Merge branch 'main' into 1484-ca7-missing-information

Author: flooie

CHANGES.md

@@ -16,6 +16,7 @@ The following changes are not yet released, but are code complete:
 
 Features:
 - Add error handling for scrapers with expected results #1447
+- Add a check to verify ACMS user data is loaded before querying attachment pages #1495
 
 Changes:
 - Expanded ACMS URL matching to support both HTTP and HTTPS protocols.
@@ -63,6 +64,7 @@ Changes:
 - Update `tenn` add backscraper #1425
 
 Fixes:
+- Add "lower_court_ids" to fields returned by OpinionSite #1432
 - Fix `va` collecting bad docket_number values #1441
 - Fix `mich` change date expected key to `date_filed` #1445
 

juriscraper/OpinionSite.py

@@ -25,6 +25,7 @@ def __init__(self, *args, **kwargs):
             "docket_numbers",
             "judges",
             "lower_courts",
+            "lower_court_ids",
             "lower_court_judges",
             "lower_court_numbers",
             "nature_of_suit",
@@ -104,6 +105,9 @@ def _get_parallel_citations(self):
     def _get_lower_courts(self):
         return None
 
+    def _get_lower_court_ids(self):
+        return None
+
     def _get_lower_court_judges(self):
         return None
 

juriscraper/OpinionSiteLinear.py

@@ -33,6 +33,7 @@ class OpinionSiteLinear(OpinionSite):
         "docket_attachment_number",
         "docket_document_number",
         "nature_of_suit",
+        "lower_court_id",
         "lower_court_number",
         "lower_court_judge",
         "author",
@@ -112,6 +113,9 @@ def _get_child_courts(self):
     def _get_lower_courts(self):
         return self._get_optional_field_by_id("lower_court")
 
+    def _get_lower_court_ids(self):
+        return self._get_optional_field_by_id("lower_court_id")
+
     def _get_lower_court_judges(self):
         return self._get_optional_field_by_id("lower_court_judge")
 

juriscraper/opinions/united_states/state/tex.py

@@ -116,8 +116,8 @@ def _process_html(self) -> None:
                 author = ""
                 per_curiam = True
 
-            lower_court, lower_court_number = self.parse_lower_court_info(
-                title
+            lower_court, lower_court_number, lower_court_id = (
+                self.parse_lower_court_info(title)
             )
 
             title_regex = r"^(?P<name>.*?)(?=; from|(?=\([^)]*\)$))"
@@ -138,22 +138,23 @@ def _process_html(self) -> None:
                     "author": author,
                     "lower_court": lower_court,
                     "lower_court_number": lower_court_number,
+                    "lower_court_id": lower_court_id,
                 }
             )
 
     @staticmethod
-    def parse_lower_court_info(title: str) -> tuple[str, str]:
+    def parse_lower_court_info(title: str) -> tuple[str, str, str]:
         """Parses lower court information from the title string
 
-        :param title string
-        :return lower_court, lower_court_number
+        :param title: a string with lower court information
+        :return: values for lower_court, lower_court_number, lower_court_id
         """
 
-        # format when appeal comes from texapp. Example:
+        # Format when appeal comes from texapp districts. Example:
         # ' from Harris County; 1st Court of Appeals District (01-22-00182-CV, 699 SW3d 20, 03-23-23)'
         texapp_regex = r" from (?P<lower_court>.*)\s*\("
 
-        # Examples:
+        # Format when appeal comes from other possible courts. Examples:
         #  "(U.S. Fifth Circuit 23-10804)"
         #  "(U.S. 5th Circuit 19-51012)"
         # "(BODA Cause No. 67623)"
@@ -162,16 +163,23 @@ def parse_lower_court_info(title: str) -> tuple[str, str]:
         if match := re.search(texapp_regex, title):
             lower_court = match.group("lower_court")
             lower_court_number = title[match.end() :].split(",")[0]
-            return lower_court, lower_court_number
+            return lower_court, lower_court_number, "texapp"
+
         elif match := re.search(other_courts_regex, title):
             lower_court = match.group("lower_court")
             lower_court_number = match.group("lower_number")
 
-            if "BODA" in lower_court_number:
+            if lower_court == "BODA":
                 lower_court = "Board of Disciplinary Appeals"
+                lower_court_id = "txboda"
+            else:
+                # if this is not a BODA match, then it can only be a
+                # Fifth Circuit match. Update this if the regex above changes
+                lower_court_id = "ca5"
+
+            return lower_court, lower_court_number, lower_court_id
 
