mirrors/DiscordChatExporter
1
0
Fork 0
mirror of https://github.com/Tyrrrz/DiscordChatExporter.git synced 2025-05-22 19:05:09 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

Parse emojis in markdown and render them using Twemoji

Browse source 
Closes #140
This commit is contained in:
Alexey Golub 2019-03-03 18:36:12 +02:00
parent a564906719
commit 28175e2110
4 changed files with 68 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

5
DiscordChatExporter.Core.Markdown/EmojiNode.cs
View file

@ -16,6 +16,11 @@
IsAnimated = isAnimated;
}
public EmojiNode(string lexeme, string name)
: this(lexeme, null, name, false)
{
}
public override string ToString() => $"<Emoji> {Name}";
}
}

55
DiscordChatExporter.Core.Markdown/Internal/Grammar.cs
View file

@ -55,7 +55,7 @@ namespace DiscordChatExporter.Core.Markdown.Internal
Parse.RegexMatch(new Regex("\\|\\|(.+?)\\|\\|", RegexOptions.Singleline))
.Select(m => new FormattedNode(m.Value, "||", TextFormatting.Spoiler, BuildTree(m.Groups[1].Value)));
// Aggregator, order matters
// Combinator, order matters
private static readonly Parser<Node> AnyFormattedNode =
ItalicBoldFormattedNode.Or(ItalicUnderlineFormattedNode)
.Or(BoldFormattedNode).Or(ItalicFormattedNode)
@ -74,7 +74,7 @@ namespace DiscordChatExporter.Core.Markdown.Internal
Parse.RegexMatch(new Regex("```(?:(\\w*?)?(?:\\s*?\\n))?(.+)```", RegexOptions.Singleline))
.Select(m => new MultilineCodeBlockNode(m.Value, m.Groups[1].Value, m.Groups[2].Value));
// Aggregator, order matters
// Combinator, order matters
private static readonly Parser<Node> AnyCodeBlockNode = MultilineCodeBlockNode.Or(InlineCodeBlockNode);
/* Mentions */
@ -95,18 +95,30 @@ namespace DiscordChatExporter.Core.Markdown.Internal
private static readonly Parser<Node> RoleMentionNode = Parse.RegexMatch("<@&(\\d+)>")
.Select(m => new MentionNode(m.Value, m.Groups[1].Value, MentionType.Role));
// Aggregator, order matters
// Combinator, order matters
private static readonly Parser<Node> AnyMentionNode =
MetaMentionNode.Or(UserMentionNode).Or(ChannelMentionNode).Or(RoleMentionNode);
/* Emojis */
// Matches all standard unicode emojis
private static readonly Parser<Node> StandardEmojiNode = Parse.RegexMatch(
"([\\u2700-\\u27bf]|" +
"(?:\\ud83c[\\udde6-\\uddff]){2}|" +
"[\\ud800-\\udbff][\\udc00-\\udfff]|" +
"[\\u0023-\\u0039]\\u20e3|" +
"\\u3299|\\u3297|\\u303d|\\u3030|\\u24c2|\\ud83c[\\udd70-\\udd71]|\\ud83c[\\udd7e-\\udd7f]|\\ud83c\\udd8e|\\ud83c[\\udd91-\\udd9a]|\\ud83c[\\udde6-\\uddff]|" +
"[\\ud83c[\\ude01-\\ude02]|\\ud83c\\ude1a|\\ud83c\\ude2f|[\\ud83c[\\ude32-\\ude3a]|[\\ud83c[\\ude50-\\ude51]|\\u203c|\\u2049|[\\u25aa-\\u25ab]|" +
"\\u25b6|\\u25c0|[\\u25fb-\\u25fe]|\\u00a9|\\u00ae|\\u2122|\\u2139|\\ud83c\\udc04|[\\u2600-\\u26FF]|\\u2b05|\\u2b06|\\u2b07|\\u2b1b|\\u2b1c|\\u2b50|" +
"\\u2b55|\\u231a|\\u231b|\\u2328|\\u23cf|[\\u23e9-\\u23f3]|[\\u23f8-\\u23fa]|\\ud83c\\udccf|\\u2934|\\u2935|[\\u2190-\\u21ff])")
.Select(m => new EmojiNode(m.Value, m.Groups[1].Value));
// <:lul:123456> or <a:lul:123456>
private static readonly Parser<Node> EmojiNode = Parse.RegexMatch("<(a)?:(.+):(\\d+)>")
private static readonly Parser<Node> CustomEmojiNode = Parse.RegexMatch("<(a)?:(.+):(\\d+)>")
.Select(m => new EmojiNode(m.Value, m.Groups[3].Value, m.Groups[2].Value, m.Groups[1].Value.IsNotBlank()));
// Aggregator, order matters
private static readonly Parser<Node> AnyEmojiNode = EmojiNode;
// Combinator, order matters
private static readonly Parser<Node> AnyEmojiNode = StandardEmojiNode.Or(CustomEmojiNode);
/* Links */
@ -114,15 +126,23 @@ namespace DiscordChatExporter.Core.Markdown.Internal
private static readonly Parser<Node> TitledLinkNode = Parse.RegexMatch("\\[(.+)\\]\\((.+)\\)")
.Select(m => new LinkNode(m.Value, m.Groups[2].Value, m.Groups[1].Value));
// Matches text that represents a link
private static readonly Parser<Match> AutoLinkMatch = Parse.RegexMatch("(https?://\\S*[^\\.,:;\"\'\\s])");
// Starts with http:// or https://, stops at the last non-whitespace character followed by whitespace or punctuation character
private static readonly Parser<Node> AutoLinkNode = Parse.RegexMatch("(https?://\\S*[^\\.,:;\"\'\\s])")
.Select(m => new LinkNode(m.Value, m.Groups[1].Value));
private static readonly Parser<Node> AutoLinkNode =
AutoLinkMatch.Select(m => new LinkNode(m.Value, m.Groups[1].Value));
// Autolink surrounded by angular brackets
private static readonly Parser<Node> HiddenLinkNode = Parse.RegexMatch("<(https?://\\S*[^\\.,:;\"\'\\s])>")
.Select(m => new LinkNode(m.Value, m.Groups[1].Value));
private static readonly Parser<Node> HiddenLinkNode =
from open in Parse.Char('<')
from linkMatch in AutoLinkMatch
from close in Parse.Char('>')
let url = linkMatch.Groups[1].Value
let lexeme = $"{open}{url}{close}"
select new LinkNode(lexeme, url);
// Aggregator, order matters
// Combinator, order matters
private static readonly Parser<Node> AnyLinkNode = TitledLinkNode.Or(HiddenLinkNode).Or(AutoLinkNode);
/* Text */
@ -131,12 +151,21 @@ namespace DiscordChatExporter.Core.Markdown.Internal
private static readonly Parser<Node> ShrugTextNode =
Parse.String("¯\\_(ツ)_/¯").Text().Select(s => new TextNode(s));
// Backslash escapes any following unicode surrogate pair
private static readonly Parser<Node> EscapedSurrogateTextNode =
from slash in Parse.Char('\\')
from high in Parse.AnyChar.Where(char.IsHighSurrogate)
from low in Parse.AnyChar
let lexeme = $"\\{high}{low}"
let text = $"{high}{low}"
select new TextNode(lexeme, text);
// Backslash escapes any following non-whitespace character except for digits and latin letters
private static readonly Parser<Node> EscapedTextNode =
Parse.RegexMatch("\\\\([^a-zA-Z0-9\\s])").Select(m => new TextNode(m.Value, m.Groups[1].Value));
// Aggregator, order matters
private static readonly Parser<Node> AnyTextNode = ShrugTextNode.Or(EscapedTextNode);
// Combinator, order matters
private static readonly Parser<Node> AnyTextNode = ShrugTextNode.Or(EscapedSurrogateTextNode).Or(EscapedTextNode);
/* Aggregator and fallback */

