Merge branch 'develop'

# Conflicts:
#	manifest-beta.json
#	src/settings.ts
#	versions.json

package.json

@@ -1,6 +1,6 @@
 {
   "name": "scambier.obsidian-search",
-  "version": "1.14.2",
+  "version": "1.15.0-beta.2",
   "description": "A search engine for Obsidian",
   "main": "dist/main.js",
   "scripts": {

src/cache-manager.ts

@@ -10,7 +10,9 @@ import {
   getAliasesFromMetadata,
   getTagsFromMetadata,
   isFileCanvas,
+  isFileFromDataloomPlugin,
   isFilePlaintext,
+  isFilenameIndexable,
   logDebug,
   makeMD5,
   removeDiacritics,
@@ -18,6 +20,7 @@ import {
 import type { CanvasData } from 'obsidian/canvas'
 import type { AsPlainObject } from 'minisearch'
 import type MiniSearch from 'minisearch'
+import { settings } from './settings'
 
 /**
  * This function is responsible for extracting the text from a file and
@@ -58,11 +61,38 @@ async function getAndMapIndexedDocument(
     content = texts.join('\r\n')
   }
 
+  // ** Dataloom plugin **
+  else if (settings.dataloomIndexing && isFileFromDataloomPlugin(path)) {
+    try {
+      const data = JSON.parse(await app.vault.cachedRead(file))
+      // data is a json object, we recursively iterate the keys
+      // and concatenate the values if the key is "markdown"
+      const texts: string[] = []
+      const iterate = (obj: any) => {
+        for (const key in obj) {
+          if (typeof obj[key] === 'object') {
+            iterate(obj[key])
+          } else if (key === 'markdown') {
+            texts.push(obj[key])
+          }
+        }
+      }
+      iterate(data)
+      content = texts.join('\r\n')
+    } catch (e) {
+      console.error('Omnisearch: Error while parsing Dataloom file', path)
+      console.error(e)
+    }
+  }
+
   // ** Image or PDF **
   else if (extractor?.canFileBeExtracted(path)) {
     content = await extractor.extractText(file)
-  } else {
+  }
-    throw new Error(`Unsupported file type: "${path}"`)
+  
+  // ** Unsupported files **
+  else if (isFilenameIndexable(path)) {
+    content = file.path
   }
 
   if (content === null || content === undefined) {

src/components/ModalInFile.svelte

@@ -18,7 +18,6 @@
   import { Query } from 'src/search/query'
   import { openNote } from 'src/tools/notes'
   import { searchEngine } from 'src/search/omnisearch'
-  import { cacheManager } from 'src/cache-manager'
 
   export let modal: OmnisearchInFileModal
   export let parent: OmnisearchVaultModal | null = null

src/file-loader.ts

@@ -1,50 +0,0 @@
-import { isFileImage, isFilePDF, isFilePlaintext } from './tools/utils'
-import type { TFile } from 'obsidian'
-import type { IndexedDocument } from './globals'
-import { cacheManager } from './cache-manager'
-
-/**
- * Return all plaintext files as IndexedDocuments
- */
-export async function getPlainTextFiles(): Promise<IndexedDocument[]> {
-  const allFiles = app.vault.getFiles().filter(f => isFilePlaintext(f.path))
-  const data: IndexedDocument[] = []
-  for (const file of allFiles) {
-    const doc = await cacheManager.getDocument(file.path)
-    data.push(doc)
-    // await cacheManager.updateLiveDocument(file.path, doc)
-  }
-  return data
-}
-
-/**
- * Return all PDFs as IndexedDocuments.
- */
-export async function getPDFAsDocuments(): Promise<IndexedDocument[]> {
-  const files = app.vault.getFiles().filter(f => isFilePDF(f.path))
-  return await getBinaryFiles(files)
-}
-
-/**
- * Return all imageas as IndexedDocuments.
- */
-export async function getImagesAsDocuments(): Promise<IndexedDocument[]> {
-  const files = app.vault.getFiles().filter(f => isFileImage(f.path))
-  return await getBinaryFiles(files)
-}
-
-async function getBinaryFiles(files: TFile[]): Promise<IndexedDocument[]> {
-  const data: IndexedDocument[] = []
-  const input = []
-  for (const file of files) {
-    input.push(
-      new Promise(async (resolve, _reject) => {
-        const doc = await cacheManager.getDocument(file.path)
-        data.push(doc)
-        return resolve(null)
-      })
-    )
-  }
-  await Promise.all(input)
-  return data
-}

