Merge pull request #972 from thelounge/bonuspunkt/message-parser

New IRC message parser
This commit is contained in:
Jérémie Astori 2017-04-21 13:40:16 -04:00 committed by GitHub
commit 6c0cf7cfcc
14 changed files with 1414 additions and 121 deletions

View file

@ -0,0 +1,12 @@
"use strict";
// Return true if any section of "a" or "b" parts (defined by their start/end
// markers) intersect each other, false otherwise.
function anyIntersection(a, b) {
return a.start <= b.start && b.start < a.end ||
a.start < b.end && b.end <= a.end ||
b.start <= a.start && a.start < b.end ||
b.start < a.end && a.end <= b.end;
}
module.exports = anyIntersection;

View file

@ -0,0 +1,34 @@
"use strict";
// Create plain text entries corresponding to areas of the text that match no
// existing entries. Returns an empty array if all parts of the text have been
// parsed into recognizable entries already.
function fill(existingEntries, text) {
let position = 0;
// Fill inner parts of the text. For example, if text is `foobarbaz` and both
// `foo` and `baz` have matched into an entry, this will return a dummy entry
// corresponding to `bar`.
const result = existingEntries.reduce((acc, textSegment) => {
if (textSegment.start > position) {
acc.push({
start: position,
end: textSegment.start
});
}
position = textSegment.end;
return acc;
}, []);
// Complete the unmatched end of the text with a dummy entry
if (position < text.length) {
result.push({
start: position,
end: text.length
});
}
return result;
}
module.exports = fill;

View file

@ -0,0 +1,43 @@
"use strict";
// Escapes the RegExp special characters "^", "$", "", ".", "*", "+", "?", "(",
// ")", "[", "]", "{", "}", and "|" in string.
// See https://lodash.com/docs/#escapeRegExp
const escapeRegExp = require("lodash/escapeRegExp");
// Given an array of channel prefixes (such as "#" and "&") and an array of user
// modes (such as "@" and "+"), this function extracts channels and nicks from a
// text.
// It returns an array of objects for each channel found with their start index,
// end index and channel name.
function findChannels(text, channelPrefixes, userModes) {
// `userModePattern` is necessary to ignore user modes in /whois responses.
// For example, a voiced user in #thelounge will have a /whois response of:
// > foo is on the following channels: +#thelounge
// We need to explicitly ignore user modes to parse such channels correctly.
const userModePattern = userModes.map(escapeRegExp).join("");
const channelPrefixPattern = channelPrefixes.map(escapeRegExp).join("");
const channelPattern = `(?:^|\\s)[${userModePattern}]*([${channelPrefixPattern}][^ \u0007]+)`;
const channelRegExp = new RegExp(channelPattern, "g");
const result = [];
let match;
do {
// With global ("g") regexes, calling `exec` multiple times will find
// successive matches in the same string.
match = channelRegExp.exec(text);
if (match) {
result.push({
start: match.index + match[0].length - match[1].length,
end: match.index + match[0].length,
channel: match[1]
});
}
} while (match);
return result;
}
module.exports = findChannels;

View file

@ -0,0 +1,56 @@
"use strict";
const URI = require("urijs");
// Known schemes to detect in a text. If a text contains `foo...bar://foo.com`,
// the parsed scheme should be `foo...bar` but if it contains
// `foo...http://foo.com`, we assume the scheme to extract will be `http`.
const commonSchemes = [
"http", "https",
"ftp", "sftp",
"smb", "file",
"irc", "ircs",
"svn", "git",
"steam", "mumble", "ts3server",
"svn+ssh", "ssh",
];
function findLinks(text) {
let result = [];
// URI.withinString() identifies URIs within text, e.g. to translate them to
// <a>-Tags.
// See https://medialize.github.io/URI.js/docs.html#static-withinString
// In our case, we store each URI encountered in a result array.
URI.withinString(text, function(url, start, end) {
// Extract the scheme of the URL detected, if there is one
const parsedScheme = URI(url).scheme().toLowerCase();
// Check if the scheme of the detected URL matches a common one above.
// In a URL like `foo..http://example.com`, the scheme would be `foo..http`,
// so we need to clean up the end of the scheme and filter out the rest.
const matchedScheme = commonSchemes.find(scheme => parsedScheme.endsWith(scheme));
// A known scheme was found, extract the unknown part from the URL
if (matchedScheme) {
const prefix = parsedScheme.length - matchedScheme.length;
start += prefix;
url = url.slice(prefix);
}
// The URL matched but does not start with a scheme (`www.foo.com`), add it
if (!parsedScheme.length) {
url = "http://" + url;
}
result.push({
start: start,
end: end,
link: url
});
});
return result;
}
module.exports = findLinks;

