WIP document mapping

src/file-loader.ts

@@ -10,6 +10,7 @@ import {
 import type { TFile } from 'obsidian'
 import type { IndexedDocument } from './globals'
 import { getImageText, getPdfText } from 'obsidian-text-extract'
+import { cacheManager } from "./cache-manager";
 
 /**
  * Return all plaintext files as IndexedDocuments
@@ -18,7 +19,7 @@ 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 getIndexedDocument(file.path)
+    const doc = await cacheManager.getDocument(file.path)
     data.push(doc)
     // await cacheManager.updateLiveDocument(file.path, doc)
   }
@@ -47,7 +48,7 @@ async function getBinaryFiles(files: TFile[]): Promise<IndexedDocument[]> {
   for (const file of files) {
     input.push(
       new Promise(async (resolve, _reject) => {
-        const doc = await getIndexedDocument(file.path)
+        const doc = await cacheManager.getDocument(file.path)
         data.push(doc)
         return resolve(null)
       })
@@ -57,61 +58,6 @@ async function getBinaryFiles(files: TFile[]): Promise<IndexedDocument[]> {
   return data
 }
 
-export async function getIndexedDocument(
-  path: string
-): Promise<IndexedDocument> {
-  const file = app.vault.getFiles().find(f => f.path === path)
-  if (!file) throw new Error(`Invalid file path: "${path}"`)
-  let content: string
-  if (isFilePlaintext(path)) {
-    content = await app.vault.cachedRead(file)
-  } else if (isFilePDF(path)) {
-    content = await getPdfText(file)
-  } else if (isFileImage(file.path)) {
-    content = await getImageText(file)
-  } else {
-    throw new Error('Invalid file format: ' + file.path)
-  }
-  content = removeDiacritics(content)
-  const metadata = app.metadataCache.getFileCache(file)
-
-  // Look for links that lead to non-existing files,
-  // and add them to the index.
-  if (metadata) {
-    // // FIXME: https://github.com/scambier/obsidian-omnisearch/issues/129
-    // const nonExisting = getNonExistingNotes(file, metadata)
-    // for (const name of nonExisting.filter(
-    //   o => !cacheManager.getLiveDocument(o)
-    // )) {
-    //   NotesIndex.addNonExistingToIndex(name, file.path)
-    // }
-
-    // EXCALIDRAW
-    // Remove the json code
-    if (metadata.frontmatter?.['excalidraw-plugin']) {
-      const comments =
-        metadata.sections?.filter(s => s.type === 'comment') ?? []
-      for (const { start, end } of comments.map(c => c.position)) {
-        content =
-          content.substring(0, start.offset - 1) + content.substring(end.offset)
-      }
-    }
-  }
-
-  return {
-    basename: removeDiacritics(file.basename),
-    content,
-    path: file.path,
-    mtime: file.stat.mtime,
-
-    tags: getTagsFromMetadata(metadata),
-    aliases: getAliasesFromMetadata(metadata).join(''),
-    headings1: metadata ? extractHeadingsFromCache(metadata, 1).join(' ') : '',
-    headings2: metadata ? extractHeadingsFromCache(metadata, 2).join(' ') : '',
-    headings3: metadata ? extractHeadingsFromCache(metadata, 3).join(' ') : '',
-  }
-}
-
 /**
  * Convert a file into an IndexedDocument.
  * Will use the cache if possible.