Update az generator 'users' list

Add version variables

.gitignore

@@ -1 +1,2 @@
-*.env
+prod/**/*.env
+.claude

Taskfile.yml

@@ -21,33 +21,24 @@ tasks:
     run-consumer:
         cmd: go run ./cmd/consumer {{.CLI_ARGS}}
         dotenv:
-            - config/app/.consumer.env
-            - config/app/.mongodb.env
+            - config/app/consumer.env
+            - config/app/mongodb.env
 
     run-feedgen-az:
         cmd: go run ./cmd/feedgen/az {{.CLI_ARGS}}
         dotenv:
-            - config/app/feedgen/.az.env
-            - config/app/.mongodb.env
+            - config/app/feedgen/az.env
+            - config/app/mongodb.env
 
     run-api:
         cmd: go run ./cmd/api {{.CLI_ARGS}}
         dotenv:
-            - config/app/.api.env
-            - config/app/.mongodb.env
+            - config/app/api.env
+            - config/app/mongodb.env
 
     run-manager:
         cmd: go run cmd/manager/main.go {{.CLI_ARGS}}
 
-    generate-env:
-        desc: Generate env files from templates
-        cmds:
-            - cp config/app/consumer.env.example config/app/.consumer.env
-            - cp config/app/api.env.example config/app/.api.env
-            - cp config/app/mongodb.env.example config/app/.mongodb.env
-            - cp config/app/feedgen/az.env.example config/app/feedgen/.az.env
-            - cp config/mongodb/env.example config/mongodb/.env
-
     docker-publish-all:
         desc: Publish docker images for all services
         cmds:

cmd/api/version.go

@@ -1,3 +1,3 @@
 package main
 
-const version = "0.1.0"
+const version = "0.2.201"

cmd/feedgen/az/version.go

@@ -1,3 +1,3 @@
 package main
 
-const version = "0.1.0"
+const version = "0.1.106"

config/app/consumer.env

@@ -0,0 +1,3 @@
+POST_MAX_DATE=720h # Save only posts created in the last month
+POST_COLLECTION_CUTOFF_CRON_DELAY=30m # 30 minutes
+POST_COLLECTION_CUTOFF_CRON_MAX_DOCUMENT=1000000 # Delete post documents after 1 million

pkg/api/base.go

@@ -23,6 +23,7 @@ func Run(
 		return err
 	}
 	feedHandler := handler.NewFeedHandler(feeds, apiConfig.FeedgenPublisherDID)
+	generatorHandler := handler.NewGeneratorHandler()
 
 	authMiddleware := middleware.NewAuth(apiConfig.ServiceDID)
 
@@ -34,6 +35,9 @@ func Run(
 		"GET /xrpc/app.bsky.feed.getFeedSkeleton",
 		authMiddleware.JWTAuthMiddleware(http.HandlerFunc(feedHandler.GetFeedSkeleton)),
 	)
+	mux.HandleFunc("GET /{feed}/users", generatorHandler.GetAllUsers)
+	mux.HandleFunc("GET /{feed}/users/valid/", generatorHandler.GetValidUsers)
+	mux.HandleFunc("GET /{feed}/users/invalid/", generatorHandler.GetInvalidUsers)
 
 	httpServer := &http.Server{
 		Addr:    fmt.Sprintf(":%d", apiConfig.APIPort),

pkg/api/handler/generator.go

@@ -0,0 +1,57 @@
+package handler
+
+import (
+	"net/http"
+
+	"github.com/aykhans/bsky-feedgen/pkg/api/response"
+	generatorAz "github.com/aykhans/bsky-feedgen/pkg/generator/az"
+)
+
+type GeneratorHandler struct{}
+
+func NewGeneratorHandler() *GeneratorHandler {
+	return &GeneratorHandler{}
+}
+
+func (handler *GeneratorHandler) GetValidUsers(w http.ResponseWriter, r *http.Request) {
+	feed := r.PathValue("feed")
+
+	validUsers := make([]string, 0)
+	switch feed {
+	case "AzPulse":
+		validUsers = generatorAz.Users.GetValidUsers()
+	}
+
+	response.JSON(w, 200, response.M{
+		"feed":  feed,
+		"users": validUsers,
+	})
+}
+
+func (handler *GeneratorHandler) GetInvalidUsers(w http.ResponseWriter, r *http.Request) {
+	feed := r.PathValue("feed")
+
+	invalidUsers := make([]string, 0)
+	switch feed {
+	case "AzPulse":
+		invalidUsers = generatorAz.Users.GetInvalidUsers()
+	}
+
+	response.JSON(w, 200, response.M{
+		"feed":  feed,
+		"users": invalidUsers,
+	})
+}
+
+func (handler *GeneratorHandler) GetAllUsers(w http.ResponseWriter, r *http.Request) {
+	feed := r.PathValue("feed")
+
+	responseData := response.M{"feed": feed}
+	switch feed {
+	case "AzPulse":
+		responseData["valid_users"] = generatorAz.Users.GetValidUsers()
+		responseData["invalid_users"] = generatorAz.Users.GetInvalidUsers()
+	}
+
+	response.JSON(w, 200, responseData)
+}

