3ffebd0bf2
Implements detection and facet generation for cashtags (stock tickers) like $AAPL and $BTC. Cashtags are identified by a dollar sign followed by 1-5 alphanumeric characters, with the first character being a letter. Detected cashtags are normalized to uppercase and included as tag facets in the rich text output. Co-authored-by: Claude Haiku 4.5 <noreply@anthropic.com>
457 lines
14 KiB
TypeScript
457 lines
14 KiB
TypeScript
import { AtpAgent, RichText, RichTextSegment } from '../src'
|
||
import {
|
||
isLink,
|
||
isMention,
|
||
isTag,
|
||
} from '../src/client/types/app/bsky/richtext/facet'
|
||
|
||
describe('detectFacets', () => {
|
||
const agent = new AtpAgent({ service: 'http://localhost' })
|
||
|
||
// Mock handle resolution
|
||
agent.com.atproto.identity.resolveHandle = async (params) => ({
|
||
success: true,
|
||
headers: {},
|
||
data: { did: `did:fake:${params?.handle}` },
|
||
})
|
||
|
||
const inputs = [
|
||
'no mention',
|
||
'@handle.com middle end',
|
||
'start @handle.com end',
|
||
'start middle @handle.com',
|
||
'@handle.com @handle.com @handle.com',
|
||
'@full123-chars.test',
|
||
'not@right',
|
||
'@handle.com!@#$chars',
|
||
'@handle.com\n@handle.com',
|
||
'parenthetical (@handle.com)',
|
||
'👨👩👧👧 @handle.com 👨👩👧👧',
|
||
|
||
'start https://middle.com end',
|
||
'start https://middle.com/foo/bar end',
|
||
'start https://middle.com/foo/bar?baz=bux end',
|
||
'start https://middle.com/foo/bar?baz=bux#hash end',
|
||
'https://start.com/foo/bar?baz=bux#hash middle end',
|
||
'start middle https://end.com/foo/bar?baz=bux#hash',
|
||
'https://newline1.com\nhttps://newline2.com',
|
||
'👨👩👧👧 https://middle.com 👨👩👧👧',
|
||
|
||
'start middle.com end',
|
||
'start middle.com/foo/bar end',
|
||
'start middle.com/foo/bar?baz=bux end',
|
||
'start middle.com/foo/bar?baz=bux#hash end',
|
||
'start.com/foo/bar?baz=bux#hash middle end',
|
||
'start middle end.com/foo/bar?baz=bux#hash',
|
||
'newline1.com\nnewline2.com',
|
||
'a example.com/index.php php link',
|
||
'a trailing bsky.app: colon',
|
||
|
||
'not.. a..url ..here',
|
||
'e.g.',
|
||
'something-cool.jpg',
|
||
'website.com.jpg',
|
||
'e.g./foo',
|
||
'website.com.jpg/foo',
|
||
|
||
'Classic article https://socket3.wordpress.com/2018/02/03/designing-windows-95s-user-interface/',
|
||
'Classic article https://socket3.wordpress.com/2018/02/03/designing-windows-95s-user-interface/ ',
|
||
'https://foo.com https://bar.com/whatever https://baz.com',
|
||
'punctuation https://foo.com, https://bar.com/whatever; https://baz.com.',
|
||
'parenthentical (https://foo.com)',
|
||
'except for https://foo.com/thing_(cool)',
|
||
]
|
||
const outputs: string[][][] = [
|
||
[['no mention']],
|
||
[['@handle.com', 'did:fake:handle.com'], [' middle end']],
|
||