View file

@ -0,0 +1,60 @@
"use strict";
const anyIntersection = require("./anyIntersection");
const fill = require("./fill");
let Object_assign = Object.assign;
if (typeof Object_assign !== "function") {
Object_assign = function(target) {
Array.prototype.slice.call(arguments, 1).forEach(function(obj) {
Object.keys(obj).forEach(function(key) {
target[key] = obj[key];
});
});
return target;
};
}
// Merge text part information within a styling fragment
function assign(textPart, fragment) {
const fragStart = fragment.start;
const start = Math.max(fragment.start, textPart.start);
const end = Math.min(fragment.end, textPart.end);
return Object_assign({}, fragment, {
start: start,
end: end,
text: fragment.text.slice(start - fragStart, end - fragStart)
});
}
// Merge the style fragments withing the text parts, taking into account
// boundaries and text sections that have not matched to links or channels.
// For example, given a string "foobar" where "foo" and "bar" have been
// identified as parts (channels, links, etc.) and "fo", "ob" and "ar" have 3
// different styles, the first resulting part will contain fragments "fo" and
// "o", and the second resulting part will contain "b" and "ar". "o" and "b"
// fragments will contain duplicate styling attributes.
function merge(textParts, styleFragments) {
// Re-build the overall text (without control codes) from the style fragments
const cleanText = styleFragments.reduce((acc, frag) => acc + frag.text, "");
// Every section of the original text that has not been captured in a "part"
// is filled with "text" parts, dummy objects with start/end but no extra
// metadata.
const allParts = textParts
.concat(fill(textParts, cleanText))
.sort((a, b) => a.start - b.start);
// Distribute the style fragments within the text parts
return allParts.map(textPart => {
textPart.fragments = styleFragments
.filter(fragment => anyIntersection(textPart, fragment))
.map(fragment => assign(textPart, fragment));
return textPart;
});
}
module.exports = merge;

View file

@ -0,0 +1,161 @@
"use strict";
// Styling control codes
const BOLD = "\x02";
const COLOR = "\x03";
const RESET = "\x0f";
const REVERSE = "\x16";
const ITALIC = "\x1d";
const UNDERLINE = "\x1f";
// Color code matcher, with format `XX,YY` where both `XX` and `YY` are
// integers, `XX` is the text color and `YY` is an optional background color.
const colorRx = /^(\d{1,2})(?:,(\d{1,2}))?/;
// Represents all other control codes that to be ignored/filtered from the text
const controlCodesRx = /[\u0000-\u001F]/g;
// Converts a given text into an array of objects, each of them representing a
// similarly styled section of the text. Each object carries the `text`, style
// information (`bold`, `textColor`, `bgcolor`, `reverse`, `italic`,
// `underline`), and `start`/`end` cursors.
function parseStyle(text) {
const result = [];
let start = 0;
let position = 0;
// At any given time, these carry style information since last time a styling
// control code was met.
let colorCodes, bold, textColor, bgColor, reverse, italic, underline;
const resetStyle = () => {
bold = false;
textColor = undefined;
bgColor = undefined;
reverse = false;
italic = false;
underline = false;
};
resetStyle();
// When called, this "closes" the current fragment by adding an entry to the
// `result` array using the styling information set last time a control code
// was met.
const emitFragment = () => {
// Uses the text fragment starting from the last control code position up to
// the current position
const textPart = text.slice(start, position);
// Filters out all non-style related control codes present in this text
const processedText = textPart.replace(controlCodesRx, "");
if (processedText.length) {
// Current fragment starts where the previous one ends, or at 0 if none
const fragmentStart = result.length ? result[result.length - 1].end : 0;
result.push({
bold,
textColor,
bgColor,
reverse,
italic,
underline,
text: processedText,
start: fragmentStart,
end: fragmentStart + processedText.length
});
}
// Now that a fragment has been "closed", the next one will start after that
start = position + 1;
};
// This loop goes through each character of the given text one by one by
// bumping the `position` cursor. Every time a new special "styling" character
// is met, an object gets created (with `emitFragment()`)information on text
// encountered since the previous styling character.
while (position < text.length) {
switch (text[position]) {
case RESET:
emitFragment();
resetStyle();
break;
// Meeting a BOLD character means that the ongoing text is either going to
// be in bold or that the previous one was in bold and the following one
// must be reset.
// This same behavior applies to COLOR, REVERSE, ITALIC, and UNDERLINE.
case BOLD:
emitFragment();
bold = !bold;
break;
case COLOR:
emitFragment();
// Go one step further to find the corresponding color
colorCodes = text.slice(position + 1).match(colorRx);
if (colorCodes) {
textColor = Number(colorCodes[1]);
if (colorCodes[2]) {
bgColor = Number(colorCodes[2]);
}
// Color code length is > 1, so bump the current position cursor by as
// much (and reset the start cursor for the current text block as well)
position += colorCodes[0].length;
start = position + 1;
} else {
// If no color codes were found, toggles back to no colors (like BOLD).
textColor = undefined;
bgColor = undefined;
}
break;
case REVERSE:
emitFragment();
reverse = !reverse;
break;
case ITALIC:
emitFragment();
italic = !italic;
break;
case UNDERLINE:
emitFragment();
underline = !underline;
break;
}
// Evaluate the next character at the next iteration
position += 1;
}
// The entire text has been parsed, so we finalize the current text fragment.
emitFragment();
return result;
}
const properties = ["bold", "textColor", "bgColor", "italic", "underline", "reverse"];
function prepare(text) {
return parseStyle(text)
// This optimizes fragments by combining them together when all their values
// for the properties defined above are equal.
.reduce((prev, curr) => {
if (prev.length) {
const lastEntry = prev[prev.length - 1];
if (properties.every(key => curr[key] === lastEntry[key])) {
lastEntry.text += curr.text;
lastEntry.end += curr.text.length;
return prev;
}
}
return prev.concat([curr]);
}, []);
}
module.exports = prepare;

