@atproto/api 0.10.2 → 0.10.4

package/dist/rich-text/util.d.ts

@@ -0,0 +1,4 @@
+export declare const MENTION_REGEX: RegExp;
+export declare const URL_REGEX: RegExp;
+export declare const TRAILING_PUNCTUATION_REGEX: RegExp;
+export declare const TAG_REGEX: RegExp;

package/dist/util.d.ts

@@ -0,0 +1 @@
+export declare function sanitizeMutedWordValue(value: string): string;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@atproto/api",
-  "version": "0.10.2",
+  "version": "0.10.4",
   "license": "MIT",
   "description": "Client library for atproto and Bluesky",
   "keywords": [
@@ -28,7 +28,7 @@
   "devDependencies": {
     "common-tags": "^1.8.2",
     "@atproto/lex-cli": "^0.3.1",
-    "@atproto/dev-env": "^0.2.34"
+    "@atproto/dev-env": "^0.2.36"
   },
   "scripts": {
     "codegen": "pnpm docgen && node ./scripts/generate-code.mjs && lex gen-api ./src/client ../../lexicons/com/atproto/*/* ../../lexicons/app/bsky/*/*",

package/src/bsky-agent.ts

@@ -13,6 +13,7 @@ import {
   BskyThreadViewPreference,
   BskyInterestsPreference,
 } from './types'
+import { sanitizeMutedWordValue } from './util'
 
 const FEED_VIEW_PREF_DEFAULTS = {
   hideReplies: false,
@@ -565,16 +566,108 @@ export class BskyAgent extends AtpAgent {
     })
   }
 
-  async upsertMutedWords(mutedWords: AppBskyActorDefs.MutedWord[]) {
-    await updateMutedWords(this, mutedWords, 'upsert')
+  async upsertMutedWords(newMutedWords: AppBskyActorDefs.MutedWord[]) {
+    await updatePreferences(this, (prefs: AppBskyActorDefs.Preferences) => {
+      let mutedWordsPref = prefs.findLast(
+        (pref) =>
+          AppBskyActorDefs.isMutedWordsPref(pref) &&
+          AppBskyActorDefs.validateMutedWordsPref(pref).success,
+      )
+
+      if (mutedWordsPref && AppBskyActorDefs.isMutedWordsPref(mutedWordsPref)) {
+        for (const updatedWord of newMutedWords) {
+          let foundMatch = false
+          const sanitizedUpdatedValue = sanitizeMutedWordValue(
+            updatedWord.value,
+          )
+
+          // was trimmed down to an empty string e.g. single `#`
+          if (!sanitizedUpdatedValue) continue
+
+          for (const existingItem of mutedWordsPref.items) {
+            if (existingItem.value === sanitizedUpdatedValue) {
+              existingItem.targets = Array.from(
+                new Set([...existingItem.targets, ...updatedWord.targets]),
+              )
+              foundMatch = true
+              break
+            }
+          }
+
+          if (!foundMatch) {
+            mutedWordsPref.items.push({
+              ...updatedWord,
+              value: sanitizedUpdatedValue,
+            })
+          }
+        }
+      } else {
+        // if the pref doesn't exist, create it
+        mutedWordsPref = {
+          items: newMutedWords.map((w) => ({
+            ...w,
+            value: sanitizeMutedWordValue(w.value),
+          })),
+        }
+      }
+
+      return prefs
+        .filter((p) => !AppBskyActorDefs.isMutedWordsPref(p))
+        .concat([
+          { ...mutedWordsPref, $type: 'app.bsky.actor.defs#mutedWordsPref' },
+        ])
+    })
   }
 
   async updateMutedWord(mutedWord: AppBskyActorDefs.MutedWord) {
-    await updateMutedWords(this, [mutedWord], 'update')
+    await updatePreferences(this, (prefs: AppBskyActorDefs.Preferences) => {
+      let mutedWordsPref = prefs.findLast(
+        (pref) =>
+          AppBskyActorDefs.isMutedWordsPref(pref) &&
+          AppBskyActorDefs.validateMutedWordsPref(pref).success,
+      )
+
+      if (mutedWordsPref && AppBskyActorDefs.isMutedWordsPref(mutedWordsPref)) {
+        for (const existingItem of mutedWordsPref.items) {
+          if (existingItem.value === mutedWord.value) {
+            existingItem.targets = mutedWord.targets
+            break
+          }
+        }
+      }
+
+      return prefs
+        .filter((p) => !AppBskyActorDefs.isMutedWordsPref(p))
+        .concat([
+          { ...mutedWordsPref, $type: 'app.bsky.actor.defs#mutedWordsPref' },
+        ])
+    })
   }
 
