ositah 25.6.dev1__py3-none-any.whl → 25.9.dev1__py3-none-any.whl

ositah/utils/projects.py

@@ -1,1178 +1,1178 @@
-# Helper functions related to projects and time declarations
-
-import re
-from datetime import datetime
-from typing import List, Tuple
-
-import numpy as np
-import pandas as pd
-from sqlalchemy.orm import joinedload
-
-from ositah.utils.agents import get_agents
-from ositah.utils.cache import clear_cached_data
-from ositah.utils.exceptions import InvalidDataSource, InvalidHitoProjectName
-from ositah.utils.hito_db import get_db
-from ositah.utils.period import get_validation_period_data
-from ositah.utils.utils import (
-    DAY_HOURS,
-    TEAM_LIST_ALL_AGENTS,
-    TIME_UNIT_HOURS,
-    TIME_UNIT_HOURS_EN,
-    TIME_UNIT_HOURS_FR,
-    TIME_UNIT_WEEKS,
-    TIME_UNIT_WEEKS_EN,
-    TIME_UNIT_WEEKS_FR,
-    WEEK_HOURS,
-    GlobalParams,
-)
-
-CATEGORY_DEFAULT = "nsip_project"
-
-DATA_SOURCE_HITO = "hito"
-DATA_SOURCE_OSITAH = "ositah"
-
-NSIP_CLASS_OTHER_ACTIVITY = "activitensipreferentiel"
-NSIP_CLASS_PROJECT = "projetnsipreferentiel"
-
-MASTERPROJECT_DELETED_ACTIVITY = "Disabled"
-MASTERPROJECT_LOCAL_PROJECT = "Local Projects"
-
-NSIP_PROJECT_ORDER = 1
-LOCAL_PROJECT_ORDER = 2
-NSIP_ACIVITY_ORDER = 3
-DISABLED_ACTIVITY_ORDER = 9999
-
-
-def hito2ositah_project_name(hito_name):
-    """
-    Split a Hito project name into a masterprojet and project name
-
-    :param hito_name: Hito name with  masterprojet and project name separated by a /
-    :return: masterprojet and project name
-    """
-    masterproject, project_name = hito_name.split(" / ", 2)
-    return masterproject, project_name
-
-
-def ositah2hito_project_name(masterproject, project):
-    """
-    Build the Hito/NSIP project name from the masterproject and project
-
-    :param masterproject: masterproject name
-    :param project: project name
-    :return: Hito/NSIP project fullname
-    """
-    return " / ".join([masterproject, project])
-
-
-def nsip2ositah_project_name(masterproject, project):
-    """
-    Build the OSITAH project name from the NSIP project name, removing the master project
-    name if it is at the head of the NSIP project name, except if the master project name
-    and the project name are identical.
-
-    :param masterproject: masterproject name
-    :param project: project name
-    :return: OSITAH project name (without the masterproject name)
-    """
-
-    if project != masterproject:
-        m = re.match(rf"{masterproject}\s+\-\s+(?P<project>.*)", project)
-        if m:
-            project = m.group("project")
-
-    return project
-
-
-def category_from_activity(category_patterns, activity) -> str:
-    """
-    Return the activity category if the activity matches the pattern. Else an empty string.
-    Called as a lambda to build the category column.
-
-    :param category_patterns: category patterns to match against the activity (dict where
-                              the key is the pattern and the value is the category)
-    :param activity: activity name
-    :return: category or np.Nan
-    """
-
-    for pattern, category in category_patterns.items():
-        if re.match(pattern.lower(), activity.lower()):
-            return category
-
-    return np.NaN
-
-
-def activity_from_project(project):
-    """
-    Return the activity the project belongs to.
-
-    :param project: project name
-    :return: activity name
-    """
-
-    global_params = GlobalParams()
-
-    for activity, pattern in global_params.project_categories.items():
-        if re.match(pattern, project):
-            return activity
-
-    return CATEGORY_DEFAULT
-
-
-def reference_masterproject(reference_type):
-    """
-    Return an OSITAH masterproject for a NSIP reference, based on its type.
-    Masterprojects for each type is defined in the configuration. A reference type
-    without a match or with an empty value is ignored (np.NaN returned).
-
-    :param reference_type: NSIP reference type
-    :return: matching master project
-    """
-
-    global_params = GlobalParams()
-
-    for type_pattern, masterproject in global_params.reference_masterprojects.items():
-        if re.match(type_pattern.lower(), reference_type.lower()):
-            if len(masterproject) > 0:
-                return masterproject
-            else:
-                return np.NaN
-
-    return np.NaN
-
-
-def time_unit(category, short=False, english=True, parenthesis=False) -> str:
-    """
-    Return the time unit as defined in the configuration as a string. If the category/column is
-    not in the configuration, return an empty string.
-
-    :param category: project category/class
-    :param short: if true, return abbreviated unit names
-    :param english: return english unit names if true. Also implies short=False
-    :param parenthesis: if True, enclose the string in ()
-    :return: time unit for the category as a string
-    """
-
-    global_params = GlobalParams()
-
-    if english:
-        unit_w = TIME_UNIT_WEEKS_EN
-        unit_h = TIME_UNIT_HOURS_EN
-    else:
-        if short:
-            unit_w = "sem."
-            unit_h = "h"
-        else:
-            unit_w = TIME_UNIT_WEEKS_FR
-            unit_h = TIME_UNIT_HOURS_FR
-
-    if category in global_params.time_unit:
-        if global_params.time_unit[category] == TIME_UNIT_WEEKS:
-            unit_str = unit_w
-        elif global_params.time_unit[category] == TIME_UNIT_HOURS:
-            unit_str = unit_h
-        else:
-            raise Exception(
-                (
-                    f"Unsupported time unit '{global_params.time_unit[category]}'"
-                    f" for category {category}"
-                )
-            )
-    else:
-        return ""
-
-    if parenthesis:
-        return f"({unit_str})"
-    else:
-        return unit_str
-
-
-def category_time_and_unit(category, hours, short=True, english=False) -> Tuple[int, str]:
-    """
-    Return the rounded category time in the appropriate unit and the category time unit
-
-    :param category: project category/class
-    :param hours: number of hours
-    :param short: if true, return abbreviated unit names
-    :param english: return english unit names if true. Also implies short=False
-    :return: project time, project unit
-    """
-
-    global_params = GlobalParams()
-
-    unit = time_unit(category, short, english)
-
-    if global_params.time_unit[category] == "w":
-        declared_time = f"{int(round(hours / WEEK_HOURS))}"
-    else:
-        declared_time = f"{int(round(hours))}"
-
-    return declared_time, unit
-
-
-def project_time(project, hours):
-    """
-    Return the rounded project time in the appropriate unit and the project time unit
-
-    :param project: project name
-    :param hours: number of hours
-    :return: project time, abbreviated project unit
-    """
-
-    return category_time_and_unit(activity_from_project(project), hours)
-
-
-def get_team_projects(
-    team,
-    team_selection_date,
-    period_date: datetime,
-    source=DATA_SOURCE_HITO,
-    use_cache: bool = True,
-):
-    """
-    Query the Hito database and return a dataframe will all the project contributions for a given
-    team. The dataframe has one row for each each agent contribution to each project.
-
-    :param team: selected team or TEAM_LIST_ALL_AGENTS for all teams
-    :param team_selection_date: last time the team selection was changed
-    :param period_date: a date that must be inside the declaration period
-    :param source: whether to use Hito (non validated) or OSITAH (validated) as a data source
-    :param use_cache: if true, use the cache if defined and up-to-date or update it with the
-                      new declarations
-    :return: dataframe or None if the query returned no entry
-    """
-
-    from ositah.utils.hito_db_model import (
-        ActiviteDetail,
-        Agent,
-        OSITAHProjectDeclaration,
-        OSITAHValidation,
-        Projet,
-        Team,
-    )
-
-    global_params = GlobalParams()
-    columns = global_params.columns
-    session_data = global_params.session_data
-    db = get_db()
-
-    validation_period = get_validation_period_data(period_date)
-
-    # Check if there is a cached version
-    if session_data.project_declarations is not None and use_cache:
-        if (
-            session_data.project_declarations_source is None
-            or source != session_data.project_declarations_source
-            or datetime.fromisoformat(team_selection_date) > session_data.cache_date
-        ):
-            # Cache must be refreshed if the selected source doesn't match the cached one or if
-            # the team has been modified since the cache was loaded (required for multi-worker
-            # configurations as the team selection does not necessarily happen on the same worker
-            # than the later processing). In a multi-worker configuration is used it may also
-            # happen that the declaration source is not defined if it was initially initialised
-            # on another worker
-            clear_cached_data()
-        else:
-            return session_data.project_declarations
-
-    if source == DATA_SOURCE_OSITAH:
-        # The query relies on the fact that only one validation entry can be in the validated
-        # state for a given period, something enforced by the declaration validation.
-        # When a team is specified, display all projects from this team and the children teams
-        query = (
-            OSITAHProjectDeclaration.query.join(
-                OSITAHValidation,
-                OSITAHProjectDeclaration.validation_id == OSITAHValidation.id,
-            )
-            .join(Agent, Agent.id == OSITAHValidation.agent_id)
-            .join(Team, Team.id == Agent.team_id)
-            .add_entity(Agent)
-            .add_entity(Team)
-            .add_entity(OSITAHValidation)
-            .filter(OSITAHValidation.validated)
-            .filter(OSITAHValidation.period_id == validation_period.id)
-        )
-        if team != TEAM_LIST_ALL_AGENTS:
-            query = query.filter(Team.nom.ilike(f"{team}%"))
-        declarations = pd.read_sql(query.statement, db.session.bind)
-        if len(declarations) == 0:
-            return None
-        declarations.rename(columns={"id": columns["activity_id"]}, inplace=True)
-        declarations.rename(columns={"id_1": columns["agent_id"]}, inplace=True)
-        declarations.rename(columns={"nom_1": columns["team"]}, inplace=True)
-        declarations.rename(columns={"hours": columns["hours"]}, inplace=True)
-        declarations.rename(columns={"id_3": "validation_id"}, inplace=True)
-        declarations.rename(columns={"timestamp": "validation_time"}, inplace=True)
-        # Drop statut column to avoid conflicts in future merge with the Agent table
-        declarations.drop(columns=["statut"], inplace=True)
-        # Ensure that email_auth is defined and if it is not, replace it by the email.