View file

@ -1,126 +1,71 @@
"use strict"; "use strict";
const Handlebars = require("handlebars/runtime"); const Handlebars = require("handlebars/runtime");
const URI = require("urijs"); const parseStyle = require("./ircmessageparser/parseStyle");
const findChannels = require("./ircmessageparser/findChannels");
const findLinks = require("./ircmessageparser/findLinks");
const merge = require("./ircmessageparser/merge");
module.exports = function(text) { // Create an HTML `span` with styling information for a given fragment
text = Handlebars.Utils.escapeExpression(text); function createFragment(fragment) {
text = colors(text); let classes = [];
text = channels(text); if (fragment.bold) {
text = uri(text); classes.push("irc-bold");
return text; }
if (fragment.textColor !== undefined) {
classes.push("irc-fg" + fragment.textColor);
}
if (fragment.bgColor !== undefined) {
classes.push("irc-bg" + fragment.bgColor);
}
if (fragment.italic) {
classes.push("irc-italic");
}
if (fragment.underline) {
classes.push("irc-underline");
}
const escapedText = Handlebars.Utils.escapeExpression(fragment.text);
if (classes.length) {
return `<span class="${classes.join(" ")}">${escapedText}</span>`;
}
return escapedText;
}
// Transform an IRC message potentially filled with styling control codes, URLs
// and channels into a string of HTML elements to display on the client.
module.exports = function parse(text) {
// Extract the styling information and get the plain text version from it
const styleFragments = parseStyle(text);
const cleanText = styleFragments.map(fragment => fragment.text).join("");
// On the plain text, find channels and URLs, returned as "parts". Parts are
// arrays of objects containing start and end markers, as well as metadata
// depending on what was found (channel or link).
const channelPrefixes = ["#", "&"]; // TODO Channel prefixes should be RPL_ISUPPORT.CHANTYPES
const userModes = ["!", "@", "%", "+"]; // TODO User modes should be RPL_ISUPPORT.PREFIX
const channelParts = findChannels(cleanText, channelPrefixes, userModes);
const linkParts = findLinks(cleanText);
// Sort all parts identified based on their position in the original text
const parts = channelParts
.concat(linkParts)
.sort((a, b) => a.start - b.start);
// Merge the styling information with the channels / URLs / text objects and
// generate HTML strings with the resulting fragments
return merge(parts, styleFragments).map(textPart => {
// Create HTML strings with styling information
const fragments = textPart.fragments.map(createFragment).join("");
// Wrap these potentially styled fragments with links and channel buttons
if (textPart.link) {
const escapedLink = Handlebars.Utils.escapeExpression(textPart.link);
return `<a href="${escapedLink}" target="_blank" rel="noopener">${fragments}</a>`;
} else if (textPart.channel) {
const escapedChannel = Handlebars.Utils.escapeExpression(textPart.channel);
return `<span class="inline-channel" role="button" tabindex="0" data-chan="${escapedChannel}">${fragments}</span>`;
}
return fragments;
}).join("");
}; };
function uri(text) {
return URI.withinString(text, function(url) {
if (url.indexOf("javascript:") === 0) {
return url;
}
var split = url.split("<");
url = "<a href='" + split[0].replace(/^www/, "http://www") + "' target='_blank' rel='noopener'>" + split[0] + "</a>";
if (split.length > 1) {
url += "<" + split.slice(1).join("<");
}
return url;
});
}
/**
* Channels names are strings of length up to fifty (50) characters.