[['start '], ['@handle.com', 'did:fake:handle.com'], [' end']],
|
||
[['start middle '], ['@handle.com', 'did:fake:handle.com']],
|
||
[
|
||
['@handle.com', 'did:fake:handle.com'],
|
||
[' '],
|
||
['@handle.com', 'did:fake:handle.com'],
|
||
[' '],
|
||
['@handle.com', 'did:fake:handle.com'],
|
||
],
|
||
[['@full123-chars.test', 'did:fake:full123-chars.test']],
|
||
[['not@right']],
|
||
[['@handle.com', 'did:fake:handle.com'], ['!@#$chars']],
|
||
[
|
||
['@handle.com', 'did:fake:handle.com'],
|
||
['\n'],
|
||
['@handle.com', 'did:fake:handle.com'],
|
||
],
|
||
[['parenthetical ('], ['@handle.com', 'did:fake:handle.com'], [')']],
|
||
[['👨👩👧👧 '], ['@handle.com', 'did:fake:handle.com'], [' 👨👩👧👧']],
|
||
|
||
[['start '], ['https://middle.com', 'https://middle.com'], [' end']],
|
||
[
|
||
['start '],
|
||
['https://middle.com/foo/bar', 'https://middle.com/foo/bar'],
|
||
[' end'],
|
||
],
|
||
[
|
||
['start '],
|
||
[
|
||
'https://middle.com/foo/bar?baz=bux',
|
||
'https://middle.com/foo/bar?baz=bux',
|
||
],
|
||
[' end'],
|
||
],
|
||
[
|
||
['start '],
|
||
[
|
||
'https://middle.com/foo/bar?baz=bux#hash',
|
||
'https://middle.com/foo/bar?baz=bux#hash',
|
||
],
|
||
[' end'],
|
||
],
|
||
[
|
||
[
|
||
'https://start.com/foo/bar?baz=bux#hash',
|
||
'https://start.com/foo/bar?baz=bux#hash',
|
||
],
|
||
[' middle end'],
|
||
],
|
||
[
|
||
['start middle '],
|
||
[
|
||
'https://end.com/foo/bar?baz=bux#hash',
|
||
'https://end.com/foo/bar?baz=bux#hash',
|
||
],
|
||
],
|
||
[
|
||
['https://newline1.com', 'https://newline1.com'],
|
||
['\n'],
|
||
['https://newline2.com', 'https://newline2.com'],
|
||
],
|
||
[['👨👩👧👧 '], ['https://middle.com', 'https://middle.com'], [' 👨👩👧👧']],
|
||
|
||
[['start '], ['middle.com', 'https://middle.com'], [' end']],
|
||
[
|
||
['start '],
|
||
['middle.com/foo/bar', 'https://middle.com/foo/bar'],
|
||
[' end'],
|
||
],
|
||
[
|
||
['start '],
|
||
['middle.com/foo/bar?baz=bux', 'https://middle.com/foo/bar?baz=bux'],
|
||
[' end'],
|
||
],
|
||
[
|
||
['start '],
|
||
[
|
||
'middle.com/foo/bar?baz=bux#hash',
|
||
'https://middle.com/foo/bar?baz=bux#hash',
|
||
],
|
||
[' end'],
|
||
],
|
||
[
|
||
[
|
||
'start.com/foo/bar?baz=bux#hash',
|
||
'https://start.com/foo/bar?baz=bux#hash',
|
||
],
|
||
[' middle end'],
|
||
],
|
||
[
|
||
['start middle '],
|
||
['end.com/foo/bar?baz=bux#hash', 'https://end.com/foo/bar?baz=bux#hash'],
|
||
],
|
||
[
|
||
['newline1.com', 'https://newline1.com'],
|
||
['\n'],
|
||
['newline2.com', 'https://newline2.com'],
|
||
],
|
||
[
|
||
['a '],
|
||
['example.com/index.php', 'https://example.com/index.php'],
|
||
[' php link'],
|
||
],
|
||
[['a trailing '], ['bsky.app', 'https://bsky.app'], [': colon']],
|
||
|
||
[['not.. a..url ..here']],
|
||
[['e.g.']],
