rename model Page to Snapshot

archivebox/cli/logging.py

@@ -218,6 +218,7 @@ def log_archiving_paused(num_links: int, idx: int, timestamp: str):
         timestamp=timestamp,
         total=num_links,
     ))
+    print()
     print('    To view your archive, open:')
     print('        {}/index.html'.format(OUTPUT_DIR))
     print('    Continue archiving where you left off by running:')
@@ -244,8 +245,11 @@ def log_archiving_finished(num_links: int):
     print('    - {} links skipped'.format(_LAST_RUN_STATS.skipped))
     print('    - {} links updated'.format(_LAST_RUN_STATS.succeeded))
     print('    - {} links had errors'.format(_LAST_RUN_STATS.failed))
+    print()
     print('    To view your archive, open:')
     print('        {}/index.html'.format(OUTPUT_DIR))
+    print('    Or run the built-in webserver:')
+    print('        archivebox server')
 
 
 def log_link_archiving_started(link: Link, link_dir: str, is_new: bool):
@@ -378,15 +382,15 @@ def log_shell_welcome_msg():
     from . import list_subcommands
 
     print('{green}# ArchiveBox Imports{reset}'.format(**ANSI))
-    print('{green}from archivebox.core.models import Page, User{reset}'.format(**ANSI))
+    print('{green}from archivebox.core.models import Snapshot, User{reset}'.format(**ANSI))
     print('{green}from archivebox import *\n    {}{reset}'.format("\n    ".join(list_subcommands().keys()), **ANSI))
     print()
     print('[i] Welcome to the ArchiveBox Shell!')
     print('    https://github.com/pirate/ArchiveBox/wiki/Usage#Shell-Usage')
     print()
     print('    {lightred}Hint:{reset} Example use:'.format(**ANSI))
-    print('        print(Page.objects.filter(is_archived=True).count())')
-    print('        Page.objects.get(url="https://example.com").as_json()')
+    print('        print(Snapshot.objects.filter(is_archived=True).count())')
+    print('        Snapshot.objects.get(url="https://example.com").as_json()')
     print('        add("https://example.com/some/new/url")')
 
 

archivebox/core/admin.py

@@ -1,9 +1,9 @@
 from django.contrib import admin
 
-from core.models import Page
+from core.models import Snapshot
 
 
-class PageAdmin(admin.ModelAdmin):
+class SnapshotAdmin(admin.ModelAdmin):
     list_display = ('timestamp', 'short_url', 'title', 'is_archived', 'num_outputs', 'added', 'updated', 'url_hash')
     readonly_fields = ('num_outputs', 'is_archived', 'added', 'updated', 'bookmarked')
     fields = ('url', 'timestamp', 'title', 'tags', *readonly_fields)
@@ -14,4 +14,4 @@ class PageAdmin(admin.ModelAdmin):
     def updated(self, obj):
         return obj.isoformat()
 
-admin.site.register(Page, PageAdmin)
+admin.site.register(Snapshot, SnapshotAdmin)

archivebox/core/migrations/0001_initial.py

@@ -1,4 +1,4 @@
-# Generated by Django 2.2 on 2019-04-17 06:46
+# Generated by Django 2.2 on 2019-05-01 03:27
 
 from django.db import migrations, models
 import uuid
@@ -13,15 +13,14 @@ class Migration(migrations.Migration):
 
     operations = [
         migrations.CreateModel(
-            name='Page',
+            name='Snapshot',
             fields=[
                 ('id', models.UUIDField(default=uuid.uuid4, editable=False, primary_key=True, serialize=False)),
-                ('url', models.URLField()),
-                ('timestamp', models.CharField(default=None, max_length=32, null=True)),
+                ('url', models.URLField(unique=True)),
+                ('timestamp', models.CharField(default=None, max_length=32, null=True, unique=True)),
                 ('title', models.CharField(default=None, max_length=128, null=True)),
                 ('tags', models.CharField(default=None, max_length=256, null=True)),
                 ('added', models.DateTimeField(auto_now_add=True)),
-                ('bookmarked', models.DateTimeField()),
                 ('updated', models.DateTimeField(default=None, null=True)),
             ],
         ),