   async removeMutedWord(mutedWord: AppBskyActorDefs.MutedWord) {
-    await updateMutedWords(this, [mutedWord], 'remove')
+    await updatePreferences(this, (prefs: AppBskyActorDefs.Preferences) => {
+      let mutedWordsPref = prefs.findLast(
+        (pref) =>
+          AppBskyActorDefs.isMutedWordsPref(pref) &&
+          AppBskyActorDefs.validateMutedWordsPref(pref).success,
+      )
+
+      if (mutedWordsPref && AppBskyActorDefs.isMutedWordsPref(mutedWordsPref)) {
+        for (let i = 0; i < mutedWordsPref.items.length; i++) {
+          const existing = mutedWordsPref.items[i]
+          if (existing.value === mutedWord.value) {
+            mutedWordsPref.items.splice(i, 1)
+            break
+          }
+        }
+      }
+
+      return prefs
+        .filter((p) => !AppBskyActorDefs.isMutedWordsPref(p))
+        .concat([
+          { ...mutedWordsPref, $type: 'app.bsky.actor.defs#mutedWordsPref' },
+        ])
+    })
   }
 
   async hidePost(postUri: string) {
@@ -646,76 +739,6 @@ async function updateFeedPreferences(
   return res
 }
 
-/**
- * A helper specifically for updating muted words preferences
- */
-async function updateMutedWords(
-  agent: BskyAgent,
-  mutedWords: AppBskyActorDefs.MutedWord[],
-  action: 'upsert' | 'update' | 'remove',
-) {
-  const sanitizeMutedWord = (word: AppBskyActorDefs.MutedWord) => ({
-    value: word.value.replace(/^#/, ''),
-    targets: word.targets,
-  })
-
-  await updatePreferences(agent, (prefs: AppBskyActorDefs.Preferences) => {
-    let mutedWordsPref = prefs.findLast(
-      (pref) =>
-        AppBskyActorDefs.isMutedWordsPref(pref) &&
-        AppBskyActorDefs.validateMutedWordsPref(pref).success,
-    )
-
-    if (mutedWordsPref && AppBskyActorDefs.isMutedWordsPref(mutedWordsPref)) {
-      if (action === 'upsert' || action === 'update') {
-        for (const word of mutedWords) {
-          let foundMatch = false
-
-          for (const existingItem of mutedWordsPref.items) {
-            if (existingItem.value === sanitizeMutedWord(word).value) {
-              existingItem.targets =
-                action === 'upsert'
-                  ? Array.from(
-                      new Set([...existingItem.targets, ...word.targets]),
-                    )
-                  : word.targets
-              foundMatch = true
-              break
-            }
-          }
-
-          if (action === 'upsert' && !foundMatch) {
-            mutedWordsPref.items.push(sanitizeMutedWord(word))
-          }
-        }
-      } else if (action === 'remove') {
-        for (const word of mutedWords) {
-          for (let i = 0; i < mutedWordsPref.items.length; i++) {
-            const existing = mutedWordsPref.items[i]
-            if (existing.value === sanitizeMutedWord(word).value) {
-              mutedWordsPref.items.splice(i, 1)
-              break
-            }
-          }
-        }
-      }
-    } else {
-      // if the pref doesn't exist, create it
-      if (action === 'upsert') {
-        mutedWordsPref = {
-          items: mutedWords.map(sanitizeMutedWord),
-        }
-      }
-    }
-
-    return prefs
-      .filter((p) => !AppBskyActorDefs.isMutedWordsPref(p))
-      .concat([
-        { ...mutedWordsPref, $type: 'app.bsky.actor.defs#mutedWordsPref' },
-      ])
-  })
-}
-
 async function updateHiddenPost(
   agent: BskyAgent,
   postUri: string,