-        declarations.loc[declarations[columns["email_auth"]].isna(), columns["email_auth"]] = (
-            declarations[columns["email"]]
-        )
-        declarations[columns["activity"]] = declarations.apply(
-            lambda row: ositah2hito_project_name(
-                row[columns["masterproject"]], row[columns["project"]]
-            ),
-            axis=1,
-        )
-
-    elif source == DATA_SOURCE_HITO:
-        # For team names, we want to keep the agent team name instead of the team_name in
-        # activity_details so it must be specified explicitely in the join with Team table
-        query = (
-            ActiviteDetail.query.join(Projet)
-            .join(Agent)
-            .join(Team, Team.id == Agent.team_id)
-            .add_entity(Projet)
-            .add_entity(Agent)
-            .add_entity(Team)
-            .filter(
-                ActiviteDetail.date >= validation_period.start_date,
-                ActiviteDetail.date <= validation_period.end_date,
-            )
-        )
-        if team != TEAM_LIST_ALL_AGENTS:
-            query = query.filter(Team.nom.ilike(f"{team}%"))
-        daily_declarations = pd.read_sql(query.statement, db.session.bind)
-        if len(daily_declarations) == 0:
-            return None
-        # Pandas add a suffix to duplicate column names, the first one being unchanged, the
-        # second being suffixed _1...
-        daily_declarations.drop(columns=["id", "id_1", "id_2"], inplace=True)
-        daily_declarations.rename(columns={"agent_id": columns["agent_id"]}, inplace=True)
-        daily_declarations.rename(columns={"libelle": columns["activity"]}, inplace=True)
-        daily_declarations.rename(columns={"nom_1": columns["team"]}, inplace=True)
-        for column in [columns["hours"], columns["percent"]]:
-            daily_declarations[column] = daily_declarations[column].astype(float)
-        # Ensure that email_auth is defined and if it is not, replace it by the email. If left
-        # undefined, the entries will not be present in the pivot table as it is part of the index.
-        daily_declarations.loc[
-            daily_declarations[columns["email_auth"]].isna(), columns["email_auth"]
-        ] = daily_declarations[columns["email"]]
-        # Rebuild agent quotite by comparing the time declared with the percent computed by Hito
-        # based on the quotite
-        daily_declarations[columns["quotite"]] = (
-            daily_declarations[columns["hours"]] / DAY_HOURS * 100
-        ) / daily_declarations[columns["percent"]]
-        global_declarations_pt = pd.pivot_table(
-            daily_declarations,
-            index=[
-                columns["lastname"],
-                columns["firstname"],
-                columns["activity"],
-                columns["activity_id"],
-                columns["team"],
-                columns["agent_id"],
-                columns["email_auth"],
-                columns["email"],
-            ],
-            values=[columns["hours"], columns["quotite"]],
-            aggfunc={columns["hours"]: "sum", columns["quotite"]: "mean"},
-        )
-        declarations = pd.DataFrame(global_declarations_pt.to_records())
-        declarations[[columns["masterproject"], columns["project"]]] = declarations[
-            columns["activity"]
-        ].str.split(" / ", n=1, expand=True)
-        # An entry in the pseudo master project MASTERPROJECT_DELETED_ACTIVITY is a special
-        # case corresponding to deleted NSIP projects: the real name is in the project part that
-        # must be parsed as for any other project
-        declarations["project_saved"] = np.NaN
-        declarations["project_saved"] = declarations["project_saved"].astype("object")
-        declarations.loc[
-            declarations[columns["masterproject"]] == MASTERPROJECT_DELETED_ACTIVITY,
-            "project_saved",
-        ] = declarations[columns["project"]]
-        # Not sure why the following line doesn't work (masterproject and project set to NaN
-        # if no row matches the indexing condition... An issue has been open:
-        # https://github.com/pandas-dev/pandas/issues/44726.
-        # declarations.loc[
-        #     declarations.project_saved.notna(),
-        #     [columns["masterproject"], columns["project"]],
-        # ] = declarations.project_saved.str.split(" / ", n=1, expand=True)
-        #
-        # The following workaround fails if project_saved contains only np.NaN. It is a known
-        # issue in Panda 1.3.4, see https://github.com/pandas-dev/pandas/issues/35807
-        # declarations[
-        #     ["newmaster", "newproject"]
-        # ] = declarations.project_saved.str.split(" / ", n=1, expand=True)
-        #
-        # Workaround based on
-        # https://github.com/pandas-dev/pandas/issues/35807#issuecomment-676912441. If no row
-        # matches the condition, only one column is created thus the need to check they all
-        # exist.
-        tmp_columns = ["newmaster", "newproject"]
-        saved_projects = (
-            declarations["project_saved"]
-            .str.split("/", expand=True, n=len(tmp_columns) - 1)
-            .rename(columns={k: name for k, name in enumerate(tmp_columns)})
-        )
-        for column in tmp_columns:
-            if column not in saved_projects.columns:
-                saved_projects[column] = np.NaN
-        declarations = declarations.join(saved_projects)
-        declarations.loc[declarations.project_saved.notna(), columns["masterproject"]] = (
-            declarations.newmaster
-        )
-        declarations.loc[declarations.project_saved.notna(), columns["project"]] = (
-            declarations.newproject
-        )
-        declarations.drop(columns=["newmaster", "newproject"], inplace=True)
-        declarations.loc[declarations.project_saved.notna(), columns["activity"]] = (
-            declarations.project_saved
-        )
-
-        # Detect project names not matching the format "masterproject / project"
-        invalid_hito_projects = declarations.loc[declarations[columns["project"]].isnull()]
-        if not invalid_hito_projects.empty:
-            raise InvalidHitoProjectName(
-                pd.Series(invalid_hito_projects[columns["masterproject"]]).unique()
-            )
-
-    else:
-        raise InvalidDataSource(source)
-
-    declarations[columns["fullname"]] = declarations[
-        [columns["lastname"], columns["firstname"]]
-    ].agg(" ".join, axis=1)
-    declarations[columns["category"]] = declarations.apply(
-        lambda row: category_from_activity(
-            global_params.category_patterns, row[columns["activity"]]
-        ),
-        axis=1,
-    )
-    declarations.loc[declarations[columns["category"]].isna(), "category"] = CATEGORY_DEFAULT
-
-    # Check quotite < 50% and flag the entry as suspect (generally means confusion between quotite
-    # and percent during declaration)
-    declarations["suspect"] = declarations[columns["quotite"]] < 0.5
-
-    if use_cache:
-        session_data.set_project_declarations(declarations, source)
-
-    return declarations
-
-
-def get_all_hito_activities(project_activity: bool):
-    """
-    Retrieve all projects or all activities defined in Hito with their associated teams
-
-    :param project_activity: if true, return all projects, else all Hito activities
-    :return: dataframe
-    """
-
-    from ositah.utils.hito_db_model import Activite, Projet
-
-    global_params = GlobalParams()
-    session_data = global_params.session_data
-    db = get_db()
-
-    # Check if there is a cached version
-    if session_data.get_hito_activities(project_activity) is not None:
-        return session_data.get_hito_activities(project_activity)
-
-    else:
-        if project_activity:
-            Activity = Projet
-        else:
-            Activity = Activite
-
-        query = Activity.query.options(joinedload(Activity.teams))
-        activities = pd.read_sql(query.statement, db.session.bind)
-        activities[["masterproject", "project"]] = activities.libelle.str.split(
-            " / ", n=1, expand=True
-        )
-        activities.rename(columns={"description_1": "team_description"}, inplace=True)
-        activities.rename(columns={"nom": "team_name"}, inplace=True)
-
-        session_data.set_hito_activities(activities, project_activity)
-
-        return activities
-
-
-def build_projects_data(team, team_selection_date, period_date: str, source):
-    """
-    Build the project list contributed by the selected team and return it as a dataframe
-
-    :param team: selected team
-    :param team_selection_date: last time the team selection was changed
-    :param period_date: a date that must be inside the declaration period
-    :param source: whether to use Hito (non validated) or OSITAH (validated) as a data source
-    :return: dataframe with projects data, dataframe with agent declarations
-    """
-
-    global_params = GlobalParams()
-    columns = global_params.columns
-    session_data = global_params.session_data
-
-    declaration_list = get_team_projects(team, team_selection_date, period_date, source)
-    if declaration_list is None:
-        return None, None
-
-    projects_data = session_data.projects_data
-    if projects_data is None:
-        projects_data_pt = pd.pivot_table(
-            declaration_list,
-            index=[
-                columns["masterproject"],
-                columns["project"],
-                columns["activity"],
-                columns["category"],
-            ],
-            values=[columns["hours"]],
-            aggfunc={columns["hours"]: "sum"},
-        )
-        projects_data = pd.DataFrame(projects_data_pt.to_records())
-        projects_data[columns["hours"]] = np.round(projects_data[columns["hours"]]).astype("int")
-        projects_data[columns["weeks"]] = np.round(projects_data[columns["hours"]] / WEEK_HOURS, 1)
-        short_name_len = 25
-        projects_data["project_short"] = projects_data[columns["project"]]
-        projects_data.loc[
-            projects_data["project_short"].str.len() > short_name_len, "project_short"
-        ] = projects_data["project_short"].str.slice_replace(start=short_name_len - 4, repl="...")
-        session_data.projects_data = projects_data
-
-    return projects_data, declaration_list
-
-
-def get_hito_nsip_activities(project_activity: bool = True):
-    """
-    Return a dataframe with all the NSIP activities defined in Hito. Activities can be
-    either projects or "references" (other activities). An activity is considered as a
-    NSIP activity if it has a matching entry in Hito referentiel.
