is stored in concat mode, indexed lookups will cause
// convertion to flat mode.
if (!silent) {
if (pmax >= 0 && state.pending.charCodeAt(pmax) === 32) {
if (pmax >= 1 && state.pending.charCodeAt(pmax - 1) === 32) {
// Find whitespaces tail of pending chars.
let ws = pmax - 1;
while (ws >= 1 && state.pending.charCodeAt(ws - 1) === 32) ws--;
state.pending = state.pending.slice(0, ws);
state.push("hardbreak", "br", 0);
} else {
state.pending = state.pending.slice(0, -1);
state.push("softbreak", "br", 0);
}
} else {
state.push("softbreak", "br", 0);
}
}
pos++;
// skip heading spaces for next line
while (pos < max && isSpace(state.src.charCodeAt(pos))) {
pos++;
}
state.pos = pos;
return true;
}
// Process escaped chars and hardbreaks
const ESCAPED = [];
for (let i = 0; i < 256; i++) {
ESCAPED.push(0);
}
"\\!\"#$%&'()*+,./:;<=>?@[]^_`{|}~-".split("").forEach((function(ch) {
ESCAPED[ch.charCodeAt(0)] = 1;
}));
function escape(state, silent) {
let pos = state.pos;
const max = state.posMax;
if (state.src.charCodeAt(pos) !== 92 /* \ */) return false;
pos++;
// '\' at the end of the inline block
if (pos >= max) return false;
let ch1 = state.src.charCodeAt(pos);
if (ch1 === 10) {
if (!silent) {
state.push("hardbreak", "br", 0);
}
pos++;
// skip leading whitespaces from next line
while (pos < max) {
ch1 = state.src.charCodeAt(pos);
if (!isSpace(ch1)) break;
pos++;
}
state.pos = pos;
return true;
}
let escapedStr = state.src[pos];
if (ch1 >= 55296 && ch1 <= 56319 && pos + 1 < max) {
const ch2 = state.src.charCodeAt(pos + 1);
if (ch2 >= 56320 && ch2 <= 57343) {
escapedStr += state.src[pos + 1];
pos++;
}
}
const origStr = "\\" + escapedStr;
if (!silent) {
const token = state.push("text_special", "", 0);
if (ch1 < 256 && ESCAPED[ch1] !== 0) {
token.content = escapedStr;
} else {
token.content = origStr;
}
token.markup = origStr;
token.info = "escape";
}
state.pos = pos + 1;
return true;
}
// Parse backticks
function backtick(state, silent) {
let pos = state.pos;
const ch = state.src.charCodeAt(pos);
if (ch !== 96 /* ` */) {
return false;
}
const start = pos;
pos++;
const max = state.posMax;
// scan marker length
while (pos < max && state.src.charCodeAt(pos) === 96 /* ` */) {
pos++;
}
const marker = state.src.slice(start, pos);
const openerLength = marker.length;
if (state.backticksScanned && (state.backticks[openerLength] || 0) <= start) {
if (!silent) state.pending += marker;
state.pos += openerLength;
return true;
}
let matchEnd = pos;
let matchStart;
// Nothing found in the cache, scan until the end of the line (or until marker is found)
while ((matchStart = state.src.indexOf("`", matchEnd)) !== -1) {
matchEnd = matchStart + 1;
// scan marker length
while (matchEnd < max && state.src.charCodeAt(matchEnd) === 96 /* ` */) {
matchEnd++;
}
const closerLength = matchEnd - matchStart;
if (closerLength === openerLength) {
// Found matching closer length.
if (!silent) {
const token = state.push("code_inline", "code", 0);
token.markup = marker;
token.content = state.src.slice(pos, matchStart).replace(/\n/g, " ").replace(/^ (.+) $/, "$1");
}
state.pos = matchEnd;
return true;
}
// Some different length found, put it in cache as upper limit of where closer can be found
state.backticks[closerLength] = matchStart;
}
// Scanned through the end, didn't find anything
state.backticksScanned = true;
if (!silent) state.pending += marker;
state.pos += openerLength;
return true;
}
// ~~strike through~~
// Insert each marker as a separate text token, and add it to delimiter list
function strikethrough_tokenize(state, silent) {
const start = state.pos;
const marker = state.src.charCodeAt(start);
if (silent) {
return false;
}
if (marker !== 126 /* ~ */) {
return false;
}
const scanned = state.scanDelims(state.pos, true);
let len = scanned.length;
const ch = String.fromCharCode(marker);
if (len < 2) {
return false;
}
let token;
if (len % 2) {
token = state.push("text", "", 0);
token.content = ch;
len--;
}
for (let i = 0; i < len; i += 2) {
token = state.push("text", "", 0);
token.content = ch + ch;
state.delimiters.push({
marker: marker,
length: 0,
// disable "rule of 3" length checks meant for emphasis
token: state.tokens.length - 1,
end: -1,
open: scanned.can_open,
close: scanned.can_close
});
}
state.pos += scanned.length;
return true;
}
function postProcess$1(state, delimiters) {
let token;
const loneMarkers = [];
const max = delimiters.length;
for (let i = 0; i < max; i++) {
const startDelim = delimiters[i];
if (startDelim.marker !== 126 /* ~ */) {
continue;
}
if (startDelim.end === -1) {
continue;
}
const endDelim = delimiters[startDelim.end];
token = state.tokens[startDelim.token];
token.type = "s_open";
token.tag = "s";
token.nesting = 1;
token.markup = "~~";
token.content = "";
token = state.tokens[endDelim.token];
token.type = "s_close";
token.tag = "s";
token.nesting = -1;
token.markup = "~~";
token.content = "";
if (state.tokens[endDelim.token - 1].type === "text" && state.tokens[endDelim.token - 1].content === "~") {
loneMarkers.push(endDelim.token - 1);
}
}
// If a marker sequence has an odd number of characters, it's splitted
// like this: `~~~~~` -> `~` + `~~` + `~~`, leaving one marker at the
// start of the sequence.
// So, we have to move all those markers after subsequent s_close tags.
while (loneMarkers.length) {
const i = loneMarkers.pop();
let j = i + 1;
while (j < state.tokens.length && state.tokens[j].type === "s_close") {
j++;
}
j--;
if (i !== j) {
token = state.tokens[j];
state.tokens[j] = state.tokens[i];
state.tokens[i] = token;
}
}
}
// Walk through delimiter list and replace text tokens with tags
function strikethrough_postProcess(state) {
const tokens_meta = state.tokens_meta;
const max = state.tokens_meta.length;
postProcess$1(state, state.delimiters);
for (let curr = 0; curr < max; curr++) {
if (tokens_meta[curr] && tokens_meta[curr].delimiters) {
postProcess$1(state, tokens_meta[curr].delimiters);
}
}
}
var r_strikethrough = {
tokenize: strikethrough_tokenize,
postProcess: strikethrough_postProcess
};
// Process *this* and _that_
// Insert each marker as a separate text token, and add it to delimiter list
function emphasis_tokenize(state, silent) {
const start = state.pos;
const marker = state.src.charCodeAt(start);
if (silent) {
return false;
}
if (marker !== 95 /* _ */ && marker !== 42 /* * */) {
return false;
}
const scanned = state.scanDelims(state.pos, marker === 42);
for (let i = 0; i < scanned.length; i++) {
const token = state.push("text", "", 0);
token.content = String.fromCharCode(marker);
state.delimiters.push({
// Char code of the starting marker (number).
marker: marker,
// Total length of these series of delimiters.
length: scanned.length,
// A position of the token this delimiter corresponds to.
token: state.tokens.length - 1,
// If this delimiter is matched as a valid opener, `end` will be
// equal to its position, otherwise it's `-1`.
end: -1,
// Boolean flags that determine if this delimiter could open or close
// an emphasis.
open: scanned.can_open,
close: scanned.can_close
});
}
state.pos += scanned.length;
return true;
}
function postProcess(state, delimiters) {
const max = delimiters.length;
for (let i = max - 1; i >= 0; i--) {
const startDelim = delimiters[i];
if (startDelim.marker !== 95 /* _ */ && startDelim.marker !== 42 /* * */) {
continue;
}
// Process only opening markers
if (startDelim.end === -1) {
continue;
}
const endDelim = delimiters[startDelim.end];
// If the previous delimiter has the same marker and is adjacent to this one,
// merge those into one strong delimiter.
// `<em><em>whatever</em></em>` -> `<strong>whatever</strong>`
const isStrong = i > 0 && delimiters[i - 1].end === startDelim.end + 1 &&
// check that first two markers match and adjacent
delimiters[i - 1].marker === startDelim.marker && delimiters[i - 1].token === startDelim.token - 1 &&
// check that last two markers are adjacent (we can safely assume they match)
delimiters[startDelim.end + 1].token === endDelim.token + 1;
const ch = String.fromCharCode(startDelim.marker);
const token_o = state.tokens[startDelim.token];
token_o.type = isStrong ? "strong_open" : "em_open";
token_o.tag = isStrong ? "strong" : "em";
token_o.nesting = 1;
token_o.markup = isStrong ? ch + ch : ch;
token_o.content = "";
const token_c = state.tokens[endDelim.token];
token_c.type = isStrong ? "strong_close" : "em_close";
token_c.tag = isStrong ? "strong" : "em";
token_c.nesting = -1;
token_c.markup = isStrong ? ch + ch : ch;
token_c.content = "";
if (isStrong) {
state.tokens[delimiters[i - 1].token].content = "";
state.tokens[delimiters[startDelim.end + 1].token].content = "";
i--;
}
}
}
// Walk through delimiter list and replace text tokens with tags
function emphasis_post_process(state) {
const tokens_meta = state.tokens_meta;
const max = state.tokens_meta.length;
postProcess(state, state.delimiters);
for (let curr = 0; curr < max; curr++) {
if (tokens_meta[curr] && tokens_meta[curr].delimiters) {
postProcess(state, tokens_meta[curr].delimiters);
}
}
}
var r_emphasis = {
tokenize: emphasis_tokenize,
postProcess: emphasis_post_process
};
// Process [link](<to> "stuff")
function link(state, silent) {
let code, label, res, ref;
let href = "";
let title = "";
let start = state.pos;
let parseReference = true;
if (state.src.charCodeAt(state.pos) !== 91 /* [ */) {
return false;
}
const oldPos = state.pos;
const max = state.posMax;
const labelStart = state.pos + 1;
const labelEnd = state.md.helpers.parseLinkLabel(state, state.pos, true);
// parser failed to find ']', so it's not a valid link
if (labelEnd < 0) {
return false;
}
let pos = labelEnd + 1;
if (pos < max && state.src.charCodeAt(pos) === 40 /* ( */) {
// Inline link
// might have found a valid shortcut link, disable reference parsing
parseReference = false;
// [link]( <href> "title" )
// ^^ skipping these spaces
pos++;
for (;pos < max; pos++) {
code = state.src.charCodeAt(pos);
if (!isSpace(code) && code !== 10) {
break;
}
}
if (pos >= max) {
return false;
}
// [link]( <href> "title" )
// ^^^^^^ parsing link destination
start = pos;
res = state.md.helpers.parseLinkDestination(state.src, pos, state.posMax);
if (res.ok) {
href = state.md.normalizeLink(res.str);
if (state.md.validateLink(href)) {
pos = res.pos;
} else {
href = "";
}
// [link]( <href> "title" )
// ^^ skipping these spaces
start = pos;
for (;pos < max; pos++) {
code = state.src.charCodeAt(pos);
if (!isSpace(code) && code !== 10) {
break;
}
}
// [link]( <href> "title" )
// ^^^^^^^ parsing link title
res = state.md.helpers.parseLinkTitle(state.src, pos, state.posMax);
if (pos < max && start !== pos && res.ok) {
title = res.str;
pos = res.pos;
// [link]( <href> "title" )
// ^^ skipping these spaces
for (;pos < max; pos++) {
code = state.src.charCodeAt(pos);
if (!isSpace(code) && code !== 10) {
break;
}
}
}
}
if (pos >= max || state.src.charCodeAt(pos) !== 41 /* ) */) {
// parsing a valid shortcut link failed, fallback to reference
parseReference = true;
}
pos++;
}
if (parseReference) {
// Link reference
if (typeof state.env.references === "undefined") {
return false;
}
if (pos < max && state.src.charCodeAt(pos) === 91 /* [ */) {
start = pos + 1;
pos = state.md.helpers.parseLinkLabel(state, pos);
if (pos >= 0) {
label = state.src.slice(start, pos++);
} else {
pos = labelEnd + 1;
}
} else {
pos = labelEnd + 1;
}
// covers label === '' and label === undefined
// (collapsed reference link and shortcut reference link respectively)
if (!label) {
label = state.src.slice(labelStart, labelEnd);
}
ref = state.env.references[normalizeReference(label)];
if (!ref) {
state.pos = oldPos;
return false;
}
href = ref.href;
title = ref.title;
}
// We found the end of the link, and know for a fact it's a valid link;
// so all that's left to do is to call tokenizer.