src/globals.ts

@@ -107,4 +107,4 @@ export function isCacheEnabled(): boolean {
 }
 
 export const SPACE_OR_PUNCTUATION =
-  /[|\n\r -#%-*,-/:;?@[-\]_{}\u00A0\u00A1\u00A7\u00AB\u00B6\u00B7\u00BB\u00BF\u037E\u0387\u055A-\u055F\u0589\u058A\u05BE\u05C0\u05C3\u05C6\u05F3\u05F4\u0609\u060A\u060C\u060D\u061B\u061E\u061F\u066A-\u066D\u06D4\u0700-\u070D\u07F7-\u07F9\u0830-\u083E\u085E\u0964\u0965\u0970\u09FD\u0A76\u0AF0\u0C77\u0C84\u0DF4\u0E4F\u0E5A\u0E5B\u0F04-\u0F12\u0F14\u0F3A-\u0F3D\u0F85\u0FD0-\u0FD4\u0FD9\u0FDA\u104A-\u104F\u10FB\u1360-\u1368\u1400\u166E\u1680\u169B\u169C\u16EB-\u16ED\u1735\u1736\u17D4-\u17D6\u17D8-\u17DA\u1800-\u180A\u1944\u1945\u1A1E\u1A1F\u1AA0-\u1AA6\u1AA8-\u1AAD\u1B5A-\u1B60\u1BFC-\u1BFF\u1C3B-\u1C3F\u1C7E\u1C7F\u1CC0-\u1CC7\u1CD3\u2000-\u200A\u2010-\u2029\u202F-\u2043\u2045-\u2051\u2053-\u205F\u207D\u207E\u208D\u208E\u2308-\u230B\u2329\u232A\u2768-\u2775\u27C5\u27C6\u27E6-\u27EF\u2983-\u2998\u29D8-\u29DB\u29FC\u29FD\u2CF9-\u2CFC\u2CFE\u2CFF\u2D70\u2E00-\u2E2E\u2E30-\u2E4F\u3000-\u3003\u3008-\u3011\u3014-\u301F\u3030\u303D\u30A0\u30FB\uA4FE\uA4FF\uA60D-\uA60F\uA673\uA67E\uA6F2-\uA6F7\uA874-\uA877\uA8CE\uA8CF\uA8F8-\uA8FA\uA8FC\uA92E\uA92F\uA95F\uA9C1-\uA9CD\uA9DE\uA9DF\uAA5C-\uAA5F\uAADE\uAADF\uAAF0\uAAF1\uABEB\uFD3E\uFD3F\uFE10-\uFE19\uFE30-\uFE52\uFE54-\uFE61\uFE63\uFE68\uFE6A\uFE6B\uFF01-\uFF03\uFF05-\uFF0A\uFF0C-\uFF0F\uFF1A\uFF1B\uFF1F\uFF20\uFF3B-\uFF3D\uFF3F\uFF5B\uFF5D\uFF5F-\uFF65]+/u
+  /[|\n\r -#%-*,.\/:;?@[-\]_{}\u00A0\u00A1\u00A7\u00AB\u00B6\u00B7\u00BB\u00BF\u037E\u0387\u055A-\u055F\u0589\u058A\u05BE\u05C0\u05C3\u05C6\u05F3\u05F4\u0609\u060A\u060C\u060D\u061B\u061E\u061F\u066A-\u066D\u06D4\u0700-\u070D\u07F7-\u07F9\u0830-\u083E\u085E\u0964\u0965\u0970\u09FD\u0A76\u0AF0\u0C77\u0C84\u0DF4\u0E4F\u0E5A\u0E5B\u0F04-\u0F12\u0F14\u0F3A-\u0F3D\u0F85\u0FD0-\u0FD4\u0FD9\u0FDA\u104A-\u104F\u10FB\u1360-\u1368\u1400\u166E\u1680\u169B\u169C\u16EB-\u16ED\u1735\u1736\u17D4-\u17D6\u17D8-\u17DA\u1800-\u180A\u1944\u1945\u1A1E\u1A1F\u1AA0-\u1AA6\u1AA8-\u1AAD\u1B5A-\u1B60\u1BFC-\u1BFF\u1C3B-\u1C3F\u1C7E\u1C7F\u1CC0-\u1CC7\u1CD3\u2000-\u200A\u2010-\u2029\u202F-\u2043\u2045-\u2051\u2053-\u205F\u207D\u207E\u208D\u208E\u2308-\u230B\u2329\u232A\u2768-\u2775\u27C5\u27C6\u27E6-\u27EF\u2983-\u2998\u29D8-\u29DB\u29FC\u29FD\u2CF9-\u2CFC\u2CFE\u2CFF\u2D70\u2E00-\u2E2E\u2E30-\u2E4F\u3000-\u3003\u3008-\u3011\u3014-\u301F\u3030\u303D\u30A0\u30FB\uA4FE\uA4FF\uA60D-\uA60F\uA673\uA67E\uA6F2-\uA6F7\uA874-\uA877\uA8CE\uA8CF\uA8F8-\uA8FA\uA8FC\uA92E\uA92F\uA95F\uA9C1-\uA9CD\uA9DE\uA9DF\uAA5C-\uAA5F\uAADE\uAADF\uAAF0\uAAF1\uABEB\uFD3E\uFD3F\uFE10-\uFE19\uFE30-\uFE52\uFE54-\uFE61\uFE63\uFE68\uFE6A\uFE6B\uFF01-\uFF03\uFF05-\uFF0A\uFF0C-\uFF0F\uFF1A\uFF1B\uFF1F\uFF20\uFF3B-\uFF3D\uFF3F\uFF5B\uFF5D\uFF5F-\uFF65]+/u