* The only restriction on a channel name is that it SHALL NOT contain
* any spaces (' '), a control G (^G or ASCII 7), a comma (',').
* Channel prefix '&' is handled as '&amp;' because this parser is executed
* after entities in the message have been escaped. This prevents a couple of bugs.
*/
function channels(text) {
return text.replace(
/(^|\s|\x07|,)((?:#|&amp;)[^\x07\s,]{1,49})/g,
'$1<span class="inline-channel" role="button" tabindex="0" data-chan="$2">$2</span>'
);
}
/**
* MIRC compliant colour and style parser
* Unfortuanately this is a non trivial operation
* See this branch for source and tests
* https://github.com/megawac/irc-style-parser/tree/shout
*/
var styleCheck_Re = /[\x00-\x1F]/,
back_re = /^([0-9]{1,2})(,([0-9]{1,2}))?/,
colourKey = "\x03",
// breaks all open styles ^O (\x0F)
styleBreak = "\x0F";
function styleTemplate(settings) {
return "<span class='" + settings.style + "'>" + settings.text + "</span>";
}
var styles = [
["normal", "\x00", ""], ["underline", "\x1F"],
["bold", "\x02"], ["italic", "\x1D"]
].map(function(style) {
var escaped = encodeURI(style[1]).replace("%", "\\x");
return {
name: style[0],
style: style[2] ? style[2] : "irc-" + style[0],
key: style[1],
keyregex: new RegExp(escaped + "(.*?)(" + escaped + "|$)")
};
});
function colors(line) {
// http://www.mirc.com/colors.html
// http://www.aviran.org/stripremove-irc-client-control-characters/
// https://github.com/perl6/mu/blob/master/examples/rules/Grammar-IRC.pm
// regexs are cruel to parse this thing
// already done?
if (!styleCheck_Re.test(line)) {
return line;
}
// split up by the irc style break character ^O
if (line.indexOf(styleBreak) >= 0) {
return line.split(styleBreak).map(colors).join("");
}
var result = line;
var parseArr = result.split(colourKey);
var text, match, colour, background = "";
for (var i = 0; i < parseArr.length; i++) {
text = parseArr[i];
match = text.match(back_re);
if (!match) {
// ^C (no colour) ending. Escape current colour and carry on
background = "";
continue;
}
colour = "irc-fg" + +match[1];
// set the background colour
if (match[3]) {
background = " irc-bg" + +match[3];
}
// update the parsed text result
result = result.replace(colourKey + text, styleTemplate({
style: colour + background,
text: text.slice(match[0].length)
}));
}
// Matching styles (italics/bold/underline)
// if only colours were this easy...
styles.forEach(function(style) {
if (result.indexOf(style.key) < 0) {
return;
}
result = result.replace(style.keyregex, function(matchedTrash, matchedText) {
return styleTemplate({
style: style.style,
text: matchedText
});
});
});
return result;
}

View file

@ -0,0 +1,30 @@
"use strict";
const expect = require("chai").expect;
const anyIntersection = require("../../../../../../client/js/libs/handlebars/ircmessageparser/anyIntersection");
describe("anyIntersection", () => {
it("should not intersect on edges", () => {
const a = {start: 1, end: 2};
const b = {start: 2, end: 3};
expect(anyIntersection(a, b)).to.equal(false);
expect(anyIntersection(b, a)).to.equal(false);
});
it("should intersect on overlapping", () => {
const a = {start: 0, end: 3};
const b = {start: 1, end: 2};
expect(anyIntersection(a, b)).to.equal(true);
expect(anyIntersection(b, a)).to.equal(true);
});
it("should not intersect", () => {
const a = {start: 0, end: 1};
const b = {start: 2, end: 3};
expect(anyIntersection(a, b)).to.equal(false);
expect(anyIntersection(b, a)).to.equal(false);
});
});

