psr-factory 4.1.0b5__py3-none-manylinux_2_28_x86_64.whl

psr/cloud/cloud.py

@@ -0,0 +1,1079 @@
+# PSR Cloud. Copyright (C) PSR, Inc - All Rights Reserved
+# Unauthorized copying of this file, via any medium is strictly prohibited
+# Proprietary and confidential
+
+import copy
+import functools
+import hashlib
+import logging
+import os
+import re
+import subprocess
+import sys
+import warnings
+import xml.etree.ElementTree as ET
+from datetime import datetime, timedelta
+from pathlib import Path
+from time import sleep, time
+from typing import List, Optional, Union
+
+import pefile
+import zeep
+from filelock import FileLock
+
+from .data import Case, CloudError, CloudInputError
+from .desktop import import_case
+from .log import enable_log_timestamp, get_logger
+from .status import FAULTY_TERMINATION_STATUS, STATUS_MAP_TEXT, ExecutionStatus
+from .tempfile import CreateTempFile
+from .version import __version__
+from .xml import create_case_xml
+
+INTERFACE_VERSION = "PyCloud " + __version__ + ", binding for " + sys.version
+
+
+def thread_safe():
+    """
+    Decorator to make a function thread-safe using filelock.
+    :param lock_file: Path to the lock file. If None, it will be automatically generated.
+    """
+
+    def decorator(func):
+        @functools.wraps(func)
+        def wrapper(*args, **kwargs):
+            with FileLock("pycloud.lock"):
+                return func(*args, **kwargs)
+
+        return wrapper
+
+    return decorator
+
+
+def _md5sum(value: str) -> str:
+    return hashlib.md5(value.encode("utf-8")).hexdigest()  # nosec
+
+
+def hash_password(password: str) -> str:
+    return _md5sum(password).upper()
+
+
+def _check_for_errors(
+    xml: ET.ElementTree, logger: Optional[logging.Logger] = None
+) -> None:
+    error = xml.find("./Parametro[@nome='erro']")
+    if error is not None:
+        if logger is not None:
+            logger.error(error.text)
+        raise CloudError(error.text)
+
+
+def _hide_password(params: str) -> str:
+    pattern = r'(<Parametro nome="senha"[^>]*>)(.*?)(</Parametro>)'
+    result = re.sub(pattern, r"\1********\3", params)
+    return result
+
+
+def _xml_to_str(xml_content: ET.ElementTree) -> str:
+    # Remove <Parametro nome="senha" ...> tag, if found in the xml_content
+    xml_str = ET.tostring(
+        xml_content.getroot(), encoding="utf-8", method="xml"
+    ).decode()
+    return _hide_password(xml_str)
+
+
+def _handle_relative_path(path: str) -> str:
+    if not os.path.isabs(path):
+        return os.path.abspath(path)
+    return path
+
+
+_PSRCLOUD_PATH = r"C:\PSR\PSRCloud"
+
+_CONSOLE_REL_PARENT_PATH = r"Oper\Console"
+
+_CONSOLE_APP = r"FakeConsole.exe"
+
+_ALLOWED_PROGRAMS = ["SDDP", "OPTGEN", "PSRIO", "GRAF"]
+
+if os.name == "nt":
+    _PSRCLOUD_CREDENTIALS_PATH = os.path.expandvars(
+        os.path.join("%appdata%", "PSR", "PSRCloud", "EPSRConfig.xml")
+    )
+else:
+    _PSRCLOUD_CREDENTIALS_PATH = ""
+
+_PSRCLOUD_USER_ENV_VAR = "PSR_CLOUD_USER"
+_PSRCLOUD_PASSWORD_HASH_ENV_VAR = "PSR_CLOUD_PASSWORD_HASH"  # nosec
+_PSRCLOUD_CONSOLE_ENV_VAR = "PSR_CLOUD_CONSOLE_PATH"
+
+_auth_error_message = f"Please set {_PSRCLOUD_USER_ENV_VAR} and {_PSRCLOUD_PASSWORD_HASH_ENV_VAR} environment variables."
+
+# FIXME uninspired name
+_DEFAULT_GET_CASES_SINCE_DAYS = 7
+
+
+_DEFAULT_CLUSTER = {
+    "name": "PSR-US",
+    "pretty_name": "External",
+    "url": "https://psrcloud.psr-inc.com/CamadaGerenciadoraServicoWeb/DespachanteWS.asmx",
+}
+
+
+class Client:
+    def __init__(self, **kwargs) -> None:
+        self.cwd = Path.cwd()
+
+        # Caches (avoiding multiple soap requests)
+        self._cloud_version_xml_cache = None
+        self._cloud_clusters_xml_cache = None
+        self._instance_type_map = None
+
+        # Options
+        self._selected_cluster = kwargs.get("cluster", _DEFAULT_CLUSTER["pretty_name"])
+        self._import_desktop = kwargs.get("import_desktop", True)
+        self._debug_mode = kwargs.get("debug", False)
+        self._dry_run = kwargs.get("dry_run", False)
+        self._timeout = kwargs.get("timeout", None)
+
+        # Client version
+        self.application_version = kwargs.get("application_version", None)
+
+        # Logging setup
+        self._quiet = kwargs.get("quiet", False)
+        self._verbose = kwargs.get("verbose", False)
+        if self._debug_mode:
+            self._quiet = False
+            self._verbose = True
+        log_id = id(self)
+        self._logger = get_logger(
+            log_id, quiet=self._quiet, debug_mode=self._debug_mode
+        )
+
+        self._logger.info(f"Client uid {log_id} initialized.")
