feat: Add receipt parsing functionality with image and PDF support, including Tesseract OCR integration
This commit is contained in:
@@ -19,6 +19,12 @@ FROM node:22-alpine AS runner
|
||||
WORKDIR /app
|
||||
ENV NODE_ENV=production
|
||||
|
||||
# Tesseract OCR systempaket (krävs av tesseract.js)
|
||||
RUN apk add --no-cache \
|
||||
tesseract-ocr \
|
||||
tesseract-ocr-data-swe \
|
||||
tesseract-ocr-data-eng
|
||||
|
||||
COPY --from=builder /app/package.json ./package.json
|
||||
COPY --from=builder /app/node_modules ./node_modules
|
||||
COPY --from=builder /app/dist ./dist
|
||||
|
||||
@@ -1,13 +1,24 @@
|
||||
import { Module } from '@nestjs/common';
|
||||
import { Module, Controller, Get } from '@nestjs/common';
|
||||
import { WebScrapingModule } from './web-scraping-service/web-scraping.module';
|
||||
import { RecipesModule } from './recipes/recipes.module';
|
||||
import { DocumentServiceModule } from './document-service/document-service.module';
|
||||
import { ReceiptParsingModule } from './receipt-parsing/receipt-parsing.module';
|
||||
|
||||
@Controller('health')
|
||||
class HealthController {
|
||||
@Get()
|
||||
check() {
|
||||
return { status: 'ok' };
|
||||
}
|
||||
}
|
||||
|
||||
@Module({
|
||||
imports: [
|
||||
DocumentServiceModule,
|
||||
WebScrapingModule,
|
||||
RecipesModule,
|
||||
ReceiptParsingModule,
|
||||
],
|
||||
controllers: [HealthController],
|
||||
})
|
||||
export class AppModule {}
|
||||
|
||||
@@ -0,0 +1,56 @@
|
||||
import {
|
||||
Controller,
|
||||
HttpCode,
|
||||
Post,
|
||||
UploadedFile,
|
||||
UseInterceptors,
|
||||
BadRequestException,
|
||||
} from '@nestjs/common';
|
||||
import { FileInterceptor } from '@nestjs/platform-express';
|
||||
import { memoryStorage } from 'multer';
|
||||
import { ReceiptParsingService, ParsedReceiptItemRaw } from './receipt-parsing.service';
|
||||
|
||||
const ALLOWED_MIMES = new Set([
|
||||
'image/jpeg',
|
||||
'image/jpg',
|
||||
'image/png',
|
||||
'image/webp',
|
||||
'image/heic',
|
||||
'image/heif',
|
||||
'application/pdf',
|
||||
'application/octet-stream',
|
||||
]);
|
||||
|
||||
@Controller('receipt-import')
|
||||
export class ReceiptParsingController {
|
||||
constructor(private readonly receiptParsingService: ReceiptParsingService) {}
|
||||
|
||||
@Post('parse')
|
||||
@HttpCode(200)
|
||||
@UseInterceptors(
|
||||
FileInterceptor('file', {
|
||||
storage: memoryStorage(),
|
||||
limits: { fileSize: 15 * 1024 * 1024 },
|
||||
fileFilter: (_req, file, cb) => {
|
||||
if (ALLOWED_MIMES.has(file.mimetype)) {
|
||||
cb(null, true);
|
||||
} else {
|
||||
cb(
|
||||
new BadRequestException(
|
||||
`Filtypen "${file.mimetype}" är inte tillåten. Använd JPEG, PNG, WebP, HEIC eller PDF.`,
|
||||
),
|
||||
false,
|
||||
);
|
||||
}
|
||||
},
|
||||
}),
|
||||
)
|
||||
async parseReceipt(
|
||||
@UploadedFile() file: Express.Multer.File,
|
||||
): Promise<ParsedReceiptItemRaw[]> {
|
||||
if (!file) {
|
||||
throw new BadRequestException('Ingen fil bifogades. Skicka en bild eller PDF under fältnamnet "file".');
|
||||
}
|
||||
return this.receiptParsingService.parseReceipt(file);
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,9 @@
|
||||
import { Module } from '@nestjs/common';
|
||||
import { ReceiptParsingController } from './receipt-parsing.controller';
|
||||
import { ReceiptParsingService } from './receipt-parsing.service';
|
||||
|
||||
@Module({
|
||||
controllers: [ReceiptParsingController],
|
||||
providers: [ReceiptParsingService],
|
||||
})
|
||||
export class ReceiptParsingModule {}
|
||||
@@ -0,0 +1,192 @@
|
||||
import {
|
||||
BadRequestException,
|
||||
Injectable,
|
||||
Logger,
|
||||
ServiceUnavailableException,
|
||||
} from '@nestjs/common';
|
||||
import * as pdfParse from 'pdf-parse';
|
||||
|
||||
const MISTRAL_API_URL = 'https://api.mistral.ai/v1/chat/completions';
|
||||
const RECEIPT_IMPORT_MODEL = 'mistral-small-2603';
|
||||
const MAX_RETRIES = 3;
|
||||
|
||||
const IMAGE_PROMPT = `Du är en kvittoläsare. Analysera detta kvitto och returnera ENDAST en JSON-array med alla köpta varor.