src/search/omnisearch.ts

@@ -12,6 +12,7 @@ import {
   logDebug,
   removeDiacritics,
   splitCamelCase,
+  splitHyphens,
   stringsToRegex,
   stripMarkdownCharacters,
   warnDebug,
@@ -22,17 +23,22 @@ import { cacheManager } from '../cache-manager'
 import { sortBy } from 'lodash-es'
 
 const tokenize = (text: string): string[] => {
-  const tokens = text.split(SPACE_OR_PUNCTUATION)
+  let tokens = text.split(SPACE_OR_PUNCTUATION)
+
+  // When enabled, we only use the chsSegmenter,
+  // and not the other custom tokenizers
   const chsSegmenter = getChsSegmenter()
   if (chsSegmenter) {
-    return tokens.flatMap(word =>
+    tokens = tokens.flatMap(word =>
       chsRegex.test(word) ? chsSegmenter.cut(word) : [word]
     )
   } else {
-    if (settings.splitCamelCase)
+    // Split camelCase tokens into "camel" and "case
-      return [...tokens, ...tokens.flatMap(splitCamelCase)]
+    tokens = [...tokens, ...tokens.flatMap(splitCamelCase)]
-    return tokens
+    // Split hyphenated tokens
+    tokens = [...tokens, ...tokens.flatMap(splitHyphens)]
   }
+  return tokens
 }
 
 export class Omnisearch {
@@ -182,12 +188,26 @@ export class Omnisearch {
 
     logDebug('Starting search for', query)
 
+    let fuzziness: number
+    switch (settings.fuzziness) {
+      case '0':
+        fuzziness = 0
+        break
+      case '1':
+        fuzziness = 0.1
+        break
+      default:
+        fuzziness = 0.2
+        break
+    }
+
     let results = this.minisearch.search(query.segmentsToStr(), {
       prefix: term => term.length >= options.prefixLength,
       // length <= 3: no fuzziness
       // length <= 5: fuzziness of 10%
       // length > 5: fuzziness of 20%
-      fuzzy: term => (term.length <= 3 ? 0 : term.length <= 5 ? 0.1 : 0.2),
+      fuzzy: term =>
+        term.length <= 3 ? 0 : term.length <= 5 ? fuzziness / 2 : fuzziness,
       combineWith: 'AND',
       boost: {
         basename: settings.weightBasename,

src/search/query.ts

@@ -9,11 +9,10 @@ type Keywords = {
 } & { text: string[] }
 
 export class Query {
-  query: Keywords & { exclude: Keywords }
+  query: Keywords & {
-  /**
+    exclude: Keywords
-   * @deprecated
+  }
-   */
+  #inQuotes: string[]
-  extensions: string[] = []
 