|
||
[['something-cool.jpg']],
|
||
[['website.com.jpg']],
|
||
[['e.g./foo']],
|
||
[['website.com.jpg/foo']],
|
||
|
||
[
|
||
['Classic article '],
|
||
[
|
||
'https://socket3.wordpress.com/2018/02/03/designing-windows-95s-user-interface/',
|
||
'https://socket3.wordpress.com/2018/02/03/designing-windows-95s-user-interface/',
|
||
],
|
||
],
|
||
[
|
||
['Classic article '],
|
||
[
|
||
'https://socket3.wordpress.com/2018/02/03/designing-windows-95s-user-interface/',
|
||
'https://socket3.wordpress.com/2018/02/03/designing-windows-95s-user-interface/',
|
||
],
|
||
[' '],
|
||
],
|
||
[
|
||
['https://foo.com', 'https://foo.com'],
|
||
[' '],
|
||
['https://bar.com/whatever', 'https://bar.com/whatever'],
|
||
[' '],
|
||
['https://baz.com', 'https://baz.com'],
|
||
],
|
||
[
|
||
['punctuation '],
|
||
['https://foo.com', 'https://foo.com'],
|
||
[', '],
|
||
['https://bar.com/whatever', 'https://bar.com/whatever'],
|
||
['; '],
|
||
['https://baz.com', 'https://baz.com'],
|
||
['.'],
|
||
],
|
||
[['parenthentical ('], ['https://foo.com', 'https://foo.com'], [')']],
|
||
[
|
||
['except for '],
|
||
['https://foo.com/thing_(cool)', 'https://foo.com/thing_(cool)'],
|
||
],
|
||
]
|
||
it('correctly handles a set of text inputs', async () => {
|
||
for (let i = 0; i < inputs.length; i++) {
|
||
const input = inputs[i]
|
||
const rt = new RichText({ text: input })
|
||
await rt.detectFacets(agent)
|
||
expect(Array.from(rt.segments(), segmentToOutput)).toEqual(outputs[i])
|
||
}
|
||
})
|
||
|
||
describe('correctly detects tags inline', () => {
|
||
const inputs: [
|
||
string,
|
||
string[],
|
||
{ byteStart: number; byteEnd: number }[],
|
||
][] = [
|
||
['#a', ['a'], [{ byteStart: 0, byteEnd: 2 }]],
|
||
[
|
||
'#a #b',
|
||
['a', 'b'],
|
||
[
|
||
{ byteStart: 0, byteEnd: 2 },
|
||
{ byteStart: 3, byteEnd: 5 },
|
||
],
|
||
],
|
||
['#1', [], []],
|
||
['#1a', ['1a'], [{ byteStart: 0, byteEnd: 3 }]],
|
||
['#tag', ['tag'], [{ byteStart: 0, byteEnd: 4 }]],
|
||
['body #tag', ['tag'], [{ byteStart: 5, byteEnd: 9 }]],
|
||
['#tag body', ['tag'], [{ byteStart: 0, byteEnd: 4 }]],
|
||
['body #tag body', ['tag'], [{ byteStart: 5, byteEnd: 9 }]],
|
||
['body #1', [], []],
|
||
['body #1a', ['1a'], [{ byteStart: 5, byteEnd: 8 }]],
|
||
['body #a1', ['a1'], [{ byteStart: 5, byteEnd: 8 }]],
|
||
['#', [], []],
|
||
['#?', [], []],
|
||
['text #', [], []],
|
||
['text # text', [], []],
|
||
[
|
||
'body #thisisa64characterstring_aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa',
|
||
['thisisa64characterstring_aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa'],
|
||
[{ byteStart: 5, byteEnd: 70 }],
|
||
],
|
||
[
|
||
'body #thisisa65characterstring_aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaab',
|
||
[],