|
||||
Varje vara ska ha följande fält:
|
||||
- "rawName": varans namn som det står på kvittot (sträng)
|
||||
- "quantity": antal eller mängd som ett tal (t.ex. 1, 2, 0.5)
|
||||
- "unit": enhet — välj ett av: "st", "kg", "g", "l", "dl", "cl", "ml", "förp", "pak", "burk", "flaska"
|
||||
- "price": pris i SEK som ett tal, eller null
|
||||
- "brand": märke eller leverantör om det tydligt framgår av varunamnet (t.ex. "Arla", "ICA", "Oatly"), annars null
|
||||
- "origin": ursprungsland om det framgår av varunamnet (t.ex. "Brasilien", "Sverige", "Italien"), annars null
|
||||
|
||||
Returnera BARA JSON-arrayen utan markdown-formatering.`;
|
||||
|
||||
const TEXT_PROMPT = (text: string) =>
|
||||
`Du är en kvittoläsare. Nedan följer rå text från ett kvitto. Analysera texten och returnera ENDAST en JSON-array med alla köpta varor.
|
||||
Varje vara ska ha följande fält:
|
||||
- "rawName": varans namn som det står på kvittot (sträng)
|
||||
- "quantity": antal eller mängd som ett tal (t.ex. 1, 2, 0.5)
|
||||
- "unit": enhet — välj ett av: "st", "kg", "g", "l", "dl", "cl", "ml", "förp", "pak", "burk", "flaska"
|
||||
- "price": pris i SEK som ett tal, eller null
|
||||
- "brand": märke eller leverantör om det tydligt framgår av varunamnet (t.ex. "Arla", "ICA", "Oatly"), annars null
|
||||
- "origin": ursprungsland om det framgår av varunamnet (t.ex. "Brasilien", "Sverige", "Italien"), annars null
|
||||
|
||||
Returnera BARA JSON-arrayen utan markdown-formatering.
|
||||
|
||||
Kvittotext:
|
||||
${text}`;
|
||||
|
||||
export interface ParsedReceiptItemRaw {
|
||||
rawName: string;
|
||||
quantity: number;
|
||||
unit: string;
|
||||
price?: number | null;
|
||||
brand?: string | null;
|
||||
origin?: string | null;
|
||||
}
|
||||
|
||||
@Injectable()
|
||||
export class ReceiptParsingService {
|
||||
private readonly logger = new Logger(ReceiptParsingService.name);
|
||||
|
||||
async parseReceipt(file: Express.Multer.File): Promise<ParsedReceiptItemRaw[]> {
|
||||
const apiKey = process.env.MISTRAL_API_KEY;
|
||||
if (!apiKey) {
|
||||
throw new ServiceUnavailableException('MISTRAL_API_KEY är inte konfigurerad');
|
||||
}
|
||||
|
||||
const isPdf =
|
||||
file.mimetype === 'application/pdf' ||
|
||||
file.mimetype === 'application/octet-stream' ||
|
||||
file.originalname?.toLowerCase().endsWith('.pdf');
|
||||
|
||||
if (isPdf) {
|
||||
return this.parseReceiptFromPdf(file.buffer, apiKey);
|
||||
}
|
||||
return this.parseReceiptFromImage(file.buffer, file.mimetype, apiKey);
|
||||
}
|
||||
|
||||
private async callMistralWithRetry(body: object, apiKey: string, source: string): Promise<Response> {
|
||||
for (let attempt = 1; attempt <= MAX_RETRIES; attempt++) {
|
||||
const response = await fetch(MISTRAL_API_URL, {
|
||||
method: 'POST',
|
||||
headers: {
|
||||
'Content-Type': 'application/json',
|
||||
Authorization: `Bearer ${apiKey}`,
|
||||
},
|
||||
body: JSON.stringify(body),
|
||||
});
|
||||
|
||||
if (response.status === 503 || response.status === 429) {
|
||||
const err = await response.text();
|
||||
this.logger.warn(`Mistral ${response.status} (${source}, försök ${attempt}/${MAX_RETRIES}): ${err}`);
|
||||
if (attempt < MAX_RETRIES) {
|
||||
await new Promise((r) => setTimeout(r, attempt * 2000));
|
||||
continue;
|
||||
}
|
||||