   constructor(text = '') {
     if (settings.ignoreDiacritics) {
@@ -44,7 +43,10 @@ export class Query {
       }
     }
     this.query = parsed
-    this.extensions = this.query.ext ?? []
+
+    // Get strings in quotes, and remove the quotes
+    this.#inQuotes =
+      text.match(/"([^"]+)"/g)?.map(o => o.replace(/"/g, '')) ?? []
   }
 
   public isEmpty(): boolean {
@@ -72,6 +74,11 @@ export class Query {
   }
 
   public getExactTerms(): string[] {
-    return this.query.text.filter(o => o.split(' ').length > 1)
+    return [
+      ...new Set([
+        ...this.query.text.filter(o => o.split(' ').length > 1),
+        ...this.#inQuotes,
+      ]),
+    ]
   }
 }

src/settings.ts

@@ -34,8 +34,12 @@ export interface OmnisearchSettings extends WeightingSettings {
   indexedFileTypes: string[]
   /** Enable PDF indexing */
   PDFIndexing: boolean
-  /** Enable PDF indexing */
+  /** Enable Images indexing */
   imagesIndexing: boolean
+  /** Enable Dataloom indexing */
+  dataloomIndexing: boolean
+  /** Enable indexing of unknown files */
+  unsupportedFilesIndexing: 'yes' | 'no' | 'default'
   /** Activate the small 🔍 button on Obsidian's ribbon */
   ribbonIcon: boolean
   /** Display the small contextual excerpt in search results */
@@ -54,6 +58,7 @@ export interface OmnisearchSettings extends WeightingSettings {
   splitCamelCase: boolean
   openInNewPane: boolean
   verboseLogging: boolean
+  fuzziness: '0' | '1' | '2'
 }
 
 /**
@@ -144,28 +149,66 @@ export class SettingsTab extends PluginSettingTab {
       )
       .setDisabled(!getTextExtractor())
 
-    // Additional files to index
+    // Dataloom Indexing
+    const indexDataLoomDesc = new DocumentFragment()
+    indexDataLoomDesc.createSpan({}, span => {
+      span.innerHTML = `Include <a href="https://github.com/trey-wallis/obsidian-dataloom">DataLoom</a> <pre style="display:inline">.loom</pre> files in search results
+      <br/>${needsARestart}`
+    })
+    new Setting(containerEl)
+      .setName('DataLoom indexing (beta)')
+      .setDesc(indexDataLoomDesc)
+      .addToggle(toggle =>
+        toggle.setValue(settings.dataloomIndexing).onChange(async v => {
+          settings.dataloomIndexing = v
+          await saveSettings(this.plugin)
+        })
+      )
+      .setDisabled(!getTextExtractor())
+
+    // Additional text files to index
     const indexedFileTypesDesc = new DocumentFragment()
     indexedFileTypesDesc.createSpan({}, span => {
-      span.innerHTML = `In addition to standard <code>md</code> files, Omnisearch can also index other <strong style="color: var(--text-accent)">plaintext</strong> files.<br/>
+      span.innerHTML = `In addition to standard <code>md</code> files, Omnisearch can also index other <strong style="color: var(--text-accent)">PLAINTEXT</strong> files.<br/>
-      Add extensions separated by a space, without the dot. Example: "<code>txt org</code>".<br />
+      Add extensions separated by a space, without the dot. Example: "<code>txt org csv</code>".<br />
       ⚠️ <span style="color: var(--text-accent)">Using extensions of non-plaintext files (like .docx or .pptx) WILL cause crashes,
       because Omnisearch will try to index their content.</span><br />
       ${needsARestart}`
     })
     new Setting(containerEl)
-      .setName('Additional files to index')
+      .setName('Additional TEXT files to index')
       .setDesc(indexedFileTypesDesc)
       .addText(component => {
         component
           .setValue(settings.indexedFileTypes.join(' '))
-          .setPlaceholder('Example: txt org')
+          .setPlaceholder('Example: txt org csv')
           .onChange(async v => {
             settings.indexedFileTypes = v.split(' ')
             await saveSettings(this.plugin)
           })
       })
 