+
+        self._console_path_setup(**kwargs)
+
+        self._credentials_setup(**kwargs)
+
+        self._cluster_setup(self._selected_cluster)
+
+    def _console_path_setup(self, **kwargs) -> None:
+        # For common users - provide PSR Cloud install path
+        if "psrcloud_path" in kwargs:
+            psrcloud_path = Path(kwargs["psrcloud_path"])
+            self._console_path = psrcloud_path / _CONSOLE_REL_PARENT_PATH / _CONSOLE_APP
+            if not os.path.exists(self._console_path):
+                err_msg = (
+                    f"PSR Cloud application not found at {self._console_path} "
+                    f"Make sure the path is correct and PSR Cloud is installed."
+                )
+                self._logger.error(err_msg)
+                self._logger.info("Provided psrcloud_path: " + str(psrcloud_path))
+                raise CloudError(err_msg)
+        # For advanced users or tests - provide full FakeConsole.exe path.
+        elif "fakeconsole_path" in kwargs:
+            self._console_path = Path(kwargs["fakeconsole_path"])
+            if not os.path.exists(self._console_path):
+                err_msg = (
+                    f"PSR Cloud application not found at {self._console_path} "
+                    f"Make sure the path is correct and PSR Cloud is installed."
+                )
+                self._logger.error(err_msg)
+                self._logger.info(
+                    "Provided fakeconsole_path: " + str(self._console_path)
+                )
+                raise CloudError(err_msg)
+        # For advanced users or tests - provide PSR Cloud console path as environment variable.
+        elif _PSRCLOUD_CONSOLE_ENV_VAR in os.environ:
+            self._console_path = Path(os.environ[_PSRCLOUD_CONSOLE_ENV_VAR]).resolve()
+            if not os.path.exists(self._console_path):
+                err_msg = (
+                    f"PSR Cloud application not found at {self._console_path} "
+                    f"Make sure the path is correct and PSR Cloud is installed."
+                )
+                self._logger.error(err_msg)
+                self._logger.info("Provided console path: " + str(self._console_path))
+                raise CloudError(err_msg)
+        else:
+            self._console_path = (
+                Path(_PSRCLOUD_PATH) / _CONSOLE_REL_PARENT_PATH / _CONSOLE_APP
+            )
+            if not os.path.exists(self._console_path):
+                err_msg = (
+                    f"PSR Cloud application not found at {self._console_path} "
+                    f"Make sure the path is correct and PSR Cloud is installed."
+                )
+                self._logger.error(err_msg)
+                self._logger.info("Using default console path.")
+                raise CloudError(err_msg)
+
+        self._logger.info(f"PSR Cloud console path: {self._console_path}")
+        self._logger.info(f"PSR Cloud console version: {self._get_console_version()}")
+
+    def _credentials_setup(self, **kwargs) -> None:
+        self.username = kwargs.get("username", None)
+        self.__password = None
+        if self.username is not None:
+            self.username = kwargs["username"]
+            self.__password = hash_password(kwargs["password"])
+            self._logger.info(
+                "Using provided credentials from PSR Cloud console arguments."
+            )
+            self._logger.warning(
+                "For security reasons, it is highly recommended to use environment variables to store your credentials.\n"
+                + f"({_PSRCLOUD_USER_ENV_VAR}, {_PSRCLOUD_PASSWORD_HASH_ENV_VAR})"
+            )
+        else:
+            if (
+                _PSRCLOUD_USER_ENV_VAR in os.environ
+                and _PSRCLOUD_PASSWORD_HASH_ENV_VAR in os.environ
+            ):
+                self.username = os.environ[_PSRCLOUD_USER_ENV_VAR]
+                self.__password = os.environ[_PSRCLOUD_PASSWORD_HASH_ENV_VAR].upper()
+                self._logger.info("Using credentials from environment variables")
+            elif os.path.exists(_PSRCLOUD_CREDENTIALS_PATH):
+                self._logger.info(
+                    "Environment variables for Cloud credentials not found"
+                )
+                xml = ET.parse(
+                    _PSRCLOUD_CREDENTIALS_PATH, parser=ET.XMLParser(encoding="utf-16")
+                )
+                root = xml.getroot()
+                username = None
+                _password = None
+                for elem in root.iter("Aplicacao"):
+                    username = elem.attrib.get("SrvUsuario")
+                    _password = elem.attrib.get("SrvSenha")
+                    break
+                if username is None or _password is None:
+                    err_msg = "Credentials not provided. " + _auth_error_message
+                    self._logger.info(
+                        "Loading credentials from file: " + _PSRCLOUD_CREDENTIALS_PATH
+                    )
+                    self._logger.error(err_msg)
+                    raise CloudInputError(err_msg)
+                self.username = username
+                self.__password = _password
+                self._logger.info("Using credentials from PSR Cloud Desktop cache")
+            else:
+                err_msg = "Username and password not provided." + _auth_error_message
+                self._logger.info(
+                    "Trying to get credentials from environment variables."