throw new ServiceUnavailableException('Mistral API: Tjänsten tillfälligt otillgänglig');
|
||||
}
|
||||
|
||||
return response;
|
||||
}
|
||||
throw new ServiceUnavailableException('Kunde inte nå Mistral API efter flera försök');
|
||||
}
|
||||
|
||||
private async parseReceiptFromImage(
|
||||
buffer: Buffer,
|
||||
mimeType: string,
|
||||
apiKey: string,
|
||||
): Promise<ParsedReceiptItemRaw[]> {
|
||||
const effectiveMime = mimeType === 'application/octet-stream' ? 'image/jpeg' : mimeType;
|
||||
const base64 = buffer.toString('base64');
|
||||
const response = await this.callMistralWithRetry(
|
||||
{
|
||||
model: RECEIPT_IMPORT_MODEL,
|
||||
messages: [
|
||||
{
|
||||
role: 'user',
|
||||
content: [
|
||||
{
|
||||
type: 'image_url',
|
||||
image_url: { url: `data:${effectiveMime};base64,${base64}` },
|
||||
},
|
||||
{ type: 'text', text: IMAGE_PROMPT },
|
||||
],
|
||||
},
|
||||
],
|
||||
max_tokens: 2000,
|
||||
temperature: 0.1,
|
||||
},
|
||||
apiKey,
|
||||
'bild',
|
||||
);
|
||||
return this.extractItemsFromMistralResponse(response, 'bild');
|
||||
}
|
||||
|
||||
private async parseReceiptFromPdf(
|
||||
buffer: Buffer,
|
||||
apiKey: string,
|
||||
): Promise<ParsedReceiptItemRaw[]> {
|
||||
let pdfText: string;
|
||||
try {
|
||||
const parsed = await pdfParse(buffer);
|
||||
pdfText = parsed.text?.trim();
|
||||
} catch {
|
||||
throw new BadRequestException('Kunde inte läsa PDF-filen. Kontrollera att filen inte är skadad.');
|
||||
}
|
||||
|
||||
if (!pdfText || pdfText.length < 20) {
|
||||
throw new BadRequestException(
|
||||
'PDF-filen verkar inte innehålla läsbar text. Prova att fotografera kvittot istället.',
|
||||
);
|
||||
}
|
||||
|
||||
this.logger.log(`PDF-text extraherad (${pdfText.length} tecken)`);
|
||||
|
||||
const response = await this.callMistralWithRetry(
|
||||
{
|
||||
model: RECEIPT_IMPORT_MODEL,
|
||||
messages: [{ role: 'user', content: TEXT_PROMPT(pdfText) }],
|
||||
max_tokens: 2000,
|
||||
temperature: 0.1,
|
||||
},
|
||||
apiKey,
|
||||
'PDF',
|
||||
);
|
||||
return this.extractItemsFromMistralResponse(response, 'PDF');
|
||||
}
|
||||
|
||||
private async extractItemsFromMistralResponse(
|
||||
response: Response,
|
||||
source: string,
|
||||
): Promise<ParsedReceiptItemRaw[]> {
|
||||
if (!response.ok) {
|
||||
const err = await response.text();
|
||||
this.logger.error(`Mistral API svarade ${response.status} (${source}): ${err}`);
|
||||
const hint =
|
||||
response.status === 401
|
||||
? 'Ogiltig API-nyckel (401)'
|
||||
: response.status === 429
|
||||
? 'För många förfrågningar (429)'
|
||||
: `HTTP ${response.status}`;
|
||||
throw new ServiceUnavailableException(`Mistral API returnerade ett fel: ${hint}`);
|
||||
}
|
||||
|
||||
const data = (await response.json()) as {
|
||||
choices: { message: { content: string } }[];
|
||||
};
|
||||
const content = data.choices?.[0]?.message?.content ?? '[]';
|
||||
|
||||
try {
|
||||
const clean = content.replace(/```(?:json)?/gi, '').trim();
|
||||
const items = JSON.parse(clean);
|
||||
if (!Array.isArray(items)) throw new Error('Inte en array');
|
||||
return items as ParsedReceiptItemRaw[];
|
||||
} catch {
|
||||
this.logger.error(`Kunde inte parsa Mistral-svar (${source}):`, content);
|
||||
throw new BadRequestException(
|
||||
