.github/workflows/container.yml

@@ -10,7 +10,7 @@ env:
 
 jobs:
   check:
-    name: chequear typescript del sitio
+    name: chequear typescript
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v3
@@ -29,7 +29,6 @@ jobs:
         working-directory: ./sitio
 
   build-and-push-sitio:
-    name: Compilar contenedor del sitio
     needs: check
     runs-on: ubuntu-latest
     permissions:
@@ -38,32 +37,26 @@ jobs:
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
       - name: Log in to the Container registry
-        uses: docker/login-action@v3
+        uses: docker/login-action@65b78e6e13532edd9afa3aa52ac7964289d1a9c1
         with:
           registry: ${{ env.REGISTRY }}
           username: ${{ github.actor }}
           password: ${{ secrets.GITHUB_TOKEN }}
       - name: Extract metadata (tags, labels) for Docker
         id: meta
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@9ec57ed1fcdbf14dcef7dfbe97b2010124a938b7
         with:
           images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}/sitio
       - name: Build and push Docker image
-        uses: docker/build-push-action@v5
+        uses: docker/build-push-action@f2a1d5e99d037542a71f64918e516c093c6f3fc4
         with:
           context: .
-          file: Dockerfile
           push: true
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
-          cache-from: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}/sitio:buildcache
-          cache-to: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}/sitio:buildcache,mode=max
 
   build-and-push-scraper:
-    name: Compilar contenedor del scraper
     runs-on: ubuntu-latest
     permissions:
       contents: read
@@ -102,5 +95,5 @@ jobs:
           push: true
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
-          cache-from: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}/scraper:buildcache
-          cache-to: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}/scraper:buildcache,mode=max
+          cache-from: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:buildcache
+          cache-to: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:buildcache,mode=max

Dockerfile

@@ -1,26 +1,29 @@
-FROM cgr.dev/chainguard/wolfi-base AS base
+FROM docker.io/oven/bun:1-alpine AS base
 WORKDIR /usr/src/app
 
 FROM base as build
-RUN apk add --no-cache nodejs npm
-RUN npm install --global pnpm
-COPY . .
-COPY db-datos/drizzle .
-RUN cd sitio && \
-    pnpm install && \
-    pnpm build
-
-FROM base
 ENV NODE_ENV=production
-RUN apk add --no-cache nodejs npm jq sqlite
+RUN apk add --no-cache nodejs
+COPY . .
+RUN bun install --frozen-lockfile
+RUN cd sitio && \
+    bun run build
+RUN bun build scraper/cli.ts --target=bun --outfile=/tmp/cli.build.js
+
+FROM cgr.dev/chainguard/wolfi-base
+RUN apk add --no-cache nodejs npm jq bun sqlite
 
 # Sitio
 COPY --from=build /usr/src/app/sitio/package.json package.real.json
 RUN sh -c 'echo {\"name\":\"sitio\",\"type\":\"module\",\"dependencies\":$(jq .dependencies < package.real.json)} > package.json' && npm install
 COPY --from=build /usr/src/app/db-datos node_modules/db-datos
 COPY --from=build /usr/src/app/sitio/build .
-COPY --from=build /usr/src/app/db-datos/drizzle .
 
+# Scraper
+COPY --from=build /tmp/cli.build.js /bin/scraper
+COPY --from=build /usr/src/app/db-datos/drizzle /bin/drizzle
+
+ENV NODE_ENV=production
 ENV DB_PATH=/db/db.db
 EXPOSE 3000
 

db-datos/db.js

@@ -2,7 +2,10 @@
 import Database from "better-sqlite3";
 import { drizzle } from "drizzle-orm/better-sqlite3";
 import { DB_PATH } from "./drizzle.config.js";
+import { migrateDb } from "./migrate.js";
 import * as schema from "./schema.js";
 
