feat(job): source from current website api ; chore: deploy to prod

DESIGN.md

@@ -52,14 +52,16 @@
   - [ ] list failed domain
   - [ ] rescan an instance
   - [ ] allow to blacklist a domain
-- [ ] package
-  - [ ] docker image with : s6 or tini + crond + python(pyinstaller) + sqlpage
+  - [ ] remove duplicate stats ; keep one per day per instance
+- [x] package
+  - [x] docker image with : s6 + crond + python + sqlpage
 - [ ] deploy on `beta-instances.mobilizon.org`
-  - [ ] DNS
-  - [ ] HC ping
-
-
-
+  - [x] DNS -> LB3
+  - [x] HC ping
+  - [x] upload image
+  - [x] docker compose config
+  - [x] upload database
+  - [x] start service
 
 
 # impl

Dockerfile

@@ -1,6 +1,6 @@
 FROM python:3.13-alpine
 
-RUN apk add --no-cache s6 bash wget unzip
+RUN apk add --no-cache s6 bash wget unzip curl
 
 COPY requirements.txt .
 

docker-compose.template.yml

@@ -0,0 +1,19 @@
+# put reverse proxy in the project network
+
+services:
+  web:
+    restart: unless-stopped
+    image: mobilizon-instances:0.1.0
+    environment:
+      SQLPAGE_ENVIRONMENT: production
+      SQLPAGE_WEB_ROOT: /app/webroot
+      SQLPAGE_DATABASE_URL: sqlite:///data/sqlpage.db
+      SCHEDULE: "2 5 * * *"  # UTC time
+      HC_PING: https://<healcheck-instance-ping-domain>/<uuid>
+      DATADIR: /data
+      SOURCE: remote  # source on instances.joinmobilizon.org
+    volumes:
+      - data:/data
+
+volumes:
+  data:

src/instances.py

@@ -9,6 +9,30 @@ from model import Model
 
 logger = logging.getLogger(__name__)
 
+UA = 'Mozilla/5.0 (X11; Linux x86_64; rv:139.0) Gecko/20100101 Firefox/139.0'
+
+def yield_from_joinmbz():
+    # load from jmbz api
+    import requests
+    url = 'https://instances.joinmobilizon.org/api/v1/instances?start=0&count=1000'
+    joinres = requests.get(url, headers={'user-agent': UA}).json().get('data')
+    # load from db
+    with sqlite3.connect(f'{os.environ.get("DATADIR")}/sqlpage.db', timeout=15.0, isolation_level='IMMEDIATE') as db:
+        rows = db.execute('select rowid, domain from instances').fetchall()
+        max_rowid = max(row[0] for row in rows)
+        domains = {row[1] for row in rows}
+        for r in joinres:
+            if r['host'] not in domains:
+                logging.debug(f'''new: {r['id']}, {r['host']}''')
+                # insert new records
+                db.execute('''insert into instances(rowid, domain, failure, "createdAt") values (?,?,?,?)''', (
+                   r['id']
+                 , r['host']
+                 , -1
+                 , int(time.time())
+                ))
+    yield from yield_from_db()
+
 
 def yield_from_db():
     with sqlite3.connect(f'file:{os.environ.get("DATADIR")}/sqlpage.db?mode=ro', uri=True) as db:
@@ -16,13 +40,19 @@ def yield_from_db():
             yield row
 
 def yield_tests():
-    #yield (105, 'keskonfai.fr',)
+    yield (105, 'keskonfai.fr',)
     yield (7, 'mobilizon.fr',)
 
+SOURCES = {
+    "test": yield_tests,
+    "local": yield_from_db,
+    "remote": yield_from_joinmbz,
+}
+
 class Fedator(Spider):
     name = "fedator"
     custom_settings = {
-        "USER_AGENT": "Mozilla/5.0 (X11; Linux x86_64; rv:100.0) Gecko/20100101 Firefox/100.0",
+        "USER_AGENT": UA,
         "ROBOTSTXT_OBEY": False,
         "REFERER_ENABLED": False,
         "COOKIES_ENABLED": False,
@@ -40,9 +70,8 @@ class Fedator(Spider):
         gql_query = open('src/query.gql', 'r').read()
         from json import dumps
         body = dumps({"query": gql_query})
-        # bbody = bytes(body, encoding='utf-8')
-        # sys.stdout.buffer.write(bbody)
-        for row in yield_tests():
+        source = SOURCES.get(os.environ.get("SOURCE","test"))
+        for row in source():
             domain = row[1]
             yield Request(
                 url = f"https://{domain}/api",
@@ -55,7 +84,7 @@ class Fedator(Spider):
         res:Model = Model.model_validate_json(response.body)
         meta = response.request.meta
         instance_id = meta.get('record')[0]
-        with sqlite3.connect(f'file:{os.environ.get("DATADIR")}/sqlpage.db', timeout=15.0, isolation_level='IMMEDIATE') as db:
+        with sqlite3.connect(f'{os.environ.get("DATADIR")}/sqlpage.db', timeout=15.0, isolation_level='IMMEDIATE') as db:
             # insert new stats
             stats = res.data.statistics
             db.execute('''insert into stats(insertedAt, instance_id, users, local_groups, total_groups, local_events, total_events, local_comments, total_comments, following, followers) values (?,?,?,?,?,?,?,?,?,?,?)

src/model.py

@@ -7,12 +7,12 @@ from pydantic import BaseModel, Field
 
 class Config(BaseModel):
     country_code: Any = Field(..., alias='countryCode')
-    description: str
+    description: Optional[str] = None
     languages: List[str]
-    long_description: str = Field(..., alias='longDescription')
+    long_description: Optional[str] = Field(..., alias='longDescription')
     name: str
     registrations_open: bool = Field(..., alias='registrationsOpen')
-    slogan: str
+    slogan: Optional[str] = None
     version: str
 
 
@@ -20,8 +20,8 @@ class Statistics(BaseModel):
     number_of_comments: int = Field(..., alias='numberOfComments')
     number_of_events: int = Field(..., alias='numberOfEvents')
     number_of_groups: int = Field(..., alias='numberOfGroups')
-    number_of_instance_followers: int = Field(..., alias='numberOfInstanceFollowers')
-    number_of_instance_followings: int = Field(..., alias='numberOfInstanceFollowings')
+    number_of_instance_followers: Optional[int] = Field(..., alias='numberOfInstanceFollowers')
+    number_of_instance_followings: Optional[int] = Field(..., alias='numberOfInstanceFollowings')
     number_of_local_comments: int = Field(..., alias='numberOfLocalComments')
     number_of_local_events: int = Field(..., alias='numberOfLocalEvents')
     number_of_local_groups: int = Field(..., alias='numberOfLocalGroups')