`Kvittot kunde inte tolkas. Försök med en tydligare ${source === 'PDF' ? 'PDF' : 'bild'}.`,
|
||||
);
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -1,4 +1,13 @@
|
||||
import { Controller, Post, Body } from '@nestjs/common';
|
||||
import {
|
||||
Controller,
|
||||
Post,
|
||||
Body,
|
||||
HttpCode,
|
||||
UploadedFile,
|
||||
UseInterceptors,
|
||||
} from '@nestjs/common';
|
||||
import { FileInterceptor } from '@nestjs/platform-express';
|
||||
import { memoryStorage } from 'multer';
|
||||
import { QuickImportService, QuickImportResult } from '../services/quick-import.service';
|
||||
|
||||
@Controller('quick-import')
|
||||
@@ -6,9 +15,20 @@ export class QuickImportController {
|
||||
constructor(private readonly quickImportService: QuickImportService) {}
|
||||
|
||||
@Post()
|
||||
@HttpCode(200)
|
||||
@UseInterceptors(
|
||||
FileInterceptor('file', {
|
||||
storage: memoryStorage(),
|
||||
limits: { fileSize: 10 * 1024 * 1024 },
|
||||
}),
|
||||
)
|
||||
async importFromInput(
|
||||
@Body() body: { input: string }
|
||||
@Body() body: { input?: string },
|
||||
@UploadedFile() file?: Express.Multer.File,
|
||||
): Promise<QuickImportResult> {
|
||||
return this.quickImportService.importFromInput(body.input);
|
||||
if (file) {
|
||||
return this.quickImportService.importFromUpload(file);
|
||||
}
|
||||
return this.quickImportService.importFromInput(body.input ?? '');
|
||||
}
|
||||
}
|
||||
|
||||
@@ -12,6 +12,7 @@ export interface ParsedRecipe {
|
||||
note?: string;
|
||||
}>;
|
||||
instructions?: string;
|
||||
imageUrl?: string;
|
||||
}
|
||||
|
||||
export abstract class RecipeParser {
|
||||
|
||||
@@ -76,11 +76,24 @@ export class IcaRecipeParser extends RecipeParser {
|
||||
}
|
||||
}
|
||||
|
||||
// Extrahera bild-URL
|
||||
let imageUrl: string | undefined;
|
||||
if (recipe.image) {
|
||||
if (typeof recipe.image === 'string') {
|
||||
imageUrl = recipe.image;
|
||||
} else if (Array.isArray(recipe.image) && recipe.image.length > 0) {
|
||||
imageUrl = typeof recipe.image[0] === 'string' ? recipe.image[0] : recipe.image[0]?.url;
|
||||
} else if (typeof recipe.image === 'object' && recipe.image.url) {
|
||||
imageUrl = recipe.image.url;
|
||||
}
|
||||
}
|
||||
|
||||
return {
|
||||
name,
|
||||
description,
|
||||
ingredients,
|
||||
instructions,
|
||||
imageUrl,
|
||||
};
|
||||
}
|
||||
|
||||
|
||||
@@ -1,51 +1,85 @@
|
||||
import { Injectable, BadRequestException } from '@nestjs/common';
|
||||
import { IcaRecipeParser } from './parsers/ica.parser';
|
||||
import { GenericRecipeParser } from './parsers/generic.parser';
|
||||
import { RecipeParser } from './parsers/base.parser';
|
||||
import { Injectable, BadRequestException, Logger } from '@nestjs/common';
|
||||
import { IcaRecipeParser } from '../parsers/ica.parser';
|
||||
import { GenericRecipeParser } from '../parsers/generic.parser';
|
||||
import { RecipeParser } from '../parsers/base.parser';
|
||||
import { createWorker } from 'tesseract.js';
|
||||
import * as pdfParse from 'pdf-parse';
|
||||
|
||||
export interface QuickImportResult {
|
||||
markdown: string;
|
||||
source: 'ica' | 'pdf' | 'other';
|
||||
source: 'ica' | 'pdf' | 'image' | 'other';
|
||||
imageUrl?: string;
|
||||
imageWarning?: string;
|
||||
}
|
||||
|
||||
@Injectable()
|
||||
export class QuickImportService {