if (!silent) {
state.pos = labelStart;
state.posMax = labelEnd;
const token_o = state.push("link_open", "a", 1);
const attrs = [ [ "href", href ] ];
token_o.attrs = attrs;
if (title) {
attrs.push([ "title", title ]);
}
state.linkLevel++;
state.md.inline.tokenize(state);
state.linkLevel--;
state.push("link_close", "a", -1);
}
state.pos = pos;
state.posMax = max;
return true;
}
// Process ![image](<src> "title")
function image(state, silent) {
let code, content, label, pos, ref, res, title, start;
let href = "";
const oldPos = state.pos;
const max = state.posMax;
if (state.src.charCodeAt(state.pos) !== 33 /* ! */) {
return false;
}
if (state.src.charCodeAt(state.pos + 1) !== 91 /* [ */) {
return false;
}
const labelStart = state.pos + 2;
const labelEnd = state.md.helpers.parseLinkLabel(state, state.pos + 1, false);
// parser failed to find ']', so it's not a valid link
if (labelEnd < 0) {
return false;
}
pos = labelEnd + 1;
if (pos < max && state.src.charCodeAt(pos) === 40 /* ( */) {
// Inline link
// [link]( <href> "title" )
// ^^ skipping these spaces
pos++;
for (;pos < max; pos++) {
code = state.src.charCodeAt(pos);
if (!isSpace(code) && code !== 10) {
break;
}
}
if (pos >= max) {
return false;
}
// [link]( <href> "title" )
// ^^^^^^ parsing link destination
start = pos;
res = state.md.helpers.parseLinkDestination(state.src, pos, state.posMax);
if (res.ok) {
href = state.md.normalizeLink(res.str);
if (state.md.validateLink(href)) {
pos = res.pos;
} else {
href = "";
}
}
// [link]( <href> "title" )
// ^^ skipping these spaces
start = pos;
for (;pos < max; pos++) {
code = state.src.charCodeAt(pos);
if (!isSpace(code) && code !== 10) {
break;
}
}
// [link]( <href> "title" )
// ^^^^^^^ parsing link title
res = state.md.helpers.parseLinkTitle(state.src, pos, state.posMax);
if (pos < max && start !== pos && res.ok) {
title = res.str;
pos = res.pos;
// [link]( <href> "title" )
// ^^ skipping these spaces
for (;pos < max; pos++) {
code = state.src.charCodeAt(pos);
if (!isSpace(code) && code !== 10) {
break;
}
}
} else {
title = "";
}
if (pos >= max || state.src.charCodeAt(pos) !== 41 /* ) */) {
state.pos = oldPos;
return false;
}
pos++;
} else {
// Link reference
if (typeof state.env.references === "undefined") {
return false;
}
if (pos < max && state.src.charCodeAt(pos) === 91 /* [ */) {
start = pos + 1;
pos = state.md.helpers.parseLinkLabel(state, pos);
if (pos >= 0) {
label = state.src.slice(start, pos++);
} else {
pos = labelEnd + 1;
}
} else {
pos = labelEnd + 1;
}
// covers label === '' and label === undefined
// (collapsed reference link and shortcut reference link respectively)
if (!label) {
label = state.src.slice(labelStart, labelEnd);
}
ref = state.env.references[normalizeReference(label)];
if (!ref) {
state.pos = oldPos;
return false;
}
href = ref.href;
title = ref.title;
}
// We found the end of the link, and know for a fact it's a valid link;
// so all that's left to do is to call tokenizer.
if (!silent) {
content = state.src.slice(labelStart, labelEnd);
const tokens = [];
state.md.inline.parse(content, state.md, state.env, tokens);
const token = state.push("image", "img", 0);
const attrs = [ [ "src", href ], [ "alt", "" ] ];
token.attrs = attrs;
token.children = tokens;
token.content = content;
if (title) {
attrs.push([ "title", title ]);
}
}
state.pos = pos;
state.posMax = max;
return true;
}
// Process autolinks '<protocol:...>'
/* eslint max-len:0 */ const EMAIL_RE = /^([a-zA-Z0-9.!#$%&'*+/=?^_`{|}~-]+@[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\.[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)*)$/;
/* eslint-disable-next-line no-control-regex */ const AUTOLINK_RE = /^([a-zA-Z][a-zA-Z0-9+.-]{1,31}):([^<>\x00-\x20]*)$/;
function autolink(state, silent) {
let pos = state.pos;
if (state.src.charCodeAt(pos) !== 60 /* < */) {
return false;
}
const start = state.pos;
const max = state.posMax;
for (;;) {
if (++pos >= max) return false;
const ch = state.src.charCodeAt(pos);
if (ch === 60 /* < */) return false;
if (ch === 62 /* > */) break;
}
const url = state.src.slice(start + 1, pos);
if (AUTOLINK_RE.test(url)) {
const fullUrl = state.md.normalizeLink(url);
if (!state.md.validateLink(fullUrl)) {
return false;
}
if (!silent) {
const token_o = state.push("link_open", "a", 1);
token_o.attrs = [ [ "href", fullUrl ] ];
token_o.markup = "autolink";
token_o.info = "auto";
const token_t = state.push("text", "", 0);
token_t.content = state.md.normalizeLinkText(url);
const token_c = state.push("link_close", "a", -1);
token_c.markup = "autolink";
token_c.info = "auto";
}
state.pos += url.length + 2;
return true;
}
if (EMAIL_RE.test(url)) {
const fullUrl = state.md.normalizeLink("mailto:" + url);
if (!state.md.validateLink(fullUrl)) {
return false;
}
if (!silent) {
const token_o = state.push("link_open", "a", 1);
token_o.attrs = [ [ "href", fullUrl ] ];
token_o.markup = "autolink";
token_o.info = "auto";
const token_t = state.push("text", "", 0);
token_t.content = state.md.normalizeLinkText(url);
const token_c = state.push("link_close", "a", -1);
token_c.markup = "autolink";
token_c.info = "auto";
}
state.pos += url.length + 2;
return true;
}
return false;
}
// Process html tags
function isLinkOpen(str) {
return /^<a[>\s]/i.test(str);
}
function isLinkClose(str) {
return /^<\/a\s*>/i.test(str);
}
function isLetter(ch) {
/* eslint no-bitwise:0 */
const lc = ch | 32;
// to lower case
return lc >= 97 /* a */ && lc <= 122 /* z */;
}
function html_inline(state, silent) {
if (!state.md.options.html) {
return false;
}
// Check start
const max = state.posMax;
const pos = state.pos;
if (state.src.charCodeAt(pos) !== 60 /* < */ || pos + 2 >= max) {
return false;
}
// Quick fail on second char
const ch = state.src.charCodeAt(pos + 1);
if (ch !== 33 /* ! */ && ch !== 63 /* ? */ && ch !== 47 /* / */ && !isLetter(ch)) {
return false;
}
const match = state.src.slice(pos).match(HTML_TAG_RE);
if (!match) {
return false;
}
if (!silent) {
const token = state.push("html_inline", "", 0);
token.content = match[0];
if (isLinkOpen(token.content)) state.linkLevel++;
if (isLinkClose(token.content)) state.linkLevel--;
}
state.pos += match[0].length;
return true;
}
// Process html entity - &#123;, &#xAF;, &quot;, ...
const DIGITAL_RE = /^&#((?:x[a-f0-9]{1,6}|[0-9]{1,7}));/i;
const NAMED_RE = /^&([a-z][a-z0-9]{1,31});/i;
function entity(state, silent) {
const pos = state.pos;
const max = state.posMax;
if (state.src.charCodeAt(pos) !== 38 /* & */) return false;
if (pos + 1 >= max) return false;
const ch = state.src.charCodeAt(pos + 1);
if (ch === 35 /* # */) {
const match = state.src.slice(pos).match(DIGITAL_RE);
if (match) {
if (!silent) {
const code = match[1][0].toLowerCase() === "x" ? parseInt(match[1].slice(1), 16) : parseInt(match[1], 10);
const token = state.push("text_special", "", 0);
token.content = isValidEntityCode(code) ? fromCodePoint(code) : fromCodePoint(65533);
token.markup = match[0];
token.info = "entity";
}
state.pos += match[0].length;
return true;
}
} else {
const match = state.src.slice(pos).match(NAMED_RE);
if (match) {
const decoded = decodeHTML(match[0]);
if (decoded !== match[0]) {
if (!silent) {
const token = state.push("text_special", "", 0);
token.content = decoded;
token.markup = match[0];
token.info = "entity";
}
state.pos += match[0].length;
return true;
}
}
}
return false;
}
// For each opening emphasis-like marker find a matching closing one
function processDelimiters(delimiters) {
const openersBottom = {};
const max = delimiters.length;
if (!max) return;
// headerIdx is the first delimiter of the current (where closer is) delimiter run
let headerIdx = 0;
let lastTokenIdx = -2;
// needs any value lower than -1
const jumps = [];
for (let closerIdx = 0; closerIdx < max; closerIdx++) {
const closer = delimiters[closerIdx];
jumps.push(0);
// markers belong to same delimiter run if:
// - they have adjacent tokens
// - AND markers are the same
if (delimiters[headerIdx].marker !== closer.marker || lastTokenIdx !== closer.token - 1) {
headerIdx = closerIdx;
}
lastTokenIdx = closer.token;
// Length is only used for emphasis-specific "rule of 3",
// if it's not defined (in strikethrough or 3rd party plugins),
// we can default it to 0 to disable those checks.