+migrateDb();
+
 export const sqlite = new Database(DB_PATH);
 export const db = drizzle(sqlite, { schema });

scraper-rs/Cargo.lock

@@ -902,15 +902,6 @@ dependencies = [
  "unicode-ident",
 ]
 
-[[package]]
-name = "quick-xml"
-version = "0.31.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1004a344b30a54e2ee58d66a71b32d2db2feb0a31f9a2d302bf0536f15de2a33"
-dependencies = [
- "memchr",
-]
-
 [[package]]
 name = "quote"
 version = "1.0.35"
@@ -1141,7 +1132,6 @@ dependencies = [
  "futures",
  "itertools",
  "nanoid",
- "quick-xml",
  "rand 0.8.5",
  "reqwest",
  "rusqlite",

scraper-rs/Cargo.toml

@@ -14,7 +14,6 @@ deadpool-sqlite = "0.7.0"
 futures = "0.3.30"
 itertools = "0.12.0"
 nanoid = "0.4.0"
-quick-xml = "0.31.0"
 rand = "0.8.5"
 reqwest = { version = "0.11.23", default-features = false, features = [
     "rustls-tls",

scraper-rs/src/sites/vtex.rs

@@ -107,34 +107,14 @@ pub fn in_stock_from_meta(dom: &VDom) -> anyhow::Result<bool> {
 
 pub fn parse_urls_from_sitemap(sitemap: &str) -> anyhow::Result<Vec<String>> {
     let dom = tl::parse(sitemap, tl::ParserOptions::default())?;
-    dom.query_selector("loc")
+    Ok(dom
+        .query_selector("loc")
         .unwrap()
         .filter_map(|h| h.get(dom.parser()))
         .filter_map(|n| n.as_tag())
         .map(|t| t.inner_text(dom.parser()))
-        .map(|s| -> anyhow::Result<String> {
-            Ok(quick_xml::escape::unescape(s.as_ref())?.to_string())
-        })
-        .try_collect()
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn test_decode_url() -> anyhow::Result<()> {
-        let links = parse_urls_from_sitemap(
-            r#"
-<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9" xmlns:xhtml="http://www.w3.org/1999/xhtml">
-<url>
-    <loc>https://www.carrefour.com.ar/postre-danette-mousse-dulce-de-leche-80-g&#x200B;-684952/p</loc>
-    <lastmod>2024-01-12T10:41:25.962Z</lastmod>
-</url>"#,
-        )?;
-        assert_eq!(links[0], "https://www.carrefour.com.ar/postre-danette-mousse-dulce-de-leche-80-g\u{200b}-684952/p");
-        Ok(())
-    }
+        .map(|s| s.to_string())
+        .collect())
 }
 
 pub async fn get_urls_from_sitemap(sitemaps: Vec<&str>) -> anyhow::Result<Vec<String>> {

sitio/src/lib/server/db.ts

@@ -1,2 +1,10 @@
-export { db } from "db-datos/db.js";
+import Database from "better-sqlite3";
+import { drizzle } from "drizzle-orm/better-sqlite3";
+import * as schema from "db-datos/schema.js";
+import { env } from "$env/dynamic/private";
+
+const sqlite = new Database(env.DB_PATH ?? "../scraper/sqlite.db");
+const db = drizzle(sqlite, { schema });
+
+export { db };
 export * as schema from "db-datos/schema.js";

sitio/tsconfig.json

@@ -9,9 +9,7 @@
     "skipLibCheck": true,
     "sourceMap": true,
     "strict": true,
-    "moduleResolution": "bundler",
-    "maxNodeModuleJsDepth": 1,
-    "declaration": true
+    "moduleResolution": "bundler"
   }
   // Path aliases are handled by https://kit.svelte.dev/docs/configuration#alias
   //

tsconfig.json

@@ -11,7 +11,6 @@
     "strict": true,
     "skipLibCheck": true,
     "esModuleInterop": true,
-    "allowJs": true,
     "checkJs": true,
     "noEmit": true,
     "forceConsistentCasingInFileNames": true