|
||
[],
|
||
],
|
||
[
|
||
'body #thisisa64characterstring_aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa!',
|
||
['thisisa64characterstring_aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa'],
|
||
[{ byteStart: 5, byteEnd: 70 }],
|
||
],
|
||
[
|
||
'its a #double#rainbow',
|
||
['double#rainbow'],
|
||
[{ byteStart: 6, byteEnd: 21 }],
|
||
],
|
||
['##hashash', ['#hashash'], [{ byteStart: 0, byteEnd: 9 }]],
|
||
['##', [], []],
|
||
['some #n0n3s@n5e!', ['n0n3s@n5e'], [{ byteStart: 5, byteEnd: 15 }]],
|
||
[
|
||
'works #with,punctuation',
|
||
['with,punctuation'],
|
||
[{ byteStart: 6, byteEnd: 23 }],
|
||
],
|
||
[
|
||
'strips trailing #punctuation, #like. #this!',
|
||
['punctuation', 'like', 'this'],
|
||
[
|
||
{ byteStart: 16, byteEnd: 28 },
|
||
{ byteStart: 30, byteEnd: 35 },
|
||
{ byteStart: 37, byteEnd: 42 },
|
||
],
|
||
],
|
||
[
|
||
'strips #multi_trailing___...',
|
||
['multi_trailing'],
|
||
[{ byteStart: 7, byteEnd: 22 }],
|
||
],
|
||
[
|
||
'works with #🦋 emoji, and #butter🦋fly',
|
||
['🦋', 'butter🦋fly'],
|
||
[
|
||
{ byteStart: 11, byteEnd: 16 },
|
||
{ byteStart: 28, byteEnd: 42 },
|
||
],
|
||
],
|
||
[
|
||
'#same #same #but #diff',
|
||
['same', 'same', 'but', 'diff'],
|
||
[
|
||
{ byteStart: 0, byteEnd: 5 },
|
||
{ byteStart: 6, byteEnd: 11 },
|
||
{ byteStart: 12, byteEnd: 16 },
|
||
{ byteStart: 17, byteEnd: 22 },
|
||
],
|
||
],
|
||
['this #️⃣tag should not be a tag', [], []],
|
||
[
|
||
'this ##️⃣tag should be a tag',
|
||
['#️⃣tag'],
|
||
[
|
||
{
|
||
byteStart: 5,
|
||
byteEnd: 16,
|
||
},
|
||
],
|
||
],
|
||
[
|
||
'this #t\nag should be a tag',
|
||
['t'],
|
||
[
|
||
{
|
||
byteStart: 5,
|
||
byteEnd: 7,
|
||
},
|
||
],
|
||
],
|
||
['no match (\\u200B): #', [], []],
|
||
['no match (\\u200Ba): #a', [], []],
|
||
['match (a\\u200Bb): #ab', ['a'], [{ byteStart: 18, byteEnd: 20 }]],
|
||
['match (ab\\u200B): #ab', ['ab'], [{ byteStart: 18, byteEnd: 21 }]],
|
||
['no match (\\u20e2tag): #⃢tag', [], []],
|
||
['no match (a\\u20e2b): #a⃢b', ['a'], [{ byteStart: 21, byteEnd: 23 }]],
|
||
[
|
||
'match full width number sign (tag): ＃tag',
|
||
['tag'],
|
||
[{ byteStart: 36, byteEnd: 42 }],
|
||
],
|
||
[
|
||
'match full width number sign (tag): ＃#️⃣tag',
|
||
['#️⃣tag'],
|
||
[{ byteStart: 36, byteEnd: 49 }],
|
||
],
|
||
['no match 1?: #1?', [], []],
|
||
]
|
||
|
||
it.each(inputs)('%s', async (input, tags, indices) => {
|
||
const rt = new RichText({ text: input })
|
||
await rt.detectFacets(agent)
|
||
|
||
const detectedTags: string[] = []
|
||
const detectedIndices: { byteStart: number; byteEnd: number }[] = []
|
||
|
||
for (const { facet } of rt.segments()) {
|
||
if (!facet) continue
|
||
for (const feature of facet.features) {
|
||
if (isTag(feature)) {
|
||