closer.length = closer.length || 0;
if (!closer.close) continue;
// Previously calculated lower bounds (previous fails)
// for each marker, each delimiter length modulo 3,
// and for whether this closer can be an opener;
// https://github.com/commonmark/cmark/commit/34250e12ccebdc6372b8b49c44fab57c72443460
/* eslint-disable-next-line no-prototype-builtins */ if (!openersBottom.hasOwnProperty(closer.marker)) {
openersBottom[closer.marker] = [ -1, -1, -1, -1, -1, -1 ];
}
const minOpenerIdx = openersBottom[closer.marker][(closer.open ? 3 : 0) + closer.length % 3];
let openerIdx = headerIdx - jumps[headerIdx] - 1;
let newMinOpenerIdx = openerIdx;
for (;openerIdx > minOpenerIdx; openerIdx -= jumps[openerIdx] + 1) {
const opener = delimiters[openerIdx];
if (opener.marker !== closer.marker) continue;
if (opener.open && opener.end < 0) {
let isOddMatch = false;
// from spec:
// If one of the delimiters can both open and close emphasis, then the
// sum of the lengths of the delimiter runs containing the opening and
// closing delimiters must not be a multiple of 3 unless both lengths
// are multiples of 3.
if (opener.close || closer.open) {
if ((opener.length + closer.length) % 3 === 0) {
if (opener.length % 3 !== 0 || closer.length % 3 !== 0) {
isOddMatch = true;
}
}
}
if (!isOddMatch) {
// If previous delimiter cannot be an opener, we can safely skip
// the entire sequence in future checks. This is required to make
// sure algorithm has linear complexity (see *_*_*_*_*_... case).
const lastJump = openerIdx > 0 && !delimiters[openerIdx - 1].open ? jumps[openerIdx - 1] + 1 : 0;
jumps[closerIdx] = closerIdx - openerIdx + lastJump;
jumps[openerIdx] = lastJump;
closer.open = false;
opener.end = closerIdx;
opener.close = false;
newMinOpenerIdx = -1;
// treat next token as start of run,
// it optimizes skips in **<...>**a**<...>** pathological case
lastTokenIdx = -2;
break;
}
}
}
if (newMinOpenerIdx !== -1) {
// If match for this delimiter run failed, we want to set lower bound for
// future lookups. This is required to make sure algorithm has linear
// complexity.
// See details here:
// https://github.com/commonmark/cmark/issues/178#issuecomment-270417442
openersBottom[closer.marker][(closer.open ? 3 : 0) + (closer.length || 0) % 3] = newMinOpenerIdx;
}
}
}
function link_pairs(state) {
const tokens_meta = state.tokens_meta;
const max = state.tokens_meta.length;
processDelimiters(state.delimiters);
for (let curr = 0; curr < max; curr++) {
if (tokens_meta[curr] && tokens_meta[curr].delimiters) {
processDelimiters(tokens_meta[curr].delimiters);
}
}
}
// Clean up tokens after emphasis and strikethrough postprocessing:
// merge adjacent text nodes into one and re-calculate all token levels
// This is necessary because initially emphasis delimiter markers (*, _, ~)
// are treated as their own separate text tokens. Then emphasis rule either
// leaves them as text (needed to merge with adjacent text) or turns them
// into opening/closing tags (which messes up levels inside).
function fragments_join(state) {
let curr, last;
let level = 0;
const tokens = state.tokens;
const max = state.tokens.length;
for (curr = last = 0; curr < max; curr++) {
// re-calculate levels after emphasis/strikethrough turns some text nodes
// into opening/closing tags
if (tokens[curr].nesting < 0) level--;
// closing tag
tokens[curr].level = level;
if (tokens[curr].nesting > 0) level++;
// opening tag
if (tokens[curr].type === "text" && curr + 1 < max && tokens[curr + 1].type === "text") {
// collapse two adjacent text nodes
tokens[curr + 1].content = tokens[curr].content + tokens[curr + 1].content;
} else {
if (curr !== last) {
tokens[last] = tokens[curr];
}
last++;
}
}
if (curr !== last) {
tokens.length = last;
}
}
/** internal
* class ParserInline
*
* Tokenizes paragraph content.
**/
// Parser rules
const _rules = [ [ "text", text ], [ "linkify", linkify ], [ "newline", newline ], [ "escape", escape ], [ "backticks", backtick ], [ "strikethrough", r_strikethrough.tokenize ], [ "emphasis", r_emphasis.tokenize ], [ "link", link ], [ "image", image ], [ "autolink", autolink ], [ "html_inline", html_inline ], [ "entity", entity ] ];
// `rule2` ruleset was created specifically for emphasis/strikethrough
// post-processing and may be changed in the future.
// Don't use this for anything except pairs (plugins working with `balance_pairs`).
const _rules2 = [ [ "balance_pairs", link_pairs ], [ "strikethrough", r_strikethrough.postProcess ], [ "emphasis", r_emphasis.postProcess ],
// rules for pairs separate '**' into its own text tokens, which may be left unused,
// rule below merges unused segments back with the rest of the text
[ "fragments_join", fragments_join ] ];
/**
* new ParserInline()
**/ function ParserInline() {
/**
* ParserInline#ruler -> Ruler
*
* [[Ruler]] instance. Keep configuration of inline rules.
**/
this.ruler = new Ruler;
for (let i = 0; i < _rules.length; i++) {
this.ruler.push(_rules[i][0], _rules[i][1]);
}
/**
* ParserInline#ruler2 -> Ruler
*
* [[Ruler]] instance. Second ruler used for post-processing
* (e.g. in emphasis-like rules).
**/ this.ruler2 = new Ruler;
for (let i = 0; i < _rules2.length; i++) {
this.ruler2.push(_rules2[i][0], _rules2[i][1]);
}
}
// Skip single token by running all rules in validation mode;
// returns `true` if any rule reported success
ParserInline.prototype.skipToken = function(state) {
const pos = state.pos;
const rules = this.ruler.getRules("");
const len = rules.length;
const maxNesting = state.md.options.maxNesting;
const cache = state.cache;
if (typeof cache[pos] !== "undefined") {
state.pos = cache[pos];
return;
}
let ok = false;
if (state.level < maxNesting) {
for (let i = 0; i < len; i++) {
// Increment state.level and decrement it later to limit recursion.
// It's harmless to do here, because no tokens are created. But ideally,
// we'd need a separate private state variable for this purpose.
state.level++;
ok = rules[i](state, true);
state.level--;
if (ok) {
if (pos >= state.pos) {
throw new Error("inline rule didn't increment state.pos");
}
break;
}
}
} else {
// Too much nesting, just skip until the end of the paragraph.
// NOTE: this will cause links to behave incorrectly in the following case,
// when an amount of `[` is exactly equal to `maxNesting + 1`:
// [[[[[[[[[[[[[[[[[[[[[foo]()
// TODO: remove this workaround when CM standard will allow nested links
// (we can replace it by preventing links from being parsed in
// validation mode)
state.pos = state.posMax;
}
if (!ok) {
state.pos++;
}
cache[pos] = state.pos;
};
// Generate tokens for input range
ParserInline.prototype.tokenize = function(state) {
const rules = this.ruler.getRules("");
const len = rules.length;
const end = state.posMax;
const maxNesting = state.md.options.maxNesting;
while (state.pos < end) {
// Try all possible rules.
// On success, rule should:
// - update `state.pos`
// - update `state.tokens`
// - return true
const prevPos = state.pos;
let ok = false;
if (state.level < maxNesting) {
for (let i = 0; i < len; i++) {
ok = rules[i](state, false);
if (ok) {
if (prevPos >= state.pos) {
throw new Error("inline rule didn't increment state.pos");
}
break;
}
}
}
if (ok) {
if (state.pos >= end) {
break;
}
continue;
}
state.pending += state.src[state.pos++];
}
if (state.pending) {
state.pushPending();
}
};
/**
* ParserInline.parse(str, md, env, outTokens)
*
* Process input string and push inline tokens into `outTokens`
**/ ParserInline.prototype.parse = function(str, md, env, outTokens) {
const state = new this.State(str, md, env, outTokens);
this.tokenize(state);
const rules = this.ruler2.getRules("");
const len = rules.length;
for (let i = 0; i < len; i++) {
rules[i](state);
}
};
ParserInline.prototype.State = StateInline;
function reFactory(opts) {
const re = {};
opts = opts || {};
re.src_Any = Any.source;
re.src_Cc = Cc.source;
re.src_Z = Z.source;
re.src_P = P.source;
// \p{\Z\P\Cc\CF} (white spaces + control + format + punctuation)
re.src_ZPCc = [ re.src_Z, re.src_P, re.src_Cc ].join("|");
// \p{\Z\Cc} (white spaces + control)
re.src_ZCc = [ re.src_Z, re.src_Cc ].join("|");
// Experimental. List of chars, completely prohibited in links
// because can separate it from other part of text
const text_separators = "[><\uff5c]";
// All possible word characters (everything without punctuation, spaces & controls)
// Defined via punctuation & spaces to save space
// Should be something like \p{\L\N\S\M} (\w but without `_`)
re.src_pseudo_letter = "(?:(?!" + text_separators + "|" + re.src_ZPCc + ")" + re.src_Any + ")";
// The same as abothe but without [0-9]
// var src_pseudo_letter_non_d = '(?:(?![0-9]|' + src_ZPCc + ')' + src_Any + ')';
re.src_ip4 = "(?:(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\\.){3}(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)";
// Prohibit any of "@/[]()" in user/pass to avoid wrong domain fetch.
re.src_auth = "(?:(?:(?!" + re.src_ZCc + "|[@/\\[\\]()]).)+@)?";
re.src_port = "(?::(?:6(?:[0-4]\\d{3}|5(?:[0-4]\\d{2}|5(?:[0-2]\\d|3[0-5])))|[1-5]?\\d{1,4}))?";
re.src_host_terminator = "(?=$|" + text_separators + "|" + re.src_ZPCc + ")" + "(?!" + (opts["---"] ? "-(?!--)|" : "-|") + "_|:\\d|\\.-|\\.(?!$|" + re.src_ZPCc + "))";
re.src_path = "(?:" + "[/?#]" + "(?:" + "(?!" + re.src_ZCc + "|" + text_separators + "|[()[\\]{}.,\"'?!\\-;]).|" + "\\[(?:(?!" + re.src_ZCc + "|\\]).)*\\]|" + "\\((?:(?!" + re.src_ZCc + "|[)]).)*\\)|" + "\\{(?:(?!" + re.src_ZCc + "|[}]).)*\\}|" + '\\"(?:(?!' + re.src_ZCc + '|["]).)+\\"|' + "\\'(?:(?!" + re.src_ZCc + "|[']).)+\\'|" +
// allow `I'm_king` if no pair found
"\\'(?=" + re.src_pseudo_letter + "|[-])|" +
// google has many dots in "google search" links (#66, #81).
