this repo has no description
feat: add html parsing
+108
-8
+3
-8
src/features/services/check_hn.ts
+3
-8
src/features/services/check_hn.ts
···
11
11
and,
12
12
isNull,
13
13
lt,
14
-
gte,
15
14
notInArray,
16
15
not,
17
16
count,
18
17
inArray,
19
18
} from "drizzle-orm";
20
-
import {
21
-
getNewStories,
22
-
getItems,
23
-
type Story,
24
-
getTopStories,
25
-
} from "../../libs/hackernews";
19
+
import { getItems, type Story, getTopStories } from "../../libs/hackernews";
26
20
import { addDays } from "../../libs/time";
27
21
import type { AnyMessageBlock } from "slack-edge";
28
22
import { sqlite } from "../../libs/db";
23
+
import { htmlToSlackMarkdown } from "../../libs/slackmd";
29
24
30
25
// Constants
31
26
const TOP_STORIES_LIMIT = 30; // Front page is considered the top 30 stories
···
756
751
type: "section",
757
752
text: {
758
753
type: "mrkdwn",
759
-
text: `*<${story.url || `https://news.ycombinator.com/item?id=${story.id}`}|${story.title}>*\n${story.text || ""}`,
754
+
text: `*<${story.url || `https://news.ycombinator.com/item?id=${story.id}`}|${story.title}>*\n${htmlToSlackMarkdown(story.text || "")}`,
760
755
},
761
756
},
762
757
{
+105
src/libs/slackmd.ts
+105
src/libs/slackmd.ts
···
1
+
/**
2
+
* Converts HTML to Slack Markdown format
3
+
* Uses regular expressions instead of DOM manipulation for Node.js environments
4
+
* @param html The HTML string to convert
5
+
* @return The converted Slack Markdown string
6
+
*/
7
+
export function htmlToSlackMarkdown(html: string): string {
8
+
// Basic tag replacements
9
+
let markdown = html.trim();
10
+
11
+
// Handle paragraphs
12
+
markdown = markdown.replace(/<p>(.*?)<\/p>/gs, "$1\n\n");
13
+
14
+
// Handle line breaks
15
+
markdown = markdown.replace(/<br\s*\/?>/gi, "\n");
16
+
17
+
// Handle bold text
18
+
markdown = markdown.replace(/<(strong|b)>(.*?)<\/(strong|b)>/gs, "*$2*");
19
+
20
+
// Handle italic text
21
+
markdown = markdown.replace(/<(em|i)>(.*?)<\/(em|i)>/gs, "_$2_");
22
+
23
+
// Handle code blocks
24
+
markdown = markdown.replace(/<code>(.*?)<\/code>/gs, "`$1`");
25
+
26
+
// Handle preformatted text
27
+
markdown = markdown.replace(/<pre>(.*?)<\/pre>/gs, "```\n$1\n```\n");
28
+
29
+
// Handle blockquotes
30
+
markdown = markdown.replace(
31
+
/<blockquote>(.*?)<\/blockquote>/gs,
32
+
(_match, content: string) => {
33
+
return (
34
+
content
35
+
.split("\n")
36
+
.map((line: string) => "> " + line)
37
+
.join("\n") + "\n"
38
+
);
39
+
},
40
+
);
41
+
42
+
// Handle unordered lists
43
+
markdown = markdown.replace(/<ul>(.*?)<\/ul>/gs, "$1\n");
44
+
45
+
// Handle ordered lists
46
+
markdown = markdown.replace(/<ol>(.*?)<\/ol>/gs, "$1\n");
47
+
48
+
// Handle list items
49
+
// We need to track if we're in an ordered list and what number to use
50
+
let listItemCounter = 0;
51
+
let inOrderedList = false;
52
+
53
+
markdown = markdown.replace(
54
+
/<(ul|ol|li)(?:\s[^>]*)?>|<\/(ul|ol|li)>/gs,
55
+
(match) => {
56
+
if (match.startsWith("</ul>") || match.startsWith("</ol>")) {
57
+
inOrderedList = false;
58
+
listItemCounter = 0;
59
+
return "";
60
+
} else if (match.startsWith("<ul")) {
61
+
inOrderedList = false;
62
+
return "";
63
+
} else if (match.startsWith("<ol")) {
64
+
inOrderedList = true;
65
+
listItemCounter = 0;
66
+
return "";
67
+
} else if (match.startsWith("<li")) {
68
+
if (inOrderedList) {
69
+
listItemCounter++;
70
+
return `${listItemCounter}. `;
71
+
} else {
72
+
return "• ";
73
+
}
74
+
} else if (match.startsWith("</li>")) {
75
+
return "\n";
76
+
}
77
+
return "";
78
+
},
79
+
);
80
+
81
+
// Handle links
82
+
markdown = markdown.replace(
83
+
/<a\s+(?:[^>]*?\s+)?href="([^"]*)"(?:\s+[^>]*)?>(.*?)<\/a>/gs,
84
+
(_match, href, text) => {
85
+
return `<${href}|${text}>`;
86
+
},
87
+
);
88
+
89
+
// Handle headings (h1-h6)
90
+
markdown = markdown.replace(/<h[1-6]>(.*?)<\/h[1-6]>/gs, "*$1*\n\n");
91
+
92
+
// Handle horizontal rules
93
+
markdown = markdown.replace(/<hr\s*\/?>/gi, "---\n");
94
+
95
+
// Remove all image tags
96
+
markdown = markdown.replace(/<img\s+[^>]*>/g, "");
97
+
98
+
// Clean up any remaining HTML tags
99
+
markdown = markdown.replace(/<[^>]+>/g, "");
100
+
101
+
// Fix multiple line breaks
102
+
markdown = markdown.replace(/\n{3,}/g, "\n\n");
103
+
104
+
return markdown;
105
+
}