-
-    :param project_activity: if true, return projects, else references
-    :return: dataframe
-    """
-
-    from ositah.utils.hito_db_model import Projet, Referentiel
-
-    db = get_db()
-
-    if project_activity:
-        project_join_id = Projet.projet_nsip_referentiel_id
-        referentiel_class = "projetnsipreferentiel"
-    else:
-        project_join_id = Projet.activite_nsip_referentiel_id
-        referentiel_class = "activitensipreferentiel"
-
-    activity_query = (
-        Projet.query.join(Referentiel, Referentiel.id == project_join_id)
-        .add_entity(Referentiel)
-        .filter(
-            Referentiel.object_class == referentiel_class,
-        )
-    )
-
-    activities = pd.read_sql(activity_query.statement, db.session.bind)
-
-    activities.drop(
-        columns=[
-            "ordre",
-            "projet_nsip_referentiel_id",
-            "activite_nsip_referentiel_id",
-        ],
-        inplace=True,
-    )
-    activities.rename(columns={"id_1": "referentiel_id"}, inplace=True)
-    activities.rename(columns={"libelle_1": "nsip_name_id"}, inplace=True)
-
-    if activities.empty:
-        activities[["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]] = [
-            np.NaN,
-            np.NaN,
-            np.NaN,
-            np.NaN,
-        ]
-    else:
-        activities[["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]] = (
-            activities.apply(
-                lambda v: nsip_activity_name_id(v["nsip_name_id"], v["class"]),
-                axis=1,
-                result_type="expand",
-            )
-        )
-        activities["nsip_project_id"] = activities["nsip_project_id"].astype(int)
-        activities["nsip_reference_id"] = activities["nsip_reference_id"].astype(int)
-
-    return activities
-
-
-def get_hito_projects():
-    """
-    Return a dataframe with the information about all projects with validated declarations
-    in the current declaration period defined in Hito and their relationship to NSIP, if relevant.
-
-    :return: Hito project dataframe
-    """
-
-    from ositah.utils.hito_db_model import (
-        OSITAHProjectDeclaration,
-        OSITAHValidation,
-        Projet,
-        Referentiel,
-    )
-
-    db = get_db()
-
-    projects_query = (
-        Projet.query.join(Referentiel, Referentiel.id == Projet.projet_nsip_referentiel_id)
-        .join(OSITAHProjectDeclaration)
-        .join(OSITAHValidation)
-        .add_entity(Referentiel)
-        .add_entity(OSITAHValidation)
-        .filter(
-            Referentiel.object_class == "projetnsipreferentiel",
-            OSITAHValidation.validated,
-        )
-    )
-    projects = pd.read_sql(projects_query.statement, db.session.bind)
-
-    activities_query = (
-        Projet.query.join(Referentiel, Referentiel.id == Projet.activite_nsip_referentiel_id)
-        .join(OSITAHProjectDeclaration)
-        .join(OSITAHValidation)
-        .add_entity(Referentiel)
-        .add_entity(OSITAHValidation)
-        .filter(
-            Referentiel.object_class == "activitensipreferentiel",
-            OSITAHValidation.validated,
-        )
-    )
-    activities = pd.read_sql(activities_query.statement, db.session.bind)
-
-    projects_activities = pd.concat([projects, activities], ignore_index=True)
-
-    projects_activities.drop(
-        columns=[
-            "id_2",
-            "ordre",
-            "projet_nsip_referentiel_id",
-            "activite_nsip_referentiel_id",
-            "agent_id",
-        ],
-        inplace=True,
-    )
-    projects_activities.rename(columns={"id_1": "referentiel_id"}, inplace=True)
-    projects_activities.rename(columns={"libelle_1": "nsip_name_id"}, inplace=True)
-    projects_activities.drop_duplicates(subset=["id"], inplace=True)
-
-    if projects_activities.empty:
-        projects_activities[
-            ["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]
-        ] = [np.NaN, np.NaN, np.NaN, np.NaN]
-    else:
-        projects_activities[
-            ["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]
-        ] = projects_activities.apply(
-            lambda v: nsip_activity_name_id(v["nsip_name_id"], v["class"]),
-            axis=1,
-            result_type="expand",
-        )
-        projects_activities["nsip_project_id"] = projects_activities["nsip_project_id"].astype(int)
-        projects_activities["nsip_reference_id"] = projects_activities["nsip_reference_id"].astype(
-            int
-        )
-
-    return projects_activities
-
-
-def nsip_activity_name_id(hito_name: str, type: str) -> List[str]:
-    """
-    Split the NISP activity project name in Hito referentiel in 3 parts: masterproject name,
-    project name, project ID and return the project ID as the project ID (3d value) or reference
-    ID (4th value) depending on the activity type. The unused ID is set to 0 rather than np.NaN
-    or pd.NA as the column may be used in merges.
-
-    :param hito_name: activity name in Hito referentiel
-    :param type: referentiel class
-    :return:
-    """
-
-    m = re.match(
-        r"(?P<master>.*?)\s+/\s+(?P<project>.*)\s+\(NSIP ID:\s*(?P<id>\w+)\)$",
-        hito_name,
-    )
-    if m:
-        try:
-            _ = int(m.group("id"))
-        except ValueError:
-            print(
-                (
-                    f"ERROR: invalid NSIP ID in Hito referentiel for '{m.group('master')} /"
-                    f" {m.group('project')}' (ID={m.group('id')})"
-                )
-            )
-            return m.group("master"), m.group("project"), 0, 0
-        if type == NSIP_CLASS_PROJECT:
-            project_id = m.group("id")
-            reference_id = 0
-        else:
-            project_id = 0
-            reference_id = m.group("id")
-        return m.group("master"), m.group("project"), project_id, reference_id
-    else:
-        print(
-            (
-                f"ERROR: invalid Hito referentiel entry format, cannot be parsed as"
-                f" master/project/id ({hito_name})"
-            )
-        )
-        return np.NaN, np.NaN, 0, 0
-
-
-def get_nsip_declarations(period_date: str, team: str):
-    """
-    Return the NSIP declaration list for the declaration period matching a given date (the
-    date must be included in the period) as a dataframe
-
-    :param period_date: date that must be inside the period
-    :param team: selected team
-    :return: declaration list as a dataframe
-    """
-
-    global_params = GlobalParams()
-
-    if global_params.nsip:
-        declarations = pd.json_normalize(global_params.nsip.get_declarations(period_date))
-        if declarations.empty:
-            return declarations
-
-        declarations.rename(columns={"id": "id_declaration"}, inplace=True)
-        # Set NaN to 0 in reference as np.NaN is a float and prevent casting to int. As it will
-        # be used in a merge better to have a 0 than a NaN.
-        if "project.id" in declarations.columns:
-            declarations.loc[declarations["project.id"].isna(), "project.id"] = 0
-            declarations["project.id"] = declarations["project.id"].astype(int)
-        else:
-            declarations["project.id"] = 0
-        if "reference.id" in declarations.columns:
-            declarations.loc[declarations["reference.id"].isna(), "reference.id"] = 0
-            declarations["reference.id"] = declarations["reference.id"].astype(int)
-        else:
-            declarations["reference.id"] = 0
-        declarations["nsip_fullname"] = (
-            declarations["agent.lastname"] + " " + declarations["agent.firstname"]
-        )
-
-        if team != TEAM_LIST_ALL_AGENTS:
-            team_agents = get_agents(period_date, team)
-            agent_emails = team_agents["email_auth"]
-            declarations = declarations.merge(
-                agent_emails,
-                how="inner",
-                left_on="agent.email",
-                right_on="email_auth",
-                suffixes=[None, "_agent"],
-            )
-
-        return declarations
-
-    else:
-        return None
-
-
-def get_nsip_activities(project_activity: bool):
-    """
-    Retrieve laboratory activities defined in NSIP and return them in a dataframe.
-    Activities can be either projects or references (other activities).
-
-    :param project_activity: true for projects, false for other activities
-    :return: dataframe or None if NSIP is not configured
-    """
-
-    global_params = GlobalParams()
-
-    if global_params.nsip:
-        activities = pd.json_normalize(
-            global_params.nsip.get_activities(project_activity), record_prefix=True
-        )
-        if not activities.empty:
-            if project_activity:
-                activities["ositah_name"] = activities.apply(
-                    lambda p: nsip2ositah_project_name(p["master_project.name"], p["name"]),
-                    axis=1,
-                )
-            else:
-                activities["master_project.name"] = activities.apply(
-                    lambda p: reference_masterproject(p["type"]),
-                    axis=1,
-                )
-                activities.drop(
-                    activities[activities["master_project.name"].isna()].index,
-                    inplace=True,
-                )
-                activities["ositah_name"] = activities["name"]
-
-        return activities
-    else:
-        return None
-
-
-def build_activity_libelle(
-    nsip_id: str,
-    master_project: str,
-    project: str,
-):
-    """
-    Build Hito project name and referentiel entry name from NSIP master project, project name and
-    project id.
-
-    :param nsip_id: NSIP ID for the project
-    :param master_project: master project name
-    :param project: project name
-    :return: Hito project name, Hito referentiel name
-    """
-
-    new_project_name = f"{master_project} / {project}"
-    new_referentiel_name = f"{new_project_name} (NSIP ID: {nsip_id})"
-    return new_project_name, new_referentiel_name
-
-
-def update_activity_name(
-    hito_project_id: str,
-    hito_referentiel_id: str,
-    nsip_id: str,
-    master_project: str,
-    project: str,
-):
-    """
-    Update a project name in Hito, both in the referentiel and in the project/activity table.