+                )
+                self._logger.error(err_msg)
+                raise CloudInputError(err_msg)
+        self._logger.info(f"Logged as {self.username}")
+
+    def _cluster_setup(self, cluster_str: str) -> None:
+        """
+        Get cluster object by name.
+        If the cluster is the default one, select it directly. If not, check using default cluster to get
+        the available clusters for this user and select the one that matches the provided name.
+        """
+
+        if (
+            _DEFAULT_CLUSTER["name"].upper() == cluster_str.upper()
+            or _DEFAULT_CLUSTER["pretty_name"].capitalize() == cluster_str.capitalize()
+        ):
+            self.cluster = _DEFAULT_CLUSTER
+        else:
+            self.cluster = None
+            clusters = self._get_clusters_by_user()
+            for cluster in clusters:
+                if (
+                    cluster["name"].upper() == cluster_str.upper()
+                    or cluster["pretty_name"].capitalize() == cluster_str.capitalize()
+                ):
+                    self.cluster = cluster
+
+        if self.cluster is not None:
+            self._logger.info(
+                f"Running on Cluster {self.cluster['name']} ({self.cluster['pretty_name']})"
+            )
+        else:
+            raise CloudInputError(f"Cluster {cluster_str} not found")
+
+    def set_cluster(self, cluster_str: str) -> None:
+        self._cluster_setup(cluster_str)
+        # Clear caches
+        self._cloud_version_xml_cache = None
+        self._cloud_clusters_xml_cache = None
+        self._instance_type_map = None
+
+    def _get_console_path(self) -> Path:
+        return self._console_path
+
+    def _get_console_parent_path(self) -> Path:
+        return self._console_path.parent
+
+    def _get_console_version(self) -> str:
+        console_path = self._get_console_path()
+        pe = pefile.PE(console_path)
+        for file_info in getattr(pe, "FileInfo", []):
+            for entry in file_info:
+                for st in getattr(entry, "StringTable", []):
+                    product_version = st.entries.get(b"ProductVersion")
+                    if product_version:
+                        return product_version.decode()
+
+    @staticmethod
+    def _check_xml(xml_content: str) -> None:
+        try:
+            ET.fromstring(xml_content)
+        except ET.ParseError:
+            _hide_password(xml_content)
+            raise CloudInputError(
+                f"Invalid XML content.\n"
+                f"Contact PSR support at psrcloud@psr-inc.com with following data:\n\n{xml_content}\n\n"
+            )
+
+    def _get_clusters_by_user(self) -> list:
+        try:
+            previous_cluster = self.cluster
+            self.cluster = _DEFAULT_CLUSTER
+            xml = self._make_soap_request("listarCluster", "listaCluster")
+
+            clusters = []
+            for cluster in xml.findall("Cluster"):
+                nome = cluster.attrib.get("nome")
+                url = cluster.attrib.get("urlServico") + "/DespachanteWS.asmx"
+                pretty_name = cluster.attrib.get("legenda", nome)
+                clusters.append({"name": nome, "pretty_name": pretty_name, "url": url})
+
+            self.cluster = previous_cluster
+        except Exception as e:
+            self.cluster = previous_cluster
+            raise e
+        return clusters
+
+    def get_clusters(self) -> List[str]:
+        clusters = self._get_clusters_by_user()
+        return [cluster["pretty_name"] for cluster in clusters]
+
+    def _run_console(self, xml_content: str) -> None:
+        self._check_xml(xml_content)
+        delete_xml = not self._debug_mode
+        with CreateTempFile(
+            str(self.cwd), "psr_cloud_", xml_content, delete_xml
+        ) as xml_file:
+            xml_file.close()
+            command = [self._get_console_path(), xml_file.name]
+            command_str = " ".join(map(str, command))
+            self._logger.debug(f"Running console command {command_str}")
+            quiet_goes_to_log = subprocess.PIPE if self._debug_mode else None
+            if self._verbose:
+                proc_stdout = subprocess.PIPE
+                proc_stderr = subprocess.PIPE
+            else:
+                if self._quiet:
+                    proc_stdout = quiet_goes_to_log
+                    proc_stderr = quiet_goes_to_log
+                else:
+                    proc_stdout = subprocess.PIPE
+                    proc_stderr = None
+            try:
+                process = subprocess.Popen(
+                    command, stdout=proc_stdout, stderr=proc_stderr, shell=False
+                )
+                enable_log_timestamp(self._logger, False)
+                if proc_stdout is not None:
+                    with process.stdout:
+                        for line in iter(process.stdout.readline, b""):
+                            if self._verbose:
+                                self._logger.info(line.decode().strip())
+                            else:
+                                self._logger.debug(line.decode().strip())
+                if proc_stderr is not None:
+                    with process.stderr:
+                        for line in iter(process.stderr.readline, b""):
+                            self._logger.error(line.decode().strip())
+                enable_log_timestamp(self._logger, True)
+                result = process.wait(timeout=self._timeout)
+
+                if result != 0:
+                    err_msg = (
+                        f"PSR Cloud console command failed with return code {result}"
+                    )