// github has ... in commit range links,
// Restrict to
// - english
// - percent-encoded
// - parts of file path
// - params separator
// until more examples found.
"\\.{2,}[a-zA-Z0-9%/&]|" + "\\.(?!" + re.src_ZCc + "|[.]|$)|" + (opts["---"] ? "\\-(?!--(?:[^-]|$))(?:-*)|" : "\\-+|") +
// allow `,,,` in paths
",(?!" + re.src_ZCc + "|$)|" +
// allow `;` if not followed by space-like char
";(?!" + re.src_ZCc + "|$)|" +
// allow `!!!` in paths, but not at the end
"\\!+(?!" + re.src_ZCc + "|[!]|$)|" + "\\?(?!" + re.src_ZCc + "|[?]|$)" + ")+" + "|\\/" + ")?";
// Allow anything in markdown spec, forbid quote (") at the first position
// because emails enclosed in quotes are far more common
re.src_email_name = '[\\-;:&=\\+\\$,\\.a-zA-Z0-9_][\\-;:&=\\+\\$,\\"\\.a-zA-Z0-9_]*';
re.src_xn = "xn--[a-z0-9\\-]{1,59}";
// More to read about domain names
// http://serverfault.com/questions/638260/
re.src_domain_root =
// Allow letters & digits (http://test1)
"(?:" + re.src_xn + "|" + re.src_pseudo_letter + "{1,63}" + ")";
re.src_domain = "(?:" + re.src_xn + "|" + "(?:" + re.src_pseudo_letter + ")" + "|" + "(?:" + re.src_pseudo_letter + "(?:-|" + re.src_pseudo_letter + "){0,61}" + re.src_pseudo_letter + ")" + ")";
re.src_host = "(?:" +
// Don't need IP check, because digits are already allowed in normal domain names
// src_ip4 +
// '|' +
"(?:(?:(?:" + re.src_domain + ")\\.)*" + re.src_domain /* _root */ + ")" + ")";
re.tpl_host_fuzzy = "(?:" + re.src_ip4 + "|" + "(?:(?:(?:" + re.src_domain + ")\\.)+(?:%TLDS%))" + ")";
re.tpl_host_no_ip_fuzzy = "(?:(?:(?:" + re.src_domain + ")\\.)+(?:%TLDS%))";
re.src_host_strict = re.src_host + re.src_host_terminator;
re.tpl_host_fuzzy_strict = re.tpl_host_fuzzy + re.src_host_terminator;
re.src_host_port_strict = re.src_host + re.src_port + re.src_host_terminator;
re.tpl_host_port_fuzzy_strict = re.tpl_host_fuzzy + re.src_port + re.src_host_terminator;
re.tpl_host_port_no_ip_fuzzy_strict = re.tpl_host_no_ip_fuzzy + re.src_port + re.src_host_terminator;
// Main rules
// Rude test fuzzy links by host, for quick deny
re.tpl_host_fuzzy_test = "localhost|www\\.|\\.\\d{1,3}\\.|(?:\\.(?:%TLDS%)(?:" + re.src_ZPCc + "|>|$))";
re.tpl_email_fuzzy = "(^|" + text_separators + '|"|\\(|' + re.src_ZCc + ")" + "(" + re.src_email_name + "@" + re.tpl_host_fuzzy_strict + ")";
re.tpl_link_fuzzy =
// Fuzzy link can't be prepended with .:/\- and non punctuation.
// but can start with > (markdown blockquote)
"(^|(?![.:/\\-_@])(?:[$+<=>^`|\uff5c]|" + re.src_ZPCc + "))" + "((?![$+<=>^`|\uff5c])" + re.tpl_host_port_fuzzy_strict + re.src_path + ")";
re.tpl_link_no_ip_fuzzy =
// Fuzzy link can't be prepended with .:/\- and non punctuation.
// but can start with > (markdown blockquote)
"(^|(?![.:/\\-_@])(?:[$+<=>^`|\uff5c]|" + re.src_ZPCc + "))" + "((?![$+<=>^`|\uff5c])" + re.tpl_host_port_no_ip_fuzzy_strict + re.src_path + ")";
return re;
}
// Helpers
// Merge objects
function assign(obj /* from1, from2, from3, ... */) {
const sources = Array.prototype.slice.call(arguments, 1);
sources.forEach((function(source) {
if (!source) {
return;
}
Object.keys(source).forEach((function(key) {
obj[key] = source[key];
}));
}));
return obj;
}
function _class(obj) {
return Object.prototype.toString.call(obj);
}
function isString(obj) {
return _class(obj) === "[object String]";
}
function isObject(obj) {
return _class(obj) === "[object Object]";
}
function isRegExp(obj) {
return _class(obj) === "[object RegExp]";
}
function isFunction(obj) {
return _class(obj) === "[object Function]";
}
function escapeRE(str) {
return str.replace(/[.?*+^$[\]\\(){}|-]/g, "\\$&");
}
const defaultOptions = {
fuzzyLink: true,
fuzzyEmail: true,
fuzzyIP: false
};
function isOptionsObj(obj) {
return Object.keys(obj || {}).reduce((function(acc, k) {
/* eslint-disable-next-line no-prototype-builtins */
return acc || defaultOptions.hasOwnProperty(k);
}), false);
}
const defaultSchemas = {
"http:": {
validate: function(text, pos, self) {
const tail = text.slice(pos);
if (!self.re.http) {
// compile lazily, because "host"-containing variables can change on tlds update.
self.re.http = new RegExp("^\\/\\/" + self.re.src_auth + self.re.src_host_port_strict + self.re.src_path, "i");
}
if (self.re.http.test(tail)) {
return tail.match(self.re.http)[0].length;
}
return 0;
}
},
"https:": "http:",
"ftp:": "http:",
"//": {
validate: function(text, pos, self) {
const tail = text.slice(pos);
if (!self.re.no_http) {
// compile lazily, because "host"-containing variables can change on tlds update.
self.re.no_http = new RegExp("^" + self.re.src_auth +
// Don't allow single-level domains, because of false positives like '//test'
// with code comments
"(?:localhost|(?:(?:" + self.re.src_domain + ")\\.)+" + self.re.src_domain_root + ")" + self.re.src_port + self.re.src_host_terminator + self.re.src_path, "i");
}
if (self.re.no_http.test(tail)) {
// should not be `://` & `///`, that protects from errors in protocol name
if (pos >= 3 && text[pos - 3] === ":") {
return 0;
}
if (pos >= 3 && text[pos - 3] === "/") {
return 0;
}
return tail.match(self.re.no_http)[0].length;
}
return 0;
}
},
"mailto:": {
validate: function(text, pos, self) {
const tail = text.slice(pos);
if (!self.re.mailto) {
self.re.mailto = new RegExp("^" + self.re.src_email_name + "@" + self.re.src_host_strict, "i");
}
if (self.re.mailto.test(tail)) {
return tail.match(self.re.mailto)[0].length;
}
return 0;
}
}
};
// RE pattern for 2-character tlds (autogenerated by ./support/tlds_2char_gen.js)
/* eslint-disable-next-line max-len */ const tlds_2ch_src_re = "a[cdefgilmnoqrstuwxz]|b[abdefghijmnorstvwyz]|c[acdfghiklmnoruvwxyz]|d[ejkmoz]|e[cegrstu]|f[ijkmor]|g[abdefghilmnpqrstuwy]|h[kmnrtu]|i[delmnoqrst]|j[emop]|k[eghimnprwyz]|l[abcikrstuvy]|m[acdeghklmnopqrstuvwxyz]|n[acefgilopruz]|om|p[aefghklmnrstwy]|qa|r[eosuw]|s[abcdeghijklmnortuvxyz]|t[cdfghjklmnortvwz]|u[agksyz]|v[aceginu]|w[fs]|y[et]|z[amw]";
// DON'T try to make PRs with changes. Extend TLDs with LinkifyIt.tlds() instead
const tlds_default = "biz|com|edu|gov|net|org|pro|web|xxx|aero|asia|coop|info|museum|name|shop|\u0440\u0444".split("|");
function resetScanCache(self) {
self.__index__ = -1;
self.__text_cache__ = "";
}
function createValidator(re) {
return function(text, pos) {
const tail = text.slice(pos);
if (re.test(tail)) {
return tail.match(re)[0].length;
}
return 0;
};
}
function createNormalizer() {
return function(match, self) {
self.normalize(match);
};
}
// Schemas compiler. Build regexps.
function compile(self) {
// Load & clone RE patterns.
const re = self.re = reFactory(self.__opts__);
// Define dynamic patterns
const tlds = self.__tlds__.slice();
self.onCompile();
if (!self.__tlds_replaced__) {
tlds.push(tlds_2ch_src_re);
}
tlds.push(re.src_xn);
re.src_tlds = tlds.join("|");
function untpl(tpl) {
return tpl.replace("%TLDS%", re.src_tlds);
}
re.email_fuzzy = RegExp(untpl(re.tpl_email_fuzzy), "i");
re.link_fuzzy = RegExp(untpl(re.tpl_link_fuzzy), "i");
re.link_no_ip_fuzzy = RegExp(untpl(re.tpl_link_no_ip_fuzzy), "i");
re.host_fuzzy_test = RegExp(untpl(re.tpl_host_fuzzy_test), "i");
// Compile each schema
const aliases = [];
self.__compiled__ = {};
// Reset compiled data
function schemaError(name, val) {
throw new Error('(LinkifyIt) Invalid schema "' + name + '": ' + val);
}
Object.keys(self.__schemas__).forEach((function(name) {
const val = self.__schemas__[name];
// skip disabled methods
if (val === null) {
return;
}
const compiled = {
validate: null,
link: null
};
self.__compiled__[name] = compiled;
if (isObject(val)) {
if (isRegExp(val.validate)) {
compiled.validate = createValidator(val.validate);
} else if (isFunction(val.validate)) {
compiled.validate = val.validate;
} else {
schemaError(name, val);
}
if (isFunction(val.normalize)) {
compiled.normalize = val.normalize;
} else if (!val.normalize) {
compiled.normalize = createNormalizer();
} else {
schemaError(name, val);
}
return;
}
if (isString(val)) {
aliases.push(name);
return;
}
schemaError(name, val);
}));
// Compile postponed aliases
aliases.forEach((function(alias) {
if (!self.__compiled__[self.__schemas__[alias]]) {
// Silently fail on missed schemas to avoid errons on disable.