View file

@ -0,0 +1,50 @@
"use strict";
const expect = require("chai").expect;
const fill = require("../../../../../../client/js/libs/handlebars/ircmessageparser/fill");
describe("fill", () => {
const text = "01234567890123456789";
it("should return an entry for the unmatched end of string", () => {
const existingEntries = [
{start: 0, end: 10},
{start: 5, end: 15},
];
const expected = [
{start: 15, end: 20},
];
const actual = fill(existingEntries, text);
expect(actual).to.deep.equal(expected);
});
it("should return an entry per unmatched areas of the text", () => {
const existingEntries = [
{start: 0, end: 5},
{start: 10, end: 15},
];
const expected = [
{start: 5, end: 10},
{start: 15, end: 20},
];
const actual = fill(existingEntries, text);
expect(actual).to.deep.equal(expected);
});
it("should not return anything when entries match all text", () => {
const existingEntries = [
{start: 0, end: 10},
{start: 10, end: 20},
];
const actual = fill(existingEntries, text);
expect(actual).to.be.empty;
});
});

View file

@ -0,0 +1,123 @@
"use strict";
const expect = require("chai").expect;
const findChannels = require("../../../../../../client/js/libs/handlebars/ircmessageparser/findChannels");
describe("findChannels", () => {
it("should find single letter channel", () => {
const input = "#a";
const expected = [{
channel: "#a",
start: 0,
end: 2
}];
const actual = findChannels(input, ["#"], ["@", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should find utf8 channels", () => {
const input = "#äöü";
const expected = [{
channel: "#äöü",
start: 0,
end: 4
}];
const actual = findChannels(input, ["#"], ["@", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should find inline channel", () => {
const input = "inline #channel text";
const expected = [{
channel: "#channel",
start: 7,
end: 15
}];
const actual = findChannels(input, ["#"], ["@", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should stop at \\0x07", () => {
const input = "#chan\x07nel";
const expected = [{
channel: "#chan",
start: 0,
end: 5
}];
const actual = findChannels(input, ["#"], ["@", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should allow classics pranks", () => {
const input = "#1,000";
const expected = [{
channel: "#1,000",
start: 0,
end: 6
}];
const actual = findChannels(input, ["#"], ["@", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should work with whois reponses", () => {
const input = "@#a";
const expected = [{
channel: "#a",
start: 1,
end: 3
}];
const actual = findChannels(input, ["#"], ["@", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should work with IRCv3.1 multi-prefix", () => {
const input = "!@%+#a";
const expected = [{
channel: "#a",
start: 4,
end: 6
}];
const actual = findChannels(input, ["#"], ["!", "@", "%", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should work with custom channelPrefixes", () => {
const input = "@a";
const expected = [{
channel: "@a",
start: 0,
end: 2
}];
const actual = findChannels(input, ["@"], ["#", "+"]);
expect(actual).to.deep.equal(expected);
});
it("should handle multiple channelPrefix correctly", () => {
const input = "##test";
const expected = [{
channel: "##test",
start: 0,
end: 6
}];
const actual = findChannels(input, ["#"], ["@", "+"]);
expect(actual).to.deep.equal(expected);
});
});