21
DiscordChatExporter.Core/Models/Emoji.cs
View file

@ -1,10 +1,12 @@
using Tyrrrz.Extensions;
using System.Collections.Generic;
using System.Linq;
using Tyrrrz.Extensions;
namespace DiscordChatExporter.Core.Models
{
// https://discordapp.com/developers/docs/resources/emoji#emoji-object
public class Emoji
public partial class Emoji
{
public string Id { get; }
@ -28,8 +30,7 @@ namespace DiscordChatExporter.Core.Models
}
// Standard unicode emoji (via twemoji)
var codePoint = char.ConvertToUtf32(Name, 0).ToString("x");
return $"https://twemoji.maxcdn.com/2/72x72/{codePoint}.png";
return $"https://twemoji.maxcdn.com/2/72x72/{GetTwemojiName(Name)}.png";
}
}
@ -40,4 +41,16 @@ namespace DiscordChatExporter.Core.Models
IsAnimated = isAnimated;
}
}
public partial class Emoji
{
private static IEnumerable<int> GetCodePoints(string emoji)
{
for (var i = 0; i < emoji.Length; i += char.IsHighSurrogate(emoji[i]) ? 2 : 1)
yield return char.ConvertToUtf32(emoji, i);
}
private static string GetTwemojiName(string emoji)
=> GetCodePoints(emoji).Select(i => i.ToString("x")).JoinToString("-");
}
}

6
DiscordChatExporter.Core/Services/ExportService.TemplateModel.cs
View file

@ -201,12 +201,14 @@ namespace DiscordChatExporter.Core.Services
else if (node is EmojiNode emojiNode)
{
buffer.Append($"<img class=\"emoji\" title=\"{emojiNode.Name}\" src=\"https://cdn.discordapp.com/emojis/{emojiNode.Id}.png\" />");
var emoji = new Emoji(emojiNode.Id, emojiNode.Name, emojiNode.IsAnimated);
buffer.Append($"<img class=\"emoji\" title=\"{emoji.Name}\" src=\"{emoji.ImageUrl}\" />");
}
else if (node is LinkNode linkNode)
{
buffer.Append($"<a href=\"{Uri.EscapeUriString(linkNode.Url)}\">{linkNode.Title.HtmlEncode()}</a>");
var escapedUrl = Uri.EscapeUriString(linkNode.Url);
buffer.Append($"<a href=\"{escapedUrl}\">{linkNode.Title.HtmlEncode()}</a>");
}
}