pkg/feed/az.go

@@ -39,7 +39,7 @@ func (f *FeedAz) Describe(_ context.Context) bsky.FeedDescribeFeedGenerator_Feed
 
 func (f *FeedAz) GetPage(
 	ctx context.Context,
-	_ string,
+	_ string, // user did
 	limit int64,
 	cursor string,
 ) ([]*bsky.FeedDefs_SkeletonFeedPost, *string, error) {

pkg/generator/az/base.go

@@ -110,16 +110,24 @@ func (generator *Generator) Start(ctx context.Context, cursorOption types.Genera
 }
 
 func (generator *Generator) IsValid(post *collections.Post) bool {
+	// Skip posts that are deep replies (not direct replies to original posts)
 	if post.Reply != nil && post.Reply.RootURI != post.Reply.ParentURI {
 		return false
 	}
 
-	if isValidUser := users.IsValid(post.DID); isValidUser != nil {
+	// Check if the user who created this post is in our pre-defined list
+	// This allows for explicit inclusion/exclusion of specific users
+	if isValidUser := Users.IsValid(post.DID); isValidUser != nil {
 		return *isValidUser
 	}
 
-	if (slices.Contains(post.Langs, "az") && len(post.Langs) < 3) || // Posts in Azerbaijani language with fewer than 3 languages
-		generator.textRegex.MatchString(post.Text) { // Posts containing Azerbaijan-related keywords
+	// A post is considered valid if it meets either of the following criteria:
+	// 1. It's primarily in Azerbaijani (language code "az") with less than 3 detected languages
+	//    (to filter out multi-language spam)
+	// 2. It contains Azerbaijan-related keywords in the text AND has at least one valid language
+	//    from our approved language list
+	if (slices.Contains(post.Langs, "az") && len(post.Langs) < 3) ||
+		(generator.textRegex.MatchString(post.Text) && Langs.IsExistsAny(post.Langs)) {
 		return true
 	}
 

pkg/generator/az/lists.go

@@ -2,7 +2,7 @@ package az
 
 import "github.com/aykhans/bsky-feedgen/pkg/generator"
 
-var users = generator.Users{
+var Users = generator.Users{
 	// Invalid
 	"did:plc:5zww7zorx2ajw7hqrhuix3ba": false,
 	"did:plc:c4vhz47h566t2ntgd7gtawen": false,
@@ -14,6 +14,23 @@ var users = generator.Users{
 	"did:plc:5vwjnzaibnwscbbcvkzhy57v": false,
 	"did:plc:6mfp3coadoobuvlg6w2avw6x": false,
 	"did:plc:lm2uhaoqoe6yo76oeihndfyi": false,
+	"did:plc:vizwdor43adw3277u2kkrssd": false,
+	"did:plc:oqatvbgbhvqbjl2w2o63ehgi": false,
+	"did:plc:gy7yilnydusx5hy2z3dltynp": false,
+	"did:plc:xk7cs24wk6njv42azm2yd7dv": false,
+	"did:plc:ijmt7f4p3dcfqtg3j3zshimn": false,
+	"did:plc:2q5dx6whenn7pnsrfn3jpd6h": false,
+	"did:plc:s2waw3gkmn7h2nn6od44apng": false,
+	"did:plc:4hm6gb7dzobynqrpypif3dck": false,
+	"did:plc:odvarii7w7soygxet3xvzop7": false,
+	"did:plc:5cbkdchsxjvz5fog2oo7m4le": false,
+	"did:plc:ooeuisen5rtr4rojmz7gkbrh": false,
+	"did:plc:6bvhdvgeqkj7nol2zodtqmww": false,
+	"did:plc:k6sxlkd5ssq2uaylzisap2tw": false,
+	"did:plc:uxljnh22mmfzmr4i3oien6mx": false,
+	"did:plc:w5gg2zgwcyfevphehdcmavev": false,
+	"did:plc:ckawbibgmrwg3lbskfppwtlw": false,
+	"did:plc:43fdk46qa5gsokzygzildsaq": false,
 
