test
reorganize, improve parser
+3
-106
-102
facet.js
-102
facet.js
···
5
5
// 2: list of "windows" {start:Number, end:Number, spans:Span[]} - do not overlap (like facets)
6
6
// 3: list of facets {index:ByteRange, features:Feature[]}
7
7
8
-
export class MarkupParser {
9
-
constructor(big_regex, process_match) {
10
-
this.process_match = process_match
11
-
this.big_regex = big_regex
12
-
}
13
-
text_to_spans(text) {
14
-
let spans = []
15
-
for (let match of String(text).matchAll(this.big_regex)) {
16
-
let span = this.process_match(match)
17
-
if (span) {
18
-
span.start = match.index
19
-
span.end = match.index+match[0].length
20
-
spans.push(span)
21
-
}
22
-
}
23
-
return spans
24
-
}
25
-
}
26
-
27
-
let r = String.raw
28
-
let whitespace = r`\s\u00AD\u2060\u200A\u200B\u200C\u200D\u20e2`
29
-
let url = r`[-\w/%&=#+~@$*'!?,.;:]*`
30
-
let url_final = r`[-\w/%&=#+~@$*']`
31
-
let big_regex = RegExp([
32
-
r`\b(?<link>https?://${url}${url_final}([(]${url}[)](${url}${url_final})?)?)(?:\[(?<link_text>.*?)\])?`,
33
-
r`(?<=^|\s)[#＃](?<hashtag>(?!\uFE0F)[^${whitespace}]*[^\p{P}${whitespace}])`, // note: must filter out #123
34
-
r`(?<=^|\s|[(])@(?<mention>[-a-zA-Z0-9]+([.][-a-zA-Z0-9]+)+)\b`,
35
-
].join("|"), 'gu')
36
-
function process_match(match) {
37
-
let g = match.groups
38
-
if (g.mention!=null) {
39
-
return {
40
-
strip: [0, 0],
41
-
feature: {
42
-
$type: 'app.bsky.richtext.facet#mention',
43
-
did: g.mention, // hack, DID must be resolved afterwards
44
-
},
45
-
}
46
-
}
47
-
if (g.link!=null) {
48
-
if (g.link_text) {
49
-
return {
50
-
strip: [g.link.length+1, 1],
51
-
feature: {
52
-
$type: 'app.bsky.richtext.facet#link',
53
-
uri: g.link,
54
-
},
55
-
}
56
-
} else {
57
-
return {
58
-
strip: [0, 0],
59
-
feature: {
60
-
$type: 'app.bsky.richtext.facet#link',
61
-
uri: g.link,
62
-
},
63
-
}
64
-
}
65
-
}
66
-
if (g.hashtag!=null && !/^#\d+$/.test(g.hashtag)) {
67
-
return {
68
-
strip: [0,0],
69
-
feature: {
70
-
$type: 'app.bsky.richtext.facet#tag',
71
-
tag: g.hashtag,
72
-
},
73
-
}
74
-
}
75
-
}
76
-
export let markup1 = new MarkupParser(big_regex, process_match)
77
-
78
-
export function markup_italic(text) {
79
-
let spans = []
80
-
for (let m of text.matchAll(/[/].*?[/]/g)) {
81
-
spans.push({
82
-
start: m.index,
83
-
end: m.index+m[0].length,
84
-
strip: [1,1],
85
-
feature: {
86
-
$type: "com.example.richtext.facet#markup",
87
-
style: "italic"
88
-
},
89
-
})
90
-
}
91
-
return spans
92
-
}
93
-
94
-
export function markup_bold(text) {
95
-
let spans = []
96
-
for (let m of text.matchAll(/[*][*].*?[*][*]/g)) {
97
-
spans.push({
98
-
start: m.index,
99
-
end: m.index+m[0].length,
100
-
strip: [2,2],
101
-
feature: {
102
-
$type: "com.example.richtext.facet#markup",
103
-
style: "bold"
104
-
},
105
-
})
106
-
}
107
-
return spans
108
-
}
109
-
110
8
// the strip field is used internally to process things.
111
9
// SOME Features have support for doing it on the client side (this way we can leave the extra syntax chars in the post text, for clients which dont recognize those Features), so we write their strip values to the output
112
10
// while other Features (i.e. the builtin ones) don't, so we want to pre-apply their strip before outputting
+3
-4
test3.html
+3
-4
test3.html
···
12
12
</style>
13
13
14
14
<script type=module>
15
-
import {MarkedText, markup_italic, markup1, markup_bold} from './facet.js'
15
+
import {MarkedText} from './facet.js'
16
16
import {index_to_byteindex, byteindex_to_index} from './byteindex.js'
17
+
import {markup_main} from './markup.js'
17
18
import {HTML} from './template.js'
18
19
19
20
function *post_to_segments(post) {
···
117
118
118
119
let text = new MarkedText(inputtext)
119
120
text.spans = [
120
-
...markup_italic(text.text),
121
-
...markup1.text_to_spans(text.text),
122
-
...markup_bold(text.text),
121
+
...markup_main(text.text),
123
122
]
124
123
125
124
text.make_windows()