-
-    :param hito_project_id: Hito project ID
-    :param hito_referentiel_id: Hito referentiel ID for the project
-    :param nsip_id: NSIP ID for the project
-    :param master_project: master project name
-    :param project: project name
-    :return: 0 if update succeeded, non-zero if an error occured, error_msg if
-             an error occured
-    """
-
-    from ositah.utils.hito_db_model import Projet, Referentiel
-
-    db = get_db()
-
-    status = 0  # Assume success
-    error_msg = ""
-    new_project_name, new_referentiel_name = build_activity_libelle(
-        nsip_id,
-        master_project,
-        project,
-    )
-
-    try:
-        referentiel_entry = Referentiel.query.filter(Referentiel.id == hito_referentiel_id).first()
-        project_entry = Projet.query.filter(Projet.id == hito_project_id).first()
-        referentiel_entry.libelle = new_referentiel_name
-        project_entry.libelle = new_project_name
-        change_log_msg = f"Modifié le {datetime.now()}"
-        if project_entry.description:
-            project_entry.description += f"; {change_log_msg}"
-        else:
-            project_entry.description = change_log_msg
-        db.session.commit()
-    except Exception as e:
-        status = 1
-        error_msg = getattr(e, "message", repr(e))
-        db.session.rollback()
-
-    return status, error_msg
-
-
-def add_activity(
-    nsip_id: str,
-    master_project: str,
-    project: str,
-    activity_teams: List[str],
-    project_activity: bool,
-):
-    """
-    Adds a new project in Hito referenciel and in Hito project/activity table
-
-    :param nsip_id: NSIP ID for the project
-    :param master_project: master project name
-    :param project: project name
-    :param activity_teams: list of team IDs associated with the project
-    :param project_activity: if True it is a NSIP project, else a NSIP activity
-    :return: 0 if update succeeded, non-zero if an error occured, error_msg if
-             an error occured
-    """
-
-    from ositah.utils.hito_db_model import Projet, Referentiel, Team
-
-    db = get_db()
-
-    status = 0  # Assume success
-    error_msg = ""
-    project_name, referentiel_name = build_activity_libelle(
-        nsip_id,
-        master_project,
-        project,
-    )
-
-    if project_activity:
-        entry_class = "projetnsipreferentiel"
-        entry_order = NSIP_PROJECT_ORDER
-    else:
-        entry_class = "activitensipreferentiel"
-        entry_order = NSIP_ACIVITY_ORDER
-
-    try:
-        referentiel_entry = Referentiel(
-            libelle=referentiel_name,
-            object_class=entry_class,
-            ordre=entry_order,
-        )
-        activity_entry = Projet(
-            libelle=project_name,
-            description=f"Créé le {datetime.now()}",
-            ordre=entry_order,
-        )
-        if activity_teams:
-            activity_entry.teams = Team.query.filter(Team.id.in_(activity_teams)).all()
-        db.session.add(referentiel_entry)
-        db.session.add(activity_entry)
-        db.session.commit()
-        # Define relationship between activity and referentiel entry after creating them so that
-        # the referentiel ID generated by the DB server can be accessed
-        if project_activity:
-            activity_entry.projet_nsip_referentiel_id = referentiel_entry.id
-        else:
-            activity_entry.activite_nsip_referentiel_id = referentiel_entry.id
-        db.session.commit()
-
-    except Exception as e:
-        status = 1
-        error_msg = getattr(e, "message", repr(e))
-        db.session.rollback()
-
-    return status, error_msg
-
-
-def remove_activity(
-    hito_project_id: str,
-    hito_referentiel_id: str,
-    nsip_id: str,
-    project_activity: bool,
-):
-    """
-    Remove the association between a Hito activity (project or reference) and NSIP. The Hito
-    activity is kept as it may be referenced by other objects but its description is updated
-    to mention that it is no longer in NSIP. The project name is updated so that it appears in the
-    pseudo-masterproject NSIP_DELETED_MASTERPROJECT. Associated teams are removed.
-
-    :param hito_project_id: Hito project ID
-    :param hito_referentiel_id: Hito referentiel ID for the project
-    :param nsip_id: NSIP ID for the project
-    :param project_activity: if True it is a NSIP project, else a NSIP activity
-    :return: 0 if update succeeded, non-zero if an error occured, error_msg if
-             an error occured
-    """
-
-    from ositah.utils.hito_db_model import Projet, Referentiel
-
-    db = get_db()
-
-    status = 0  # Assume success
-    error_msg = ""
-
-    try:
-        referentiel_entry = Referentiel.query.filter(Referentiel.id == hito_referentiel_id).first()
-        db.session.query()
-        activity_entry = Projet.query.filter(Projet.id == hito_project_id).first()
-        if project_activity:
-            activity_entry.projet_nsip_referentiel_id = None
-        else:
-            activity_entry.activite_nsip_referentiel_id = None
-        change_log_msg = f"Desactivé le {datetime.now()} (NSIP ID={nsip_id})"
-        if activity_entry.description:
-            activity_entry.description += f"; {change_log_msg}"
-        else:
-            activity_entry.description = change_log_msg
-        activity_entry.libelle = f"{MASTERPROJECT_DELETED_ACTIVITY} / {activity_entry.libelle}"
-        activity_entry.ordre = DISABLED_ACTIVITY_ORDER
-        if len(activity_entry.teams) > 0:
-            activity_entry.teams.clear()
-        db.session.delete(referentiel_entry)
-        db.session.commit()
-    except Exception as e:
-        status = 1
-        error_msg = getattr(e, "message", repr(e))
-        db.session.rollback()
-
-    return status, error_msg
-
-
-def add_activity_teams(
-    masterproject: str, project: str, team_list: List[str], project_activity: bool
-):
-    """
-    Add teams to an activity.
-
-    :param masterproject: activity masterproject name
-    :param project: activity project name
-    :param team_list: list of team names to add
-    :param project_activity: if true, an Hito project else an Hito activity
-    :return: status (0 if success), error_msg (empty if success)
-    """
-
-    from ositah.utils.hito_db_model import Activite, Projet, Team
-
-    db = get_db()
-
-    status = 0  # Assume success
-    error_msg = ""
-
-    if project_activity:
-        Activity = Projet
-    else:
-        Activity = Activite
-
-    activity_name = ositah2hito_project_name(masterproject, project)
-    activity = Activity.query.filter(Activity.libelle == activity_name).first()
-
-    if activity:
-        try:
-            for team in team_list:
-                team_object = Team.query.filter(Team.nom == team).first()
-                activity.teams.append(team_object)
-            db.session.commit()
-        except Exception as e:
-            status = 1
-            error_msg = getattr(e, "message", repr(e))
-            db.session.rollback()
-
-    return status, error_msg
-
-
-def remove_activity_teams(
-    masterproject: str, project: str, team_list: List[str], project_activity: bool
-):
-    """
-    Remove teams from an activity. If the team is not present in teams list, silently
-    ignore it.
-
-    :param masterproject: activity masterproject name
-    :param project: activity project name
-    :param team_list: list of team names to remove
-    :param project_activity: if true, an Hito project else an Hito activity
-    :return: status (0 if success), error_msg (empty if success)
-    """
-
-    from ositah.utils.hito_db_model import Activite, Projet, Team
-
-    db = get_db()
-
-    status = 0  # Assume success
-    error_msg = ""
-
-    if project_activity:
-        Activity = Projet
-    else:
-        Activity = Activite
-
-    activity_name = ositah2hito_project_name(masterproject, project)
-    activity = Activity.query.filter(Activity.libelle == activity_name).first()
-
-    if activity:
-        try:
-            for team in team_list:
-                team_object = Team.query.filter(Team.nom == team).first()
-                if team_object in activity.teams:
-                    activity.teams.remove(team_object)
-            db.session.commit()
-        except Exception as e:
-            status = 1
-            error_msg = getattr(e, "message", repr(e))
-            db.session.rollback()
-
-    return status, error_msg
-
-
-def reenable_activity(activity_name: str, project_activity: bool, name_prefix: str = None):
-    """
-    Reenable a disabled activity. This involves:
-    - Updating master project to match the original one
-    - If it was an NSIP project, recreate the referentiel entry
-
-    :param activity_name: activity name
-    :param project_activity: if true, an Hito project else an Hito activity
-    :param name_prefix: activity name prefix for deleted or local activities
-    :return: status and error message if any
-    """
-
-    from ositah.utils.hito_db_model import Activite, Projet, Referentiel
-
-    db = get_db()
-
-    status = 0  # Assume success
-    error_msg = ""
-
-    if project_activity:
-        Activity = Projet
-    else:
-        Activity = Activite
-
-    # Retrieve activity attributes and NSIP ID if present in description
-    if name_prefix:
-        activity_full_name = f"{name_prefix} / {activity_name}"
-    else:
-        activity_full_name = activity_name
-    activity_entry = Activity.query.filter(Activity.libelle == activity_full_name).first()
-
-    m = re.search(r"\(NSIP ID\=(?P<id>\d+)\)$", activity_entry.description)
-    if m:
-        nsip_id = m.group("id")
-    else:
-        nsip_id = None
-
-    # Check if an entry exist in the referentiel for the NSIP ID: if not, create it
-    nsip_entry = Referentiel.query.filter(
-        Referentiel.libelle.ilike(f"%(NSIP ID = {nsip_id})")
-    ).first()
-    if not nsip_entry:
-        master_project, activity = hito2ositah_project_name(activity_name)
-        project_name, referentiel_name = build_activity_libelle(
-            nsip_id,
-            master_project,
-            activity,
-        )
-
-        if project_activity:
-            entry_class = "projetnsipreferentiel"
-            entry_order = NSIP_PROJECT_ORDER
-        else:
-            entry_class = "activitensipreferentiel"
-            entry_order = NSIP_ACIVITY_ORDER
-
-        referentiel_entry = Referentiel(
-            libelle=referentiel_name,
-            object_class=entry_class,
-            ordre=entry_order,
-        )
-    else:
-        referentiel_entry = None
-
-    # Create referentiel entry if necessary and update activity
-    try:
-        if referentiel_entry:
-            db.session.add(referentiel_entry)
-        activity_entry.libelle = activity_name
-        activity_entry.description = f"Modifié le {datetime.now()}"
-        db.session.commit()
-        # Define relationship between activity and referentiel entry after creating them so that
-        # the referentiel ID generated by the DB server can be accessed
-        if project_activity:
-            activity_entry.projet_nsip_referentiel_id = referentiel_entry.id
-        else:
-            activity_entry.activite_nsip_referentiel_id = referentiel_entry.id
-        db.session.commit()
-
-    except Exception as e:
-        status = 1
-        error_msg = getattr(e, "message", repr(e))
-        db.session.rollback()
-
-    # Clear cached data to force a refresh of project list
-    clear_cached_data()
-
-    return status, error_msg
+# Helper functions related to projects and time declarations
+
+import re
+from datetime import datetime
+from typing import List, Tuple
+
+import numpy as np
+import pandas as pd
+from sqlalchemy.orm import joinedload
+
+from ositah.utils.agents import get_agents
+from ositah.utils.cache import clear_cached_data
+from ositah.utils.exceptions import InvalidDataSource, InvalidHitoProjectName
+from ositah.utils.hito_db import get_db
+from ositah.utils.period import get_validation_period_data
+from ositah.utils.utils import (
+    DAY_HOURS,
+    TEAM_LIST_ALL_AGENTS,
+    TIME_UNIT_HOURS,
+    TIME_UNIT_HOURS_EN,
+    TIME_UNIT_HOURS_FR,
+    TIME_UNIT_WEEKS,
+    TIME_UNIT_WEEKS_EN,
+    TIME_UNIT_WEEKS_FR,
+    WEEK_HOURS,
+    GlobalParams,
+)
+
+CATEGORY_DEFAULT = "nsip_project"
+
+DATA_SOURCE_HITO = "hito"
+DATA_SOURCE_OSITAH = "ositah"
+
+NSIP_CLASS_OTHER_ACTIVITY = "activitensipreferentiel"
+NSIP_CLASS_PROJECT = "projetnsipreferentiel"
+
+MASTERPROJECT_DELETED_ACTIVITY = "Disabled"
+MASTERPROJECT_LOCAL_PROJECT = "Local Projects"
+
+NSIP_PROJECT_ORDER = 1
+LOCAL_PROJECT_ORDER = 2
+NSIP_ACIVITY_ORDER = 3
+DISABLED_ACTIVITY_ORDER = 9999
+
+
+def hito2ositah_project_name(hito_name):
+    """
+    Split a Hito project name into a masterprojet and project name
+
+    :param hito_name: Hito name with  masterprojet and project name separated by a /
+    :return: masterprojet and project name
+    """
+    masterproject, project_name = hito_name.split(" / ", 2)
+    return masterproject, project_name
+
+
+def ositah2hito_project_name(masterproject, project):
+    """
+    Build the Hito/NSIP project name from the masterproject and project
+
+    :param masterproject: masterproject name
+    :param project: project name
+    :return: Hito/NSIP project fullname
+    """
+    return " / ".join([masterproject, project])
+
+
+def nsip2ositah_project_name(masterproject, project):
+    """
+    Build the OSITAH project name from the NSIP project name, removing the master project
+    name if it is at the head of the NSIP project name, except if the master project name
+    and the project name are identical.