 	// Valid
 	"did:plc:jbt4qi6psd7rutwzedtecsq7": true,
@@ -22,6 +39,16 @@ var users = generator.Users{
 	"did:plc:phtq2rhgbwipyx5ie3apw44j": true,
 	"did:plc:jfdvklrs5n5qv7f25v6swc5h": true,
 	"did:plc:u5ez5w6qslh6advti4wyddba": true,
-	"did:plc:cs2cbzojm6hmx5lfxiuft3mq": true,
 	"did:plc:x7alwnnjygt2aqcwblhazko7": true,
+	"did:plc:mgciyhgfn65z7iazxuar6o6a": true,
+	"did:plc:ay2f5go4lxq2hspiaqohegac": true,
+	"did:plc:ftoopigdpuzqt2kpeyqxsofx": true,
+	"did:plc:cs2cbzojm6hmx5lfxiuft3mq": true,
+}
+
+var Langs = generator.Langs{
+	"az": true,
+	"en": true,
+	"tr": true,
+	"ru": true,
 }

pkg/generator/base.go

@@ -4,6 +4,15 @@ import "github.com/aykhans/bsky-feedgen/pkg/utils"
 
 type Users map[string]bool
 
+// IsValid checks if a given DID exists in the Users map and returns its validity status.
+//
+// Parameters:
+//
+//	did: The Decentralized Identifier string to check
+//
+// Returns:
+//   - *bool: A pointer to the validity status if the DID exists in the map
+//   - nil: If the DID does not exist in the map
 func (u Users) IsValid(did string) *bool {
 	isValid, ok := u[did]
 	if ok == false {
@@ -12,3 +21,69 @@ func (u Users) IsValid(did string) *bool {
 
 	return utils.ToPtr(isValid)
 }
+
+// GetValidUsers returns a slice of DIDs that are marked as valid in the Users map.
+//
+// Returns:
+//   - []string: A slice of valid DIDs, limited by the specified parameters
+func (u Users) GetValidUsers() []string {
+	validUsers := make([]string, 0)
+
+	for did, isValid := range u {
+		if isValid {
+			validUsers = append(validUsers, did)
+		}
+	}
+
+	return validUsers
+}
+
+// GetInvalidUsers returns a slice of DIDs that are marked as invalid in the Users map.
+//
+// Returns:
+//   - []string: A slice of invalid DIDs, limited by the specified parameters
+func (u Users) GetInvalidUsers() []string {
+	invalidUsers := make([]string, 0)
+
+	for did, isValid := range u {
+		if !isValid {
+			invalidUsers = append(invalidUsers, did)
+		}
+	}
+
+	return invalidUsers
+}
+
+// GetAll returns a slice of all DIDs in the Users map, regardless of validity status.
+//
+// Returns:
+//   - []string: A slice containing all DIDs in the map
+func (u Users) GetAll() []string {
+	allUsers := make([]string, 0, len(u))
+
+	for did := range u {
+		allUsers = append(allUsers, did)
+	}
+
+	return allUsers
+}
+
+type Langs map[string]bool
+
+// IsExistsAny checks if any of the given language codes exist in the Langs map.
+//
+// Parameters:
+//   - langs: A slice of language code strings to check for existence
+//
+// Returns:
+//   - bool: true if at least one language code from the input slice exists in the map,
+//           false if none of the provided language codes exist
+func (l Langs) IsExistsAny(langs []string) bool {
+	for _, lang := range langs {
+		if _, ok := l[lang]; ok {
+			return true
+		}
+	}
+
+	return false
+}