+                    self._logger.error(err_msg)
+                    raise CloudError(err_msg)
+            except subprocess.CalledProcessError as e:
+                err_msg = f"PSR Cloud console command failed with exception: {str(e)}"
+                self._logger.error(err_msg)
+                raise CloudError(err_msg)
+
+    def _validate_case(self, case: "Case") -> "Case":
+        if not case.program:
+            raise CloudInputError("Program not provided")
+        elif case.program not in self.get_programs():
+            raise CloudInputError(
+                f"Program {case.program} not found. Available programs are: {', '.join(self.get_programs())}"
+            )
+
+        if not case.memory_per_process_ratio:
+            raise CloudInputError("Memory per process ratio not provided")
+        elif case.memory_per_process_ratio not in self.get_memory_per_process_ratios():
+            raise CloudInputError(
+                f"Memory per process ratio {case.memory_per_process_ratio} not found. Available ratios are: {', '.join(self.get_memory_per_process_ratios())}"
+            )
+
+        if case.number_of_processes < 1 or case.number_of_processes > 512:
+            raise CloudInputError("Number of processes must be between 1 and 512")
+
+        if case.data_path and not Path(case.data_path).exists():
+            raise CloudInputError("Data path does not exist")
+
+        if case.parent_case_id is None:
+            case.parent_case_id = 0
+
+        def validate_selection(
+            selection, available_options, selection_name, program_name
+        ):
+            if selection is None:
+                raise CloudInputError(
+                    f"{selection_name} of program {program_name} not provided"
+                )
+            elif isinstance(selection, str):
+                if selection not in available_options.values():
+                    raise CloudInputError(
+                        f"{selection_name} {selection} of program {program_name} not found. Available {selection_name.lower()}s are: {', '.join(available_options.values())}"
+                    )
+                return next(
+                    key
+                    for key, value in available_options.items()
+                    if value == selection
+                )
+            elif selection not in available_options:
+                raise CloudInputError(
+                    f"{selection_name} id {selection} of program {program_name} not found. Available {selection_name.lower()} ids are: {', '.join(map(str,available_options.keys()))}"
+                )
+            return selection
+
+        program_versions = self.get_program_versions(case.program)
+        case.program_version = validate_selection(
+            case.program_version, program_versions, "Version", case.program
+        )
+
+        execution_types = self.get_execution_types(case.program, case.program_version)
+        case.execution_type = validate_selection(
+            case.execution_type, execution_types, "Execution type", case.program
+        )
+
+        instance_type_map = self._get_instance_type_map()
+        if all(value[1] == False for value in instance_type_map.values()):
+            is_spot_disabled = True
+        else:
+            is_spot_disabled = False
+
+        if case.price_optimized == True and is_spot_disabled == True:
+            raise CloudError("Price Optimized is temporarily unavailable.")
+
+        repository_durations = self.get_repository_durations()
+        case.repository_duration = validate_selection(
+            case.repository_duration,
+            repository_durations,
+            "Repository duration",
+            case.program,
+        )
+
+        if case.budget:
+            budgets = self.get_budgets()
+            match_list = _budget_matches_list(case.budget, budgets)
+            if len(match_list) == 0:
+                raise CloudInputError(
+                    f'Budget "{case.budget}" not found. Get a list of available budgets using Client().get_budgets().'
+                )
+            elif len(match_list) > 1:
+                raise CloudInputError(
+                    f'Multiple budgets found for "{case.budget}". Please use the budget id instead of the name.\n'
+                    "\n".join([f' - "{budget}"' for budget in match_list])
+                )
+            else:
+                # Replace partial with complete budget name
+                case.budget = match_list[0]
+
+        return case
+
+    def _pre_process_graph(self, path: str, case_id: int) -> None:
+        # This method is only used for testing the graf cloud execution.
+        # Error handling is already done on the tests module.
+        parameters = {
+            "urlServico": self.cluster["url"],
+            "usuario": self.username,
+            "senha": self.__password,
+            "idioma": "3",
+            "modelo": "Graf",
+            "comando": "PreProcessamento",
+            "cluster": self.cluster["name"],
+            "repositorioId": str(case_id),
+            "diretorioDestino": path,
+            "tipoExecucao": "1",
+        }
+
+        xml_content = create_case_xml(parameters)
+        self._run_console(xml_content)
+
+    def _check_until_status(
+        self, case_id: int, requested_status: "ExecutionStatus", timeout: int = 60 * 60
+    ) -> bool:
+        """
+        Check the status of a case until the requested status is reached or timeout occurs.
+
+        :param case_id: The ID of the case to check.
+        :param requested_status: The status to wait for.
+        :param timeout: The maximum time to wait in seconds (default is 3600 seconds or 1 hour).