-            return lower_court, lower_court_number
-        return "", ""
+        return "", "", ""
 
     @staticmethod
     def extract_type(link: etree.Element) -> str:

juriscraper/pacer/acms_attachment_page.py

@@ -95,6 +95,13 @@ def query(self, case_id: str, entry_id: str):
         :param entry_id: The unique identifier of the specific docket entry to
                          retrieve.
         """
+        # Ensure ACMS user data is available before fetching attachments.
+        # Attachment requests include user-specific data in the request body,
+        # so the session must have the necessary authentication data
+        # initialized beforehand.
+        if not self.session.acms_user_data:
+            self.session.get_acms_auth_object(self.court_id)
+
         # Fetch Docket Entry Details
         docket_info = self.api_client.get_docket_entries(case_id)
 

tests/examples/opinions/united_states/tex_example.compare.json

@@ -10,6 +10,7 @@
     "docket_numbers": "23-0808",
     "judges": "Blacklock",
     "lower_courts": "Harris County; 14th Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "14-22-00013-CV",
     "case_name_shorts": "",
     "authors": "Blacklock",
@@ -27,6 +28,7 @@
     "docket_numbers": "23-0607",
     "judges": "Bland, Blacklock, Devine",
     "lower_courts": "Hidalgo County; 13th Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "13-22-00374-CV",
     "case_name_shorts": "",
     "authors": "Bland",
@@ -44,6 +46,7 @@
     "docket_numbers": "23-0607",
     "judges": "Busby",
     "lower_courts": "Hidalgo County; 13th Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "13-22-00374-CV",
     "case_name_shorts": "",
     "authors": "Busby",
@@ -61,6 +64,7 @@
     "docket_numbers": "23-0460",
     "judges": "Bland, Devine",
     "lower_courts": "Tarrant County; 2nd Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "02-22-00319-CV",
     "case_name_shorts": "",
     "authors": "Bland",
@@ -78,6 +82,7 @@
     "docket_numbers": "23-0460",
     "judges": "Huddle, Blacklock, Lehrmann, Boyd, Busby, Young, Sullivan",
     "lower_courts": "Tarrant County; 2nd Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "02-22-00319-CV",
     "case_name_shorts": "",
     "authors": "Huddle",
@@ -95,6 +100,7 @@
     "docket_numbers": "23-1035",
     "judges": "",
     "lower_courts": "Dallas County; 5th Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "05-23-00984-CV",
     "case_name_shorts": "",
     "authors": "",
@@ -112,6 +118,7 @@
     "docket_numbers": "23-0927",
     "judges": "Sullivan",
     "lower_courts": "Loving County; 8th Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "08-22-00129-CV",
     "case_name_shorts": "",
     "authors": "Sullivan",
@@ -129,6 +136,7 @@
     "docket_numbers": "24-0428",
     "judges": "",
     "lower_courts": "Harris County; 14th Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "14-23-00087-CV",
     "case_name_shorts": "",
     "authors": "",
@@ -146,6 +154,7 @@
     "docket_numbers": "24-0616",
     "judges": "Lehrmann",
     "lower_courts": "U.S. Fifth Circuit",
+    "lower_court_ids": "ca5",
     "lower_court_numbers": "23-10072",
     "case_name_shorts": "",
     "authors": "Lehrmann",
@@ -163,6 +172,7 @@
     "docket_numbers": "24-0759",
     "judges": "Sullivan",
     "lower_courts": "U.S. Fifth Circuit",
+    "lower_court_ids": "ca5",
     "lower_court_numbers": "23-10804",
     "case_name_shorts": "",
     "authors": "Sullivan",
@@ -180,6 +190,7 @@
     "docket_numbers": "23-0384",
     "judges": "Sullivan",
     "lower_courts": "Harris County; 1st Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "01-20-00055-CV",
     "case_name_shorts": "",
     "authors": "Sullivan",
@@ -197,6 +208,7 @@
     "docket_numbers": "24-0273",
     "judges": "Bland",
     "lower_courts": "Harris County; 14th Court of Appeals District",
+    "lower_court_ids": "texapp",
     "lower_court_numbers": "14-23-00090-CV",
     "case_name_shorts": "",
     "authors": "Bland",