Automatically add tags to bookmarks based on URL pattern (sissbruecke…

…r#736)

* [WIP] DSL

* upd

* upd

* upd

* upd

* upd

* upd

* upd

* upd

* upd

* upd

* upd

* dsl2

* full feature

* upd

* upd

* upd

* upd

* rename to auto_tagging_rules

* update migration after rebase

* add REST API tests

* improve settings view

---------

Co-authored-by: Sascha Ißbrücker <[email protected]>

bookmarks/migrations/0036_userprofile_auto_tagging_rules.py

@@ -0,0 +1,18 @@
+# Generated by Django 5.0.3 on 2024-05-17 07:09
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ("bookmarks", "0035_userprofile_tag_grouping"),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name="userprofile",
+            name="auto_tagging_rules",
+            field=models.TextField(blank=True),
+        ),
+    ]

bookmarks/models.py

@@ -415,6 +415,7 @@ class UserProfile(models.Model):
     display_remove_bookmark_action = models.BooleanField(default=True, null=False)
     permanent_notes = models.BooleanField(default=False, null=False)
     custom_css = models.TextField(blank=True, null=False)
+    auto_tagging_rules = models.TextField(blank=True, null=False)
     search_preferences = models.JSONField(default=dict, null=False)
     enable_automatic_html_snapshots = models.BooleanField(default=True, null=False)
     default_mark_unread = models.BooleanField(default=False, null=False)
@@ -445,6 +446,7 @@ class Meta:
             "permanent_notes",
             "default_mark_unread",
             "custom_css",
+            "auto_tagging_rules",
         ]
 
 

bookmarks/services/auto_tagging.py

@@ -0,0 +1,70 @@
+from urllib.parse import urlparse, parse_qs
+import re
+import idna
+
+
+def get_tags(script: str, url: str):
+    parsed_url = urlparse(url.lower())
+    result = set()
+
+    for line in script.lower().split("\n"):
+        if "#" in line:
+            i = line.index("#")
+            line = line[:i]
+
+        parts = line.split()
+        if len(parts) < 2:
+            continue
+
+        domain_pattern = re.sub("^https?://", "", parts[0])
+        path_pattern = None
+        qs_pattern = None
+
+        if "/" in domain_pattern:
+            i = domain_pattern.index("/")
+            path_pattern = domain_pattern[i:]
+            domain_pattern = domain_pattern[:i]
+
+        if path_pattern and "?" in path_pattern:
+            i = path_pattern.index("?")
+            qs_pattern = path_pattern[i + 1 :]
+            path_pattern = path_pattern[:i]
+
+        if not _domains_matches(domain_pattern, parsed_url.netloc):
+            continue
+
+        if path_pattern and not _path_matches(path_pattern, parsed_url.path):
+            continue
+
+        if qs_pattern and not _qs_matches(qs_pattern, parsed_url.query):
+            continue
+
+        for tag in parts[1:]:
+            result.add(tag)
+
+    return result
+
+
+def _path_matches(expected_path: str, actual_path: str) -> bool:
+    return actual_path.startswith(expected_path)
+
+
+def _domains_matches(expected_domain: str, actual_domain: str) -> bool:
+    expected_domain = idna.encode(expected_domain)
+    actual_domain = idna.encode(actual_domain)
+
+    return actual_domain.endswith(expected_domain)
+
+
+def _qs_matches(expected_qs: str, actual_qs: str) -> bool:
+    expected_qs = parse_qs(expected_qs, keep_blank_values=True)
+    actual_qs = parse_qs(actual_qs, keep_blank_values=True)
+
+    for key in expected_qs:
+        if key not in actual_qs:
+            return False
+        for value in expected_qs[key]:
+            if value != "" and value not in actual_qs[key]:
+                return False
+
+    return True

bookmarks/services/bookmarks.py

@@ -10,6 +10,7 @@
 from bookmarks.models import Bookmark, BookmarkAsset, parse_tag_string
 from bookmarks.services import tasks
 from bookmarks.services import website_loader
+from bookmarks.services import auto_tagging
 from bookmarks.services.tags import get_or_create_tags
 
 logger = logging.getLogger(__name__)
@@ -242,6 +243,15 @@ def _update_website_metadata(bookmark: Bookmark):
 