+
+    :param masterproject: masterproject name
+    :param project: project name
+    :return: OSITAH project name (without the masterproject name)
+    """
+
+    if project != masterproject:
+        m = re.match(rf"{masterproject}\s+\-\s+(?P<project>.*)", project)
+        if m:
+            project = m.group("project")
+
+    return project
+
+
+def category_from_activity(category_patterns, activity) -> str:
+    """
+    Return the activity category if the activity matches the pattern. Else an empty string.
+    Called as a lambda to build the category column.
+
+    :param category_patterns: category patterns to match against the activity (dict where
+                              the key is the pattern and the value is the category)
+    :param activity: activity name
+    :return: category or np.Nan
+    """
+
+    for pattern, category in category_patterns.items():
+        if re.match(pattern.lower(), activity.lower()):
+            return category
+
+    return np.NaN
+
+
+def activity_from_project(project):
+    """
+    Return the activity the project belongs to.
+
+    :param project: project name
+    :return: activity name
+    """
+
+    global_params = GlobalParams()
+
+    for activity, pattern in global_params.project_categories.items():
+        if re.match(pattern, project):
+            return activity
+
+    return CATEGORY_DEFAULT
+
+
+def reference_masterproject(reference_type):
+    """
+    Return an OSITAH masterproject for a NSIP reference, based on its type.
+    Masterprojects for each type is defined in the configuration. A reference type
+    without a match or with an empty value is ignored (np.NaN returned).
+
+    :param reference_type: NSIP reference type
+    :return: matching master project
+    """
+
+    global_params = GlobalParams()
+
+    for type_pattern, masterproject in global_params.reference_masterprojects.items():
+        if re.match(type_pattern.lower(), reference_type.lower()):
+            if len(masterproject) > 0:
+                return masterproject
+            else:
+                return np.NaN
+
+    return np.NaN
+
+
+def time_unit(category, short=False, english=True, parenthesis=False) -> str:
+    """
+    Return the time unit as defined in the configuration as a string. If the category/column is
+    not in the configuration, return an empty string.
+
+    :param category: project category/class
+    :param short: if true, return abbreviated unit names
+    :param english: return english unit names if true. Also implies short=False
+    :param parenthesis: if True, enclose the string in ()
+    :return: time unit for the category as a string
+    """
+
+    global_params = GlobalParams()
+
+    if english:
+        unit_w = TIME_UNIT_WEEKS_EN
+        unit_h = TIME_UNIT_HOURS_EN
+    else:
+        if short:
+            unit_w = "sem."
+            unit_h = "h"
+        else:
+            unit_w = TIME_UNIT_WEEKS_FR
+            unit_h = TIME_UNIT_HOURS_FR
+
+    if category in global_params.time_unit:
+        if global_params.time_unit[category] == TIME_UNIT_WEEKS:
+            unit_str = unit_w
+        elif global_params.time_unit[category] == TIME_UNIT_HOURS:
+            unit_str = unit_h
+        else:
+            raise Exception(
+                (
+                    f"Unsupported time unit '{global_params.time_unit[category]}'"
+                    f" for category {category}"
+                )
+            )
+    else:
+        return ""
+
+    if parenthesis:
+        return f"({unit_str})"
+    else:
+        return unit_str
+
+
+def category_time_and_unit(category, hours, short=True, english=False) -> Tuple[int, str]:
+    """
+    Return the rounded category time in the appropriate unit and the category time unit
+
+    :param category: project category/class
+    :param hours: number of hours
+    :param short: if true, return abbreviated unit names
+    :param english: return english unit names if true. Also implies short=False
+    :return: project time, project unit
+    """
+
+    global_params = GlobalParams()
+
+    unit = time_unit(category, short, english)
+
+    if global_params.time_unit[category] == "w":
+        declared_time = f"{int(round(hours / WEEK_HOURS))}"
+    else:
+        declared_time = f"{int(round(hours))}"
+
+    return declared_time, unit
+
+
+def project_time(project, hours):
+    """
+    Return the rounded project time in the appropriate unit and the project time unit
+
+    :param project: project name
+    :param hours: number of hours
+    :return: project time, abbreviated project unit
+    """
+
+    return category_time_and_unit(activity_from_project(project), hours)
+
+
+def get_team_projects(
+    team,
+    team_selection_date,
+    period_date: datetime,
+    source=DATA_SOURCE_HITO,
+    use_cache: bool = True,
+):
+    """
+    Query the Hito database and return a dataframe will all the project contributions for a given
+    team. The dataframe has one row for each each agent contribution to each project.
+
+    :param team: selected team or TEAM_LIST_ALL_AGENTS for all teams
+    :param team_selection_date: last time the team selection was changed
+    :param period_date: a date that must be inside the declaration period
+    :param source: whether to use Hito (non validated) or OSITAH (validated) as a data source
+    :param use_cache: if true, use the cache if defined and up-to-date or update it with the
+                      new declarations
+    :return: dataframe or None if the query returned no entry
+    """
+
+    from ositah.utils.hito_db_model import (
+        ActiviteDetail,
+        Agent,
+        OSITAHProjectDeclaration,
+        OSITAHValidation,
+        Projet,
+        Team,
+    )
+
+    global_params = GlobalParams()
+    columns = global_params.columns
+    session_data = global_params.session_data
+    db = get_db()
+
+    validation_period = get_validation_period_data(period_date)
+
+    # Check if there is a cached version
+    if session_data.project_declarations is not None and use_cache:
+        if (
+            session_data.project_declarations_source is None
+            or source != session_data.project_declarations_source
+            or datetime.fromisoformat(team_selection_date) > session_data.cache_date
+        ):
+            # Cache must be refreshed if the selected source doesn't match the cached one or if
+            # the team has been modified since the cache was loaded (required for multi-worker
+            # configurations as the team selection does not necessarily happen on the same worker
+            # than the later processing). In a multi-worker configuration is used it may also
+            # happen that the declaration source is not defined if it was initially initialised
+            # on another worker
+            clear_cached_data()
+        else:
+            return session_data.project_declarations
+
+    if source == DATA_SOURCE_OSITAH:
+        # The query relies on the fact that only one validation entry can be in the validated
+        # state for a given period, something enforced by the declaration validation.
+        # When a team is specified, display all projects from this team and the children teams
+        query = (
+            OSITAHProjectDeclaration.query.join(
+                OSITAHValidation,
+                OSITAHProjectDeclaration.validation_id == OSITAHValidation.id,
+            )
+            .join(Agent, Agent.id == OSITAHValidation.agent_id)
+            .join(Team, Team.id == Agent.team_id)
+            .add_entity(Agent)
+            .add_entity(Team)
+            .add_entity(OSITAHValidation)
+            .filter(OSITAHValidation.validated)
+            .filter(OSITAHValidation.period_id == validation_period.id)
+        )
+        if team != TEAM_LIST_ALL_AGENTS:
+            query = query.filter(Team.nom.ilike(f"{team}%"))
+        declarations = pd.read_sql(query.statement, db.session.bind)
+        if len(declarations) == 0:
+            return None
+        declarations.rename(columns={"id": columns["activity_id"]}, inplace=True)
+        declarations.rename(columns={"id_1": columns["agent_id"]}, inplace=True)
+        declarations.rename(columns={"nom_1": columns["team"]}, inplace=True)
+        declarations.rename(columns={"hours": columns["hours"]}, inplace=True)
+        declarations.rename(columns={"id_3": "validation_id"}, inplace=True)
+        declarations.rename(columns={"timestamp": "validation_time"}, inplace=True)
+        # Drop statut column to avoid conflicts in future merge with the Agent table
+        declarations.drop(columns=["statut"], inplace=True)
+        # Ensure that email_auth is defined and if it is not, replace it by the email.