// schemaError(alias, self.__schemas__[alias]);
return;
}
self.__compiled__[alias].validate = self.__compiled__[self.__schemas__[alias]].validate;
self.__compiled__[alias].normalize = self.__compiled__[self.__schemas__[alias]].normalize;
}));
// Fake record for guessed links
self.__compiled__[""] = {
validate: null,
normalize: createNormalizer()
};
// Build schema condition
const slist = Object.keys(self.__compiled__).filter((function(name) {
// Filter disabled & fake schemas
return name.length > 0 && self.__compiled__[name];
})).map(escapeRE).join("|");
// (?!_) cause 1.5x slowdown
self.re.schema_test = RegExp("(^|(?!_)(?:[><\uff5c]|" + re.src_ZPCc + "))(" + slist + ")", "i");
self.re.schema_search = RegExp("(^|(?!_)(?:[><\uff5c]|" + re.src_ZPCc + "))(" + slist + ")", "ig");
self.re.schema_at_start = RegExp("^" + self.re.schema_search.source, "i");
self.re.pretest = RegExp("(" + self.re.schema_test.source + ")|(" + self.re.host_fuzzy_test.source + ")|@", "i");
// Cleanup
resetScanCache(self);
}
/**
* class Match
*
* Match result. Single element of array, returned by [[LinkifyIt#match]]
**/ function Match(self, shift) {
const start = self.__index__;
const end = self.__last_index__;
const text = self.__text_cache__.slice(start, end);
/**
* Match#schema -> String
*
* Prefix (protocol) for matched string.
**/ this.schema = self.__schema__.toLowerCase();
/**
* Match#index -> Number
*
* First position of matched string.
**/ this.index = start + shift;
/**
* Match#lastIndex -> Number
*
* Next position after matched string.
**/ this.lastIndex = end + shift;
/**
* Match#raw -> String
*
* Matched string.
**/ this.raw = text;
/**
* Match#text -> String
*
* Notmalized text of matched string.
**/ this.text = text;
/**
* Match#url -> String
*
* Normalized url of matched string.
**/ this.url = text;
}
function createMatch(self, shift) {
const match = new Match(self, shift);
self.__compiled__[match.schema].normalize(match, self);
return match;
}
/**
* class LinkifyIt
**/
/**
* new LinkifyIt(schemas, options)
* - schemas (Object): Optional. Additional schemas to validate (prefix/validator)
* - options (Object): { fuzzyLink|fuzzyEmail|fuzzyIP: true|false }
*
* Creates new linkifier instance with optional additional schemas.
* Can be called without `new` keyword for convenience.
*
* By default understands:
*
* - `http(s)://...` , `ftp://...`, `mailto:...` & `//...` links
* - "fuzzy" links and emails (example.com, foo@bar.com).
*
* `schemas` is an object, where each key/value describes protocol/rule:
*
* - __key__ - link prefix (usually, protocol name with `:` at the end, `skype:`
* for example). `linkify-it` makes shure that prefix is not preceeded with
* alphanumeric char and symbols. Only whitespaces and punctuation allowed.
* - __value__ - rule to check tail after link prefix
* - _String_ - just alias to existing rule
* - _Object_
* - _validate_ - validator function (should return matched length on success),
* or `RegExp`.
* - _normalize_ - optional function to normalize text & url of matched result
* (for example, for @twitter mentions).
*
* `options`:
*
* - __fuzzyLink__ - recognige URL-s without `http(s):` prefix. Default `true`.
* - __fuzzyIP__ - allow IPs in fuzzy links above. Can conflict with some texts
* like version numbers. Default `false`.
* - __fuzzyEmail__ - recognize emails without `mailto:` prefix.
*
**/ function LinkifyIt(schemas, options) {
if (!(this instanceof LinkifyIt)) {
return new LinkifyIt(schemas, options);
}
if (!options) {
if (isOptionsObj(schemas)) {
options = schemas;
schemas = {};
}
}
this.__opts__ = assign({}, defaultOptions, options);
// Cache last tested result. Used to skip repeating steps on next `match` call.
this.__index__ = -1;
this.__last_index__ = -1;
// Next scan position
this.__schema__ = "";
this.__text_cache__ = "";
this.__schemas__ = assign({}, defaultSchemas, schemas);
this.__compiled__ = {};
this.__tlds__ = tlds_default;
this.__tlds_replaced__ = false;
this.re = {};
compile(this);
}
/** chainable
* LinkifyIt#add(schema, definition)
* - schema (String): rule name (fixed pattern prefix)
* - definition (String|RegExp|Object): schema definition
*
* Add new rule definition. See constructor description for details.
**/ LinkifyIt.prototype.add = function add(schema, definition) {
this.__schemas__[schema] = definition;
compile(this);
return this;
};
/** chainable
* LinkifyIt#set(options)
* - options (Object): { fuzzyLink|fuzzyEmail|fuzzyIP: true|false }
*
* Set recognition options for links without schema.
**/ LinkifyIt.prototype.set = function set(options) {
this.__opts__ = assign(this.__opts__, options);
return this;
};
/**
* LinkifyIt#test(text) -> Boolean
*
* Searches linkifiable pattern and returns `true` on success or `false` on fail.
**/ LinkifyIt.prototype.test = function test(text) {
// Reset scan cache
this.__text_cache__ = text;
this.__index__ = -1;
if (!text.length) {
return false;
}
let m, ml, me, len, shift, next, re, tld_pos, at_pos;
// try to scan for link with schema - that's the most simple rule
if (this.re.schema_test.test(text)) {
re = this.re.schema_search;
re.lastIndex = 0;
while ((m = re.exec(text)) !== null) {
len = this.testSchemaAt(text, m[2], re.lastIndex);
if (len) {
this.__schema__ = m[2];
this.__index__ = m.index + m[1].length;
this.__last_index__ = m.index + m[0].length + len;
break;
}
}
}
if (this.__opts__.fuzzyLink && this.__compiled__["http:"]) {
// guess schemaless links
tld_pos = text.search(this.re.host_fuzzy_test);
if (tld_pos >= 0) {
// if tld is located after found link - no need to check fuzzy pattern
if (this.__index__ < 0 || tld_pos < this.__index__) {
if ((ml = text.match(this.__opts__.fuzzyIP ? this.re.link_fuzzy : this.re.link_no_ip_fuzzy)) !== null) {
shift = ml.index + ml[1].length;
if (this.__index__ < 0 || shift < this.__index__) {
this.__schema__ = "";
this.__index__ = shift;
this.__last_index__ = ml.index + ml[0].length;
}
}
}
}
}
if (this.__opts__.fuzzyEmail && this.__compiled__["mailto:"]) {
// guess schemaless emails
at_pos = text.indexOf("@");
if (at_pos >= 0) {
// We can't skip this check, because this cases are possible:
// 192.168.1.1@gmail.com, my.in@example.com
if ((me = text.match(this.re.email_fuzzy)) !== null) {
shift = me.index + me[1].length;
next = me.index + me[0].length;
if (this.__index__ < 0 || shift < this.__index__ || shift === this.__index__ && next > this.__last_index__) {
this.__schema__ = "mailto:";
this.__index__ = shift;
this.__last_index__ = next;
}
}
}
}
return this.__index__ >= 0;
};
/**
* LinkifyIt#pretest(text) -> Boolean
*
* Very quick check, that can give false positives. Returns true if link MAY BE
* can exists. Can be used for speed optimization, when you need to check that
* link NOT exists.
**/ LinkifyIt.prototype.pretest = function pretest(text) {
return this.re.pretest.test(text);
};
/**
* LinkifyIt#testSchemaAt(text, name, position) -> Number
* - text (String): text to scan
* - name (String): rule (schema) name
* - position (Number): text offset to check from
*
* Similar to [[LinkifyIt#test]] but checks only specific protocol tail exactly
* at given position. Returns length of found pattern (0 on fail).
**/ LinkifyIt.prototype.testSchemaAt = function testSchemaAt(text, schema, pos) {
// If not supported schema check requested - terminate
if (!this.__compiled__[schema.toLowerCase()]) {
return 0;
}
return this.__compiled__[schema.toLowerCase()].validate(text, pos, this);
};
/**
* LinkifyIt#match(text) -> Array|null
*
* Returns array of found link descriptions or `null` on fail. We strongly
* recommend to use [[LinkifyIt#test]] first, for best speed.
*
* ##### Result match description
*
* - __schema__ - link schema, can be empty for fuzzy links, or `//` for
* protocol-neutral links.
* - __index__ - offset of matched text
* - __lastIndex__ - index of next char after mathch end
* - __raw__ - matched text
* - __text__ - normalized text
* - __url__ - link, generated from matched text
**/ LinkifyIt.prototype.match = function match(text) {
const result = [];
let shift = 0;
// Try to take previous element from cache, if .test() called before
if (this.__index__ >= 0 && this.__text_cache__ === text) {
result.push(createMatch(this, shift));
shift = this.__last_index__;
}
// Cut head if cache was used
let tail = shift ? text.slice(shift) : text;
// Scan string until end reached
while (this.test(tail)) {
result.push(createMatch(this, shift));
tail = tail.slice(this.__last_index__);
shift += this.__last_index__;
}
if (result.length) {
return result;
}
return null;
};
/**
* LinkifyIt#matchAtStart(text) -> Match|null
*
* Returns fully-formed (not fuzzy) link if it starts at the beginning
* of the string, and null otherwise.
**/ LinkifyIt.prototype.matchAtStart = function matchAtStart(text) {
// Reset scan cache
this.__text_cache__ = text;
this.__index__ = -1;
if (!text.length) return null;
const m = this.re.schema_at_start.exec(text);
if (!m) return null;
const len = this.testSchemaAt(text, m[2], m[0].length);
if (!len) return null;
this.__schema__ = m[2];
this.__index__ = m.index + m[1].length;
this.__last_index__ = m.index + m[0].length + len;
return createMatch(this, 0);
};
/** chainable
* LinkifyIt#tlds(list [, keepOld]) -> this
* - list (Array): list of tlds
* - keepOld (Boolean): merge with current list if `true` (`false` by default)
*
* Load (or merge) new tlds list. Those are user for fuzzy links (without prefix)
* to avoid false positives. By default this algorythm used:
*
* - hostname with any 2-letter root zones are ok.
* - biz|com|edu|gov|net|org|pro|web|xxx|aero|asia|coop|info|museum|name|shop|рф
* are ok.
* - encoded (`xn--...`) root zones are ok.
*
* If list is replaced, then exact match for 2-chars root zones will be checked.
**/ LinkifyIt.prototype.tlds = function tlds(list, keepOld) {
list = Array.isArray(list) ? list : [ list ];
if (!keepOld) {
this.__tlds__ = list.slice();
this.__tlds_replaced__ = true;
compile(this);
return this;
}
this.__tlds__ = this.__tlds__.concat(list).sort().filter((function(el, idx, arr) {
return el !== arr[idx - 1];
})).reverse();
compile(this);
return this;
};
/**
* LinkifyIt#normalize(match)
*
* Default normalizer (if schema does not define it's own).
**/ LinkifyIt.prototype.normalize = function normalize(match) {
// Do minimal possible changes by default. Need to collect feedback prior
// to move forward https://github.com/markdown-it/linkify-it/issues/1
if (!match.schema) {
match.url = "http://" + match.url;
}
if (match.schema === "mailto:" && !/^mailto:/i.test(match.url)) {
match.url = "mailto:" + match.url;
}
};
/**
* LinkifyIt#onCompile()
*
* Override to modify basic RegExp-s.
