angristan · angristan · Aug 23, 2020 · Apr 5, 2020
diff --git a/.env.example b/.env.example
@@ -40,3 +40,7 @@ AWS_S3_ENDPOINT=https://minio.feedbin.domain.tld
 AWS_S3_PATH_STYLE="true"
 ENTRY_IMAGE_HOST=minio.feedbin.domain.tld
 
+# Extract service (needs to run behind HTTPS:443)
+EXTRACT_HOST=extract.feedbin.domain.tld
+EXTRACT_USER=username
+EXTRACT_SECRET=secret
diff --git a/README.md b/README.md
@@ -12,6 +12,7 @@ Here is a breakdown of all the containers:
 * `workers`: some of the sidekiq workers for background processing
 * `refresher`: sidekiq worker for refreshing feeds
 * `image`: sidekiq worker to find thumbnails
+* `extract`: nodejs service to extract article content from full web pages
 * `camo`: a node reverse proxy to prevent mixed content
 * `minio`: object storage for images, favicons, imports
 * `redis`: cache, store sidekiq queues and stats

diff --git a/caddy/example.Caddyfile b/caddy/example.Caddyfile
@@ -18,3 +18,10 @@ minio.feedbin.domain.tld {
     transparent
   }
 }
+
+extract.feedbin.domain.tld {
+  gzip
+  proxy / http://feedbin-extract:3000 {
+    transparent
+  }
+}
diff --git a/docker-compose-example.yml b/docker-compose-example.yml
@@ -141,3 +141,9 @@ services:
     restart: always
     mem_limit: 512m
 
+  feedbin-extract:
+    build: ./feedbin-extract
+    env_file:
+      - ./.env
+    restart: always
+    mem_limit: 512m
diff --git a/feedbin-extract/Dockerfile b/feedbin-extract/Dockerfile
@@ -0,0 +1,11 @@
+FROM node:12
+
+WORKDIR /app
+
+RUN git clone https://github.com/feedbin/extract.git /app \
+    && cd /app \
+    && npm install \
+    && mkdir users \
+    && echo "secret" > users/username
+
+CMD ["node", "app/server.js"]