Austin/reader
1
0
Fork 0
mirror of https://github.com/intergalacticalvariable/reader.git synced 2024-11-16 11:42:32 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

remove alt-text service

Browse Source
This commit is contained in:
Harsh Gupta 2024-08-14 14:46:37 +05:30
parent 4c957adbce
commit cbe4fa94c1
2 changed files with 3 additions and 96 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
backend/functions/src/cloud-functions/crawler.ts
View File

@ -10,7 +10,7 @@ import _ from 'lodash';
import { PageSnapshot, PuppeteerControl, ScrappingOptions } from '../services/puppeteer';
import { Request, Response } from 'express';
const pNormalizeUrl = import("@esm2cjs/normalize-url");
import { AltTextService } from '../services/alt-text';
// import { AltTextService } from '../services/alt-text';
import TurndownService from 'turndown';
// import { Crawled } from '../db/crawled';
import { cleanAttribute } from '../utils/misc';
@ -74,7 +74,7 @@ export class CrawlerHost extends RPCHost {
protected globalLogger: Logger,
protected puppeteerControl: PuppeteerControl,
protected jsdomControl: JSDomControl,
protected altTextService: AltTextService,
// protected altTextService: AltTextService,
// protected pdfExtractor: PDFExtractor,
protected firebaseObjectStorage: FirebaseStorageBucketControl,
protected threadLocal: AsyncContext,
@ -399,10 +399,7 @@ export class CrawlerHost extends RPCHost {
const urlToAltMap: { [k: string]: string | undefined; } = {};
if (snapshot.imgs?.length && this.threadLocal.get('withGeneratedAlt')) {
const tasks = _.uniqBy((snapshot.imgs || []), 'src').map(async (x) => {
const r = await this.altTextService.getAltText(x).catch((err: any) => {
this.logger.warn(`Failed to get alt text for ${x.src}`, { err: marshalErrorLike(err) });
return undefined;
});
const r = "ALT TEXT!!!"
if (r && x.src) {
urlToAltMap[x.src.trim()] = r;
}

90
backend/functions/src/services/alt-text.ts
View File

@ -1,90 +0,0 @@
import { AssertionFailureError, AsyncService, HashManager } from 'civkit';
import { singleton } from 'tsyringe';
import { Logger } from '../shared/services/logger';
import { CanvasService } from '../shared/services/canvas';
import { ImageInterrogationManager } from '../shared/services/common-iminterrogate';
import { ImgBrief } from './puppeteer';
import { ImgAlt } from '../db/img-alt';
const md5Hasher = new HashManager('md5', 'hex');
@singleton()
export class AltTextService extends AsyncService {
altsToIgnore = 'image,img,photo,picture,pic,alt,figure,fig'.split(',');
logger = this.globalLogger.child({ service: this.constructor.name });
constructor(
protected globalLogger: Logger,
protected imageInterrogator: ImageInterrogationManager,
protected canvasService: CanvasService
) {
super(...arguments);
}
override async init() {
await this.dependencyReady();
this.emit('ready');
}
async caption(url: string) {
try {
const img = await this.canvasService.loadImage(url);
const resized = this.canvasService.fitImageToSquareBox(img, 1024);
const exported = await this.canvasService.canvasToBuffer(resized, 'image/png');
const r = await this.imageInterrogator.interrogate('blip2', {
image: exported,
// prompt: `A formal caption in one sentence, concise and in the third person: HTML <img> alt text of this image. Return "**NSFW**" if you don't feel comfortable captioning it.`
});
return r.replaceAll(/[\n\"]|(\.\s*$)/g, '').trim();
} catch (err) {
throw new AssertionFailureError({ message: `Could not generate alt text for url ${url}`, cause: err });
}
}
async getAltText(imgBrief: ImgBrief) {
if (!imgBrief.src) {
return undefined;
}
if (imgBrief.alt && !this.altsToIgnore.includes(imgBrief.alt.trim().toLowerCase())) {
return imgBrief.alt;
}
const digest = md5Hasher.hash(imgBrief.src);
const shortDigest = Buffer.from(digest, 'hex').toString('base64url');
const existing = await ImgAlt.fromFirestore(shortDigest);
if (existing) {
return existing.generatedAlt || existing.originalAlt || '';
}
let generatedCaption = '';
try {
generatedCaption = await this.caption(imgBrief.src);
} catch (err) {
this.logger.warn(`Unable to generate alt text for ${imgBrief.src}`, { err });
}
// Don't try again until the next day
const expireMixin = generatedCaption ? {} : { expireAt: new Date(Date.now() + 1000 * 3600 * 24) };
await ImgAlt.COLLECTION.doc(shortDigest).set(
{
_id: shortDigest,
src: imgBrief.src || '',
width: imgBrief.naturalWidth || 0,
height: imgBrief.naturalHeight || 0,
urlDigest: digest,
originalAlt: imgBrief.alt || '',
generatedAlt: generatedCaption || '',
createdAt: new Date(),
...expireMixin
}, { merge: true }
);
return generatedCaption;
}
}