View file

@ -0,0 +1,106 @@
"use strict";
const expect = require("chai").expect;
const findLinks = require("../../../../../../client/js/libs/handlebars/ircmessageparser/findLinks");
describe("findLinks", () => {
it("should find url", () => {
const input = "irc://freenode.net/thelounge";
const expected = [{
start: 0,
end: 28,
link: "irc://freenode.net/thelounge",
}];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
it("should find urls with www", () => {
const input = "www.nooooooooooooooo.com";
const expected = [{
start: 0,
end: 24,
link: "http://www.nooooooooooooooo.com"
}];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
it("should find urls in strings", () => {
const input = "look at https://thelounge.github.io/ for more information";
const expected = [{
link: "https://thelounge.github.io/",
start: 8,
end: 36
}];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
it("should find urls in strings starting with www", () => {
const input = "use www.duckduckgo.com for privacy reasons";
const expected = [{
link: "http://www.duckduckgo.com",
start: 4,
end: 22
}];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
it("should find urls with odd surroundings", () => {
const input = "<https://theos.kyriasis.com/~kyrias/stats/archlinux.html>";
const expected = [{
link: "https://theos.kyriasis.com/~kyrias/stats/archlinux.html",
start: 1,
end: 56
}];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
it("should find urls with starting with www. and odd surroundings", () => {
const input = ".:www.github.com:.";
const expected = [{
link: "http://www.github.com",
start: 2,
end: 16
}];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
it("should not find urls", () => {
const input = "text www. text";
const expected = [];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
it("should handle multiple www. correctly", () => {
const input = "www.www.test.com";
const expected = [{
link: "http://www.www.test.com",
start: 0,
end: 16
}];
const actual = findLinks(input);
expect(actual).to.deep.equal(expected);
});
});

View file

@ -0,0 +1,63 @@
"use strict";
const expect = require("chai").expect;
const merge = require("../../../../../../client/js/libs/handlebars/ircmessageparser/merge");
describe("merge", () => {
it("should split style information", () => {
const textParts = [{
start: 0,
end: 10,
flag1: true
}, {
start: 10,
end: 20,
flag2: true
}];
const styleFragments = [{
start: 0,
end: 5,
text: "01234"
}, {
start: 5,
end: 15,
text: "5678901234"
}, {
start: 15,
end: 20,
text: "56789"
}];
const expected = [{
start: 0,
end: 10,
flag1: true,
fragments: [{
start: 0,
end: 5,
text: "01234"
}, {
start: 5,
end: 10,
text: "56789"
}]
}, {
start: 10,
end: 20,
flag2: true,
fragments: [{
start: 10,
end: 15,
text: "01234"
}, {
start: 15,
end: 20,
text: "56789"
}]
}];
const actual = merge(textParts, styleFragments);
expect(actual).to.deep.equal(expected);
});
});

View file

@ -0,0 +1,274 @@
"use strict";
const expect = require("chai").expect;
const parseStyle = require("../../../../../../client/js/libs/handlebars/ircmessageparser/parseStyle");
describe("parseStyle", () => {
it("should skip control codes", () => {
const input = "text\x01with\x04control\x05codes";
const expected = [{
bold: false,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "textwithcontrolcodes",
start: 0,
end: 20
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should parse bold", () => {
const input = "\x02bold";
const expected = [{
bold: true,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "bold",
start: 0,
end: 4
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should parse textColor", () => {
const input = "\x038yellowText";
const expected = [{
bold: false,
textColor: 8,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "yellowText",
start: 0,
end: 10
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should parse textColor and background", () => {
const input = "\x034,8yellowBG redText";
const expected = [{
textColor: 4,
bgColor: 8,
bold: false,
reverse: false,
italic: false,
underline: false,
text: "yellowBG redText",
start: 0,
end: 16
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should parse italic", () => {
const input = "\x1ditalic";
const expected = [{
bold: false,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: true,
underline: false,
text: "italic",
start: 0,
end: 6
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should carry state corretly forward", () => {
const input = "\x02bold\x038yellow\x02nonBold\x03default";
const expected = [{
bold: true,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "bold",
start: 0,
end: 4
}, {
bold: true,
textColor: 8,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "yellow",
start: 4,
end: 10
}, {
bold: false,
textColor: 8,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "nonBold",
start: 10,
end: 17
}, {
bold: false,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "default",
start: 17,
end: 24
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should toggle bold correctly", () => {
const input = "\x02bold\x02 \x02bold\x02";
const expected = [{
bold: true,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "bold",
start: 0,
end: 4
}, {
bold: false,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: " ",
start: 4,
end: 5
}, {
bold: true,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "bold",
start: 5,
end: 9
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should reset all styles", () => {
const input = "\x02\x034\x16\x1d\x1ffull\x0fnone";
const expected = [{
bold: true,
textColor: 4,
bgColor: undefined,
reverse: true,
italic: true,
underline: true,
text: "full",
start: 0,
end: 4
}, {
bold: false,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "none",
start: 4,
end: 8
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should not emit empty fragments", () => {
const input = "\x031\x031,2\x031\x031,2\x031\x031,2\x03a";
const expected = [{
bold: false,
textColor: undefined,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: "a",
start: 0,
end: 1
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
it("should optimize fragments", () => {
const rawString = "oh hi test text";
const colorCode = "\x0312";
const input = colorCode + rawString.split("").join(colorCode);
const expected = [{
bold: false,
textColor: 12,
bgColor: undefined,
reverse: false,
italic: false,
underline: false,
text: rawString,
start: 0,
end: rawString.length
}];
const actual = parseStyle(input);
expect(actual).to.deep.equal(expected);
});
});