+    // Unsupported files
+    const indexUnsupportedDesc = new DocumentFragment()
+    indexUnsupportedDesc.createSpan({}, span => {
+      span.innerHTML = `
+      Omnisearch can index file<strong>names</strong> of "unsupported" files, such as e.g. <pre style="display:inline">.mp4</pre> or <pre style="display:inline">.xlsx</pre>.<br/>
+      "Obsidian setting" will respect the value of "<em>Files & Links > Detect all file extensions</em>".
+      <br />${needsARestart}`
+    })
+    new Setting(containerEl)
+      .setName('Index unsupported files (beta)')
+      .setDesc(indexUnsupportedDesc)
+      .addDropdown(dropdown => {
+        dropdown
+          .addOptions({ yes: 'Yes', no: 'No', default: 'Obsidian setting' })
+          .setValue(settings.unsupportedFilesIndexing)
+          .onChange(async v => {
+            ;(settings.unsupportedFilesIndexing as any) = v
+            await saveSettings(this.plugin)
+          })
+      })
+
     //#endregion Indexing
 
     //#region Behavior
@@ -264,6 +307,29 @@ export class SettingsTab extends PluginSettingTab {
         })
       )
 
+    // Fuzziness
+    new Setting(containerEl)
+      .setName('Fuzziness')
+      .setDesc(
+        "Define the level of fuzziness for the search. The higher the fuzziness, the more results you'll get."
+      )
+      .addDropdown(dropdown =>
+        dropdown
+          .addOptions({
+            0: 'Exact match',
+            1: 'Not too fuzzy',
+            2: 'Fuzzy enough',
+          })
+          .setValue(settings.fuzziness)
+          .onChange(async v => {
+            if (!['0', '1', '2'].includes(v)) {
+              v = '2'
+            }
+            settings.fuzziness = v as '0' | '1' | '2'
+            await saveSettings(this.plugin)
+          })
+      )
+
     //#endregion Behavior
 
     //#region User Interface
@@ -466,6 +532,8 @@ export const DEFAULT_SETTINGS: OmnisearchSettings = {
   indexedFileTypes: [] as string[],
   PDFIndexing: false,
   imagesIndexing: false,
+  dataloomIndexing: false,
+  unsupportedFilesIndexing: 'no',
   splitCamelCase: false,
   openInNewPane: false,
 
@@ -476,6 +544,7 @@ export const DEFAULT_SETTINGS: OmnisearchSettings = {
   highlight: true,
   showPreviousQueryResults: true,
   simpleSearch: false,
+  fuzziness: '0',
 
   weightBasename: 3,
   weightDirectory: 2,
@@ -500,5 +569,13 @@ export async function saveSettings(plugin: Plugin): Promise<void> {
 }
 
 export function isPluginDisabled(): boolean {
-  return app.loadLocalStorage(K_DISABLE_OMNISEARCH) == '1'
+  return app.loadLocalStorage(K_DISABLE_OMNISEARCH) === '1'
+}
+
+export function canIndexUnsupportedFiles(): boolean {
+  return (
+    settings.unsupportedFilesIndexing === 'yes' ||
+    (settings.unsupportedFilesIndexing === 'default' &&
+      !!app.vault.getConfig('showUnsupportedFiles'))
+  )
 }

src/tools/utils.ts

@@ -18,7 +18,7 @@ import {
   SPACE_OR_PUNCTUATION,
   type SearchMatch,
 } from '../globals'
-import { settings } from '../settings'
+import { canIndexUnsupportedFiles, settings } from '../settings'
 import { type BinaryLike, createHash } from 'crypto'
 import { md5 } from 'pure-md5'
 
@@ -26,14 +26,10 @@ export function highlighter(str: string): string {
   return `<span class="${highlightClass}">${str}</span>`
 }
 
-export function highlighterGroups(...args: any[]) {
+export function highlighterGroups(_substring: string, ...args: any[]) {
-  if (
+  // args[0] is the single char preceding args[1], which is the word we want to highlight
-    args[1] !== null &&
+  if (!!args[1].trim())
-    args[1] !== undefined &&
+    return `<span>${args[0]}</span><span class="${highlightClass}">${args[1]}</span>`
-    args[2] !== null &&
-    args[2] !== undefined
-  )
-    return `<span>${args[1]}</span><span class="${highlightClass}">${args[2]}</span>`
   return '&lt;no content&gt;'
 }
 