**/ LinkifyIt.prototype.onCompile = function onCompile() {};
/** Highest positive signed 32-bit float value */ const maxInt = 2147483647;
// aka. 0x7FFFFFFF or 2^31-1
/** Bootstring parameters */ const base = 36;
const tMin = 1;
const tMax = 26;
const skew = 38;
const damp = 700;
const initialBias = 72;
const initialN = 128;
// 0x80
const delimiter = "-";
// '\x2D'
/** Regular expressions */ const regexPunycode = /^xn--/;
const regexNonASCII = /[^\0-\x7F]/;
// Note: U+007F DEL is excluded too.
const regexSeparators = /[\x2E\u3002\uFF0E\uFF61]/g;
// RFC 3490 separators
/** Error messages */ const errors = {
overflow: "Overflow: input needs wider integers to process",
"not-basic": "Illegal input >= 0x80 (not a basic code point)",
"invalid-input": "Invalid input"
};
/** Convenience shortcuts */ const baseMinusTMin = base - tMin;
const floor = Math.floor;
const stringFromCharCode = String.fromCharCode;
/*--------------------------------------------------------------------------*/
/**
* A generic error utility function.
* @private
* @param {String} type The error type.
* @returns {Error} Throws a `RangeError` with the applicable error message.
*/ function error(type) {
throw new RangeError(errors[type]);
}
/**
* A generic `Array#map` utility function.
* @private
* @param {Array} array The array to iterate over.
* @param {Function} callback The function that gets called for every array
* item.
* @returns {Array} A new array of values returned by the callback function.
*/ function map(array, callback) {
const result = [];
let length = array.length;
while (length--) {
result[length] = callback(array[length]);
}
return result;
}
/**
* A simple `Array#map`-like wrapper to work with domain name strings or email
* addresses.
* @private
* @param {String} domain The domain name or email address.
* @param {Function} callback The function that gets called for every
* character.
* @returns {String} A new string of characters returned by the callback
* function.
*/ function mapDomain(domain, callback) {
const parts = domain.split("@");
let result = "";
if (parts.length > 1) {
// In email addresses, only the domain name should be punycoded. Leave
// the local part (i.e. everything up to `@`) intact.
result = parts[0] + "@";
domain = parts[1];
}
// Avoid `split(regex)` for IE8 compatibility. See #17.
domain = domain.replace(regexSeparators, ".");
const labels = domain.split(".");
const encoded = map(labels, callback).join(".");
return result + encoded;
}
/**
* Creates an array containing the numeric code points of each Unicode
* character in the string. While JavaScript uses UCS-2 internally,
* this function will convert a pair of surrogate halves (each of which
* UCS-2 exposes as separate characters) into a single code point,
* matching UTF-16.
* @see `punycode.ucs2.encode`
* @see <https://mathiasbynens.be/notes/javascript-encoding>
* @memberOf punycode.ucs2
* @name decode
* @param {String} string The Unicode input string (UCS-2).
* @returns {Array} The new array of code points.
*/ function ucs2decode(string) {
const output = [];
let counter = 0;
const length = string.length;
while (counter < length) {
const value = string.charCodeAt(counter++);
if (value >= 55296 && value <= 56319 && counter < length) {
// It's a high surrogate, and there is a next character.
const extra = string.charCodeAt(counter++);
if ((extra & 64512) == 56320) {
// Low surrogate.
output.push(((value & 1023) << 10) + (extra & 1023) + 65536);
} else {
// It's an unmatched surrogate; only append this code unit, in case the
// next code unit is the high surrogate of a surrogate pair.
output.push(value);
counter--;
}
} else {
output.push(value);
}
}
return output;
}
/**
* Creates a string based on an array of numeric code points.
* @see `punycode.ucs2.decode`
* @memberOf punycode.ucs2
* @name encode
* @param {Array} codePoints The array of numeric code points.
* @returns {String} The new Unicode string (UCS-2).
*/ const ucs2encode = codePoints => String.fromCodePoint(...codePoints)
/**
* Converts a basic code point into a digit/integer.
* @see `digitToBasic()`
* @private
* @param {Number} codePoint The basic numeric code point value.
* @returns {Number} The numeric value of a basic code point (for use in
* representing integers) in the range `0` to `base - 1`, or `base` if
* the code point does not represent a value.
*/;
const basicToDigit = function(codePoint) {
if (codePoint >= 48 && codePoint < 58) {
return 26 + (codePoint - 48);
}
if (codePoint >= 65 && codePoint < 91) {
return codePoint - 65;
}
if (codePoint >= 97 && codePoint < 123) {
return codePoint - 97;
}
return base;
};
/**
* Converts a digit/integer into a basic code point.
* @see `basicToDigit()`
* @private
* @param {Number} digit The numeric value of a basic code point.
* @returns {Number} The basic code point whose value (when used for
* representing integers) is `digit`, which needs to be in the range
* `0` to `base - 1`. If `flag` is non-zero, the uppercase form is
* used; else, the lowercase form is used. The behavior is undefined
* if `flag` is non-zero and `digit` has no uppercase form.
*/ const digitToBasic = function(digit, flag) {
// 0..25 map to ASCII a..z or A..Z
// 26..35 map to ASCII 0..9
return digit + 22 + 75 * (digit < 26) - ((flag != 0) << 5);
};
/**
* Bias adaptation function as per section 3.4 of RFC 3492.
* https://tools.ietf.org/html/rfc3492#section-3.4
* @private
*/ const adapt = function(delta, numPoints, firstTime) {
let k = 0;
delta = firstTime ? floor(delta / damp) : delta >> 1;
delta += floor(delta / numPoints);
for (;delta > baseMinusTMin * tMax >> 1; k += base) {
delta = floor(delta / baseMinusTMin);
}
return floor(k + (baseMinusTMin + 1) * delta / (delta + skew));
};
/**
* Converts a Punycode string of ASCII-only symbols to a string of Unicode
* symbols.
* @memberOf punycode
* @param {String} input The Punycode string of ASCII-only symbols.
* @returns {String} The resulting string of Unicode symbols.
*/ const decode = function(input) {
// Don't use UCS-2.
const output = [];
const inputLength = input.length;
let i = 0;
let n = initialN;
let bias = initialBias;
// Handle the basic code points: let `basic` be the number of input code
// points before the last delimiter, or `0` if there is none, then copy
// the first basic code points to the output.
let basic = input.lastIndexOf(delimiter);
if (basic < 0) {
basic = 0;
}
for (let j = 0; j < basic; ++j) {
// if it's not a basic code point
if (input.charCodeAt(j) >= 128) {
error("not-basic");
}
output.push(input.charCodeAt(j));
}
// Main decoding loop: start just after the last delimiter if any basic code
// points were copied; start at the beginning otherwise.
for (let index = basic > 0 ? basic + 1 : 0; index < inputLength; ) {
// `index` is the index of the next character to be consumed.
// Decode a generalized variable-length integer into `delta`,
// which gets added to `i`. The overflow checking is easier
// if we increase `i` as we go, then subtract off its starting
// value at the end to obtain `delta`.
const oldi = i;
for (let w = 1, k = base; ;k += base) {
if (index >= inputLength) {
error("invalid-input");
}
const digit = basicToDigit(input.charCodeAt(index++));
if (digit >= base) {
error("invalid-input");
}
if (digit > floor((maxInt - i) / w)) {
error("overflow");
}
i += digit * w;
const t = k <= bias ? tMin : k >= bias + tMax ? tMax : k - bias;
if (digit < t) {
break;
}
const baseMinusT = base - t;
if (w > floor(maxInt / baseMinusT)) {
error("overflow");
}
w *= baseMinusT;
}
const out = output.length + 1;
bias = adapt(i - oldi, out, oldi == 0);
// `i` was supposed to wrap around from `out` to `0`,
// incrementing `n` each time, so we'll fix that now:
if (floor(i / out) > maxInt - n) {
error("overflow");
}
n += floor(i / out);
i %= out;
// Insert `n` at position `i` of the output.
output.splice(i++, 0, n);
}
return String.fromCodePoint(...output);
};
/**
* Converts a string of Unicode symbols (e.g. a domain name label) to a
* Punycode string of ASCII-only symbols.
* @memberOf punycode
* @param {String} input The string of Unicode symbols.
* @returns {String} The resulting Punycode string of ASCII-only symbols.
*/ const encode = function(input) {
const output = [];
// Convert the input in UCS-2 to an array of Unicode code points.
input = ucs2decode(input);
// Cache the length.
const inputLength = input.length;
// Initialize the state.
let n = initialN;
let delta = 0;
let bias = initialBias;
// Handle the basic code points.
for (const currentValue of input) {
if (currentValue < 128) {
output.push(stringFromCharCode(currentValue));
}
}
const basicLength = output.length;
let handledCPCount = basicLength;
// `handledCPCount` is the number of code points that have been handled;
// `basicLength` is the number of basic code points.
// Finish the basic string with a delimiter unless it's empty.
if (basicLength) {
output.push(delimiter);
}
// Main encoding loop:
while (handledCPCount < inputLength) {
// All non-basic code points < n have been handled already. Find the next
// larger one:
let m = maxInt;
for (const currentValue of input) {
if (currentValue >= n && currentValue < m) {
m = currentValue;
}
}
// Increase `delta` enough to advance the decoder's <n,i> state to <m,0>,
// but guard against overflow.
const handledCPCountPlusOne = handledCPCount + 1;
if (m - n > floor((maxInt - delta) / handledCPCountPlusOne)) {
error("overflow");
}
delta += (m - n) * handledCPCountPlusOne;
n = m;
for (const currentValue of input) {
if (currentValue < n && ++delta > maxInt) {
error("overflow");
}
if (currentValue === n) {
// Represent delta as a generalized variable-length integer.
let q = delta;
for (let k = base; ;k += base) {
const t = k <= bias ? tMin : k >= bias + tMax ? tMax : k - bias;
if (q < t) {
break;
}
const qMinusT = q - t;
const baseMinusT = base - t;
output.push(stringFromCharCode(digitToBasic(t + qMinusT % baseMinusT, 0)));
q = floor(qMinusT / baseMinusT);
}
output.push(stringFromCharCode(digitToBasic(q, 0)));
bias = adapt(delta, handledCPCountPlusOne, handledCPCount === basicLength);
delta = 0;
++handledCPCount;
}
}
++delta;
++n;
}
return output.join("");
};
/**
* Converts a Punycode string representing a domain name or an email address
* to Unicode. Only the Punycoded parts of the input will be converted, i.e.
* it doesn't matter if you call it on a string that has already been
* converted to Unicode.
* @memberOf punycode
* @param {String} input The Punycoded domain name or email address to
* convert to Unicode.
* @returns {String} The Unicode representation of the given Punycode
* string.