package/src/index.ts

@@ -8,11 +8,13 @@ export {
 } from '@atproto/lexicon'
 export { parseLanguage } from '@atproto/common-web'
 export * from './types'
+export * from './util'
 export * from './client'
 export * from './agent'
 export * from './rich-text/rich-text'
 export * from './rich-text/sanitization'
 export * from './rich-text/unicode'
+export * from './rich-text/util'
 export * from './moderation'
 export * from './moderation/types'
 export { LABELS } from './moderation/const/labels'

package/src/rich-text/detection.ts

@@ -1,6 +1,12 @@
 import TLDs from 'tlds'
 import { AppBskyRichtextFacet } from '../client'
 import { UnicodeString } from './unicode'
+import {
+  URL_REGEX,
+  MENTION_REGEX,
+  TAG_REGEX,
+  TRAILING_PUNCTUATION_REGEX,
+} from './util'
 
 export type Facet = AppBskyRichtextFacet.Main
 
@@ -9,7 +15,7 @@ export function detectFacets(text: UnicodeString): Facet[] | undefined {
   const facets: Facet[] = []
   {
     // mentions
-    const re = /(^|\s|\()(@)([a-zA-Z0-9.-]+)(\b)/g
+    const re = MENTION_REGEX
     while ((match = re.exec(text.utf16))) {
       if (!isValidDomain(match[3]) && !match[3].endsWith('.test')) {
         continue // probably not a handle
@@ -33,8 +39,7 @@ export function detectFacets(text: UnicodeString): Facet[] | undefined {
   }
   {
     // links
-    const re =
-      /(^|\s|\()((https?:\/\/[\S]+)|((?<domain>[a-z][a-z0-9]*(\.[a-z0-9]+)+)[\S]*))/gim
+    const re = URL_REGEX
     while ((match = re.exec(text.utf16))) {
       let uri = match[2]
       if (!uri.startsWith('http')) {
@@ -70,11 +75,14 @@ export function detectFacets(text: UnicodeString): Facet[] | undefined {
     }
   }
   {
-    const re = /(^|\s)#((?!\ufe0f)[^\d\s]\S*)(?=\s)?/g
+    const re = TAG_REGEX
     while ((match = re.exec(text.utf16))) {
       let [, leading, tag] = match
 
-      tag = tag.trim().replace(/\p{P}+$/gu, '') // strip ending punctuation
+      if (!tag) continue
+
+      // strip ending punctuation and any spaces
+      tag = tag.trim().replace(TRAILING_PUNCTUATION_REGEX, '')
 
       if (tag.length === 0 || tag.length > 64) continue
 

package/src/rich-text/util.ts

@@ -0,0 +1,11 @@
+export const MENTION_REGEX = /(^|\s|\()(@)([a-zA-Z0-9.-]+)(\b)/g
+export const URL_REGEX =
+  /(^|\s|\()((https?:\/\/[\S]+)|((?<domain>[a-z][a-z0-9]*(\.[a-z0-9]+)+)[\S]*))/gim
+export const TRAILING_PUNCTUATION_REGEX = /\p{P}+$/gu
+
+/**
+ * `\ufe0f` emoji modifier
+ * `\u00AD\u2060\u200A\u200B\u200C\u200D\u20e2` zero-width spaces (likely incomplete)
+ */
+export const TAG_REGEX =
+  /(^|\s)[#＃]((?!\ufe0f)[^\s\u00AD\u2060\u200A\u200B\u200C\u200D\u20e2]*[^\d\s\p{P}\u00AD\u2060\u200A\u200B\u200C\u200D\u20e2]+[^\s\u00AD\u2060\u200A\u200B\u200C\u200D\u20e2]*)?/gu

package/src/util.ts

@@ -0,0 +1,6 @@
+export function sanitizeMutedWordValue(value: string) {
+  return value
+    .trim()
+    .replace(/^#(?!\ufe0f)/, '')
+    .replace(/[\r\n\u00AD\u2060\u200D\u200C\u200B]+/, '')
+}