View file

@ -0,0 +1,336 @@
"use strict";
const expect = require("chai").expect;
const parse = require("../../../../../client/js/libs/handlebars/parse");
describe("parse Handlebars helper", () => {
it("should not introduce xss", () => {
const testCases = [{
input: "<img onerror='location.href=\"//youtube.com\"'>",
expected: "&lt;img onerror&#x3D;&#x27;location.href&#x3D;&quot;//youtube.com&quot;&#x27;&gt;"
}, {
input: "#&\">bug",
expected: "<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#&amp;&quot;&gt;bug\">#&amp;&quot;&gt;bug</span>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should skip control codes", () => {
const testCases = [{
input: "text\x01with\x04control\x05codes",
expected: "textwithcontrolcodes"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should find urls", () => {
const testCases = [{
input: "irc://freenode.net/thelounge",
expected:
"<a href=\"irc://freenode.net/thelounge\" target=\"_blank\" rel=\"noopener\">" +
"irc://freenode.net/thelounge" +
"</a>"
}, {
input: "www.nooooooooooooooo.com",
expected:
"<a href=\"http://www.nooooooooooooooo.com\" target=\"_blank\" rel=\"noopener\">" +
"www.nooooooooooooooo.com" +
"</a>"
}, {
input: "look at https://thelounge.github.io/ for more information",
expected:
"look at " +
"<a href=\"https://thelounge.github.io/\" target=\"_blank\" rel=\"noopener\">" +
"https://thelounge.github.io/" +
"</a>" +
" for more information",
}, {
input: "use www.duckduckgo.com for privacy reasons",
expected:
"use " +
"<a href=\"http://www.duckduckgo.com\" target=\"_blank\" rel=\"noopener\">" +
"www.duckduckgo.com" +
"</a>" +
" for privacy reasons"
}, {
input: "svn+ssh://example.org",
expected:
"<a href=\"svn+ssh://example.org\" target=\"_blank\" rel=\"noopener\">" +
"svn+ssh://example.org" +
"</a>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("url with a dot parsed correctly", () => {
const input =
"bonuspunkt: your URL parser misparses this URL: https://msdn.microsoft.com/en-us/library/windows/desktop/ms644989(v=vs.85).aspx";
const correctResult =
"bonuspunkt: your URL parser misparses this URL: " +
"<a href=\"https://msdn.microsoft.com/en-us/library/windows/desktop/ms644989(v&#x3D;vs.85).aspx\" target=\"_blank\" rel=\"noopener\">" +
"https://msdn.microsoft.com/en-us/library/windows/desktop/ms644989(v&#x3D;vs.85).aspx" +
"</a>";
const actual = parse(input);
expect(actual).to.deep.equal(correctResult);
});
it("should balance brackets", () => {
const testCases = [{
input: "<https://theos.kyriasis.com/~kyrias/stats/archlinux.html>",
expected:
"&lt;" +
"<a href=\"https://theos.kyriasis.com/~kyrias/stats/archlinux.html\" target=\"_blank\" rel=\"noopener\">" +
"https://theos.kyriasis.com/~kyrias/stats/archlinux.html" +
"</a>" +
"&gt;"
}, {
input: "abc (www.example.com)",
expected:
"abc (" +
"<a href=\"http://www.example.com\" target=\"_blank\" rel=\"noopener\">" +
"www.example.com" +
"</a>" +
")"
}, {
input: "http://example.com/Test_(Page)",
expected:
"<a href=\"http://example.com/Test_(Page)\" target=\"_blank\" rel=\"noopener\">" +
"http://example.com/Test_(Page)" +
"</a>"
}, {
input: "www.example.com/Test_(Page)",
expected:
"<a href=\"http://www.example.com/Test_(Page)\" target=\"_blank\" rel=\"noopener\">" +
"www.example.com/Test_(Page)" +
"</a>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should not find urls", () => {
const testCases = [{
input: "text www. text",
expected: "text www. text"
}, {
input: "http://.",
expected: "http://."
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should find channels", () => {
const testCases = [{
input: "#a",
expected:
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#a\">" +
"#a" +
"</span>"
}, {
input: "#test",
expected:
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#test\">" +
"#test" +
"</span>"
}, {
input: "#äöü",
expected:
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#äöü\">" +