*/ const toUnicode = function(input) {
return mapDomain(input, (function(string) {
return regexPunycode.test(string) ? decode(string.slice(4).toLowerCase()) : string;
}));
};
/**
* Converts a Unicode string representing a domain name or an email address to
* Punycode. Only the non-ASCII parts of the domain name will be converted,
* i.e. it doesn't matter if you call it with a domain that's already in
* ASCII.
* @memberOf punycode
* @param {String} input The domain name or email address to convert, as a
* Unicode string.
* @returns {String} The Punycode representation of the given domain name or
* email address.
*/ const toASCII = function(input) {
return mapDomain(input, (function(string) {
return regexNonASCII.test(string) ? "xn--" + encode(string) : string;
}));
};
/*--------------------------------------------------------------------------*/
/** Define the public API */ const punycode = {
/**
* A string representing the current Punycode.js version number.
* @memberOf punycode
* @type String
*/
version: "2.3.1",
/**
* An object of methods to convert from JavaScript's internal character
* representation (UCS-2) to Unicode code points, and back.
* @see <https://mathiasbynens.be/notes/javascript-encoding>
* @memberOf punycode
* @type Object
*/
ucs2: {
decode: ucs2decode,
encode: ucs2encode
},
decode: decode,
encode: encode,
toASCII: toASCII,
toUnicode: toUnicode
};
// markdown-it default options
var cfg_default = {
options: {
// Enable HTML tags in source
html: false,
// Use '/' to close single tags (<br />)
xhtmlOut: false,
// Convert '\n' in paragraphs into <br>
breaks: false,
// CSS language prefix for fenced blocks
langPrefix: "language-",
// autoconvert URL-like texts to links
linkify: false,
// Enable some language-neutral replacements + quotes beautification
typographer: false,
// Double + single quotes replacement pairs, when typographer enabled,
// and smartquotes on. Could be either a String or an Array.
// For example, you can use '«»„“' for Russian, '„“‚‘' for German,
// and ['«\xA0', '\xA0»', '‹\xA0', '\xA0›'] for French (including nbsp).
quotes: "\u201c\u201d\u2018\u2019",
/* “”‘’ */
// Highlighter function. Should return escaped HTML,
// or '' if the source string is not changed and should be escaped externaly.
// If result starts with <pre... internal wrapper is skipped.
// function (/*str, lang*/) { return ''; }
highlight: null,
// Internal protection, recursion limit
maxNesting: 100
},
components: {
core: {},
block: {},
inline: {}
}
};
// "Zero" preset, with nothing enabled. Useful for manual configuring of simple
// modes. For example, to parse bold/italic only.
var cfg_zero = {
options: {
// Enable HTML tags in source
html: false,
// Use '/' to close single tags (<br />)
xhtmlOut: false,
// Convert '\n' in paragraphs into <br>
breaks: false,
// CSS language prefix for fenced blocks
langPrefix: "language-",
// autoconvert URL-like texts to links
linkify: false,
// Enable some language-neutral replacements + quotes beautification
typographer: false,
// Double + single quotes replacement pairs, when typographer enabled,
// and smartquotes on. Could be either a String or an Array.
// For example, you can use '«»„“' for Russian, '„“‚‘' for German,
// and ['«\xA0', '\xA0»', '‹\xA0', '\xA0›'] for French (including nbsp).
quotes: "\u201c\u201d\u2018\u2019",
/* “”‘’ */
// Highlighter function. Should return escaped HTML,
// or '' if the source string is not changed and should be escaped externaly.
// If result starts with <pre... internal wrapper is skipped.
// function (/*str, lang*/) { return ''; }
highlight: null,
// Internal protection, recursion limit
maxNesting: 20
},
components: {
core: {
rules: [ "normalize", "block", "inline", "text_join" ]
},
block: {
rules: [ "paragraph" ]
},
inline: {
rules: [ "text" ],
rules2: [ "balance_pairs", "fragments_join" ]
}
}
};
// Commonmark default options
var cfg_commonmark = {
options: {
// Enable HTML tags in source
html: true,
// Use '/' to close single tags (<br />)
xhtmlOut: true,
// Convert '\n' in paragraphs into <br>
breaks: false,
// CSS language prefix for fenced blocks
langPrefix: "language-",
// autoconvert URL-like texts to links
linkify: false,
// Enable some language-neutral replacements + quotes beautification
typographer: false,
// Double + single quotes replacement pairs, when typographer enabled,
// and smartquotes on. Could be either a String or an Array.
// For example, you can use '«»„“' for Russian, '„“‚‘' for German,
// and ['«\xA0', '\xA0»', '‹\xA0', '\xA0›'] for French (including nbsp).
quotes: "\u201c\u201d\u2018\u2019",
/* “”‘’ */
// Highlighter function. Should return escaped HTML,
// or '' if the source string is not changed and should be escaped externaly.
// If result starts with <pre... internal wrapper is skipped.
// function (/*str, lang*/) { return ''; }
highlight: null,
// Internal protection, recursion limit
maxNesting: 20
},
components: {
core: {
rules: [ "normalize", "block", "inline", "text_join" ]
},
block: {
rules: [ "blockquote", "code", "fence", "heading", "hr", "html_block", "lheading", "list", "reference", "paragraph" ]
},
inline: {
rules: [ "autolink", "backticks", "emphasis", "entity", "escape", "html_inline", "image", "link", "newline", "text" ],
rules2: [ "balance_pairs", "emphasis", "fragments_join" ]
}
}
};
// Main parser class
const config = {
default: cfg_default,
zero: cfg_zero,
commonmark: cfg_commonmark
};
// This validator can prohibit more than really needed to prevent XSS. It's a
// tradeoff to keep code simple and to be secure by default.
// If you need different setup - override validator method as you wish. Or
// replace it with dummy function and use external sanitizer.
const BAD_PROTO_RE = /^(vbscript|javascript|file|data):/;
const GOOD_DATA_RE = /^data:image\/(gif|png|jpeg|webp);/;
function validateLink(url) {
// url should be normalized at this point, and existing entities are decoded
const str = url.trim().toLowerCase();
return BAD_PROTO_RE.test(str) ? GOOD_DATA_RE.test(str) : true;
}
const RECODE_HOSTNAME_FOR = [ "http:", "https:", "mailto:" ];
function normalizeLink(url) {
const parsed = urlParse(url, true);
if (parsed.hostname) {
// Encode hostnames in urls like:
// `http://host/`, `https://host/`, `mailto:user@host`, `//host/`
// We don't encode unknown schemas, because it's likely that we encode
// something we shouldn't (e.g. `skype:name` treated as `skype:host`)
if (!parsed.protocol || RECODE_HOSTNAME_FOR.indexOf(parsed.protocol) >= 0) {
try {
parsed.hostname = punycode.toASCII(parsed.hostname);
} catch (er) {}
}
}
return encode$1(format(parsed));
}
function normalizeLinkText(url) {
const parsed = urlParse(url, true);
if (parsed.hostname) {
// Encode hostnames in urls like:
// `http://host/`, `https://host/`, `mailto:user@host`, `//host/`
// We don't encode unknown schemas, because it's likely that we encode
// something we shouldn't (e.g. `skype:name` treated as `skype:host`)
if (!parsed.protocol || RECODE_HOSTNAME_FOR.indexOf(parsed.protocol) >= 0) {
try {
parsed.hostname = punycode.toUnicode(parsed.hostname);
} catch (er) {}
}
}
// add '%' to exclude list because of https://github.com/markdown-it/markdown-it/issues/720
return decode$1(format(parsed), decode$1.defaultChars + "%");
}
/**
* class MarkdownIt
*
* Main parser/renderer class.
*
* ##### Usage
*
* ```javascript
* // node.js, "classic" way:
* var MarkdownIt = require('markdown-it'),
* md = new MarkdownIt();
* var result = md.render('# markdown-it rulezz!');
*
* // node.js, the same, but with sugar:
* var md = require('markdown-it')();
* var result = md.render('# markdown-it rulezz!');
*
* // browser without AMD, added to "window" on script load
* // Note, there are no dash.
* var md = window.markdownit();
* var result = md.render('# markdown-it rulezz!');
* ```
*
* Single line rendering, without paragraph wrap:
*
* ```javascript
* var md = require('markdown-it')();
* var result = md.renderInline('__markdown-it__ rulezz!');
* ```
**/
/**
* new MarkdownIt([presetName, options])
* - presetName (String): optional, `commonmark` / `zero`
* - options (Object)
*
* Creates parser instanse with given config. Can be called without `new`.
*
* ##### presetName
*
* MarkdownIt provides named presets as a convenience to quickly
* enable/disable active syntax rules and options for common use cases.
*
* - ["commonmark"](https://github.com/markdown-it/markdown-it/blob/master/lib/presets/commonmark.mjs) -
* configures parser to strict [CommonMark](http://commonmark.org/) mode.
* - [default](https://github.com/markdown-it/markdown-it/blob/master/lib/presets/default.mjs) -
* similar to GFM, used when no preset name given. Enables all available rules,
* but still without html, typographer & autolinker.
* - ["zero"](https://github.com/markdown-it/markdown-it/blob/master/lib/presets/zero.mjs) -
* all rules disabled. Useful to quickly setup your config via `.enable()`.
* For example, when you need only `bold` and `italic` markup and nothing else.
*
* ##### options:
*
* - __html__ - `false`. Set `true` to enable HTML tags in source. Be careful!
* That's not safe! You may need external sanitizer to protect output from XSS.
* It's better to extend features via plugins, instead of enabling HTML.
* - __xhtmlOut__ - `false`. Set `true` to add '/' when closing single tags
* (`<br />`). This is needed only for full CommonMark compatibility. In real
* world you will need HTML output.
* - __breaks__ - `false`. Set `true` to convert `\n` in paragraphs into `<br>`.
* - __langPrefix__ - `language-`. CSS language class prefix for fenced blocks.
* Can be useful for external highlighters.
* - __linkify__ - `false`. Set `true` to autoconvert URL-like text to links.
* - __typographer__ - `false`. Set `true` to enable [some language-neutral
* replacement](https://github.com/markdown-it/markdown-it/blob/master/lib/rules_core/replacements.mjs) +
* quotes beautification (smartquotes).
* - __quotes__ - `“”‘’`, String or Array. Double + single quotes replacement
* pairs, when typographer enabled and smartquotes on. For example, you can
* use `'«»„“'` for Russian, `'„“‚‘'` for German, and
* `['«\xA0', '\xA0»', '‹\xA0', '\xA0›']` for French (including nbsp).
* - __highlight__ - `null`. Highlighter function for fenced code blocks.
* Highlighter `function (str, lang)` should return escaped HTML. It can also
* return empty string if the source was not changed and should be escaped
* externaly. If result starts with <pre... internal wrapper is skipped.