|
||||
/**
|
||||
* Detekterar typ av input (URL eller filsökväg) och importerar från lämplig källa
|
||||
*/
|
||||
private readonly logger = new Logger(QuickImportService.name);
|
||||
|
||||
async importFromInput(input: string): Promise<QuickImportResult> {
|
||||
input = input.trim();
|
||||
console.log('[QuickImport] Mottog input:', input);
|
||||
this.logger.log(`Mottog input: ${input}`);
|
||||
|
||||
if (!input) {
|
||||
throw new BadRequestException('Du måste ange en URL eller filsökväg');
|
||||
}
|
||||
|
||||
// Detektera typ
|
||||
const isUrl = this.isUrl(input);
|
||||
const isPdf = this.isPdfPath(input);
|
||||
|
||||
console.log('[QuickImport] isUrl:', isUrl, 'isPdf:', isPdf);
|
||||
|
||||
if (isUrl) {
|
||||
console.log('[QuickImport] Detekterade URL, försöker scrapa...');
|
||||
if (this.isUrl(input)) {
|
||||
this.logger.log('Detekterade URL, försöker scrapa...');
|
||||
return this.scrapeRecipeFromUrl(input);
|
||||
} else if (isPdf) {
|
||||
console.log('[QuickImport] Detekterade PDF-fil');
|
||||
}
|
||||
|
||||
throw new BadRequestException(
|
||||
'PDF-import under utveckling. Försök med en URL från ICA.se eller annat receptsida.'
|
||||
);
|
||||
} else {
|
||||
console.log('[QuickImport] Input är inte URL eller PDF');
|
||||
throw new BadRequestException(
|
||||
'Ogültig input. Ange en gyltig URL (t.ex. ica.se/recept/...) eller filsökväg'
|
||||
'Ogiltig input. Ange en giltig URL (t.ex. ica.se/recept/...) eller ladda upp en fil.',
|
||||
);
|
||||
}
|
||||
|
||||
async importFromUpload(file: Express.Multer.File): Promise<QuickImportResult> {
|
||||
this.logger.log(`Fil mottagen: ${file.originalname} (${file.mimetype})`);
|
||||
|
||||
if (
|
||||
file.mimetype === 'application/pdf' ||
|
||||
file.mimetype === 'application/octet-stream' ||
|
||||
file.originalname?.toLowerCase().endsWith('.pdf')
|
||||
) {
|
||||
return this.importFromPdf(file.buffer);
|
||||
}
|
||||
|
||||
if (file.mimetype.startsWith('image/')) {
|
||||
return this.importFromImage(file.buffer);
|
||||
}
|
||||
|
||||
throw new BadRequestException('Otillåten filtyp. Använd JPEG, PNG, WebP eller PDF.');
|
||||
}
|
||||
|
||||
private async importFromPdf(buffer: Buffer): Promise<QuickImportResult> {
|
||||
try {
|
||||
this.logger.log('Parsar PDF med pdf-parse...');
|
||||
const data = await pdfParse(buffer);
|
||||
const markdown = data.text?.trim() || '(Tom PDF)';
|
||||
return { markdown, source: 'pdf' };
|
||||
} catch (err) {
|
||||
const message = err instanceof Error ? err.message : 'Okänt fel';
|
||||
this.logger.error(`PDF-parsning misslyckades: ${message}`);
|
||||
throw new BadRequestException('Kunde inte läsa PDF-filen. Kontrollera att filen inte är skadad.');
|
||||
}
|
||||
}
|
||||
|
||||
private async importFromImage(buffer: Buffer): Promise<QuickImportResult> {
|
||||
try {
|
||||
this.logger.log('Parsar bild med Tesseract OCR...');
|
||||
const worker = await createWorker('swe+eng');
|
||||
const result = await worker.recognize(buffer);
|
||||
await worker.terminate();
|
||||
const markdown = result.data.text?.trim() || '(Tom bild)';
|
||||
return { markdown, source: 'image' };
|
||||
} catch (err) {
|
||||
const message = err instanceof Error ? err.message : 'Okänt fel';
|
||||
this.logger.error(`OCR misslyckades: ${message}`);