+        declarations.loc[declarations[columns["email_auth"]].isna(), columns["email_auth"]] = (
+            declarations[columns["email"]]
+        )
+        declarations[columns["activity"]] = declarations.apply(
+            lambda row: ositah2hito_project_name(
+                row[columns["masterproject"]], row[columns["project"]]
+            ),
+            axis=1,
+        )
+
+    elif source == DATA_SOURCE_HITO:
+        # For team names, we want to keep the agent team name instead of the team_name in
+        # activity_details so it must be specified explicitely in the join with Team table
+        query = (
+            ActiviteDetail.query.join(Projet)
+            .join(Agent)
+            .join(Team, Team.id == Agent.team_id)
+            .add_entity(Projet)
+            .add_entity(Agent)
+            .add_entity(Team)
+            .filter(
+                ActiviteDetail.date >= validation_period.start_date,
+                ActiviteDetail.date <= validation_period.end_date,
+            )
+        )
+        if team != TEAM_LIST_ALL_AGENTS:
+            query = query.filter(Team.nom.ilike(f"{team}%"))
+        daily_declarations = pd.read_sql(query.statement, db.session.bind)
+        if len(daily_declarations) == 0:
+            return None
+        # Pandas add a suffix to duplicate column names, the first one being unchanged, the
+        # second being suffixed _1...
+        daily_declarations.drop(columns=["id", "id_1", "id_2"], inplace=True)
+        daily_declarations.rename(columns={"agent_id": columns["agent_id"]}, inplace=True)
+        daily_declarations.rename(columns={"libelle": columns["activity"]}, inplace=True)
+        daily_declarations.rename(columns={"nom_1": columns["team"]}, inplace=True)
+        for column in [columns["hours"], columns["percent"]]:
+            daily_declarations[column] = daily_declarations[column].astype(float)
+        # Ensure that email_auth is defined and if it is not, replace it by the email. If left
+        # undefined, the entries will not be present in the pivot table as it is part of the index.
+        daily_declarations.loc[
+            daily_declarations[columns["email_auth"]].isna(), columns["email_auth"]
+        ] = daily_declarations[columns["email"]]
+        # Rebuild agent quotite by comparing the time declared with the percent computed by Hito
+        # based on the quotite
+        daily_declarations[columns["quotite"]] = (
+            daily_declarations[columns["hours"]] / DAY_HOURS * 100
+        ) / daily_declarations[columns["percent"]]
+        global_declarations_pt = pd.pivot_table(
+            daily_declarations,
+            index=[
+                columns["lastname"],
+                columns["firstname"],
+                columns["activity"],
+                columns["activity_id"],
+                columns["team"],
+                columns["agent_id"],
+                columns["email_auth"],
+                columns["email"],
+            ],
+            values=[columns["hours"], columns["quotite"]],
+            aggfunc={columns["hours"]: "sum", columns["quotite"]: "mean"},
+        )
+        declarations = pd.DataFrame(global_declarations_pt.to_records())
+        declarations[[columns["masterproject"], columns["project"]]] = declarations[
+            columns["activity"]
+        ].str.split(" / ", n=1, expand=True)
+        # An entry in the pseudo master project MASTERPROJECT_DELETED_ACTIVITY is a special
+        # case corresponding to deleted NSIP projects: the real name is in the project part that
+        # must be parsed as for any other project
+        declarations["project_saved"] = np.NaN
+        declarations["project_saved"] = declarations["project_saved"].astype("object")
+        declarations.loc[
+            declarations[columns["masterproject"]] == MASTERPROJECT_DELETED_ACTIVITY,
+            "project_saved",
+        ] = declarations[columns["project"]]
+        # Not sure why the following line doesn't work (masterproject and project set to NaN
+        # if no row matches the indexing condition... An issue has been open:
+        # https://github.com/pandas-dev/pandas/issues/44726.
+        # declarations.loc[
+        #     declarations.project_saved.notna(),
+        #     [columns["masterproject"], columns["project"]],
+        # ] = declarations.project_saved.str.split(" / ", n=1, expand=True)
+        #
+        # The following workaround fails if project_saved contains only np.NaN. It is a known
+        # issue in Panda 1.3.4, see https://github.com/pandas-dev/pandas/issues/35807
+        # declarations[
+        #     ["newmaster", "newproject"]
+        # ] = declarations.project_saved.str.split(" / ", n=1, expand=True)
+        #
+        # Workaround based on
+        # https://github.com/pandas-dev/pandas/issues/35807#issuecomment-676912441. If no row
+        # matches the condition, only one column is created thus the need to check they all
+        # exist.
+        tmp_columns = ["newmaster", "newproject"]
+        saved_projects = (
+            declarations["project_saved"]
+            .str.split("/", expand=True, n=len(tmp_columns) - 1)
+            .rename(columns={k: name for k, name in enumerate(tmp_columns)})
+        )
+        for column in tmp_columns:
+            if column not in saved_projects.columns:
+                saved_projects[column] = np.NaN
+        declarations = declarations.join(saved_projects)
+        declarations.loc[declarations.project_saved.notna(), columns["masterproject"]] = (
+            declarations.newmaster
+        )
+        declarations.loc[declarations.project_saved.notna(), columns["project"]] = (
+            declarations.newproject
+        )
+        declarations.drop(columns=["newmaster", "newproject"], inplace=True)
+        declarations.loc[declarations.project_saved.notna(), columns["activity"]] = (
+            declarations.project_saved
+        )
+
+        # Detect project names not matching the format "masterproject / project"
+        invalid_hito_projects = declarations.loc[declarations[columns["project"]].isnull()]
+        if not invalid_hito_projects.empty:
+            raise InvalidHitoProjectName(
+                pd.Series(invalid_hito_projects[columns["masterproject"]]).unique()
+            )
+
+    else:
+        raise InvalidDataSource(source)
+
+    declarations[columns["fullname"]] = declarations[
+        [columns["lastname"], columns["firstname"]]
+    ].agg(" ".join, axis=1)
+    declarations[columns["category"]] = declarations.apply(
+        lambda row: category_from_activity(
+            global_params.category_patterns, row[columns["activity"]]
+        ),
+        axis=1,
+    )
+    declarations.loc[declarations[columns["category"]].isna(), "category"] = CATEGORY_DEFAULT
+
+    # Check quotite < 50% and flag the entry as suspect (generally means confusion between quotite
+    # and percent during declaration)
+    declarations["suspect"] = declarations[columns["quotite"]] < 0.5
+
+    if use_cache:
+        session_data.set_project_declarations(declarations, source)
+
+    return declarations
+
+
+def get_all_hito_activities(project_activity: bool):
+    """
+    Retrieve all projects or all activities defined in Hito with their associated teams
+
+    :param project_activity: if true, return all projects, else all Hito activities
+    :return: dataframe
+    """
+
+    from ositah.utils.hito_db_model import Activite, Projet
+
+    global_params = GlobalParams()
+    session_data = global_params.session_data
+    db = get_db()
+
+    # Check if there is a cached version
+    if session_data.get_hito_activities(project_activity) is not None:
+        return session_data.get_hito_activities(project_activity)
+
+    else:
+        if project_activity:
+            Activity = Projet
+        else:
+            Activity = Activite
+
+        query = Activity.query.options(joinedload(Activity.teams))
+        activities = pd.read_sql(query.statement, db.session.bind)
+        activities[["masterproject", "project"]] = activities.libelle.str.split(
+            " / ", n=1, expand=True
+        )
+        activities.rename(columns={"description_1": "team_description"}, inplace=True)
+        activities.rename(columns={"nom": "team_name"}, inplace=True)
+
+        session_data.set_hito_activities(activities, project_activity)
+
+        return activities
+
+
+def build_projects_data(team, team_selection_date, period_date: str, source):
+    """
+    Build the project list contributed by the selected team and return it as a dataframe
+
+    :param team: selected team
+    :param team_selection_date: last time the team selection was changed
+    :param period_date: a date that must be inside the declaration period
+    :param source: whether to use Hito (non validated) or OSITAH (validated) as a data source
+    :return: dataframe with projects data, dataframe with agent declarations
+    """
+
+    global_params = GlobalParams()
+    columns = global_params.columns
+    session_data = global_params.session_data
+
+    declaration_list = get_team_projects(team, team_selection_date, period_date, source)
+    if declaration_list is None:
+        return None, None
+
+    projects_data = session_data.projects_data
+    if projects_data is None:
+        projects_data_pt = pd.pivot_table(
+            declaration_list,
+            index=[
+                columns["masterproject"],
+                columns["project"],
+                columns["activity"],
+                columns["category"],
+            ],
+            values=[columns["hours"]],
+            aggfunc={columns["hours"]: "sum"},
+        )
+        projects_data = pd.DataFrame(projects_data_pt.to_records())
+        projects_data[columns["hours"]] = np.round(projects_data[columns["hours"]]).astype("int")
+        projects_data[columns["weeks"]] = np.round(projects_data[columns["hours"]] / WEEK_HOURS, 1)
+        short_name_len = 25
+        projects_data["project_short"] = projects_data[columns["project"]]
+        projects_data.loc[
+            projects_data["project_short"].str.len() > short_name_len, "project_short"
+        ] = projects_data["project_short"].str.slice_replace(start=short_name_len - 4, repl="...")
+        session_data.projects_data = projects_data
+
+    return projects_data, declaration_list
+
+
+def get_hito_nsip_activities(project_activity: bool = True):
+    """
+    Return a dataframe with all the NSIP activities defined in Hito. Activities can be
+    either projects or "references" (other activities). An activity is considered as a
+    NSIP activity if it has a matching entry in Hito referentiel.