+        :return: True if the requested status is reached, False if timeout occurs.
+        """
+        status = None
+        start_time = time()
+        original_quiet_flag = self._quiet
+        original_verbose_flag = self._verbose
+        original_debug_flag = self._debug_mode
+        self._quiet, self._verbose, self._debug_mode = True, False, False
+        try:
+            while status not in FAULTY_TERMINATION_STATUS + [
+                ExecutionStatus.SUCCESS,
+                requested_status,
+            ]:
+                if time() - start_time > timeout:
+                    self._logger.error(
+                        f"Timeout reached while waiting for status {requested_status}"
+                    )
+                    return False
+                status, _ = self.get_status(case_id)
+                sleep(20)
+        finally:
+            self._quiet = original_quiet_flag
+            self._verbose = original_verbose_flag
+            self._debug_mode = original_debug_flag
+
+        return status == requested_status
+
+    def _clean_folder(self, folder):
+        for root, dirs, files in os.walk(folder, topdown=False):
+            for file in files:
+                file_path = os.path.join(root, file)
+                os.remove(file_path)
+
+            for dir in dirs:
+                dir_path = os.path.join(root, dir)
+                os.rmdir(dir_path)
+
+    @thread_safe()
+    def run_case(self, case: "Case", **kwargs) -> int:
+        self._validate_case(case)
+        instance_type_map = self._get_instance_type_map()
+        instance_type_id = next(
+            key
+            for key, value in instance_type_map.items()
+            if value[0] == case.memory_per_process_ratio
+            and value[1] == case.price_optimized
+        )
+        case.data_path = _handle_relative_path(case.data_path)
+
+        if case.program == "GRAF":
+            wait = True
+        else:
+            wait = kwargs.get("wait", False)
+
+        if self.application_version:
+            interface_version = self.application_version + " - " + INTERFACE_VERSION
+        else:
+            interface_version = INTERFACE_VERSION
+        parameters = {
+            "urlServico": self.cluster["url"],
+            "usuario": self.username,
+            "senha": self.__password,
+            "idioma": "3",
+            "modelo": case.program,
+            "comando": "executar",
+            "cluster": self.cluster["name"],
+            "diretorioDados": case.data_path,
+            "origemDados": "LOCAL",
+            "s3Dados": "",
+            "nproc": case.number_of_processes,
+            "repositorioId": "0",
+            "repositorioPai": case.parent_case_id,
+            "instanciaTipo": instance_type_id,
+            "validacaoModelo": "True",
+            "validacaoUsuario": "False",
+            "idVersao": case.program_version,
+            "pathModelo": "C:\\PSR",
+            "idTipoExecucao": case.execution_type,
+            "nomeCaso": case.name,
+            "tipoExecucao": str(int(not wait)),
+            "deveAgendar": "False",
+            "userTag": "(Untagged)",
+            "lifecycle": case.repository_duration,
+            "versaoInterface": interface_version,
+        }
+
+        if case.budget:
+            parameters["budget"] = case.budget
+        if case.upload_only is not None:
+            parameters["saveInCloud"] = case.upload_only
+
+        xml_content = create_case_xml(parameters)
+
+        if self._dry_run:
+            return xml_content
+
+        self._run_console(xml_content)
+        xml = ET.parse(
+            f"{self._get_console_parent_path()}\\fake{case.program}_async.xml"
+        )
+        _check_for_errors(xml, self._logger)
+        id_parameter = xml.find("./Parametro[@nome='repositorioId']")
+        if id_parameter is None:
+            xml_str = _xml_to_str(xml)
+            raise CloudError(
+                f"Case id not found on returned XML response.\n"
+                f"Contact PSR support at psrcloud@psr-inc.com with following data:\n\n{xml_str}\n\n"
+            )
+
+        case_id = int(id_parameter.text)
+        if not wait:
+            self._logger.info(f"Case {case.name} started with id {case_id}")
+
+        if self._import_desktop and case.program != "GRAF":
+            try:
+                case_copy = copy.deepcopy(case)
+                case_copy.id = case_id
+                replace_case_str_values(self, case_copy)
+                import_case(case_copy, self.cluster["name"], instance_type_id)
+            except Exception as e:
+                msg = f"Failed to import case {case.name} to desktop:\n{str(e)}"
+                self._logger.error(msg)
+                warnings.warn(msg)
+
+        return case_id
+
+    def get_status(self, case_id: int) -> tuple["ExecutionStatus", str]:
+        # case = self.get_case(case_id)
+        delete_xml = not self._debug_mode
+        xml_content = ""
+        with CreateTempFile(
+            str(self.cwd), "psr_cloud_status_", xml_content, delete_xml
+        ) as xml_file:
+            status_xml_path = os.path.abspath(xml_file.name)
+
+            parameters = {
+                "urlServico": self.cluster["url"],
+                "usuario": self.username,
+                "senha": self.__password,
+                "idioma": "3",
+                "idFila": str(case_id),
+                "modelo": "SDDP",
+                "comando": "obterstatusresultados",
+                "arquivoSaida": status_xml_path,