|
||||
throw new BadRequestException('Kunde inte läsa text ur bilden. Prova en tydligare bild.');
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Kontrollerar om input är en URL
|
||||
*/
|
||||
private isUrl(input: string): boolean {
|
||||
try {
|
||||
new URL(input);
|
||||
@@ -55,28 +89,10 @@ export class QuickImportService {
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Kontrollerar om input är en PDF-filsökväg
|
||||
*/
|
||||
private isPdfPath(input: string): boolean {
|
||||
const normalized = input.toLowerCase();
|
||||
return normalized.endsWith('.pdf');
|
||||
}
|
||||
|
||||
/**
|
||||
* 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);
|
||||
this.logger.log(`Hämtar HTML från: ${url}`);
|
||||
|
||||
// Hämta HTML från URL
|
||||
const response = await fetch(url, {
|
||||
headers: {
|
||||
'User-Agent':
|
||||
@@ -84,16 +100,15 @@ export class QuickImportService {
|
||||
},
|
||||
});
|
||||
|
||||
console.log('[QuickImport] HTTP status:', response.status);
|
||||
this.logger.log(`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');
|
||||
this.logger.log(`HTML längd: ${html.length} tecken`);
|
||||
|
||||
// Välj lämplig parser
|
||||
const parsers: RecipeParser[] = [
|
||||
new IcaRecipeParser(),
|
||||
new GenericRecipeParser(),
|
||||
@@ -102,7 +117,7 @@ export class QuickImportService {
|
||||
let recipe = null;
|
||||
for (const parser of parsers) {
|
||||
if (parser.canHandle(url)) {
|
||||
console.log('[QuickImport] Använder parser:', parser.constructor.name);
|
||||
this.logger.log(`Använder parser: ${parser.constructor.name}`);
|
||||
recipe = parser.parse(html);
|
||||
break;
|
||||
}
|
||||
@@ -112,68 +127,73 @@ export class QuickImportService {
|
||||
throw new Error('Ingen parserutrustning tillgänglig');
|
||||
}
|
||||
|
||||
console.log('[QuickImport] Parsad recept:', {
|
||||
name: recipe.name,
|
||||
ingredienser: recipe.ingredients.length,
|
||||
});
|
||||
this.logger.log(`Parsad recept: ${recipe.name} (${recipe.ingredients.length} ingredienser)`);
|
||||
|
||||
if (!recipe.name) {
|
||||
throw new Error('Kunde inte hitta receptnamn på sidan. Försök med en annan länk.');
|
||||
}
|
||||
|
||||
// Konvertera till Markdown-format
|
||||
const markdown = this.recipeToMarkdown(recipe, url);
|
||||
console.log('[QuickImport] Markdown genererad, längd:', markdown.length);
|
||||
this.logger.log(`Markdown genererad, längd: ${markdown.length}`);
|
||||
|
||||
// Detektera källa från URL
|
||||
let source: 'ica' | 'pdf' | 'other' = 'other';
|
||||
let source: 'ica' | 'pdf' | 'image' | 'other' = 'other';
|
||||
if (/ica\.se/i.test(url)) {
|
||||
source = 'ica';
|
||||
}
|
||||
|
||||
return {
|
||||
markdown,
|
||||
source,
|
||||
};
|
||||
// Normalisera bild-URL från parsern (ingen download — görs i recipe-app vid sparande)
|
||||
let imageUrl: string | undefined;
|
||||
let imageWarning: string | undefined;
|
||||
if (recipe.imageUrl) {
|
||||
const normalized = this.normalizeImageUrl(recipe.imageUrl, url);
|
||||
if (normalized) {
|
||||
imageUrl = normalized;
|
||||
} else {
|
||||
imageWarning = 'Receptbild kunde inte tolkas till en giltig URL.';
|
||||
this.logger.warn(`Kunde inte normalisera bild-URL: "${recipe.imageUrl}"`);
|
||||
}
|
||||
}
|
||||
|
||||
return { markdown, source, imageUrl, imageWarning };