 def _update_bookmark_tags(bookmark: Bookmark, tag_string: str, user: User):
     tag_names = parse_tag_string(tag_string)
+
+    if user.profile.auto_tagging_rules:
+        auto_tag_names = auto_tagging.get_tags(
+            user.profile.auto_tagging_rules, bookmark.url
+        )
+        for auto_tag_name in auto_tag_names:
+            if auto_tag_name not in tag_names:
+                tag_names.append(auto_tag_name)
+
     tags = get_or_create_tags(tag_names, user)
     bookmark.tags.set(tags)
 

bookmarks/templates/settings/general.html

@@ -118,6 +118,21 @@ <h2>Profile</h2>
             If disabled, tags will not be grouped.
           </div>
         </div>
+        <div class="form-group">
+          <details {% if form.auto_tagging_rules.value %}open{% endif %}>
+            <summary>Auto Tagging</summary>
+            <label for="{{ form.auto_tagging_rules.id_for_label }}" class="text-assistive">Auto Tagging</label>
+            <div class="mt-2">
+              {{ form.auto_tagging_rules|add_class:"form-input custom-css"|attr:"rows:6" }}
+            </div>
+          </details>
+          <div class="form-input-hint">
+            Automatically adds tags to bookmarks based on predefined rules.
+            Each line is a single rule that maps a URL to one or more tags. For example:
+            <pre>youtube.com video
+reddit.com/r/Music music reddit</pre>
+          </div>
+        </div>
         <div class="form-group">
           <label for="{{ form.enable_favicons.id_for_label }}" class="form-checkbox">
             {{ form.enable_favicons }}

bookmarks/tests/test_auto_tagging.py

@@ -0,0 +1,179 @@
+from bookmarks.services import auto_tagging
+from django.test import TestCase
+
+
+class AutoTaggingTestCase(TestCase):
+    def test_auto_tag_by_domain(self):
+        script = """
+            example.com example
+            test.com test
+        """
+        url = "https://example.com/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["example"]))
+
+    def test_auto_tag_by_domain_ignores_case(self):
+        script = """
+            EXAMPLE.com example
+        """
+        url = "https://example.com/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["example"]))
+
+    def test_auto_tag_by_domain_should_add_all_tags(self):
+        script = """
+            example.com one two three
+        """
+        url = "https://example.com/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["one", "two", "three"]))
+
+    def test_auto_tag_by_domain_work_with_idn_domains(self):
+        script = """
+            रजिस्ट्री.भारत tag1
+        """
+        url = "https://www.xn--81bg3cc2b2bk5hb.xn--h2brj9c/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["tag1"]))
+
+        script = """
+            xn--81bg3cc2b2bk5hb.xn--h2brj9c tag1
+        """
+        url = "https://www.रजिस्ट्री.भारत/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["tag1"]))
+
+    def test_auto_tag_by_domain_and_path(self):
+        script = """
+            example.com/one one
+            example.com/two two
+            test.com test
+        """
+        url = "https://example.com/one/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["one"]))
+
+    def test_auto_tag_by_domain_and_path_ignores_case(self):
+        script = """
+            example.com/One one
+        """
+        url = "https://example.com/one/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["one"]))
+
+    def test_auto_tag_by_domain_and_path_matches_path_ltr(self):
+        script = """
+            example.com/one one
+            example.com/two two
+            test.com test
+        """
+        url = "https://example.com/one/two"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["one"]))
+
+    def test_auto_tag_by_domain_ignores_domain_in_path(self):
+        script = """
+            example.com example
+        """
+        url = "https://test.com/example.com"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set([]))
+
+    def test_auto_tag_by_domain_includes_subdomains(self):
+        script = """
+            example.com example
+            test.example.com test
+            some.example.com some
+        """
+        url = "https://test.example.com/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["example", "test"]))
+
+    def test_auto_tag_by_domain_matches_domain_rtl(self):
+        script = """
+            example.com example
+        """
+        url = "https://example.com.bad-website.com/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set([]))
+
+    def test_auto_tag_by_domain_ignores_schema(self):
+        script = """
+            https://example.com/ https
+            http://example.com/ http
+        """
+        url = "http://example.com/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["https", "http"]))
+
+    def test_auto_tag_by_domain_ignores_lines_with_no_tags(self):
+        script = """
+            example.com
+        """
+        url = "https://example.com/"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set([]))
+
+    def test_auto_tag_by_domain_path_and_qs(self):
+        script = """
+            example.com/page?a=b tag1     # true, matches a=b
+            example.com/page?a=c&c=d tag2 # true, matches both a=c and c=d
+            example.com/page?c=d&l=p tag3 # false, l=p doesn't exists
+            example.com/page?a=bb tag4    # false bb != b
+            example.com/page?a=b&a=c tag5 # true, matches both a=b and a=c
+            example.com/page?a=B tag6     # true, matches a=b because case insensitive
+            example.com/page?A=b tag7     # true, matches a=b because case insensitive
+        """
+        url = "https://example.com/page/some?z=x&a=b&v=b&c=d&o=p&a=c"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["tag1", "tag2", "tag5", "tag6", "tag7"]))
+
+    def test_auto_tag_by_domain_path_and_qs_with_empty_value(self):
+        script = """
+            example.com/page?a= tag1
+            example.com/page?b= tag2
+        """
+        url = "https://example.com/page/some?a=value"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["tag1"]))
+
+    def test_auto_tag_by_domain_path_and_qs_works_with_encoded_url(self):
+        script = """
+            example.com/page?a=йцу tag1
+            example.com/page?a=%D0%B9%D1%86%D1%83 tag2
+        """
+        url = "https://example.com/page?a=%D0%B9%D1%86%D1%83"
+
+        tags = auto_tagging.get_tags(script, url)
+
+        self.assertEqual(tags, set(["tag1", "tag2"]))

bookmarks/tests/test_bookmarks_api.py

@@ -440,6 +440,20 @@ def test_create_bookmark_is_not_shared_by_default(self):
         bookmark = Bookmark.objects.get(url=data["url"])
         self.assertFalse(bookmark.shared)
 
+    def test_create_bookmark_should_add_tags_from_auto_tagging(self):
+        tag1 = self.setup_tag()
+        tag2 = self.setup_tag()
+
+        self.authenticate()
+        profile = self.get_or_create_test_user().profile
+        profile.auto_tagging_rules = f"example.com {tag2.name}"
+        profile.save()
+
+        data = {"url": "https://example.com/", "tag_names": [tag1.name]}
+        self.post(reverse("bookmarks:bookmark-list"), data, status.HTTP_201_CREATED)
+        bookmark = Bookmark.objects.get(url=data["url"])
+        self.assertCountEqual(bookmark.tags.all(), [tag1, tag2])
+
     def test_get_bookmark(self):
         self.authenticate()
         bookmark = self.setup_bookmark()
@@ -512,6 +526,22 @@ def test_update_bookmark_shared_flag(self):
         updated_bookmark = Bookmark.objects.get(id=bookmark.id)
         self.assertEqual(updated_bookmark.shared, True)
 
+    def test_update_bookmark_adds_tags_from_auto_tagging(self):
+        bookmark = self.setup_bookmark()
+        tag1 = self.setup_tag()
+        tag2 = self.setup_tag()
+
+        self.authenticate()
+        profile = self.get_or_create_test_user().profile
+        profile.auto_tagging_rules = f"example.com {tag2.name}"
+        profile.save()
+
+        data = {"url": "https://example.com/", "tag_names": [tag1.name]}
+        url = reverse("bookmarks:bookmark-detail", args=[bookmark.id])
+        self.put(url, data, expected_status_code=status.HTTP_200_OK)
+        updated_bookmark = Bookmark.objects.get(id=bookmark.id)
+        self.assertCountEqual(updated_bookmark.tags.all(), [tag1, tag2])
+
     def test_patch_bookmark(self):
         self.authenticate()
         bookmark = self.setup_bookmark()
@@ -583,6 +613,22 @@ def test_patch_with_empty_payload_does_not_modify_bookmark(self):
         self.assertEqual(updated_bookmark.description, bookmark.description)
         self.assertListEqual(updated_bookmark.tag_names, bookmark.tag_names)
 
+    def test_patch_bookmark_adds_tags_from_auto_tagging(self):
+        bookmark = self.setup_bookmark()
+        tag1 = self.setup_tag()
+        tag2 = self.setup_tag()
+
+        self.authenticate()
+        profile = self.get_or_create_test_user().profile
+        profile.auto_tagging_rules = f"example.com {tag2.name}"
+        profile.save()
+
+        data = {"tag_names": [tag1.name]}
+        url = reverse("bookmarks:bookmark-detail", args=[bookmark.id])
+        self.patch(url, data, expected_status_code=status.HTTP_200_OK)
+        updated_bookmark = Bookmark.objects.get(id=bookmark.id)
+        self.assertCountEqual(updated_bookmark.tags.all(), [tag1, tag2])
+
     def test_delete_bookmark(self):
         self.authenticate()
         bookmark = self.setup_bookmark()