+
+    :param project_activity: if true, return projects, else references
+    :return: dataframe
+    """
+
+    from ositah.utils.hito_db_model import Projet, Referentiel
+
+    db = get_db()
+
+    if project_activity:
+        project_join_id = Projet.projet_nsip_referentiel_id
+        referentiel_class = "projetnsipreferentiel"
+    else:
+        project_join_id = Projet.activite_nsip_referentiel_id
+        referentiel_class = "activitensipreferentiel"
+
+    activity_query = (
+        Projet.query.join(Referentiel, Referentiel.id == project_join_id)
+        .add_entity(Referentiel)
+        .filter(
+            Referentiel.object_class == referentiel_class,
+        )
+    )
+
+    activities = pd.read_sql(activity_query.statement, db.session.bind)
+
+    activities.drop(
+        columns=[
+            "ordre",
+            "projet_nsip_referentiel_id",
+            "activite_nsip_referentiel_id",
+        ],
+        inplace=True,
+    )
+    activities.rename(columns={"id_1": "referentiel_id"}, inplace=True)
+    activities.rename(columns={"libelle_1": "nsip_name_id"}, inplace=True)
+
+    if activities.empty:
+        activities[["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]] = [
+            np.NaN,
+            np.NaN,
+            np.NaN,
+            np.NaN,
+        ]
+    else:
+        activities[["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]] = (
+            activities.apply(
+                lambda v: nsip_activity_name_id(v["nsip_name_id"], v["class"]),
+                axis=1,
+                result_type="expand",
+            )
+        )
+        activities["nsip_project_id"] = activities["nsip_project_id"].astype(int)
+        activities["nsip_reference_id"] = activities["nsip_reference_id"].astype(int)
+
+    return activities
+
+
+def get_hito_projects():
+    """
+    Return a dataframe with the information about all projects with validated declarations
+    in the current declaration period defined in Hito and their relationship to NSIP, if relevant.
+
+    :return: Hito project dataframe
+    """
+
+    from ositah.utils.hito_db_model import (
+        OSITAHProjectDeclaration,
+        OSITAHValidation,
+        Projet,
+        Referentiel,
+    )
+
+    db = get_db()
+
+    projects_query = (
+        Projet.query.join(Referentiel, Referentiel.id == Projet.projet_nsip_referentiel_id)
+        .join(OSITAHProjectDeclaration)
+        .join(OSITAHValidation)
+        .add_entity(Referentiel)
+        .add_entity(OSITAHValidation)
+        .filter(
+            Referentiel.object_class == "projetnsipreferentiel",
+            OSITAHValidation.validated,
+        )
+    )
+    projects = pd.read_sql(projects_query.statement, db.session.bind)
+
+    activities_query = (
+        Projet.query.join(Referentiel, Referentiel.id == Projet.activite_nsip_referentiel_id)
+        .join(OSITAHProjectDeclaration)
+        .join(OSITAHValidation)
+        .add_entity(Referentiel)
+        .add_entity(OSITAHValidation)
+        .filter(
+            Referentiel.object_class == "activitensipreferentiel",
+            OSITAHValidation.validated,
+        )
+    )
+    activities = pd.read_sql(activities_query.statement, db.session.bind)
+
+    projects_activities = pd.concat([projects, activities], ignore_index=True)
+
+    projects_activities.drop(
+        columns=[
+            "id_2",
+            "ordre",
+            "projet_nsip_referentiel_id",
+            "activite_nsip_referentiel_id",
+            "agent_id",
+        ],
+        inplace=True,
+    )
+    projects_activities.rename(columns={"id_1": "referentiel_id"}, inplace=True)
+    projects_activities.rename(columns={"libelle_1": "nsip_name_id"}, inplace=True)
+    projects_activities.drop_duplicates(subset=["id"], inplace=True)
+
+    if projects_activities.empty:
+        projects_activities[
+            ["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]
+        ] = [np.NaN, np.NaN, np.NaN, np.NaN]
+    else:
+        projects_activities[
+            ["nsip_master", "nsip_project", "nsip_project_id", "nsip_reference_id"]
+        ] = projects_activities.apply(
+            lambda v: nsip_activity_name_id(v["nsip_name_id"], v["class"]),
+            axis=1,
+            result_type="expand",
+        )
+        projects_activities["nsip_project_id"] = projects_activities["nsip_project_id"].astype(int)
+        projects_activities["nsip_reference_id"] = projects_activities["nsip_reference_id"].astype(
+            int
+        )
+
+    return projects_activities
+
+
+def nsip_activity_name_id(hito_name: str, type: str) -> List[str]:
+    """
+    Split the NISP activity project name in Hito referentiel in 3 parts: masterproject name,
+    project name, project ID and return the project ID as the project ID (3d value) or reference
+    ID (4th value) depending on the activity type. The unused ID is set to 0 rather than np.NaN
+    or pd.NA as the column may be used in merges.
+
+    :param hito_name: activity name in Hito referentiel
+    :param type: referentiel class
+    :return:
+    """
+
+    m = re.match(
+        r"(?P<master>.*?)\s+/\s+(?P<project>.*)\s+\(NSIP ID:\s*(?P<id>\w+)\)$",
+        hito_name,
+    )
+    if m:
+        try:
+            _ = int(m.group("id"))
+        except ValueError:
+            print(
+                (
+                    f"ERROR: invalid NSIP ID in Hito referentiel for '{m.group('master')} /"
+                    f" {m.group('project')}' (ID={m.group('id')})"
+                )
+            )
+            return m.group("master"), m.group("project"), 0, 0
+        if type == NSIP_CLASS_PROJECT:
+            project_id = m.group("id")
+            reference_id = 0
+        else:
+            project_id = 0
+            reference_id = m.group("id")
+        return m.group("master"), m.group("project"), project_id, reference_id
+    else:
+        print(
+            (
+                f"ERROR: invalid Hito referentiel entry format, cannot be parsed as"
+                f" master/project/id ({hito_name})"
+            )
+        )
+        return np.NaN, np.NaN, 0, 0
+
+
+def get_nsip_declarations(period_date: str, team: str):
+    """
+    Return the NSIP declaration list for the declaration period matching a given date (the
+    date must be included in the period) as a dataframe
+
+    :param period_date: date that must be inside the period
+    :param team: selected team
+    :return: declaration list as a dataframe
+    """
+
+    global_params = GlobalParams()
+
+    if global_params.nsip:
+        declarations = pd.json_normalize(global_params.nsip.get_declarations(period_date))
+        if declarations.empty:
+            return declarations
+
+        declarations.rename(columns={"id": "id_declaration"}, inplace=True)
+        # Set NaN to 0 in reference as np.NaN is a float and prevent casting to int. As it will
+        # be used in a merge better to have a 0 than a NaN.
+        if "project.id" in declarations.columns:
+            declarations.loc[declarations["project.id"].isna(), "project.id"] = 0
+            declarations["project.id"] = declarations["project.id"].astype(int)
+        else:
+            declarations["project.id"] = 0
+        if "reference.id" in declarations.columns:
+            declarations.loc[declarations["reference.id"].isna(), "reference.id"] = 0
+            declarations["reference.id"] = declarations["reference.id"].astype(int)
+        else:
+            declarations["reference.id"] = 0
+        declarations["nsip_fullname"] = (
+            declarations["agent.lastname"] + " " + declarations["agent.firstname"]
+        )
+
+        if team != TEAM_LIST_ALL_AGENTS:
+            team_agents = get_agents(period_date, team)
+            agent_emails = team_agents["email_auth"]
+            declarations = declarations.merge(
+                agent_emails,
+                how="inner",
+                left_on="agent.email",
+                right_on="email_auth",
+                suffixes=[None, "_agent"],
+            )
+
+        return declarations
+
+    else:
+        return None
+
+
+def get_nsip_activities(project_activity: bool):
+    """
+    Retrieve laboratory activities defined in NSIP and return them in a dataframe.
+    Activities can be either projects or references (other activities).
+
+    :param project_activity: true for projects, false for other activities
+    :return: dataframe or None if NSIP is not configured
+    """
+
+    global_params = GlobalParams()
+
+    if global_params.nsip:
+        activities = pd.json_normalize(
+            global_params.nsip.get_activities(project_activity), record_prefix=True
+        )
+        if not activities.empty:
+            if project_activity:
+                activities["ositah_name"] = activities.apply(
+                    lambda p: nsip2ositah_project_name(p["master_project.name"], p["name"]),
+                    axis=1,
+                )
+            else:
+                activities["master_project.name"] = activities.apply(
+                    lambda p: reference_masterproject(p["type"]),
+                    axis=1,
+                )
+                activities.drop(
+                    activities[activities["master_project.name"].isna()].index,
+                    inplace=True,
+                )
+                activities["ositah_name"] = activities["name"]
+
+        return activities
+    else:
+        return None
+
+
+def build_activity_libelle(
+    nsip_id: str,
+    master_project: str,
+    project: str,
+):
+    """
+    Build Hito project name and referentiel entry name from NSIP master project, project name and
+    project id.
+
+    :param nsip_id: NSIP ID for the project
+    :param master_project: master project name
+    :param project: project name
+    :return: Hito project name, Hito referentiel name
+    """
+
+    new_project_name = f"{master_project} / {project}"
+    new_referentiel_name = f"{new_project_name} (NSIP ID: {nsip_id})"
+    return new_project_name, new_referentiel_name
+
+
+def update_activity_name(
+    hito_project_id: str,
+    hito_referentiel_id: str,
+    nsip_id: str,
+    master_project: str,
+    project: str,
+):
+    """
+    Update a project name in Hito, both in the referentiel and in the project/activity table.