*
* ##### Example
*
* ```javascript
* // commonmark mode
* var md = require('markdown-it')('commonmark');
*
* // default mode
* var md = require('markdown-it')();
*
* // enable everything
* var md = require('markdown-it')({
* html: true,
* linkify: true,
* typographer: true
* });
* ```
*
* ##### Syntax highlighting
*
* ```js
* var hljs = require('highlight.js') // https://highlightjs.org/
*
* var md = require('markdown-it')({
* highlight: function (str, lang) {
* if (lang && hljs.getLanguage(lang)) {
* try {
* return hljs.highlight(str, { language: lang, ignoreIllegals: true }).value;
* } catch (__) {}
* }
*
* return ''; // use external default escaping
* }
* });
* ```
*
* Or with full wrapper override (if you need assign class to `<pre>` or `<code>`):
*
* ```javascript
* var hljs = require('highlight.js') // https://highlightjs.org/
*
* // Actual default values
* var md = require('markdown-it')({
* highlight: function (str, lang) {
* if (lang && hljs.getLanguage(lang)) {
* try {
* return '<pre><code class="hljs">' +
* hljs.highlight(str, { language: lang, ignoreIllegals: true }).value +
* '</code></pre>';
* } catch (__) {}
* }
*
* return '<pre><code class="hljs">' + md.utils.escapeHtml(str) + '</code></pre>';
* }
* });
* ```
*
**/ function MarkdownIt(presetName, options) {
if (!(this instanceof MarkdownIt)) {
return new MarkdownIt(presetName, options);
}
if (!options) {
if (!isString$1(presetName)) {
options = presetName || {};
presetName = "default";
}
}
/**
* MarkdownIt#inline -> ParserInline
*
* Instance of [[ParserInline]]. You may need it to add new rules when
* writing plugins. For simple rules control use [[MarkdownIt.disable]] and
* [[MarkdownIt.enable]].
**/ this.inline = new ParserInline;
/**
* MarkdownIt#block -> ParserBlock
*
* Instance of [[ParserBlock]]. You may need it to add new rules when
* writing plugins. For simple rules control use [[MarkdownIt.disable]] and
* [[MarkdownIt.enable]].
**/ this.block = new ParserBlock;
/**
* MarkdownIt#core -> Core
*
* Instance of [[Core]] chain executor. You may need it to add new rules when
* writing plugins. For simple rules control use [[MarkdownIt.disable]] and
* [[MarkdownIt.enable]].
**/ this.core = new Core;
/**
* MarkdownIt#renderer -> Renderer
*
* Instance of [[Renderer]]. Use it to modify output look. Or to add rendering
* rules for new token types, generated by plugins.
*
* ##### Example
*
* ```javascript
* var md = require('markdown-it')();
*
* function myToken(tokens, idx, options, env, self) {
* //...
* return result;
* };
*
* md.renderer.rules['my_token'] = myToken
* ```
*
* See [[Renderer]] docs and [source code](https://github.com/markdown-it/markdown-it/blob/master/lib/renderer.mjs).
**/ this.renderer = new Renderer;
/**
* MarkdownIt#linkify -> LinkifyIt
*
* [linkify-it](https://github.com/markdown-it/linkify-it) instance.
* Used by [linkify](https://github.com/markdown-it/markdown-it/blob/master/lib/rules_core/linkify.mjs)
* rule.
**/ this.linkify = new LinkifyIt;
/**
* MarkdownIt#validateLink(url) -> Boolean
*
* Link validation function. CommonMark allows too much in links. By default
* we disable `javascript:`, `vbscript:`, `file:` schemas, and almost all `data:...` schemas
* except some embedded image types.
*
* You can change this behaviour:
*
* ```javascript
* var md = require('markdown-it')();
* // enable everything
* md.validateLink = function () { return true; }
* ```
**/ this.validateLink = validateLink;
/**
* MarkdownIt#normalizeLink(url) -> String
*
* Function used to encode link url to a machine-readable format,
* which includes url-encoding, punycode, etc.
**/ this.normalizeLink = normalizeLink;
/**
* MarkdownIt#normalizeLinkText(url) -> String
*
* Function used to decode link url to a human-readable format`
**/ this.normalizeLinkText = normalizeLinkText;
// Expose utils & helpers for easy acces from plugins
/**
* MarkdownIt#utils -> utils
*
* Assorted utility functions, useful to write plugins. See details
* [here](https://github.com/markdown-it/markdown-it/blob/master/lib/common/utils.mjs).
**/ this.utils = utils;
/**
* MarkdownIt#helpers -> helpers
*
* Link components parser functions, useful to write plugins. See details
* [here](https://github.com/markdown-it/markdown-it/blob/master/lib/helpers).
**/ this.helpers = assign$1({}, helpers);
this.options = {};
this.configure(presetName);
if (options) {
this.set(options);
}
}
/** chainable
* MarkdownIt.set(options)
*
* Set parser options (in the same format as in constructor). Probably, you
* will never need it, but you can change options after constructor call.
*
* ##### Example
*
* ```javascript
* var md = require('markdown-it')()
* .set({ html: true, breaks: true })
* .set({ typographer, true });
* ```
*
* __Note:__ To achieve the best possible performance, don't modify a
* `markdown-it` instance options on the fly. If you need multiple configurations
* it's best to create multiple instances and initialize each with separate
* config.
**/ MarkdownIt.prototype.set = function(options) {
assign$1(this.options, options);
return this;
};
/** chainable, internal
* MarkdownIt.configure(presets)
*
* Batch load of all options and compenent settings. This is internal method,
* and you probably will not need it. But if you will - see available presets
* and data structure [here](https://github.com/markdown-it/markdown-it/tree/master/lib/presets)
*
* We strongly recommend to use presets instead of direct config loads. That
* will give better compatibility with next versions.
**/ MarkdownIt.prototype.configure = function(presets) {
const self = this;
if (isString$1(presets)) {
const presetName = presets;
presets = config[presetName];
if (!presets) {
throw new Error('Wrong `markdown-it` preset "' + presetName + '", check name');
}
}
if (!presets) {
throw new Error("Wrong `markdown-it` preset, can't be empty");
}
if (presets.options) {
self.set(presets.options);
}
if (presets.components) {
Object.keys(presets.components).forEach((function(name) {
if (presets.components[name].rules) {
self[name].ruler.enableOnly(presets.components[name].rules);
}
if (presets.components[name].rules2) {
self[name].ruler2.enableOnly(presets.components[name].rules2);
}
}));
}
return this;
};
/** chainable
* MarkdownIt.enable(list, ignoreInvalid)
* - list (String|Array): rule name or list of rule names to enable
* - ignoreInvalid (Boolean): set `true` to ignore errors when rule not found.
*
* Enable list or rules. It will automatically find appropriate components,
* containing rules with given names. If rule not found, and `ignoreInvalid`
* not set - throws exception.
*
* ##### Example
*
* ```javascript
* var md = require('markdown-it')()
* .enable(['sub', 'sup'])
* .disable('smartquotes');
* ```
**/ MarkdownIt.prototype.enable = function(list, ignoreInvalid) {
let result = [];
if (!Array.isArray(list)) {
list = [ list ];
}
[ "core", "block", "inline" ].forEach((function(chain) {
result = result.concat(this[chain].ruler.enable(list, true));
}), this);
result = result.concat(this.inline.ruler2.enable(list, true));
const missed = list.filter((function(name) {
return result.indexOf(name) < 0;
}));
if (missed.length && !ignoreInvalid) {
throw new Error("MarkdownIt. Failed to enable unknown rule(s): " + missed);
}
return this;
};
/** chainable
* MarkdownIt.disable(list, ignoreInvalid)
* - list (String|Array): rule name or list of rule names to disable.
* - ignoreInvalid (Boolean): set `true` to ignore errors when rule not found.
*
* The same as [[MarkdownIt.enable]], but turn specified rules off.
**/ MarkdownIt.prototype.disable = function(list, ignoreInvalid) {
let result = [];
if (!Array.isArray(list)) {
list = [ list ];
}
[ "core", "block", "inline" ].forEach((function(chain) {
result = result.concat(this[chain].ruler.disable(list, true));
}), this);
result = result.concat(this.inline.ruler2.disable(list, true));
const missed = list.filter((function(name) {
return result.indexOf(name) < 0;
}));
if (missed.length && !ignoreInvalid) {
throw new Error("MarkdownIt. Failed to disable unknown rule(s): " + missed);
}
return this;
};
/** chainable
* MarkdownIt.use(plugin, params)
*
* Load specified plugin with given params into current parser instance.
* It's just a sugar to call `plugin(md, params)` with curring.
*
* ##### Example
*
* ```javascript
* var iterator = require('markdown-it-for-inline');
* var md = require('markdown-it')()
* .use(iterator, 'foo_replace', 'text', function (tokens, idx) {
* tokens[idx].content = tokens[idx].content.replace(/foo/g, 'bar');
* });
* ```
**/ MarkdownIt.prototype.use = function(plugin /*, params, ... */) {
const args = [ this ].concat(Array.prototype.slice.call(arguments, 1));
plugin.apply(plugin, args);
return this;
};
/** internal
* MarkdownIt.parse(src, env) -> Array
* - src (String): source string
* - env (Object): environment sandbox
*
* Parse input string and return list of block tokens (special token type
* "inline" will contain list of inline tokens). You should not call this
* method directly, until you write custom renderer (for example, to produce
* AST).
*
* `env` is used to pass data between "distributed" rules and return additional
* metadata like reference info, needed for the renderer. It also can be used to
* inject data in specific cases. Usually, you will be ok to pass `{}`,
* and then pass updated object to renderer.
**/ MarkdownIt.prototype.parse = function(src, env) {
if (typeof src !== "string") {
throw new Error("Input data should be a String");
}
const state = new this.core.State(src, this, env);
this.core.process(state);
return state.tokens;
};
/**
* MarkdownIt.render(src [, env]) -> String
* - src (String): source string
* - env (Object): environment sandbox
*
* Render markdown string into html. It does all magic for you :).
*
* `env` can be used to inject additional metadata (`{}` by default).
* But you will not need it with high probability. See also comment
* in [[MarkdownIt.parse]].
**/ MarkdownIt.prototype.render = function(src, env) {
env = env || {};
return this.renderer.render(this.parse(src, env), this.options, env);
};
/** internal
* MarkdownIt.parseInline(src, env) -> Array
* - src (String): source string
* - env (Object): environment sandbox
*
* The same as [[MarkdownIt.parse]] but skip all block rules. It returns the
* block tokens list with the single `inline` element, containing parsed inline
* tokens in `children` property. Also updates `env` object.
**/ MarkdownIt.prototype.parseInline = function(src, env) {
const state = new this.core.State(src, this, env);
state.inlineMode = true;
this.core.process(state);
return state.tokens;
};
/**
* MarkdownIt.renderInline(src [, env]) -> String
* - src (String): source string
* - env (Object): environment sandbox
*
* Similar to [[MarkdownIt.render]] but for single paragraph content. Result
* will NOT be wrapped into `<p>` tags.
**/ MarkdownIt.prototype.renderInline = function(src, env) {
env = env || {};
return this.renderer.render(this.parseInline(src, env), this.options, env);
};
return MarkdownIt;
}));
Reference in a new issue Copy permalink