|
||||
} catch (err) {
|
||||
const message = err instanceof Error ? err.message : 'Okänt fel vid scraping';
|
||||
console.error('[QuickImport] ERROR:', message);
|
||||
this.logger.error(`Scraping misslyckades: ${message}`);
|
||||
throw new BadRequestException(
|
||||
`Kunde inte hämta recept: ${message}. Kontrollera att länken är korrekt och försök igen.`
|
||||
`Kunde inte hämta recept: ${message}. Kontrollera att länken är korrekt och försök igen.`,
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Konvertera receptobjekt till Markdown-format
|
||||
*/
|
||||
private normalizeImageUrl(rawImageUrl: string, pageUrl: string): string | null {
|
||||
const trimmed = rawImageUrl.trim();
|
||||
if (!trimmed) return null;
|
||||
const protocolNormalized = trimmed.startsWith('//') ? `https:${trimmed}` : trimmed;
|
||||
try {
|
||||
return new URL(protocolNormalized, pageUrl).toString();
|
||||
} catch {
|
||||
return null;
|
||||
}
|
||||
}
|
||||
|
||||
private recipeToMarkdown(
|
||||
recipe: {
|
||||
name: string;
|
||||
description?: string;
|
||||
ingredients: Array<{
|
||||
quantity: number;
|
||||
unit: string;
|
||||
name: string;
|
||||
note?: string;
|
||||
}>;
|
||||
ingredients: Array<{ quantity: number; unit: string; name: string; note?: string }>;
|
||||
instructions?: string;
|
||||
},
|
||||
sourceUrl?: string,
|
||||
): string {
|
||||
const lines: string[] = [];
|
||||
|
||||
// Titel
|
||||
lines.push(`# ${recipe.name}`);
|
||||
lines.push('');
|
||||
|
||||
// Beskrivning
|
||||
if (recipe.description) {
|
||||
lines.push(recipe.description);
|
||||
lines.push('');
|
||||
}
|
||||
|
||||
// Ingredienser
|
||||
if (recipe.ingredients.length > 0) {
|
||||
lines.push('## Ingredienser');
|
||||
for (const ing of recipe.ingredients) {
|
||||
@@ -185,14 +205,12 @@ export class QuickImportService {
|
||||
lines.push('');
|
||||
}
|
||||
|
||||
// Instruktioner
|
||||
if (recipe.instructions) {
|
||||
lines.push('## Tillvägagångssätt');
|
||||
lines.push(recipe.instructions);
|
||||
lines.push('');
|
||||
}
|
||||
|
||||
// Källa
|
||||
if (sourceUrl) {
|
||||
lines.push('---');
|
||||
lines.push('');
|
||||
@@ -202,3 +220,4 @@ export class QuickImportService {
|
||||
return lines.join('\n');
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@@ -1,9 +0,0 @@
|
||||
import { Module } from '@nestjs/common';
|
||||
import { QuickImportController } from './controllers/quick-import.controller';
|
||||
import { QuickImportService } from './services/quick-import.service';
|
||||
|
||||
@Module({
|
||||
controllers: [QuickImportController],
|
||||
providers: [QuickImportService],
|
||||
})
|
||||
export class WebScrapingModule {}
|
||||
@@ -1,9 +1,9 @@
|
||||
import { Module } from '@nestjs/common';
|
||||
import { QuickImportController } from './quick-import.controller';
|
||||
import { QuickImportService } from './quick-import.service';
|
||||
import { QuickImportController } from './controllers/quick-import.controller';
|
||||
import { QuickImportService } from './services/quick-import.service';
|
||||
|
||||
@Module({
|
||||
controllers: [QuickImportController],
|
||||
providers: [QuickImportService],
|
||||
})
|
||||
export class QuickImportModule {}
|
||||
export class WebScrapingModule {}
|
||||
|
||||
Reference in New Issue
Block a user