Fix filenames parameter

docassemble/AssemblyLine/data/questions/demo_search_sessions.yml

@@ -0,0 +1,59 @@
+---
+include:
+  - assembly_line.yml
+---
+mandatory: True
+code: |
+  search_keyword
+  if not matching_results:
+    no_matches
+  show_results
+  show_selected_interview
+---
+id: search_keyword
+question: |
+  What would you like to search for?
+subquestion: |
+  Searching is not case sensitive.
+fields:
+  - Keyword: search_keyword
+---
+code: |
+  matching_results = find_saved_interview(search_keyword, user_id="all")
+---
+event: no_matches
+question: |
+  We didn't find any matches
+subquestion: |
+  Check your spelling and try again.
+---
+continue button field: show_results
+question: |
+  We found ${ len(matching_results) } results that match your keyword
+subquestion: |
+  % for result in matching_results:
+  * ${ nice_interview_subtitle(result, exclude_identical=False)}
+  ${ result["modtime"]}
+  % endfor
+
+  Raw results:
+  <div class="overflow-scroll" style="height: 200px;">
+  ${ matching_results }
+  </div>
+fields:
+  - Which session do you want to load: interview_to_load
+    datatype: integer
+    code: |
+      [{idx: f"<b>{nice_interview_subtitle(answer, exclude_identical=False)}</b>" + f' ({answer.get("modtime", DAEmpty()).strftime("%B %d, %Y")})'} for idx, answer in enumerate(matching_results)]
+---
+event: show_selected_interview
+id: show the selected interview
+question: |
+  Here is some information from the session you chose
+subquestion: |
+  Filename: ${ matching_results[interview_to_load]["filename"] } [BR]
+  Session ID: ${ matching_results[interview_to_load]["key"] }
+
+  <div class="overflow-scroll" style="height: 200px;">
+  <pre>${ get_filtered_session_variables_string(filename=matching_results[interview_to_load]["filename"], session_id=matching_results[interview_to_load]["key"]) }</pre>
+  </div

docassemble/AssemblyLine/sessions.py

@@ -56,6 +56,7 @@
     "config_with_language_fallback",
     "delete_interview_sessions",
     "export_interview_variables",
+    "find_saved_interview",
     "get_filtered_session_variables_string",
     "get_filtered_session_variables",
     "get_interview_metadata",
@@ -65,6 +66,7 @@
     "is_valid_json",
     "load_interview_answers",
     "load_interview_json",
+    "nice_interview_subtitle",
     "rename_current_session",
     "rename_interview_answers",
     "save_interview_answers",
@@ -253,7 +255,7 @@ def is_file_like(obj: Any) -> bool:
 
 
 def set_interview_metadata(
-    filename: str, session_id: int, data: Dict, metadata_key_name="metadata"
+    filename: str, session_id: str, data: Dict, metadata_key_name="metadata"
 ) -> None:
     """Add searchable interview metadata for the specified filename and session ID.
     Intended to be used to add an interview title, etc.
@@ -265,7 +267,7 @@ def set_interview_metadata(
 
     Args:
         filename (str): The filename of the interview to add metadata for
-        session_id (int): The session ID of the interview to add metadata for
+        session_id (str): The session ID of the interview to add metadata for
         data (Dict): The metadata to add
         metadata_key_name (str, optional): The name of the metadata key. Defaults to "metadata".
     """
@@ -275,14 +277,14 @@ def set_interview_metadata(
 
 
 def get_interview_metadata(
-    filename: str, session_id: int, metadata_key_name: str = "metadata"
+    filename: str, session_id: str, metadata_key_name: str = "metadata"
 ) -> Dict[str, Any]:
     """Retrieve the unencrypted metadata associated with an interview.
     We implement this with the docassemble jsonstorage table and a dedicated `tag` which defaults to `metadata`.
 
     Args:
         filename (str): The filename of the interview to retrieve metadata for
-        session_id (int): The session ID of the interview to retrieve metadata for
+        session_id (str): The session ID of the interview to retrieve metadata for
         metadata_key_name (str, optional): The name of the metadata key. Defaults to "metadata".
 