package/tests/bsky-agent.test.ts

@@ -1202,13 +1202,18 @@ describe('agent', () => {
         await agent.upsertMutedWords([
           { value: 'hashtag', targets: ['content'] },
         ])
+        // is sanitized to `hashtag`
         await agent.upsertMutedWords([{ value: '#hashtag', targets: ['tag'] }])
+
         const { mutedWords } = await agent.getPreferences()
+
         expect(mutedWords.find((m) => m.value === '#hashtag')).toBeFalsy()
+        // merged with existing
         expect(mutedWords.find((m) => m.value === 'hashtag')).toStrictEqual({
           value: 'hashtag',
           targets: ['content', 'tag'],
         })
+        // only one added
         expect(mutedWords.filter((m) => m.value === 'hashtag').length).toBe(1)
       })
 
@@ -1237,15 +1242,21 @@ describe('agent', () => {
         expect(mutedWords.find((m) => m.value === 'no_exist')).toBeFalsy()
       })
 
-      it('updateMutedWord with #', async () => {
+      it('updateMutedWord with #, does not update', async () => {
+        await agent.upsertMutedWords([
+          {
+            value: '#just_a_tag',
+            targets: ['tag'],
+          },
+        ])
         await agent.updateMutedWord({
-          value: 'hashtag',
+          value: '#just_a_tag',
           targets: ['tag', 'content'],
         })
         const { mutedWords } = await agent.getPreferences()
-        expect(mutedWords.find((m) => m.value === 'hashtag')).toStrictEqual({
-          value: 'hashtag',
-          targets: ['tag', 'content'],
+        expect(mutedWords.find((m) => m.value === 'just_a_tag')).toStrictEqual({
+          value: 'just_a_tag',
+          targets: ['tag'],
         })
       })
 
@@ -1262,11 +1273,124 @@ describe('agent', () => {
         expect(mutedWords.find((m) => m.value === 'tag_then_none')).toBeFalsy()
       })
 
-      it('removeMutedWord with #', async () => {
+      it('removeMutedWord with #, no match, no removal', async () => {
         await agent.removeMutedWord({ value: '#hashtag', targets: [] })
         const { mutedWords } = await agent.getPreferences()
 
