Merge branch 'main' into scrape_citations_command

cl/citations/match_citations_queries.py

@@ -131,11 +131,13 @@ def es_case_name_query(
 
 
 def es_search_db_for_full_citation(
-    full_citation: FullCaseCitation,
+    full_citation: FullCaseCitation, query_citation: bool = False
 ) -> list[Hit]:
     """For a citation object, try to match it to an item in the database using
     a variety of heuristics.
     :param full_citation: A FullCaseCitation instance.
+    :param query_citation: Whether this is related to es_get_query_citation
+    resolution
     return: A ElasticSearch Result object with the results, or an empty list if
      no hits
     """
@@ -147,12 +149,20 @@ def es_search_db_for_full_citation(
         Q(
             "term", **{"status.raw": "Published"}
         ),  # Non-precedential documents aren't cited
-        Q("match", cluster_child="opinion"),
     ]
+
+    if query_citation:
+        # If this is related to query citation resolution, look for
+        # opinion_cluster to determine if a citation matched a single cluster.
+        filters.append(Q("match", cluster_child="opinion_cluster"))
+    else:
+        filters.append(Q("match", cluster_child="opinion"))
+
     must_not = []
     if full_citation.citing_opinion is not None:
         # Eliminate self-cites.
         must_not.append(Q("match", id=full_citation.citing_opinion.pk))
+
     # Set up filter parameters
     if full_citation.year:
         start_year = end_year = full_citation.year
@@ -204,7 +214,6 @@ def es_search_db_for_full_citation(
                 full_citation.citing_opinion,
             )
             return results
-
     # Give up.
     return []
 
@@ -225,7 +234,9 @@ def es_get_query_citation(cd: CleanData) -> Hit | None:
     matches = None
     if len(citations) == 1:
         # If it's not exactly one match, user doesn't get special help.
-        matches = es_search_db_for_full_citation(citations[0])
+        matches = es_search_db_for_full_citation(
+            citations[0], query_citation=True
+        )
         if len(matches) == 1:
             # If more than one match, don't show the tip
             return matches[0]

cl/lib/elasticsearch_utils.py

@@ -1789,6 +1789,83 @@ def merge_unavailable_fields_on_parent_document(
                         result["id"], ""
                     )
 
+        case (
+            SEARCH_TYPES.RECAP | SEARCH_TYPES.DOCKETS
+        ) if request_type == "frontend":
+            # Merge initial complaint button to the frontend search results.
+            docket_ids = {doc["docket_id"] for doc in results}
+            # This query retrieves initial complaint documents considering two
+            # possibilities:
+            # 1. For district, bankruptcy, and appellate entries where we don't know
+            #    if the entry contains attachments, it considers:
+            #    document_number=1 and attachment_number=None and document_type=PACER_DOCUMENT
+            #    This represents the main document with document_number 1.
+            # 2. For appellate entries where the attachment page has already been
+            #    merged, it considers:
+            #    document_number=1 and attachment_number=1 and document_type=ATTACHMENT
+            #    This represents document_number 1 that has been converted to an attachment.
+
+            appellate_court_ids = (
+                Court.federal_courts.appellate_pacer_courts().values_list(
+                    "pk", flat=True
+                )
+            )
+            initial_complaints = (
+                RECAPDocument.objects.filter(
+                    QObject(
+                        QObject(
+                            attachment_number=None,
+                            document_type=RECAPDocument.PACER_DOCUMENT,
+                        )
+                        | QObject(
+                            attachment_number=1,
+                            document_type=RECAPDocument.ATTACHMENT,
+                            docket_entry__docket__court_id__in=appellate_court_ids,
+                        )
+                    ),
+                    docket_entry__docket_id__in=docket_ids,
+                    document_number="1",
+                )
+                .select_related(
+                    "docket_entry",
+                    "docket_entry__docket",
+                    "docket_entry__docket__court",
+                )
+                .only(
+                    "pk",
+                    "document_type",
+                    "document_number",
+                    "attachment_number",
+                    "pacer_doc_id",
+                    "is_available",
+                    "filepath_local",
+                    "docket_entry__docket_id",
+                    "docket_entry__docket__slug",
+                    "docket_entry__docket__pacer_case_id",
+                    "docket_entry__docket__court__jurisdiction",
+                    "docket_entry__docket__court_id",
+                )
+            )
+            initial_complaints_in_page = {}
+            for initial_complaint in initial_complaints:
+                if initial_complaint.has_valid_pdf:
+                    initial_complaints_in_page[
+                        initial_complaint.docket_entry.docket_id
+                    ] = (initial_complaint.get_absolute_url(), None)
+                else:
+                    initial_complaints_in_page[
+                        initial_complaint.docket_entry.docket_id
+                    ] = (None, initial_complaint.pacer_url)
+
+            for result in results:
+                complaint_url, buy_complaint_url = (
+                    initial_complaints_in_page.get(
+                        result.docket_id, (None, None)
+                    )
+                )
+                result["initial_complaint_url"] = complaint_url
+                result["buy_initial_complaint_url"] = buy_complaint_url
+
         case SEARCH_TYPES.OPINION if request_type == "v4" and not highlight:
             # Retrieves the Opinion plain_text from the DB to fill the snippet
             # when highlighting is disabled. Considering the same prioritization