     Returns:
@@ -342,21 +344,21 @@ def get_saved_interview_list(
     # while it appears to be performant enough for real-world usage.
     # Up to ~ 1,000 sessions performs well and is higher than expected for an end-user
     get_sessions_query = text(
-        """
-           SELECT  userdict.indexno
-           ,userdict.filename as filename
-           ,num_keys
-           ,userdictkeys.user_id as user_id
-           ,userdict.modtime as modtime
-           ,userdict.key as key
-           ,jsonstorage.data->'auto_title' as auto_title
-           ,jsonstorage.data->'title' as title
-           ,jsonstorage.data->'description' as description
-           ,jsonstorage.data->'steps' as steps
-           ,jsonstorage.data->'progress' as progress
-           ,jsonstorage.data->'original_interview_filename' as original_interview_filename
-           ,jsonstorage.data->'answer_count' as answer_count
-           ,jsonstorage.data as data
+    """
+    SELECT userdict.indexno
+        ,userdict.filename as filename
+        ,num_keys
+        ,userdictkeys.user_id as user_id
+        ,userdict.modtime as modtime
+        ,userdict.key as key
+        ,jsonstorage.data->'auto_title' as auto_title
+        ,jsonstorage.data->'title' as title
+        ,jsonstorage.data->'description' as description
+        ,jsonstorage.data->'steps' as steps
+        ,jsonstorage.data->'progress' as progress
+        ,jsonstorage.data->'original_interview_filename' as original_interview_filename
+        ,jsonstorage.data->'answer_count' as answer_count
+        ,jsonstorage.data as data
     FROM userdict 
     NATURAL JOIN 
     (
@@ -371,9 +373,7 @@ def get_saved_interview_list(
     LEFT JOIN jsonstorage
     ON userdict.key = jsonstorage.key AND (jsonstorage.tags = :metadata)
     WHERE (userdictkeys.user_id = :user_id or :user_id is null)
-    
-    AND
-    (userdict.filename = :filename OR :filename is null)
+    AND (userdict.filename = :filename OR :filename is null)
     AND (userdict.filename NOT IN :filenames_to_exclude)
     AND (NOT :exclude_newly_started_sessions OR num_keys > 1)
     ORDER BY modtime desc 
@@ -436,6 +436,143 @@ def get_saved_interview_list(
     return sessions
 
 
+def find_saved_interview(
+    keyword: str,
+    metadata_column_names: Optional[List[str]] = None,
+    filenames: Optional[List[str]] = None,
+    user_id: Union[int, str, None] = None,
+    metadata_key_name: str = "metadata",
+    limit: int = 50,
+    offset: int = 0,
+    filename_to_exclude: str = "",
+    exclude_current_filename: bool = True,
+    exclude_filenames: Optional[List[str]] = None,
+    exclude_newly_started_sessions: bool = False,
+    global_search_allowed_roles:Optional[Union[Set[str],List[str]]]=None,
+) -> List[Dict[str, Any]]:
+    """Get a list of sessions where the metadata for the session matches the provided keyword search terms.
+    This function is designed to be used in a search interface where the user can search for sessions by keyword.
+    The keyword search is case-insensitive and will match any part of the metadata column values.
+
+    Args:
+        keyword (str): The keyword to search for in the metadata
+        metadata_column_names (List[str], optional): The names of the metadata columns to search. If not provided, defaults to ["title", "auto_title", "description"].
+        filenames (List[str], optional): The filename or filenames of the interviews to retrieve sessions for.
+        user_id (Union[int, str, None], optional): The user ID to retrieve sessions for. Defaults to current user. Specify "all" if you want and have the necessary privileges to search all sessions.
+        metadata_key_name (str, optional): The name of the metadata key. Defaults to "metadata".
+        limit (int, optional): The maximum number of results to return. Defaults to 50.
+        offset (int, optional): The offset to start returning results from. Defaults to 0.
+        filename_to_exclude (str, optional): The filename to exclude from the results. Defaults to "".
+        exclude_current_filename (bool, optional): Whether to exclude the current filename from the results. Defaults to True.
+        exclude_filenames (Optional[List[str]], optional): A list of filenames to exclude from the results. Defaults to None.
+        exclude_newly_started_sessions (bool, optional): Whether to exclude sessions that are still on "step 1". Defaults to False.
+        global_search_allowed_roles (Union[Set[str],List[str]], optional): A list or set of roles that are allowed to search all sessions. Defaults to {'admin','developer', 'advocate'}. 'admin' and 'developer' are always allowed to search all sessions.
+
+    Returns:
+        List[Dict[str, Any]]: A list of saved sessions for the specified filename that match the search keyword
+    """
+    if not metadata_column_names:
+        metadata_column_names = {"title", "auto_title", "description"}
+    if not global_search_allowed_roles:
+        global_search_allowed_roles = {'admin','developer', 'advocate'}
+    global_search_allowed_roles = set(global_search_allowed_roles).union({'admin','developer'})
+
+    # Construct the dynamic part of the SQL query for metadata column selection and keyword search
+    metadata_search_conditions = " OR ".join(
+        f"COALESCE(jsonstorage.data->>{repr(column)}, '') ILIKE '%' || :keyword || '%'" for column in metadata_column_names
+    )
+
+    # we retrieve the default metadata columns even if we don't search them
+    metadata_column_names = set(metadata_column_names).union({"title", "auto_title", "description"})
+
+    if filenames:
+        # Create a parameterized string with placeholders for filenames
+        filenames_placeholder = ", ".join([":filename{}".format(i) for i in range(len(filenames))])
+        filename_condition = f"userdict.filename IN ({filenames_placeholder})"
+    else:
+        filename_condition = "TRUE"  # If no filenames are provided, this condition does not filter anything.
+
+    get_sessions_query = text(
+        f"""
+        SELECT  userdict.indexno,
+                userdict.filename as filename,
+                num_keys,
+                userdictkeys.user_id as user_id,
+                userdict.modtime as modtime,
+                userdict.key as key,
+                {', '.join(f"jsonstorage.data->>{repr(column)} as {column}" for column in metadata_column_names)},
+                jsonstorage.data as data
+        FROM userdict 
+        NATURAL JOIN (
+            SELECT key, MAX(modtime) AS modtime, COUNT(key) AS num_keys
+            FROM userdict
+            GROUP BY key
+        ) mostrecent
+        LEFT JOIN userdictkeys ON userdictkeys.key = userdict.key
+        LEFT JOIN jsonstorage ON userdict.key = jsonstorage.key AND (jsonstorage.tags = :metadata)
+        WHERE (userdictkeys.user_id = :user_id OR :user_id is NULL)
+          AND {filename_condition}
+          AND (userdict.filename NOT IN :filenames_to_exclude)
+          AND (NOT :exclude_newly_started_sessions OR num_keys > 1)
+          AND ({metadata_search_conditions})
+        ORDER BY modtime DESC
+        LIMIT :limit OFFSET :offset;
+        """
+    )
+
+    if offset < 0:
+        offset = 0
+
+    if exclude_current_filename:
+        current_filename = user_info().filename
+    else:
+        current_filename = ""
+    if not filename_to_exclude:
+        filename_to_exclude = ""
+    filenames_to_exclude = []
+    if exclude_filenames:
+        filenames_to_exclude.extend(exclude_filenames)
+    filenames_to_exclude.extend([current_filename, filename_to_exclude])
+    if user_id is None:
+        if user_logged_in():
+            user_id = user_info().id
+        else:
+            log("Asked to get interview list for user that is not logged in")
+            return []
+
+    if user_id == "all":
+        if user_has_privilege(global_search_allowed_roles):
+            user_id = None
+        elif user_logged_in():
+            user_id = user_info().id
+            log(
+                f"User {user_info().email} does not have permission to list interview sessions belonging to other users"
+            )
+        else:
+            log("Asked to get interview list for user that is not logged in")
+            return []
+
+    with db.connect() as con:
+        rs = con.execute(
+            get_sessions_query,
+            {
+                "metadata": metadata_key_name,
+                "keyword": keyword,
+                "user_id": user_id,
+                "limit": limit,
+                "offset": offset,
+                "filenames_to_exclude": tuple(filenames_to_exclude),
+                "exclude_newly_started_sessions": exclude_newly_started_sessions,
+            },
+        )
+
+    sessions = []
+    for session in rs:
+        sessions.append(dict(session._mapping))
+
+    return sessions
+
+
 def delete_interview_sessions(
     user_id: Optional[int] = None,
     filename_to_exclude: str = al_session_store_default_filename,
@@ -911,7 +1048,7 @@ def session_list_html(
 