archivebox/core/migrations/0002_auto_20190417_0739.py

@@ -1,27 +0,0 @@
-# Generated by Django 2.2 on 2019-04-17 07:39
-
-from django.db import migrations, models
-
-
-class Migration(migrations.Migration):
-
-    dependencies = [
-        ('core', '0001_initial'),
-    ]
-
-    operations = [
-        migrations.RemoveField(
-            model_name='page',
-            name='bookmarked',
-        ),
-        migrations.AlterField(
-            model_name='page',
-            name='timestamp',
-            field=models.CharField(default=None, max_length=32, null=True, unique=True),
-        ),
-        migrations.AlterField(
-            model_name='page',
-            name='url',
-            field=models.URLField(unique=True),
-        ),
-    ]

archivebox/core/models.py

@@ -8,7 +8,7 @@ from ..util import parse_date
 from ..index.schema import Link
 
 
-class Page(models.Model):
+class Snapshot(models.Model):
     id = models.UUIDField(primary_key=True, default=uuid.uuid4, editable=False)
 
     url = models.URLField(unique=True)

archivebox/core/views.py

@@ -5,7 +5,7 @@ from django.shortcuts import render, redirect
 from django.http import HttpResponse
 from django.views import View, static
 
-from core.models import Page
+from core.models import Snapshot
 
 from ..index import load_main_index, load_main_index_meta
 from ..config import OUTPUT_DIR, VERSION, FOOTER_INFO
@@ -59,7 +59,7 @@ class LinkDetails(View):
         except (IndexError, ValueError):
             slug, archivefile = path.split('/', 1)[0], 'index.html'
 
-        all_pages = list(Page.objects.all())
+        all_pages = list(Snapshot.objects.all())
 
         # slug is a timestamp
         by_ts = {page.timestamp: page for page in all_pages}

archivebox/index/sql.py

@@ -13,30 +13,38 @@ from ..config import setup_django, OUTPUT_DIR
 @enforce_types
 def parse_sql_main_index(out_dir: str=OUTPUT_DIR) -> Iterator[Link]:
     setup_django(out_dir, check_db=True)
-    from core.models import Page
+    from core.models import Snapshot
 
     return (
-        Link.from_json(page.as_json(*Page.keys))
-        for page in Page.objects.all()
+        Link.from_json(page.as_json(*Snapshot.keys))
+        for page in Snapshot.objects.all()
     )
 
 @enforce_types
 def write_sql_main_index(links: List[Link], out_dir: str=OUTPUT_DIR) -> None:
     setup_django(out_dir, check_db=True)
-    from core.models import Page
+    from core.models import Snapshot
+    from django.db import transaction
 
     all_urls = {link.url: link for link in links}
-
-    for page in Page.objects.all():
-        if page.url in all_urls:
-            info = {k: v for k, v in all_urls.pop(page.url)._asdict().items() if k in Page.keys}
-            Page.objects.update(**info)
-        else:
-            page.delete()
-
-    for url, link in all_urls.items():
-        info = {k: v for k, v in link._asdict().items() if k in Page.keys}
-        Page.objects.update_or_create(url=url, defaults=info)
+    all_ts = {link.timestamp: link for link in links}
+
+    with transaction.atomic():
+        for snapshot in Snapshot.objects.all():
+            if snapshot.timestamp in all_ts:
+                info = {k: v for k, v in all_urls.pop(snapshot.url)._asdict().items() if k in Snapshot.keys}
+                snapshot.delete()
+                Snapshot.objects.create(**info)
+            if snapshot.url in all_urls:
+                info = {k: v for k, v in all_urls.pop(snapshot.url)._asdict().items() if k in Snapshot.keys}
+                snapshot.delete()
+                Snapshot.objects.create(**info)
+            else:
+                snapshot.delete()
+
+        for url, link in all_urls.items():
+            info = {k: v for k, v in link._asdict().items() if k in Snapshot.keys}
+            Snapshot.objects.update_or_create(url=url, defaults=info)