cl/search/signals.py

@@ -329,11 +329,11 @@
         RECAPDocument: {
             "self": {
                 "description": ["short_description"],
-                "document_type": ["document_type"],
+                "document_type": ["document_type", "absolute_url"],
                 "document_number": ["document_number", "absolute_url"],
                 "pacer_doc_id": ["pacer_doc_id"],
                 "plain_text": ["plain_text"],
-                "attachment_number": ["attachment_number"],
+                "attachment_number": ["attachment_number", "absolute_url"],
                 "is_available": ["is_available"],
                 "page_count": ["page_count"],
                 "filepath_local": ["filepath_local"],
@@ -364,6 +364,7 @@
                 "assigned_to_str": ["assignedTo"],
                 "referred_to_str": ["referredTo"],
                 "pacer_case_id": ["pacer_case_id"],
+                "slug": ["absolute_url"],
             }
         },
         Person: {

cl/search/templates/includes/search_result.html

@@ -207,6 +207,15 @@ <h4>
       {% endwith %}
     {% endfor %}
     <div class="col-md-offset-half">
+      {% if result.initial_complaint_url %}
+        <a href="{{ result.initial_complaint_url }}" class="initial-complaint btn-primary btn">
+          Initial Complaint
+        </a>
+      {% elif result.buy_initial_complaint_url %}
+        <a href="{{ result.buy_initial_complaint_url }}" rel="nofollow" target="_blank"  class="initial-complaint btn-primary btn">
+          Buy Initial Complaint
+        </a>
+      {% endif %}
       {% if result.child_remaining %}
         <a href="{% url "show_results" %}?type={{ type|urlencode }}&q={% if request.GET.q %}({{ request.GET.q|urlencode }})%20AND%20{% endif %}docket_id%3A{{ result.docket_id|urlencode }}" class="btn-default btn">
           View Additional Results for this Case

cl/search/tests/tests_es_opinion.py

@@ -1982,13 +1982,15 @@ def test_display_query_citation_frontend(self) -> None:
         """Confirm if the query citation alert is shown on the frontend when
         querying a single citation, and it's found into ES."""
 
+        # Cluster with citation and multiple sibling opinions is properly matched.
         with self.captureOnCommitCallbacks(execute=True):
             cluster = OpinionClusterFactory.create(
                 precedential_status=PRECEDENTIAL_STATUS.PUBLISHED,
                 docket=self.docket_1,
                 date_filed=datetime.date(2024, 8, 23),
             )
             OpinionFactory.create(cluster=cluster, plain_text="")
+            OpinionFactory.create(cluster=cluster, plain_text="")
             CitationWithParentsFactory.create(
                 volume=31,
                 reporter="Pa. D. & C.",
@@ -2009,6 +2011,39 @@ def test_display_query_citation_frontend(self) -> None:
         self.assertIn(
             "It looks like you're trying to search for", r.content.decode()
         )
+
+        # Add a new cluster for the same citation. This time, it is not
+        # possible to identify a unique case for the citation.
+        with self.captureOnCommitCallbacks(execute=True):
+            cluster_2 = OpinionClusterFactory.create(
+                case_name="Test case",
+                precedential_status=PRECEDENTIAL_STATUS.PUBLISHED,
+                docket=self.docket_1,
+                date_filed=datetime.date(2024, 8, 23),
+            )
+            OpinionFactory.create(cluster=cluster_2, plain_text="")
+            CitationWithParentsFactory.create(
+                volume=31,
+                reporter="Pa. D. & C.",
+                page="445",
+                type=2,
+                cluster=cluster_2,
+            )
+
+        search_params = {
+            "type": SEARCH_TYPES.OPINION,
+            "q": "31 Pa. D. & C. 445",
+            "order_by": "score desc",
+        }
+        r = self.client.get(
+            reverse("show_results"),
+            search_params,
+        )
+        self.assertNotIn(
+            "It looks like you're trying to search for", r.content.decode()
+        )
+
+        cluster_2.delete()
         cluster.delete()