prod/Makefile

@@ -0,0 +1,26 @@
+.PHONY: configure
+
+configure:
+	@cp config/app/api.env.example config/app/.api.env
+	@cp config/app/consumer.env.example config/app/.consumer.env
+	@cp config/app/mongodb.env.example config/app/.mongodb.env
+	@cp config/app/feedgen/az.env.example config/app/feedgen/.az.env
+	@cp config/caddy/env.example config/caddy/.env
+	@cp config/mongodb/env.example config/mongodb/.env
+
+	@read -p "Enter MongoDB username: " mongodb_user; \
+	printf "Enter MongoDB password: "; \
+	read mongodb_pass; \
+	sed -i "s/MONGO_INITDB_ROOT_USERNAME=.*/MONGO_INITDB_ROOT_USERNAME=$$mongodb_user/" config/mongodb/.env; \
+	sed -i "s/MONGO_INITDB_ROOT_PASSWORD=.*/MONGO_INITDB_ROOT_PASSWORD=$$mongodb_pass/" config/mongodb/.env; \
+	sed -i "s/MONGODB_USERNAME=.*/MONGODB_USERNAME=$$mongodb_user/" config/app/.mongodb.env; \
+	sed -i "s/MONGODB_PASSWORD=.*/MONGODB_PASSWORD=$$mongodb_pass/" config/app/.mongodb.env
+
+	@read -p "Enter domain name (e.g., feeds.bsky.example.com): " domain; \
+	read -p "Enter your AT Protocol DID: " publisher_did; \
+	sed -i "s/DOMAIN=.*/DOMAIN=$$domain/" config/caddy/.env; \
+	sed -i "s|FEEDGEN_HOSTNAME=.*|FEEDGEN_HOSTNAME=https://$$domain|" config/app/.api.env; \
+	sed -i "s/FEEDGEN_PUBLISHER_DID=.*/FEEDGEN_PUBLISHER_DID=$$publisher_did/" config/app/.api.env
+
+	@echo
+	@echo "Configuration complete! You can now run 'docker compose up -d'"

prod/README.md

@@ -0,0 +1,76 @@
+# Example Production Deployment
+
+This is an example of a production deployment for the Feed Generator.
+
+## Architecture
+
+The production setup includes the following services:
+
+- **MongoDB**: Database for storing posts and feed data
+- **Consumer**: Service that consumes AT Protocol firehose data
+- **Feed Generator (AZ)**: Generates feeds for Azerbaijan-related content
+- **API**: REST API service for serving feeds
+- **Caddy**: Reverse proxy
+
+## Quick Start
+
+1. **Configure the environment**:
+   ```bash
+   make configure
+   ```
+   This will:
+   - Copy all example configuration files
+   - Prompt for MongoDB credentials
+   - Prompt for domain name and AT Protocol DID
+   - Update configuration files with your values
+
+2. **Start the services**:
+   ```bash
+   docker compose up -d
+   ```
+
+3. **Check service status**:
+   ```bash
+   docker compose ps
+   docker compose logs
+   ```
+
+## Configuration Files
+
+### Application Configuration
+- `config/app/.api.env` - API service configuration
+- `config/app/.consumer.env` - Consumer service configuration  
+- `config/app/.mongodb.env` - MongoDB connection settings
+- `config/app/feedgen/.az.env` - Azerbaijan feed generator settings
+
+### Infrastructure Configuration
+- `config/caddy/.env` - Caddy reverse proxy settings
+- `config/caddy/Caddyfile` - Caddy server configuration
+- `config/mongodb/.env` - MongoDB initialization settings
+
+## Environment Variables
+
+### API Service
+- `FEEDGEN_HOSTNAME` - Public hostname for the feed generator
+- `FEEDGEN_PUBLISHER_DID` - Your AT Protocol DID
+- `API_PORT` - Port for the API service (default: 8421)
+
+### Consumer Service
+- `POST_MAX_DATE` - Maximum age of posts to store (default: 720h/30 days)
+- `POST_COLLECTION_CUTOFF_CRON_DELAY` - Cleanup interval (default: 30m)
+- `POST_COLLECTION_CUTOFF_CRON_MAX_DOCUMENT` - Max documents before cleanup (default: 1M)
+
+### AZ Feed Generator
+- `FEED_AZ_GENERATER_CRON_DELAY` - Feed generation interval (default: 1m)
+- `FEED_AZ_COLLECTION_CUTOFF_CRON_DELAY` - Cleanup interval (default: 30m)
+- `FEED_AZ_COLLECTION_CUTOFF_CRON_MAX_DOCUMENT` - Max documents before cleanup (default: 500K)
+
+### MongoDB
+- `MONGODB_HOST` - MongoDB hostname (default: mongodb)
+- `MONGODB_PORT` - MongoDB port (default: 27017)
+- `MONGODB_USERNAME` - Database username
+- `MONGODB_PASSWORD` - Database password
+
+### Caddy
+- `DOMAIN` - Your domain name
+- `API_HOST` - Internal API service URL (default: http://api:8421)

