enh/refac: temp chat file upload behaviour

client-side content extraction

src/lib/components/chat/Chat.svelte

@@ -1597,7 +1597,7 @@
 		let files = JSON.parse(JSON.stringify(chatFiles));
 		files.push(
 			...(userMessage?.files ?? []).filter((item) =>
-				['doc', 'file', 'note', 'collection'].includes(item.type)
+				['doc', 'text', 'file', 'note', 'collection'].includes(item.type)
 			)
 		);
 		// Remove duplicates

src/lib/components/chat/MessageInput.svelte

@@ -1,4 +1,8 @@
 <script lang="ts">
+	import * as pdfjs from 'pdfjs-dist';
+	import * as pdfWorker from 'pdfjs-dist/build/pdf.worker.mjs';
+	pdfjs.GlobalWorkerOptions.workerSrc = import.meta.url + 'pdfjs-dist/build/pdf.worker.mjs';
+
 	import DOMPurify from 'dompurify';
 	import { marked } from 'marked';
 	import heic2any from 'heic2any';
@@ -23,13 +27,15 @@
 		tools,
 		user as _user,
 		showControls,
-		TTSWorker
+		TTSWorker,
+		temporaryChatEnabled
 	} from '$lib/stores';
 
 	import {
 		blobToFile,
 		compressImage,
 		createMessagesList,
+		extractContentFromFile,
 		extractCurlyBraceWords,
 		extractInputVariables,
 		getCurrentDateTime,
@@ -529,47 +535,77 @@
 
 		files = [...files, fileItem];
 
-		try {
-			// If the file is an audio file, provide the language for STT.
-			let metadata = null;
-			if (
-				(file.type.startsWith('audio/') || file.type.startsWith('video/')) &&
-				$settings?.audio?.stt?.language
-			) {
-				metadata = {
-					language: $settings?.audio?.stt?.language
-				};
-			}
-
-			// During the file upload, file content is automatically extracted.
-			const uploadedFile = await uploadFile(localStorage.token, file, metadata);
-
-			if (uploadedFile) {
-				console.log('File upload completed:', {
-					id: uploadedFile.id,
-					name: fileItem.name,
-					collection: uploadedFile?.meta?.collection_name
-				});
-
-				if (uploadedFile.error) {
-					console.warn('File upload warning:', uploadedFile.error);
-					toast.warning(uploadedFile.error);
+		if (!$temporaryChatEnabled) {
+			try {
+				// If the file is an audio file, provide the language for STT.
+				let metadata = null;
+				if (
+					(file.type.startsWith('audio/') || file.type.startsWith('video/')) &&
+					$settings?.audio?.stt?.language
+				) {
+					metadata = {
+						language: $settings?.audio?.stt?.language
+					};
 				}
 
-				fileItem.status = 'uploaded';
-				fileItem.file = uploadedFile;
-				fileItem.id = uploadedFile.id;
-				fileItem.collection_name =
-					uploadedFile?.meta?.collection_name || uploadedFile?.collection_name;
-				fileItem.url = `${WEBUI_API_BASE_URL}/files/${uploadedFile.id}`;
+				// During the file upload, file content is automatically extracted.
+				const uploadedFile = await uploadFile(localStorage.token, file, metadata);
 
-				files = files;
-			} else {
+				if (uploadedFile) {
+					console.log('File upload completed:', {
+						id: uploadedFile.id,
+						name: fileItem.name,
+						collection: uploadedFile?.meta?.collection_name
+					});
+
+					if (uploadedFile.error) {
+						console.warn('File upload warning:', uploadedFile.error);
+						toast.warning(uploadedFile.error);
+					}
+
+					fileItem.status = 'uploaded';
+					fileItem.file = uploadedFile;
+					fileItem.id = uploadedFile.id;
+					fileItem.collection_name =
+						uploadedFile?.meta?.collection_name || uploadedFile?.collection_name;
+					fileItem.url = `${WEBUI_API_BASE_URL}/files/${uploadedFile.id}`;
+
+					files = files;
+				} else {
+					files = files.filter((item) => item?.itemId !== tempItemId);
+				}
+			} catch (e) {
+				toast.error(`${e}`);
 				files = files.filter((item) => item?.itemId !== tempItemId);
 			}
-		} catch (e) {
-			toast.error(`${e}`);
-			files = files.filter((item) => item?.itemId !== tempItemId);
+		} else {
+			// If temporary chat is enabled, we just add the file to the list without uploading it.
+
+			const content = await extractContentFromFile(file, pdfjsLib).catch((error) => {
+				toast.error(
+					$i18n.t('Failed to extract content from the file: {{error}}', { error: error })
+				);
+				return null;
+			});
+
+			if (content === null) {
+				toast.error($i18n.t('Failed to extract content from the file.'));
+				files = files.filter((item) => item?.itemId !== tempItemId);
+				return null;
+			} else {
+				console.log('Extracted content from file:', {
+					name: file.name,
+					size: file.size,
+					content: content
+				});
+
+				fileItem.status = 'uploaded';
+				fileItem.type = 'text';
+				fileItem.content = content;
+				fileItem.id = uuidv4(); // Temporary ID for the file
+
+				files = files;
+			}
 		}
 	};
 

src/lib/utils/index.ts

@@ -1507,3 +1507,74 @@ export const parseJsonValue = (value: string): any => {
 
 	return value;
 };
+
+export const extractContentFromFile = async (file, pdfjsLib = null) => {
+	// Known text file extensions for extra fallback
+	const textExtensions = [
+		'.txt',
+		'.md',
+		'.csv',
+		'.json',
+		'.js',
+		'.ts',
+		'.css',
+		'.html',
+		'.xml',
+		'.yaml',
+		'.yml',
+		'.rtf'
+	];
+
+	function getExtension(filename) {
+		const dot = filename.lastIndexOf('.');
+		return dot === -1 ? '' : filename.substr(dot).toLowerCase();
+	}
+
+	// Uses pdfjs to extract text from PDF
+	async function extractPdfText(file) {
+		if (!pdfjsLib) {
+			throw new Error('pdfjsLib is required for PDF extraction');
+		}
+
+		const arrayBuffer = await file.arrayBuffer();
+		const pdf = await pdfjsLib.getDocument({ data: arrayBuffer }).promise;
+		let allText = '';
+		for (let pageNum = 1; pageNum <= pdf.numPages; pageNum++) {
+			const page = await pdf.getPage(pageNum);
+			const content = await page.getTextContent();
+			const strings = content.items.map((item) => item.str);
+			allText += strings.join(' ') + '\n';
+		}
+		return allText;
+	}
+
+	// Reads file as text using FileReader
+	function readAsText(file) {
+		return new Promise((resolve, reject) => {
+			const reader = new FileReader();
+			reader.onload = () => resolve(reader.result);
+			reader.onerror = reject;
+			reader.readAsText(file);
+		});
+	}
+
+	const type = file.type || '';
+	const ext = getExtension(file.name);
+
+	// PDF check
+	if (type === 'application/pdf' || ext === '.pdf') {
+		return await extractPdfText(file);
+	}
+
+	// Text check (plain or common text-based)
+	if (type.startsWith('text/') || textExtensions.includes(ext)) {
+		return await readAsText(file);
+	}
+
+	// Fallback: try to read as text, if decodable
+	try {
+		return await readAsText(file);
+	} catch (err) {
+		throw new Error('Unsupported or non-text file type: ' + (file.name || type));
+	}
+};