+            }
+            run_xml_content = create_case_xml(parameters)
+
+            self._run_console(run_xml_content)
+            xml = ET.parse(status_xml_path)
+            parameter_status = xml.find("./Parametro[@nome='statusExecucao']")
+            if parameter_status is None:
+                xml_str = _xml_to_str(xml)
+                raise CloudError(
+                    f"Status not found on returned XML response.\n"
+                    f"Contact PSR support at psrcloud@psr-inc.com with following data:\n\n{xml_str}\n\n"
+                )
+            try:
+                status = ExecutionStatus(int(parameter_status.text))
+            except CloudError:
+                xml_str = _xml_to_str(xml)
+                raise CloudError(
+                    f"Unrecognized status on returned XML response.\n"
+                    f"Contact PSR support at psrcloud@psr-inc.com with following data:\n\n{xml_str}\n\n"
+                )
+
+            self._logger.info(f"Status: {STATUS_MAP_TEXT[status]}")
+            return status, STATUS_MAP_TEXT[status]
+
+    def list_download_files(self, case_id: int) -> List[str]:
+        xml_files = self._make_soap_request(
+            "prepararListaArquivosRemotaDownload",
+            "listaArquivoRemota",
+            additional_arguments={
+                "cluster": self.cluster["name"],
+                "filtro": "(.*)",
+                "diretorioRemoto": str(case_id),
+            },
+        )
+
+        files = []
+
+        for file in xml_files.findall("Arquivo"):
+            file_info = {
+                "name": file.attrib.get("nome"),
+                "filesize": file.attrib.get("filesize"),
+                "filedate": file.attrib.get("filedate"),
+            }
+            files.append(file_info)
+
+        return files
+
+    def download_results(
+        self,
+        case_id: int,
+        output_path: Union[str, Path],
+        files: Optional[List[str]] = None,
+        extensions: Optional[List[str]] = None,
+    ) -> None:
+        filter = ".*("
+
+        if not extensions and not files:
+            extensions = ["csv", "log", "hdr", "bin", "out", "ok"]
+
+        filter_elements = []
+
+        if extensions:
+            Client._validate_extensions(extensions)
+            filter_elements.extend([f".*.{ext}$" for ext in extensions])
+
+        if files:
+            filter_elements.extend(files)
+
+        filter += "|".join(filter_elements)
+        filter += ")$"
+
+        self._logger.info("Download filter: " + filter)
+        case = self.get_case(case_id)
+        output_path = _handle_relative_path(output_path)
+        parameters = {
+            "urlServico": self.cluster["url"],
+            "usuario": self.username,
+            "senha": self.__password,
+            "idioma": "3",
+            "_cluster": self.cluster["name"],
+            "modelo": case.program,
+            "comando": "download",
+            "diretorioDestino": output_path,
+            "repositorioId": str(case_id),
+            "filtroDownloadPorMascara": filter,
+        }
+
+        xml_content = create_case_xml(parameters)
+        self._run_console(xml_content)
+        self._logger.info(f"Results downloaded to {output_path}")
+
+    def cancel_case(self, case_id: int, wait: bool = False) -> bool:
+        parameters = {
+            "urlServico": self.cluster["url"],
+            "usuario": self.username,
+            "senha": self.__password,
+            "idioma": "3",
+            "modelo": "SDDP",
+            "comando": "cancelarfila",
+            "cancelamentoForcado": "False",
+            "idFila": str(case_id),
+        }
+
+        xml_content = create_case_xml(parameters)
+        self._run_console(xml_content)
+        self._logger.info(f"Request to cancel case {case_id} was sent")
+
+        if wait:
+            self._logger.info(f"Waiting for case {case_id} to be canceled")
+            if self._check_until_status(
+                case_id, ExecutionStatus.CANCELLED, timeout=60 * 10
+            ):
+                self._logger.info(f"Case {case_id} was successfully canceled")
+                return True
+            else:
+                self._logger.error(f"Failed to cancel case {case_id}")
+                return False
+        else:
+            return True
+
+    def _cases_from_xml(self, xml: ET.Element) -> List["Case"]:
+        instance_type_map = self._get_instance_type_map()
+        cases = []
+        for fila in xml.findall("Fila"):
+            try:
+                case = Case(
+                    name=fila.attrib.get("nomeCaso"),
+                    data_path=None,
+                    program=fila.attrib.get("programa"),
+                    program_version=int(fila.attrib.get("idVersao")),
+                    execution_type=int(fila.attrib.get("idTipoExecucao")),
+                    price_optimized=bool(fila.attrib.get("flagSpot")),
+                    number_of_processes=int(fila.attrib.get("numeroProcesso")),
+                    id=int(fila.attrib.get("repositorioId")),
+                    user=fila.attrib.get("usuario"),
+                    execution_date=datetime.strptime(
+                        fila.attrib.get("dataInicio"), "%d/%m/%Y %H:%M"
+                    ),
+                    parent_case_id=int(fila.attrib.get("repositorioPai"))
+                    if fila.attrib.get("repositorioPai")
+                    else 0,
+                    memory_per_process_ratio=(
+                        instance_type_map[int(fila.attrib.get("instanciaTipo"))][0]