detectedTags.push(feature.tag)
|
||
}
|
||
}
|
||
detectedIndices.push(facet.index)
|
||
}
|
||
|
||
expect(detectedTags).toEqual(tags)
|
||
expect(detectedIndices).toEqual(indices)
|
||
})
|
||
})
|
||
|
||
describe('correctly detects cashtags inline', () => {
|
||
const inputs: [
|
||
string,
|
||
string[],
|
||
{ byteStart: number; byteEnd: number }[],
|
||
][] = [
|
||
['$AAPL', ['$AAPL'], [{ byteStart: 0, byteEnd: 5 }]],
|
||
['$aapl', ['$AAPL'], [{ byteStart: 0, byteEnd: 5 }]], // normalized to uppercase
|
||
['$A', ['$A'], [{ byteStart: 0, byteEnd: 2 }]],
|
||
['$a', ['$A'], [{ byteStart: 0, byteEnd: 2 }]], // single char normalized
|
||
[
|
||
'$BTC $ETH',
|
||
['$BTC', '$ETH'],
|
||
[
|
||
{ byteStart: 0, byteEnd: 4 },
|
||
{ byteStart: 5, byteEnd: 9 },
|
||
],
|
||
],
|
||
['$100', [], []], // starts with digit - not a cashtag
|
||
['$GOOGL', ['$GOOGL'], [{ byteStart: 0, byteEnd: 6 }]], // 5 chars - max length
|
||
['$TOOLONG', [], []], // >5 chars
|
||
['check $LEGO now', ['$LEGO'], [{ byteStart: 6, byteEnd: 11 }]],
|
||
['($GOOG)', ['$GOOG'], [{ byteStart: 1, byteEnd: 6 }]],
|
||
['$AAPL.', ['$AAPL'], [{ byteStart: 0, byteEnd: 5 }]], // trailing punctuation
|
||
[
|
||
'$AAPL, $MSFT!',
|
||
['$AAPL', '$MSFT'],
|
||
[
|
||
{ byteStart: 0, byteEnd: 5 },
|
||
{ byteStart: 7, byteEnd: 12 },
|
||
],
|
||
],
|
||
['no$SPACE', [], []], // must have leading space or start
|
||
['$', [], []], // just dollar sign
|
||
['$ AAPL', [], []], // space after $
|
||
['$123ABC', [], []], // starts with digit
|
||
['$ABC12', ['$ABC12'], [{ byteStart: 0, byteEnd: 6 }]], // digits after letters OK (5 chars)
|
||
['$ABC123', [], []], // 6 chars - too long
|
||
]
|
||
|
||
it.each(inputs)('%s', (input, tags, indices) => {
|
||
const rt = new RichText({ text: input })
|
||
rt.detectFacetsWithoutResolution()
|
||
|
||
const detectedTags: string[] = []
|
||
const detectedIndices: { byteStart: number; byteEnd: number }[] = []
|
||
|
||
for (const { facet } of rt.segments()) {
|
||
if (!facet) continue
|
||
for (const feature of facet.features) {
|
||
if (isTag(feature) && feature.tag.startsWith('$')) {
|
||
detectedTags.push(feature.tag)
|
||
}
|
||
}
|
||
if (
|
||
facet.features.some(
|
||
(f) => isTag(f) && (f as any).tag?.startsWith('$'),
|
||
)
|
||
) {
|
||
detectedIndices.push(facet.index)
|
||
}
|
||
}
|
||
|
||
expect(detectedTags).toEqual(tags)
|
||
expect(detectedIndices).toEqual(indices)
|
||
})
|
||
})
|
||
})
|
||
|
||
function segmentToOutput(segment: RichTextSegment): string[] {
|
||
if (segment.facet) {
|
||
return [
|
||
segment.text,
|
||
segment.facet?.features.map((f) => {
|
||
if (isMention(f)) return f.did
|
||
if (isLink(f)) return f.uri
|
||
return undefined
|
||
})?.[0] || '',
|
||
]
|
||
}
|
||
return [segment.text]
|
||
}
|