A community-supported supercharged document management system: scan, index and archive all your documents
Python aiangulararchivingdjangodmsdocument-managementdocument-management-systemllmmachine-learningocroptical-character-recognitionpdf
Fix/performance: prevent token re-use in dropdown filtering, also a perf thing (#13804)
shamoonb4068d38bc16
diff --git a/src-ui/src/app/utils/text-search.spec.ts b/src-ui/src/app/utils/text-search.spec.tsindex a1225c1d0..7fe684042 100644--- a/src-ui/src/app/utils/text-search.spec.ts+++ b/src-ui/src/app/utils/text-search.spec.ts@@ -15,6 +15,35 @@ describe('text search utilities', () => {expect(matchesSearchText('taxes 2026', 'tax receipt')).toBeFalsy()})+ it('does not let two terms match the same word', () => {+ expect(matchesSearchText('Another Tag', 'another tag th')).toBeFalsy()+ expect(matchesSearchText('Another Tag', 'another tag ag')).toBeFalsy()+ expect(matchesSearchText('Another Tag', 'another tag e')).toBeFalsy()+ expect(matchesSearchText('Another Tag', 'another tag')).toBeTruthy()+ expect(matchesSearchText('Another Tag', 'tag another')).toBeTruthy()+ })++ it('matches a single term anywhere in the value', () => {+ expect(matchesSearchText('Another Tag', 'anoth')).toBeTruthy()+ expect(matchesSearchText('Another Tag', 'th')).toBeTruthy()+ })++ it('treats punctuation as a separator on both sides', () => {+ expect(matchesSearchText('medical-history', 'medical history')).toBeTruthy()+ expect(matchesSearchText('medical history', 'medical-history')).toBeTruthy()+ expect(matchesSearchText('medical-history', 'medical dental')).toBeFalsy()+ })++ it('matches longer terms first so they cannot be starved', () => {+ expect(matchesSearchText('tagger tag', 'tag tagger')).toBeTruthy()+ })++ it('handles a query with no usable terms', () => {+ expect(matchesSearchText('Another Tag', '')).toBeTruthy()+ // Still filters, so the dropdown can offer to create a tag named "---"+ expect(matchesSearchText('Another Tag', '---')).toBeFalsy()+ })+it('matches a large set of tag names without blocking input', () => {const tagNames = Array.from({ length: 1280 },diff --git a/src-ui/src/app/utils/text-search.ts b/src-ui/src/app/utils/text-search.tsindex f7f552408..82e29eead 100644--- a/src-ui/src/app/utils/text-search.ts+++ b/src-ui/src/app/utils/text-search.ts@@ -3,13 +3,18 @@ import { diacritics } from 'normalize-diacritics/diacritics'export type SearchTextValue =string | number | boolean | bigint | null | undefined+const NON_ASCII = /[^\x00-\x7F]/+const SEPARATORS = /[^\p{L}\p{N}]+/u+export function normalizeSearchText(value: SearchTextValue): string {- const normalized = diacritics.reduce(- (text, replacement) => {- return text.replace(replacement.diacritics, replacement.letter)- },- String(value ?? '')- )+ const text = String(value ?? '')++ // Nothing in the table matches ASCII, so skip normaliation+ if (!NON_ASCII.test(text)) return text.toLocaleLowerCase()++ const normalized = diacritics.reduce((text, replacement) => {+ return text.replace(replacement.diacritics, replacement.letter)+ }, text)return normalized.toLocaleLowerCase()}@@ -18,8 +23,28 @@ export function matchesSearchText(value: SearchTextValue,searchText: SearchTextValue): boolean {- const normalizedValue = normalizeSearchText(value)- const searchTerms = normalizeSearchText(searchText).trim().split(/\s+/)+ const query = normalizeSearchText(searchText)+ const terms = query.split(SEPARATORS).filter(Boolean)++ // Empty or punctuation-only query, nothing to split into terms+ if (terms.length === 0) {+ return normalizeSearchText(value).includes(query.trim())+ }++ const words = normalizeSearchText(value).split(SEPARATORS).filter(Boolean)+ const claimed = new Array<boolean>(words.length).fill(false)- return searchTerms.every((term) => normalizedValue.includes(term))+ // Each term takes a word of its own, longest first, so that "another tag th"+ // doesn't match "Another Tag" by finding the "th" inside "another"+ return terms+ .sort((a, b) => b.length - a.length)+ .every((term) => {+ for (let i = 0; i < words.length; i++) {+ if (!claimed[i] && words[i].includes(term)) {+ claimed[i] = true+ return true+ }+ }+ return false+ })}