prod/config/app/api.env.example

@@ -0,0 +1,3 @@
+FEEDGEN_HOSTNAME=https://feeds.bsky.example.com
+FEEDGEN_PUBLISHER_DID=did:plc:qwertyuiopp
+API_PORT=8421

prod/config/app/consumer.env.example

@@ -1,3 +1,3 @@
 POST_MAX_DATE=720h # Save only posts created in the last month
 POST_COLLECTION_CUTOFF_CRON_DELAY=30m # 30 minutes
-POST_COLLECTION_CUTOFF_CRON_MAX_DOCUMENT=10000000 # Delete post documents after 10 million
+POST_COLLECTION_CUTOFF_CRON_MAX_DOCUMENT=1000000 # Delete post documents after 1 million

prod/config/app/feedgen/az.env.example

@@ -0,0 +1,3 @@
+FEED_AZ_GENERATER_CRON_DELAY=1m # 1 minute
+FEED_AZ_COLLECTION_CUTOFF_CRON_DELAY=30m # 30 minutes
+FEED_AZ_COLLECTION_CUTOFF_CRON_MAX_DOCUMENT=500000 # Delete post documents after 500 thousand

prod/config/app/mongodb.env.example

@@ -0,0 +1,4 @@
+MONGODB_HOST=mongodb
+MONGODB_PORT=27017
+MONGODB_USERNAME=root
+MONGODB_PASSWORD=toor

prod/config/caddy/Caddyfile

@@ -0,0 +1,11 @@
+{
+    admin off
+}
+
+{$DOMAIN} {
+    request_body {
+        max_size 8MB
+    }
+
+    reverse_proxy {$API_HOST}
+}

prod/config/caddy/env.example

@@ -0,0 +1,2 @@
+DOMAIN=feeds.bsky.example.com
+API_HOST=http://api:8421

prod/config/mongodb/env.example

@@ -0,0 +1,2 @@
+MONGO_INITDB_ROOT_USERNAME=root
+MONGO_INITDB_ROOT_PASSWORD=toor

prod/docker-compose.yml

@@ -0,0 +1,63 @@
+services:
+    mongodb:
+        image: mongo:8.0.9-noble
+        restart: unless-stopped
+        # ports:
+        #     - 27017:27017
+        env_file: ./config/mongodb/.env
+        volumes:
+            - mongodb_data:/data/db
+        healthcheck:
+            test: echo 'db.runCommand("ping").ok' | mongosh --quiet
+            interval: 10s
+            timeout: 5s
+            retries: 5
+            start_period: 20s
+
+    consumer:
+        image: git.aykhans.me/bsky/feedgen-consumer:latest
+        restart: unless-stopped
+        env_file:
+            - ./config/app/.mongodb.env
+            - ./config/app/.consumer.env
+        depends_on:
+            mongodb:
+                condition: service_healthy
+
+    feedgen_az:
+        image: git.aykhans.me/bsky/feedgen-generator-az:latest
+        restart: unless-stopped
+        env_file:
+            - ./config/app/.mongodb.env
+            - ./config/app/feedgen/.az.env
+        depends_on:
+            mongodb:
+                condition: service_healthy
+
+    api:
+        image: git.aykhans.me/bsky/feedgen-api:latest
+        restart: unless-stopped
+        ports:
+            - 8421:8421
+        env_file:
+            - ./config/app/.mongodb.env
+            - ./config/app/.api.env
+        depends_on:
+            mongodb:
+                condition: service_healthy
+
+    caddy:
+        image: caddy:2.10.0-alpine
+        restart: unless-stopped
+        ports:
+            - 80:80
+            - 443:443
+            - 443:443/udp
+        env_file: ./config/caddy/.env
+        volumes:
+            - ./config/caddy/Caddyfile:/etc/caddy/Caddyfile
+            - caddy_data:/data
+
+volumes:
+    mongodb_data:
+    caddy_data: