Merge branch 'trusted-sources' of 'https://github.com/jjmerchante/gri…

…moirelab-sortinghat' Merges #928 Closes #928
chaoss · Sep 18, 2024 · de6ddf8 · de6ddf8
2 parents d547ea0 + f7292fd
commit de6ddf8
Show file tree

Hide file tree

Showing 5 changed files with 24 additions and 3 deletions.
diff --git a/config/settings/config_testing.py b/config/settings/config_testing.py
@@ -18,6 +18,8 @@
 
 SORTINGHAT_GENDERIZE_API_KEY = 'fake-key'
 
+MATCH_TRUSTED_SOURCES = ['github', 'gitlab', 'slack']
+
 DEFAULT_AUTO_FIELD = 'django.db.models.AutoField'
 
 INSTALLED_APPS = [

diff --git a/config/settings/devel.py b/config/settings/devel.py
@@ -181,3 +181,6 @@
 }
 
 MULTI_TENANT = os.environ.get('SORTINGHAT_MULTI_TENANT', 'False').lower() in ('true', '1')
+
+MATCH_TRUSTED_SOURCES = os.environ.get('SORTINGHAT_MATCH_TRUSTED_SOURCES',
+                                       'github,gitlab,slack').split(',')
diff --git a/releases/unreleased/customizable-trusted-sources-for-username-matching.yml b/releases/unreleased/customizable-trusted-sources-for-username-matching.yml
@@ -0,0 +1,9 @@
+---
+title: Customizable trusted sources for username matching
+category: added
+author: Jose Javier Merchante <jjmerchante@bitergia.com>
+issue: null
+notes: >
+  Enable customization of trusted sources for username matching in
+  settings. This feature allows to define and manage the trusted data
+  sources for username matching from the configuration.
diff --git a/sortinghat/config/settings.py b/sortinghat/config/settings.py
@@ -379,3 +379,10 @@
 
 PERMISSION_GROUPS_LIST_PATH = os.environ.get('SORTINGHAT_PERMISSION_GROUPS_LIST_PATH',
                                             os.path.join(BASE_DIR, 'config', 'permission_groups.json'))
+
+#
+# Trusted data sources for matching by username
+#
+
+MATCH_TRUSTED_SOURCES = os.environ.get('SORTINGHAT_MATCH_TRUSTED_SOURCES',
+                                       'github,gitlab,slack').split(',')
diff --git a/sortinghat/core/recommendations/matching.py b/sortinghat/core/recommendations/matching.py
@@ -26,6 +26,8 @@
 import numpy
 
 from collections import defaultdict
+
+from django.conf import settings
 from django.forms.models import model_to_dict
 
 from ..db import (find_individual_by_uuid)
@@ -39,8 +41,6 @@
 EMAIL_ADDRESS_REGEX = r"^(?P<email>[^\s@]+@[^\s@.]+\.[^\s@]+)$"
 NAME_REGEX = r"^\w+\s\w+"
 
-MATCH_USERNAME_SOURCES = ['github', 'gitlab', 'slack']
-
 
 def recommend_matches(source_uuids, target_uuids,
                       criteria, exclude=True,
@@ -199,7 +199,7 @@ def _filter_criteria(df, c, strict=True, match_source=False):
         if match_source and c == 'username':
             cols += ['source']
             cdf = df[cols]
-            cdf = cdf[cdf['source'].isin(MATCH_USERNAME_SOURCES)]
+            cdf = cdf[cdf['source'].isin(settings.MATCH_TRUSTED_SOURCES)]
         else:
             cdf = df[cols]
         cdf = cdf.dropna(subset=[c])