"#äöü" +
"</span>"
}, {
input: "inline #channel text",
expected:
"inline " +
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#channel\">" +
"#channel" +
"</span>" +
" text"
}, {
input: "#1,000",
expected:
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#1,000\">" +
"#1,000" +
"</span>"
}, {
input: "@#a",
expected:
"@" +
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#a\">" +
"#a" +
"</span>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should not find channels", () => {
const testCases = [{
input: "hi#test",
expected: "hi#test"
}, {
input: "#",
expected: "#"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should style like mirc", () => {
const testCases = [{
input: "\x02bold",
expected: "<span class=\"irc-bold\">bold</span>"
}, {
input: "\x038yellowText",
expected: "<span class=\"irc-fg8\">yellowText</span>"
}, {
input: "\x030,0white,white",
expected: "<span class=\"irc-fg0 irc-bg0\">white,white</span>"
}, {
input: "\x034,8yellowBGredText",
expected: "<span class=\"irc-fg4 irc-bg8\">yellowBGredText</span>"
}, {
input: "\x1ditalic",
expected: "<span class=\"irc-italic\">italic</span>"
}, {
input: "\x1funderline",
expected: "<span class=\"irc-underline\">underline</span>"
}, {
input: "\x02bold\x038yellow\x02nonBold\x03default",
expected:
"<span class=\"irc-bold\">bold</span>" +
"<span class=\"irc-bold irc-fg8\">yellow</span>" +
"<span class=\"irc-fg8\">nonBold</span>" +
"default"
}, {
input: "\x02bold\x02 \x02bold\x02",
expected:
"<span class=\"irc-bold\">bold</span>" +
" " +
"<span class=\"irc-bold\">bold</span>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should go bonkers like mirc", () => {
const testCases = [{
input: "\x02irc\x0f://\x1dfreenode.net\x0f/\x034,8thelounge",
expected:
"<a href=\"irc://freenode.net/thelounge\" target=\"_blank\" rel=\"noopener\">" +
"<span class=\"irc-bold\">irc</span>" +
"://" +
"<span class=\"irc-italic\">freenode.net</span>" +
"/" +
"<span class=\"irc-fg4 irc-bg8\">thelounge</span>" +
"</a>"
}, {
input: "\x02#\x038,9thelounge",
expected:
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#thelounge\">" +
"<span class=\"irc-bold\">#</span>" +
"<span class=\"irc-bold irc-fg8 irc-bg9\">thelounge</span>" +
"</span>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should optimize generated html", () => {
const testCases = [{
input: "test \x0312#\x0312\x0312\"te\x0312st\x0312\x0312\x0312\x0312\x0312\x0312\x0312\x0312\x0312\x0312\x0312a",
expected:
"test " +
"<span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"#&quot;testa\">" +
"<span class=\"irc-fg12\">#&quot;testa</span>" +
"</span>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should trim commom protocols", () => {
const testCases = [{
input: "like..http://example.com",
expected:
"like.." +
"<a href=\"http://example.com\" target=\"_blank\" rel=\"noopener\">" +
"http://example.com" +
"</a>"
}, {
input: "like..HTTP://example.com",
expected:
"like.." +
"<a href=\"HTTP://example.com\" target=\"_blank\" rel=\"noopener\">" +
"HTTP://example.com" +
"</a>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should not find channel in fragment", () => {
const testCases = [{
input: "http://example.com/#hash",
expected:
"" +
"<a href=\"http://example.com/#hash\" target=\"_blank\" rel=\"noopener\">" +
"http://example.com/#hash" +
"</a>"
}];
const actual = testCases.map(testCase => parse(testCase.input));
const expected = testCases.map(testCase => testCase.expected);
expect(actual).to.deep.equal(expected);
});
it("should not overlap parts", () => {
const input = "Url: http://example.com/path Channel: ##channel";
const actual = parse(input);
expect(actual).to.equal(
"Url: <a href=\"http://example.com/path\" target=\"_blank\" rel=\"noopener\">http://example.com/path</a> " +
"Channel: <span class=\"inline-channel\" role=\"button\" tabindex=\"0\" data-chan=\"##channel\">##channel</span>"
);
});
});