 def rename_interview_answers(
     filename: str,
-    session_id: int,
+    session_id: str,
     new_name: str,
     metadata_key_name: str = "metadata",
 ) -> None:
@@ -920,7 +1057,7 @@ def rename_interview_answers(
 
     Args:
         filename (str): The filename of the interview to rename
-        session_id (int): The session ID of the interview to rename
+        session_id (str): The session ID of the interview to rename
         new_name (str): The new name to set for the interview
         metadata_key_name (str, optional): The name of the metadata key. Defaults to "metadata".
 
@@ -1151,6 +1288,7 @@ def get_filtered_session_variables_string(
     session_id: Optional[int] = None,
     variables_to_filter: Union[Set[str], List[str], None] = None,
     additional_variables_to_filter: Optional[Union[Set[str], List[str]]] = None,
+    indent:int=4,
 ) -> str:
     """
     Returns a JSON string that represents the filtered contents of a specified filename and session ID.
@@ -1161,6 +1299,7 @@ def get_filtered_session_variables_string(
         session_id (Optional[int], optional): Session ID to retrieve variables from. Defaults to None.
         variables_to_filter (Union[Set[str], List[str], None], optional): List or set of variables to exclude. Defaults to `al_sessions_variables_to_remove`.
         additional_variables_to_filter (Union[Set[str], List[str], None], optional): List or set of additional variables to exclude. Defaults to None.
+        indent (int, optional): Number of spaces to indent the JSON string. Defaults to 4.
 
     Returns:
         str: A JSON-formatted string of filtered session variables.
@@ -1173,12 +1312,12 @@ def get_filtered_session_variables_string(
             additional_variables_to_filter=additional_variables_to_filter,
         )
     )
-    return json.dumps(simple_vars)
+    return json.dumps(simple_vars, indent=indent)
 
 
 def load_interview_answers(
     old_interview_filename: str,
-    old_session_id: int,
+    old_session_id: str,
     new_session: bool = False,
     new_interview_filename: Optional[str] = None,
     variables_to_filter: Optional[List[str]] = None,
@@ -1191,7 +1330,7 @@ def load_interview_answers(
 
     Args:
         old_interview_filename (str): Filename of the old interview.
-        old_session_id (int): Session ID of the old interview.
+        old_session_id (str): Session ID of the old interview.
         new_session (bool, optional): Determines whether to create a new session. Defaults to False.
         new_interview_filename (Optional[str], optional): Filename for the new session. Defaults to None.
         variables_to_filter (Optional[List[str]], optional): List of variables to exclude. Defaults to None.