380 lines
11 KiB
TypeScript
380 lines
11 KiB
TypeScript
import {
|
|
BadRequestException,
|
|
Injectable,
|
|
ServiceUnavailableException,
|
|
UnsupportedMediaTypeException,
|
|
} from '@nestjs/common';
|
|
import * as fs from 'node:fs/promises';
|
|
import * as path from 'node:path';
|
|
import * as pdfParse from 'pdf-parse';
|
|
import { createWorker } from 'tesseract.js';
|
|
import { IcaRecipeParser } from './parsers/ica.parser';
|
|
import { GenericRecipeParser } from './parsers/generic.parser';
|
|
import { RecipeParser } from './parsers/base.parser';
|
|
|
|
export interface QuickImportResult {
|
|
markdown: string;
|
|
source: 'ica' | 'pdf' | 'image' | 'other';
|
|
}
|
|
|
|
type UploadKind = 'pdf' | 'image';
|
|
|
|
@Injectable()
|
|
export class QuickImportService {
|
|
/**
|
|
* Detekterar typ av input (URL eller filsökväg) och importerar från lämplig källa
|
|
*/
|
|
async importFromInput(input: string): Promise<QuickImportResult> {
|
|
const trimmed = input.trim();
|
|
console.log('[QuickImport] Mottog input:', trimmed);
|
|
|
|
if (!trimmed) {
|
|
throw new BadRequestException('Du måste ange en URL eller ladda upp en fil');
|
|
}
|
|
|
|
if (this.isUrl(trimmed)) {
|
|
console.log('[QuickImport] Detekterade URL, försöker scrapa...');
|
|
return this.scrapeRecipeFromUrl(trimmed);
|
|
}
|
|
|
|
if (this.looksLikeLocalFile(trimmed)) {
|
|
console.log('[QuickImport] Försöker läsa lokal fil:', trimmed);
|
|
try {
|
|
const buffer = await fs.readFile(trimmed);
|
|
return this.importFromUpload({
|
|
buffer,
|
|
originalname: path.basename(trimmed),
|
|
mimetype: this.getMimeTypeFromExtension(trimmed),
|
|
} as Express.Multer.File);
|
|
} catch (error) {
|
|
console.error('[QuickImport] Kunde inte läsa lokal fil:', error);
|
|
throw new BadRequestException(
|
|
'Kunde inte läsa filen. Använd filuppladdning i gränssnittet eller kontrollera sökvägen.',
|
|
);
|
|
}
|
|
}
|
|
|
|
throw new BadRequestException(
|
|
'Ogiltig input. Ange en giltig URL eller ladda upp en PDF- eller bildfil.',
|
|
);
|
|
}
|
|
|
|
async importFromUpload(file: Express.Multer.File): Promise<QuickImportResult> {
|
|
if (!file?.buffer) {
|
|
throw new BadRequestException('Ingen fil skickades med.');
|
|
}
|
|
|
|
console.log('[QuickImport] Mottog uppladdad fil:', file.originalname, file.mimetype);
|
|
const kind = this.getUploadKind(file);
|
|
|
|
if (kind === 'pdf') {
|
|
const text = await this.extractTextFromPdf(file.buffer);
|
|
return {
|
|
markdown: this.normalizeImportedTextToMarkdown(text, file.originalname),
|
|
source: 'pdf',
|
|
};
|
|
}
|
|
|
|
const text = await this.extractTextFromImage(file.buffer);
|
|
return {
|
|
markdown: this.normalizeImportedTextToMarkdown(text, file.originalname),
|
|
source: 'image',
|
|
};
|
|
}
|
|
|
|
/**
|
|
* Kontrollerar om input är en URL
|
|
*/
|
|
private isUrl(input: string): boolean {
|
|
try {
|
|
new URL(input);
|
|
return true;
|
|
} catch {
|
|
return false;
|
|
}
|
|
}
|
|
|
|
private looksLikeLocalFile(input: string): boolean {
|
|
const normalized = input.toLowerCase();
|
|
return /[\\/]/.test(input) || /\.(pdf|png|jpg|jpeg|webp|bmp)$/i.test(normalized);
|
|
}
|
|
|
|
private getMimeTypeFromExtension(filename: string): string {
|
|
const ext = path.extname(filename).toLowerCase();
|
|
|
|
if (ext === '.pdf') return 'application/pdf';
|
|
if (ext === '.png') return 'image/png';
|
|
if (ext === '.jpg' || ext === '.jpeg') return 'image/jpeg';
|
|
if (ext === '.webp') return 'image/webp';
|
|
if (ext === '.bmp') return 'image/bmp';
|
|
|
|
return 'application/octet-stream';
|
|
}
|
|
|
|
private getUploadKind(
|
|
file: Pick<Express.Multer.File, 'mimetype' | 'originalname'>,
|
|
): UploadKind {
|
|
const type = (file.mimetype ?? '').toLowerCase();
|
|