+                        if fila.attrib.get("instanciaTipo") in instance_type_map
+                        else min([value[0] for value in instance_type_map.values()])
+                    ),
+                    repository_duration=int(fila.attrib.get("duracaoRepositorio")),
+                    budget=fila.attrib.get("budget"),
+                )
+                cases.append(case)
+            except (TypeError, ValueError):
+                pass
+                # case_id = fila.attrib.get("repositorioId")
+                # Optionally, log the error or handle it as needed
+                # self._logger.error(f"Error processing case with ID {case_id}: {e}")
+
+        # sort cases by execution date desc
+        cases.sort(key=lambda x: x.execution_date, reverse=True)
+        return cases
+
+    def get_all_cases_since(
+        self, since: Union[int, datetime] = _DEFAULT_GET_CASES_SINCE_DAYS
+    ) -> List["Case"]:
+        if isinstance(since, int):
+            initial_date = datetime.now() - timedelta(days=since)
+            initial_date_iso = initial_date.isoformat().replace("T", " ")[:-7]
+        else:
+            initial_date_iso = since.strftime("%Y-%m-%d %H:%M:%S")
+
+        xml = self._make_soap_request(
+            "listarFila",
+            "dados",
+            additional_arguments={"dataInicial": initial_date_iso},
+        )
+
+        return self._cases_from_xml(xml)
+
+    def get_case(self, case_id: int) -> "Case":
+        cases = self.get_cases([case_id])
+        if cases and len(cases) > 0:
+            return cases[0]
+        raise CloudInputError(f"Case {case_id} not found")
+
+    def get_cases(self, case_ids: List[int]) -> List["Case"]:
+        case_ids_str = ",".join(map(str, case_ids))
+        xml = self._make_soap_request(
+            "listarFila",
+            "dados",
+            additional_arguments={"listaRepositorio": case_ids_str},
+        )
+        return self._cases_from_xml(xml)
+
+    def get_budgets(self) -> list:
+        xml = self._make_soap_request(
+            "listarCluster",
+            "listaCluster",
+        )
+
+        budgets = []
+        for cluster in xml.findall("Cluster"):
+            if cluster.attrib.get("nome").lower() == self.cluster["name"].lower():
+                collection = cluster.findall("ColecaoBudget")[0]
+                budgets = [
+                    budget.attrib.get("nome") for budget in collection.findall("Budget")
+                ]
+                break
+        budgets.sort()
+        return budgets
+
+    def _make_soap_request(self, service: str, name: str = "", **kwargs) -> ET.Element:
+        portal_ws = zeep.Client(self.cluster["url"] + "?WSDL")
+        section = str(id(self))
+        password_md5 = _md5sum(self.username + self.__password + section).upper()
+        password_md5 = (
+            password_md5
+            if self.cluster["name"] == "PSR-US"
+            or self.cluster["name"] == "PSR-HOTFIX"
+            or self.cluster["name"] == "PSR-US_OHIO"
+            else self.__password.upper()
+        )
+        additional_arguments = kwargs.get("additional_arguments", None)
+        parameters = {
+            "sessao_id": section,
+            "tipo_autenticacao": "portal"
+            if self.cluster["name"] == "PSR-US"
+            or self.cluster["name"] == "PSR-HOTFIX"
+            or self.cluster["name"] == "PSR-US_OHIO"
+            else "bcrypt",
+            "idioma": "3",
+        }
+        if additional_arguments:
+            parameters.update(additional_arguments)
+
+        # FIXME make additional arguments work as a dictionary to work with this code
+        xml_input = create_case_xml(parameters)
+
+        try:
+            xml_output_str = portal_ws.service.despacharServico(
+                service, self.username, password_md5, xml_input
+            )
+        except zeep.exceptions.Fault as e:
+            # Log the full exception details
+            self._logger.error(f"Zeep Fault: {str(e)}")
+            raise CloudError(
+                "Failed to connect to PSR Cloud service. Contact PSR support: psrcloud@psr-inc.com"
+            )
+        # Remove control characters - this is a thing
+        xml_output_str = xml_output_str.replace("&amp;#x1F;", "")
+
+        xml_output = ET.fromstring(xml_output_str)
+
+        if name:
+            for child in xml_output:
+                if child.attrib.get("nome") == name:
+                    xml_output = ET.fromstring(child.text)
+                    break
+            else:
+                raise ValueError(
+                    f"Invalid XML response from PSR Cloud: {xml_output_str}. Please contact PSR support at psrcloud@psr-inc.com"
+                )
+        return xml_output
+
+    def _get_cloud_versions_xml(self) -> ET.Element:
+        if self._cloud_version_xml_cache is not None:
+            return self._cloud_version_xml_cache
+        self._cloud_version_xml_cache = self._make_soap_request("obterVersoes", "dados")
+        return self._cloud_version_xml_cache
+
+    def _get_cloud_clusters_xml(self) -> ET.Element:
+        if self._cloud_clusters_xml_cache is not None:
+            return self._cloud_clusters_xml_cache
+        self._cloud_clusters_xml_cache = self._make_soap_request(
+            "listarCluster", "listaCluster"
+        )
+        return self._cloud_clusters_xml_cache
+
+    def get_programs(self) -> List[str]:
+        xml = self._get_cloud_versions_xml()
+        programs = [model.attrib["nome"] for model in xml]