+
+    :param hito_project_id: Hito project ID
+    :param hito_referentiel_id: Hito referentiel ID for the project
+    :param nsip_id: NSIP ID for the project
+    :param master_project: master project name
+    :param project: project name
+    :return: 0 if update succeeded, non-zero if an error occured, error_msg if
+             an error occured
+    """
+
+    from ositah.utils.hito_db_model import Projet, Referentiel
+
+    db = get_db()
+
+    status = 0  # Assume success
+    error_msg = ""
+    new_project_name, new_referentiel_name = build_activity_libelle(
+        nsip_id,
+        master_project,
+        project,
+    )
+
+    try:
+        referentiel_entry = Referentiel.query.filter(Referentiel.id == hito_referentiel_id).first()
+        project_entry = Projet.query.filter(Projet.id == hito_project_id).first()
+        referentiel_entry.libelle = new_referentiel_name
+        project_entry.libelle = new_project_name
+        change_log_msg = f"Modifié le {datetime.now()}"
+        if project_entry.description:
+            project_entry.description += f"; {change_log_msg}"
+        else:
+            project_entry.description = change_log_msg
+        db.session.commit()
+    except Exception as e:
+        status = 1
+        error_msg = getattr(e, "message", repr(e))
+        db.session.rollback()
+
+    return status, error_msg
+
+
+def add_activity(
+    nsip_id: str,
+    master_project: str,
+    project: str,
+    activity_teams: List[str],
+    project_activity: bool,
+):
+    """
+    Adds a new project in Hito referenciel and in Hito project/activity table
+
+    :param nsip_id: NSIP ID for the project
+    :param master_project: master project name
+    :param project: project name
+    :param activity_teams: list of team IDs associated with the project
+    :param project_activity: if True it is a NSIP project, else a NSIP activity
+    :return: 0 if update succeeded, non-zero if an error occured, error_msg if
+             an error occured
+    """
+
+    from ositah.utils.hito_db_model import Projet, Referentiel, Team
+
+    db = get_db()
+
+    status = 0  # Assume success
+    error_msg = ""
+    project_name, referentiel_name = build_activity_libelle(
+        nsip_id,
+        master_project,
+        project,
+    )
+
+    if project_activity:
+        entry_class = "projetnsipreferentiel"
+        entry_order = NSIP_PROJECT_ORDER
+    else:
+        entry_class = "activitensipreferentiel"
+        entry_order = NSIP_ACIVITY_ORDER
+
+    try:
+        referentiel_entry = Referentiel(
+            libelle=referentiel_name,
+            object_class=entry_class,
+            ordre=entry_order,
+        )
+        activity_entry = Projet(
+            libelle=project_name,
+            description=f"Créé le {datetime.now()}",
+            ordre=entry_order,
+        )
+        if activity_teams:
+            activity_entry.teams = Team.query.filter(Team.id.in_(activity_teams)).all()
+        db.session.add(referentiel_entry)
+        db.session.add(activity_entry)
+        db.session.commit()
+        # Define relationship between activity and referentiel entry after creating them so that
+        # the referentiel ID generated by the DB server can be accessed
+        if project_activity:
+            activity_entry.projet_nsip_referentiel_id = referentiel_entry.id
+        else:
+            activity_entry.activite_nsip_referentiel_id = referentiel_entry.id
+        db.session.commit()
+
+    except Exception as e:
+        status = 1
+        error_msg = getattr(e, "message", repr(e))
+        db.session.rollback()
+
+    return status, error_msg
+
+
+def remove_activity(
+    hito_project_id: str,
+    hito_referentiel_id: str,
+    nsip_id: str,
+    project_activity: bool,
+):
+    """
+    Remove the association between a Hito activity (project or reference) and NSIP. The Hito
+    activity is kept as it may be referenced by other objects but its description is updated
+    to mention that it is no longer in NSIP. The project name is updated so that it appears in the
+    pseudo-masterproject NSIP_DELETED_MASTERPROJECT. Associated teams are removed.
+
+    :param hito_project_id: Hito project ID
+    :param hito_referentiel_id: Hito referentiel ID for the project
+    :param nsip_id: NSIP ID for the project
+    :param project_activity: if True it is a NSIP project, else a NSIP activity
+    :return: 0 if update succeeded, non-zero if an error occured, error_msg if
+             an error occured
+    """
+
+    from ositah.utils.hito_db_model import Projet, Referentiel
+
+    db = get_db()
+
+    status = 0  # Assume success
+    error_msg = ""
+
+    try:
+        referentiel_entry = Referentiel.query.filter(Referentiel.id == hito_referentiel_id).first()
+        db.session.query()
+        activity_entry = Projet.query.filter(Projet.id == hito_project_id).first()
+        if project_activity:
+            activity_entry.projet_nsip_referentiel_id = None
+        else:
+            activity_entry.activite_nsip_referentiel_id = None
+        change_log_msg = f"Desactivé le {datetime.now()} (NSIP ID={nsip_id})"
+        if activity_entry.description:
+            activity_entry.description += f"; {change_log_msg}"
+        else:
+            activity_entry.description = change_log_msg
+        activity_entry.libelle = f"{MASTERPROJECT_DELETED_ACTIVITY} / {activity_entry.libelle}"
+        activity_entry.ordre = DISABLED_ACTIVITY_ORDER
+        if len(activity_entry.teams) > 0:
+            activity_entry.teams.clear()
+        db.session.delete(referentiel_entry)
+        db.session.commit()
+    except Exception as e:
+        status = 1
+        error_msg = getattr(e, "message", repr(e))
+        db.session.rollback()
+
+    return status, error_msg
+
+
+def add_activity_teams(
+    masterproject: str, project: str, team_list: List[str], project_activity: bool
+):
+    """
+    Add teams to an activity.
+
+    :param masterproject: activity masterproject name
+    :param project: activity project name
+    :param team_list: list of team names to add
+    :param project_activity: if true, an Hito project else an Hito activity
+    :return: status (0 if success), error_msg (empty if success)
+    """
+
+    from ositah.utils.hito_db_model import Activite, Projet, Team
+
+    db = get_db()
+
+    status = 0  # Assume success
+    error_msg = ""
+
+    if project_activity:
+        Activity = Projet
+    else:
+        Activity = Activite
+
+    activity_name = ositah2hito_project_name(masterproject, project)
+    activity = Activity.query.filter(Activity.libelle == activity_name).first()
+
+    if activity:
+        try:
+            for team in team_list:
+                team_object = Team.query.filter(Team.nom == team).first()
+                activity.teams.append(team_object)
+            db.session.commit()
+        except Exception as e:
+            status = 1
+            error_msg = getattr(e, "message", repr(e))
+            db.session.rollback()
+
+    return status, error_msg
+
+
+def remove_activity_teams(
+    masterproject: str, project: str, team_list: List[str], project_activity: bool
+):
+    """
+    Remove teams from an activity. If the team is not present in teams list, silently
+    ignore it.
+
+    :param masterproject: activity masterproject name
+    :param project: activity project name
+    :param team_list: list of team names to remove
+    :param project_activity: if true, an Hito project else an Hito activity
+    :return: status (0 if success), error_msg (empty if success)
+    """
+
+    from ositah.utils.hito_db_model import Activite, Projet, Team
+
+    db = get_db()
+
+    status = 0  # Assume success
+    error_msg = ""
+
+    if project_activity:
+        Activity = Projet
+    else:
+        Activity = Activite
+
+    activity_name = ositah2hito_project_name(masterproject, project)
+    activity = Activity.query.filter(Activity.libelle == activity_name).first()
+
+    if activity:
+        try:
+            for team in team_list:
+                team_object = Team.query.filter(Team.nom == team).first()
+                if team_object in activity.teams:
+                    activity.teams.remove(team_object)
+            db.session.commit()
+        except Exception as e:
+            status = 1
+            error_msg = getattr(e, "message", repr(e))
+            db.session.rollback()
+
+    return status, error_msg
+
+
+def reenable_activity(activity_name: str, project_activity: bool, name_prefix: str = None):
+    """
+    Reenable a disabled activity. This involves:
+    - Updating master project to match the original one
+    - If it was an NSIP project, recreate the referentiel entry
+
+    :param activity_name: activity name
+    :param project_activity: if true, an Hito project else an Hito activity
+    :param name_prefix: activity name prefix for deleted or local activities
+    :return: status and error message if any
+    """
+
+    from ositah.utils.hito_db_model import Activite, Projet, Referentiel
+
+    db = get_db()
+
+    status = 0  # Assume success
+    error_msg = ""
+
+    if project_activity:
+        Activity = Projet
+    else:
+        Activity = Activite
+
+    # Retrieve activity attributes and NSIP ID if present in description
+    if name_prefix:
+        activity_full_name = f"{name_prefix} / {activity_name}"
+    else:
+        activity_full_name = activity_name
+    activity_entry = Activity.query.filter(Activity.libelle == activity_full_name).first()
+
+    m = re.search(r"\(NSIP ID\=(?P<id>\d+)\)$", activity_entry.description)
+    if m:
+        nsip_id = m.group("id")
+    else:
+        nsip_id = None
+
+    # Check if an entry exist in the referentiel for the NSIP ID: if not, create it
+    nsip_entry = Referentiel.query.filter(
+        Referentiel.libelle.ilike(f"%(NSIP ID = {nsip_id})")
+    ).first()
+    if not nsip_entry:
+        master_project, activity = hito2ositah_project_name(activity_name)
+        project_name, referentiel_name = build_activity_libelle(
+            nsip_id,
+            master_project,
+            activity,
+        )
+
+        if project_activity:
+            entry_class = "projetnsipreferentiel"
+            entry_order = NSIP_PROJECT_ORDER
+        else:
+            entry_class = "activitensipreferentiel"
+            entry_order = NSIP_ACIVITY_ORDER
+
+        referentiel_entry = Referentiel(
+            libelle=referentiel_name,
+            object_class=entry_class,
+            ordre=entry_order,
+        )
+    else:
+        referentiel_entry = None
+
+    # Create referentiel entry if necessary and update activity
+    try:
+        if referentiel_entry:
+            db.session.add(referentiel_entry)
+        activity_entry.libelle = activity_name
+        activity_entry.description = f"Modifié le {datetime.now()}"
+        db.session.commit()
+        # Define relationship between activity and referentiel entry after creating them so that
+        # the referentiel ID generated by the DB server can be accessed
+        if project_activity:
+            activity_entry.projet_nsip_referentiel_id = referentiel_entry.id
+        else:
+            activity_entry.activite_nsip_referentiel_id = referentiel_entry.id
+        db.session.commit()
+
+    except Exception as e:
+        status = 1
+        error_msg = getattr(e, "message", repr(e))
+        db.session.rollback()
+
+    # Clear cached data to force a refresh of project list
+    clear_cached_data()
+
+    return status, error_msg