const name = (file.originalname ?? '').toLowerCase();
|
|
|
|
if (type.includes('pdf') || name.endsWith('.pdf')) {
|
|
return 'pdf';
|
|
}
|
|
|
|
if (
|
|
type.startsWith('image/') ||
|
|
['.png', '.jpg', '.jpeg', '.webp', '.bmp'].some((ext) => name.endsWith(ext))
|
|
) {
|
|
return 'image';
|
|
}
|
|
|
|
throw new UnsupportedMediaTypeException(
|
|
'Endast PDF, PNG, JPG, JPEG, WEBP och BMP stöds.',
|
|
);
|
|
}
|
|
|
|
private async extractTextFromPdf(buffer: Buffer): Promise<string> {
|
|
try {
|
|
const result = await pdfParse(buffer);
|
|
const text = result.text?.replace(/\u0000/g, '').trim();
|
|
|
|
if (!text) {
|
|
throw new BadRequestException(
|
|
'PDF-filen saknar läsbar text. Prova bildimport om det är en skannad sida.',
|
|
);
|
|
}
|
|
|
|
return text;
|
|
} catch (error) {
|
|
if (error instanceof BadRequestException) {
|
|
throw error;
|
|
}
|
|
|
|
console.error('[QuickImport] PDF ERROR:', error);
|
|
throw new ServiceUnavailableException('PDF-importen misslyckades.');
|
|
}
|
|
}
|
|
|
|
private async extractTextFromImage(buffer: Buffer): Promise<string> {
|
|
const worker = await createWorker('swe+eng');
|
|
|
|
try {
|
|
const result = await worker.recognize(buffer);
|
|
const text = result.data.text?.trim();
|
|
|
|
if (!text) {
|
|
throw new BadRequestException('Ingen text hittades i bilden.');
|
|
}
|
|
|
|
return text;
|
|
} catch (error) {
|
|
if (error instanceof BadRequestException) {
|
|
throw error;
|
|
}
|
|
|
|
console.error('[QuickImport] OCR ERROR:', error);
|
|
throw new ServiceUnavailableException('OCR-importen misslyckades.');
|
|
} finally {
|
|
await worker.terminate();
|
|
}
|
|
}
|
|
|
|
private normalizeImportedTextToMarkdown(text: string, sourceName?: string): string {
|
|
const cleanedText = text
|
|
.replace(/\r/g, '')
|
|
.replace(/[ \t]+/g, ' ')
|
|
.replace(/\n{3,}/g, '\n\n')
|
|
.trim();
|
|
|
|
if (!cleanedText) {
|
|
throw new BadRequestException('Ingen läsbar text hittades i filen.');
|
|
}
|
|
|
|
const title = cleanedText.split('\n').find((line) => line.trim().length > 3)?.trim() ?? 'Importerat recept';
|
|
const ingredients: string[] = [];
|
|
const instructions: string[] = [];
|
|
let section: 'unknown' | 'ingredients' | 'instructions' = 'unknown';
|
|
|
|
for (const rawLine of cleanedText.split('\n')) {
|
|
const line = rawLine.trim();
|
|
if (!line || line === title) {
|
|
continue;
|
|
}
|
|
|
|
const lower = line.toLowerCase();
|
|
|
|
if (/^ingred/i.test(lower)) {
|
|
section = 'ingredients';
|
|
continue;
|
|
}
|
|
|
|
if (/^(gör så här|gor sa har|instruktioner|tillvägagångssätt|tillvagagangssatt|method|instructions)/i.test(lower)) {
|
|
section = 'instructions';
|
|
continue;
|
|
}
|
|
|
|
if (section === 'unknown') {
|
|
section = this.looksLikeIngredientLine(line) ? 'ingredients' : 'instructions';
|
|
}
|
|
|
|
if (section === 'ingredients') {
|
|
ingredients.push(line.startsWith('-') ? line : `- ${line}`);
|
|
} else {
|
|
instructions.push(line);
|
|
}
|
|
}
|
|
|
|
return [
|
|
`# ${title}`,
|
|
'',
|
|
'## Ingredienser',
|
|
...(ingredients.length > 0 ? ingredients : ['- Komplettera ingredienser manuellt']),
|
|
'',
|
|
'## Tillvägagångssätt',
|
|
...(instructions.length > 0 ? instructions : ['Komplettera tillagningsstegen manuellt.']),
|
|
'',
|
|
sourceName ? `Källa: ${sourceName}` : '',
|
|
]
|
|
.filter(Boolean)
|
|
.join('\n');
|
|
}
|
|
|
|
private looksLikeIngredientLine(line: string): boolean {
|
|
return (
|
|
/^[-*•]\s+/.test(line) ||
|
|
/^\d+[.,]?\d*\s+/.test(line) ||
|
|
/\b(g|kg|hg|mg|ml|dl|cl|l|tsk|msk|krm|st|pkt|förp|klyfta)\b/i.test(line)
|
|
);
|
|
}
|
|
|
|
/**
|
|
* Skrapar recept från en URL
|
|
*
|
|
* Använder site-specifika parsers om tillgängliga,
|
|
* annars fallback till generisk parser.