+        return [program for program in programs if program in _ALLOWED_PROGRAMS]
+
+    def get_program_versions(self, program: str) -> dict[int, str]:
+        if not isinstance(program, str):
+            raise CloudInputError("Program must be a string")
+        elif program not in self.get_programs():
+            raise CloudInputError(
+                f"Program {program} not found. Available programs: {', '.join(self.get_programs())}"
+            )
+        xml = self._get_cloud_versions_xml()
+        versions = {}
+
+        for model in xml:
+            if model.attrib["nome"] == program:
+                for version_child in model.findall(".//Versao"):
+                    version_id = int(version_child.attrib["id"])
+                    version_name = version_child.attrib["versao"]
+                    versions[version_id] = version_name
+
+        return versions
+
+    def get_execution_types(
+        self, program: str, version: Union[str, int]
+    ) -> dict[int, str]:
+        if not isinstance(program, str):
+            raise CloudInputError("Program must be a string")
+        elif program not in self.get_programs():
+            raise CloudInputError(
+                f"Program {program} not found. Available programs: {', '.join(self.get_programs())}"
+            )
+        if isinstance(version, int):
+            if version not in self.get_program_versions(program):
+                raise CloudInputError(
+                    f"Version id {version} of program {program} not found. Available version ids: {', '.join(map(str, list(self.get_program_versions(program).keys())))}"
+                )
+            version = next(
+                v for k, v in self.get_program_versions(program).items() if k == version
+            )
+        elif version not in self.get_program_versions(program).values():
+            raise CloudInputError(
+                f"Version {version} of program {program} not found. Available versions: {', '.join(self.get_program_versions(program).values())}"
+            )
+        xml = self._get_cloud_versions_xml()
+        return {
+            int(execution_type.attrib["id"]): execution_type.attrib["nome"]
+            for program_child in xml
+            if program_child.attrib["nome"] == program
+            for version_child in program_child[0][0][0]
+            if version_child.attrib["versao"] == version
+            for execution_type in version_child[0]
+        }
+
+    def get_memory_per_process_ratios(self) -> List[str]:
+        xml = self._get_cloud_clusters_xml()
+        return sorted(
+            list(
+                {
+                    f"{instance_type.attrib['memoriaPorCore']}:1"
+                    for cluster in xml
+                    if cluster.attrib["nome"] == self.cluster["name"]
+                    for colection in cluster
+                    if colection.tag == "ColecaoInstanciaTipo"
+                    for instance_type in colection
+                }
+            )
+        )
+
+    def get_repository_durations(self) -> dict[int, str]:
+        if self.cluster == "PSR-US":
+            return {
+                2: "Normal (1 month)",
+            }
+
+        else:
+            return {
+                1: "Short (1 week)",
+                2: "Normal (1 month)",
+                3: "Extended (6 months)",
+                4: "Long (2 years)",
+            }
+
+    def _get_instance_type_map(self) -> dict[int, tuple[str, bool]]:
+        if self._instance_type_map is not None:
+            return self._instance_type_map
+        xml = self._get_cloud_clusters_xml()
+        self._instance_type_map = {
+            int(instance_type.attrib["id"]): (
+                f'{instance_type.attrib["memoriaPorCore"]}:1',
+                "Price Optimized" in instance_type.attrib["descricao"],
+            )
+            for cluster in xml
+            if cluster.attrib["nome"] == self.cluster["name"]
+            for collection in cluster
+            if collection.tag == "ColecaoInstanciaTipo"
+            for instance_type in collection
+        }
+        return self._instance_type_map
+
+    @staticmethod
+    def _validate_extensions(extensions: List[str]):
+        for ext in extensions:
+            if not ext.isalnum():
+                raise CloudInputError(
+                    f"Invalid extension '{ext}' detected. Extensions must be alphanumeric."
+                )
+
+
+def _budget_matches_list(budget_part: str, all_budgets: List[str]) -> List[str]:
+    """Tests if a part of a budget name is in the list all_budgets and returns a list of matches."""
+    lowered_budget_part = budget_part.lower()
+    return [budget for budget in all_budgets if lowered_budget_part in budget.lower()]
+
+
+def replace_case_str_values(client: Client, case: Case) -> Case:
+    """Create a new case object using internal integer IDs instead of string values."""
+    # Model Version
+    if isinstance(case.program_version, str):
+        program_versions = client.get_program_versions(case.program)
+        case.program_version = next(
+            key
+            for key, value in program_versions.items()
+            if value == case.program_version
+        )
+
+    # Execution Type
+    if isinstance(case.execution_type, str):
+        execution_types = client.get_execution_types(case.program, case.program_version)
+        case.execution_type = next(
+            key
+            for key, value in execution_types.items()
+            if value == case.execution_type
+        )
+    return case