@@ -91,15 +87,20 @@ export function getAllIndices(text: string, regex: RegExp): SearchMatch[] {
  */
 export function stringsToRegex(strings: string[]): RegExp {
   if (!strings.length) return /^$/g
+
+  // sort strings by decreasing length, so that longer strings are matched first
+  strings.sort((a, b) => b.length - a.length)
+
   const joined =
     '(' +
     // Default word split is not applied if the user uses the cm-chs-patch plugin
     (getChsSegmenter()
       ? ''
       : // Split on start of line, spaces, punctuation, or capital letters (for camelCase)
+      // We also add the hyphen to the list of characters that can split words
       settings.splitCamelCase
-      ? `^|${SPACE_OR_PUNCTUATION.source}|[A-Z]`
+      ? `^|${SPACE_OR_PUNCTUATION.source}|\-|[A-Z]`
-      : `^|${SPACE_OR_PUNCTUATION.source}`) +
+      : `^|${SPACE_OR_PUNCTUATION.source}|\-`) +
     ')' +
     `(${strings.map(s => escapeRegex(s)).join('|')})`
 
@@ -262,18 +263,34 @@ export function getCtrlKeyLabel(): 'ctrl' | '⌘' {
   return Platform.isMacOS ? '⌘' : 'ctrl'
 }
 
-export function isFileIndexable(path: string): boolean {
+export function isContentIndexable(path: string): boolean {
   const hasTextExtractor = !!getTextExtractor()
   const canIndexPDF = hasTextExtractor && settings.PDFIndexing
   const canIndexImages = hasTextExtractor && settings.imagesIndexing
   return (
     isFilePlaintext(path) ||
     isFileCanvas(path) ||
+    isFileFromDataloomPlugin(path) ||
     (canIndexPDF && isFilePDF(path)) ||
     (canIndexImages && isFileImage(path))
   )
 }
 
+export function isFilenameIndexable(path: string): boolean {
+  return (
+    canIndexUnsupportedFiles() ||
+    isFilePlaintext(path) ||
+    isFileCanvas(path) ||
+    isFileFromDataloomPlugin(path) ||
+    isFilePDF(path) ||
+    isFileImage(path)    
+  )
+}
+
+export function isFileIndexable(path: string): boolean {
+  return isFilenameIndexable(path) || isContentIndexable(path)
+}
+
 export function isFileImage(path: string): boolean {
   const ext = getExtension(path)
   return ext === 'png' || ext === 'jpg' || ext === 'jpeg' || ext === 'webp'
@@ -291,6 +308,10 @@ export function isFileCanvas(path: string): boolean {
   return path.endsWith('.canvas')
 }
 
+export function isFileFromDataloomPlugin(path: string): boolean {
+  return path.endsWith('.loom') || path.endsWith('.dashboard')
+}
+
 export function getExtension(path: string): string {
   const split = path.split('.')
   return split[split.length - 1] ?? ''
@@ -318,11 +339,32 @@ export function chunkArray<T>(arr: T[], len: number): T[][] {
 }
 
 /**
- * Converts a 'fooBarBAZLorem' into ['foo', 'Bar', 'BAZ', 'Lorem]
+ * Converts a 'fooBarBAZLorem' into ['foo', 'Bar', 'BAZ', 'Lorem']
+ * If the string isn't camelCase, returns an empty array
  * @param text
  */
 export function splitCamelCase(text: string): string[] {
-  return text.replace(/([a-z](?=[A-Z]))/g, '$1 ').split(' ')
+  const split = text
+    .replace(/([a-z](?=[A-Z]))/g, '$1 ')
+    .split(' ')
+    .filter(t => t)
+  if (split.length > 1) {
+    return split
+  }
+  return []
+}
+
+/**
+ * Converts a 'foo-bar-baz' into ['foo', 'bar', 'baz']
+ * If the string isn't hyphenated, returns an empty array
+ * @param text
+ */
+export function splitHyphens(text: string): string[] {
+  const split = text.split('-').filter(t => t)
+  if (split.length > 1) {
+    return split
+  }
+  return []
 }
 
 export function logDebug(...args: any[]): void {