|
|
*
|
|
* @param url URL till receptsidan
|
|
* @returns Markdown-format
|
|
*/
|
|
private async scrapeRecipeFromUrl(url: string): Promise<QuickImportResult> {
|
|
try {
|
|
console.log('[QuickImport] Hämtar HTML från:', url);
|
|
|
|
const response = await fetch(url, {
|
|
headers: {
|
|
'User-Agent':
|
|
'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36',
|
|
},
|
|
});
|
|
|
|
console.log('[QuickImport] HTTP status:', response.status);
|
|
|
|
if (!response.ok) {
|
|
throw new Error(`HTTP ${response.status}: ${response.statusText}`);
|
|
}
|
|
|
|
const html = await response.text();
|
|
console.log('[QuickImport] HTML längd:', html.length, 'tecken');
|
|
|
|
const parsers: RecipeParser[] = [
|
|
new IcaRecipeParser(),
|
|
new GenericRecipeParser(),
|
|
];
|
|
|
|
let recipe = null;
|
|
for (const parser of parsers) {
|
|
if (parser.canHandle(url)) {
|
|
console.log('[QuickImport] Använder parser:', parser.constructor.name);
|
|
recipe = parser.parse(html);
|
|
break;
|
|
}
|
|
}
|
|
|
|
if (!recipe) {
|
|
throw new Error('Ingen parserutrustning tillgänglig');
|
|
}
|
|
|
|
console.log('[QuickImport] Parsad recept:', {
|
|
name: recipe.name,
|
|
ingredienser: recipe.ingredients.length,
|
|
});
|
|
|
|
if (!recipe.name) {
|
|
throw new Error('Kunde inte hitta receptnamn på sidan. Försök med en annan länk.');
|
|
}
|
|
|
|
const markdown = this.recipeToMarkdown(recipe, url);
|
|
console.log('[QuickImport] Markdown genererad, längd:', markdown.length);
|
|
|
|
let source: 'ica' | 'pdf' | 'image' | 'other' = 'other';
|
|
if (/ica\.se/i.test(url)) {
|
|
source = 'ica';
|
|
}
|
|
|
|
return {
|
|
markdown,
|
|
source,
|
|
};
|
|
} catch (err) {
|
|
const message = err instanceof Error ? err.message : 'Okänt fel vid scraping';
|
|
console.error('[QuickImport] ERROR:', message);
|
|
throw new BadRequestException(
|
|
`Kunde inte hämta recept: ${message}. Kontrollera att länken är korrekt och försök igen.`
|
|
);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Konvertera receptobjekt till Markdown-format
|
|
*/
|
|
private recipeToMarkdown(
|
|
recipe: {
|
|
name: string;
|
|
description?: string;
|
|
ingredients: Array<{
|
|
quantity: number;
|
|
unit: string;
|
|
name: string;
|
|
note?: string;
|
|
}>;
|
|
instructions?: string;
|
|
},
|
|
sourceUrl?: string,
|
|
): string {
|
|
const lines: string[] = [];
|
|
|
|
lines.push(`# ${recipe.name}`);
|
|
lines.push('');
|
|
|
|
if (recipe.description) {
|
|
lines.push(recipe.description);
|
|
lines.push('');
|
|
}
|
|
|
|
if (recipe.ingredients.length > 0) {
|
|
lines.push('## Ingredienser');
|
|
for (const ing of recipe.ingredients) {
|
|
const quantity = ing.quantity > 0 ? `${ing.quantity} ` : '';
|
|
const unit = ing.unit ? `${ing.unit} ` : '';
|
|
const note = ing.note ? ` (${ing.note})` : '';
|
|
lines.push(`- ${quantity}${unit}${ing.name}${note}`);
|
|
}
|
|
lines.push('');
|
|
}
|
|
|
|
if (recipe.instructions) {
|
|
lines.push('## Tillvägagångssätt');
|
|
lines.push(recipe.instructions);
|
|
lines.push('');
|
|
}
|
|
|
|
if (sourceUrl) {
|
|
lines.push('---');
|
|
lines.push('');
|
|
lines.push(`Källa: [${sourceUrl}](${sourceUrl})`);
|
|
}
|
|
|
|
return lines.join('\n');
|
|
}
|
|
}
|