-        expect(mutedWords.find((m) => m.value === 'hashtag')).toBeFalsy()
+        // was inserted with #hashtag, but we don't sanitize on remove
+        expect(mutedWords.find((m) => m.value === 'hashtag')).toBeTruthy()
+      })
+
+      it('single-hash #', async () => {
+        const prev = await agent.getPreferences()
+        const length = prev.mutedWords.length
+        await agent.upsertMutedWords([{ value: '#', targets: [] }])
+        const end = await agent.getPreferences()
+
+        // sanitized to empty string, not inserted
+        expect(end.mutedWords.length).toEqual(length)
+      })
+
+      it('multi-hash ##', async () => {
+        await agent.upsertMutedWords([{ value: '##', targets: [] }])
+        const { mutedWords } = await agent.getPreferences()
+
+        expect(mutedWords.find((m) => m.value === '#')).toBeTruthy()
+      })
+
+      it('multi-hash ##hashtag', async () => {
+        await agent.upsertMutedWords([{ value: '##hashtag', targets: [] }])
+        const a = await agent.getPreferences()
+
+        expect(a.mutedWords.find((w) => w.value === '#hashtag')).toBeTruthy()
+
+        await agent.removeMutedWord({ value: '#hashtag', targets: [] })
+        const b = await agent.getPreferences()
+
+        expect(b.mutedWords.find((w) => w.value === '#hashtag')).toBeFalsy()
+      })
+
+      it('hash emoji #️⃣', async () => {
+        await agent.upsertMutedWords([{ value: '#️⃣', targets: [] }])
+        const { mutedWords } = await agent.getPreferences()
+
+        expect(mutedWords.find((m) => m.value === '#️⃣')).toBeTruthy()
+
+        await agent.removeMutedWord({ value: '#️⃣', targets: [] })
+        const end = await agent.getPreferences()
+
+        expect(end.mutedWords.find((m) => m.value === '#️⃣')).toBeFalsy()
+      })
+
+      it('hash emoji ##️⃣', async () => {
+        await agent.upsertMutedWords([{ value: '##️⃣', targets: [] }])
+        const { mutedWords } = await agent.getPreferences()
+
+        expect(mutedWords.find((m) => m.value === '#️⃣')).toBeTruthy()
+
+        await agent.removeMutedWord({ value: '#️⃣', targets: [] })
+        const end = await agent.getPreferences()
+
+        expect(end.mutedWords.find((m) => m.value === '#️⃣')).toBeFalsy()
+      })
+
+      it('hash emoji ###️⃣', async () => {
+        await agent.upsertMutedWords([{ value: '###️⃣', targets: [] }])
+        const { mutedWords } = await agent.getPreferences()
+
+        expect(mutedWords.find((m) => m.value === '##️⃣')).toBeTruthy()
+
+        await agent.removeMutedWord({ value: '##️⃣', targets: [] })
+        const end = await agent.getPreferences()
+
+        expect(end.mutedWords.find((m) => m.value === '##️⃣')).toBeFalsy()
+      })
+
+      describe(`invalid characters`, () => {
+        it('zero width space', async () => {
+          const prev = await agent.getPreferences()
+          const length = prev.mutedWords.length
+          await agent.upsertMutedWords([{ value: '#​', targets: [] }])
+          const { mutedWords } = await agent.getPreferences()
+
+          expect(mutedWords.length).toEqual(length)
+        })
+
+        it('newline', async () => {
+          await agent.upsertMutedWords([
+            { value: 'test value\n with newline', targets: [] },
+          ])
+          const { mutedWords } = await agent.getPreferences()
+
+          expect(
+            mutedWords.find((m) => m.value === 'test value with newline'),
+          ).toBeTruthy()
+        })
+
+        it('newline(s)', async () => {
+          await agent.upsertMutedWords([
+            { value: 'test value\n\r with newline', targets: [] },
+          ])
+          const { mutedWords } = await agent.getPreferences()
+
+          expect(
+            mutedWords.find((m) => m.value === 'test value with newline'),
+          ).toBeTruthy()
+        })
+
+        it('empty space', async () => {
+          await agent.upsertMutedWords([{ value: ' ', targets: [] }])
+          const { mutedWords } = await agent.getPreferences()
+
+          expect(mutedWords.find((m) => m.value === ' ')).toBeFalsy()
+        })
+
+        it('leading/trailing space', async () => {
+          await agent.upsertMutedWords([{ value: ' trim ', targets: [] }])
+          const { mutedWords } = await agent.getPreferences()
+
+          expect(mutedWords.find((m) => m.value === 'trim')).toBeTruthy()
+        })
       })
     })
 

package/tests/rich-text-detection.test.ts

@@ -218,7 +218,7 @@ describe('detectFacets', () => {
     }
   })
 
-  it('correctly detects tags inline', async () => {
+  describe('correctly detects tags inline', () => {
     const inputs: [
       string,
       string[],
@@ -234,11 +234,13 @@ describe('detectFacets', () => {
         ],
       ],
       ['#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 }]],
       ['#', [], []],
       ['#?', [], []],
@@ -254,12 +256,18 @@ describe('detectFacets', () => {
         [],
         [],
       ],
+      [
+        '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',
@@ -309,9 +317,36 @@ describe('detectFacets', () => {
           },
         ],
       ],
+      [
+        'this #t\nag should be a tag',
+        ['t'],
+        [
+          {
+            byteStart: 5,
+            byteEnd: 7,
+          },
+        ],
+      ],
+      ['no match (\\u200B): #​', [], []],
+      ['no match (\\u200Ba): #​a', [], []],
+      ['match (a\\u200Bb): #a​b', ['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?', [], []],
     ]
 
-    for (const [input, tags, indices] of inputs) {
+    it.each(inputs)('%s', async (input, tags, indices) => {
       const rt = new RichText({ text: input })
       await rt.detectFacets(agent)
 
@@ -330,7 +365,7 @@ describe('detectFacets', () => {
 
       expect(detectedTags).toEqual(tags)
       expect(detectedIndices).toEqual(indices)
-    }
+    })
   })
 })