machineconfig 1.95__py3-none-any.whl → 1.96__py3-none-any.whl

machineconfig/utils/scheduling.py

@@ -1,222 +1,312 @@
-"""Task scheduler
-"""
-
-from crocodile.file_management import P, Read, str2timedelta, Save
-# from crocodile.meta import Terminal
-from machineconfig.utils.utils import get_shell_script_executing_python_file
-from dataclasses import dataclass
-from datetime import datetime, timedelta
-import platform
-import subprocess
-from typing import Optional
-# from crocodile.meta import Scheduler
-from rich.console import Console
-from rich.panel import Panel
-
-
-SCHEDULER_DEFAULT_ROOT = P.home().joinpath("dotfiles/scripts/.scheduler")
-SUCCESS = "success"
-DEFAULT_CONFIG = """
-[specs]
-frequency = 30d
-start = 2024-01-01 01:00
-
-[runtime]
-venv = ve
-"""
-
-
-class Register:
-    def __init__(self, root: str):
-        self.root = P(root).expanduser().absolute()
-
-    def register_runtime(self, frequency_months: int = 1):
-        start, end = self.get_report_start_end_datetimes(frequency_months=frequency_months)
-        report = Report(name="runtime", start=start, end=end, status="success")
-        report.to_path(self.root.joinpath("runtime.ini"))
-        return report
-
-    def read_runtime(self):
-        return Report.from_path(self.root.joinpath("runtime.ini"))
-
-    @staticmethod
-    def format_date(date: datetime):
-        return str(date.year)[2:] + "-" + str(date.month).zfill(2)
-    @staticmethod
-    def get_report_start_end_datetimes(frequency_months: int):
-        now = datetime.now()
-        import numpy as np
-        chunks = np.arange(start=1, stop=12, step=frequency_months)
-        chunk_now_start_month: int = chunks[chunks <= now.month][-1]
-        chunk_now_start = datetime(year=now.year, month=chunk_now_start_month, day=1)
-        from dateutil.relativedelta import relativedelta
-        previous_chunk_start = chunk_now_start - relativedelta(months=frequency_months)
-        return previous_chunk_start, chunk_now_start
-
-
-@dataclass
-class Report:
-    name: str
-    start: datetime
-    end: datetime
-    status: str
-
-    @classmethod
-    def from_path(cls, path: P, return_default_if_not_found: bool=False):
-        if not path.exists():
-            if return_default_if_not_found:
-                return Report(name=path.parent.name, start=datetime(year=2000, month=1, day=1), end=datetime(year=2000, month=1, day=1), status="NA")
-            else:
-                raise ValueError(f"Could not find report at {path}")
-        ini = Read.ini(path)['report']
-        return cls(
-            name=ini["name"],
-            start=datetime.fromisoformat(ini["start"]),
-            end=datetime.fromisoformat(ini["end"]),
-            status=ini["status"],
-        )
-    def to_path(self, path: P):
-        Save.ini(path=path, obj={'report': {
-            'name': self.name,
-            'start': self.start.isoformat(),
-            'end': self.end.isoformat(),
-            'status': str(self.status),
-        }})
-
-
-@dataclass
-class Task:
-    name: str
-    task_root: P
-    frequency: timedelta
-    start: datetime
-    venv: str
-    @property
-    def report_path(self):
-        return self.task_root.joinpath("report.ini")
-
-
-def read_task_from_dir(path: P):
-    tasks_config = Read.ini(path.joinpath("config.ini"))
-    task = Task(name=path.name,
-                task_root=path,
-                frequency=str2timedelta(tasks_config["specs"]["frequency"]),
-                start=datetime.fromisoformat(tasks_config["specs"]["start"]),
-                venv=tasks_config["runtime"]["venv"],
-                # output_dir=P(a_task_section["output_dir"]).expanduser().absolute(),
-                )
-    return task
-
-
-def main(root: Optional[str] = None, ignore_conditions: bool=True):
-    if root is None: root_resolved = SCHEDULER_DEFAULT_ROOT
-    else: root_resolved = P(root).expanduser().absolute()
-    tasks_dirs = root_resolved.search(files=False, folders=True).filter(lambda x: x.joinpath("task.py").exists())
-
-    # Print a fancy box using rich
-    console = Console()
-    console.print(Panel("TASK SCHEDULER INITIALIZED", title="Status", expand=False))
-
-    print(f"📁 Root directory resolved: {root_resolved}")
-
-    tasks: list[Task] = []
-    for a_dir in tasks_dirs:
-        tasks.append(read_task_from_dir(a_dir))
-
-    from machineconfig.utils.utils import choose_multiple_options
-    import pandas as pd
-    df_res = pd.DataFrame([Report.from_path(path=a_task.report_path).__dict__ for a_task in tasks])
-    tasks_chosen_raw = choose_multiple_options(df_res.to_markdown().splitlines(), "📋 Choose tasks to run")
-    tasks_chosen = [tasks[int(a_task_chosen.split("|")[1])] for a_task_chosen in tasks_chosen_raw]
-
-    print(f"""
-🎯 Selected Tasks:
-{df_res}
-""")
-
-    result: list[Report] = []
-    for a_task in tasks_chosen:
-        if not ignore_conditions:
-            answer, report = should_task_run(a_task)
-        else:
-            answer, report = True, None
-        if answer: report = run_task(a_task)
-        else:
-            assert report is not None
-        result.append(report)
-
-    df_res = pd.DataFrame([r.__dict__ for r in result])
-    print(f"""
-✅ Task Execution Results:
-{df_res}
-""")
-    return ""
-
-
-def should_task_run(task: Task, tolerance_mins: int = 1440) -> tuple[bool, Optional[Report]]:
-    if not task.report_path.exists():
-        print(f"""
-⚠️  Task {task.name} has no record of being run before. Running now...
-""")
-        return True, None
-    old_report = Report.from_path(task.report_path)
-    time_since_execution = datetime.now() - old_report.end
-    if time_since_execution > task.frequency:
-        print(f"""
-⚠️  Task {task.name} has not been run for {time_since_execution}. It is meant to run every {task.frequency}. Running now if time is okay...
-""")
-    elif old_report.status != SUCCESS:
-        print(f"""
-⚠️  Task {task.name} last run failed. Running now if time is okay...
-""")
-    else:
-        print(f"""
-✅ Task `{task.name}` was run successfully {time_since_execution} ago. Skipping...
-""")
-        return False, old_report
-
-    suitable_run_time = task.start.time()
-    time_now = datetime.now().time()
-    min_diff = abs(suitable_run_time.hour - time_now.hour) * 60 + abs(suitable_run_time.minute - time_now.minute)
-    if not min_diff < tolerance_mins:
-        status = f"⌚ Time now is not suitable for running task {task.name} (Ideally, it should be run at {suitable_run_time})"
-        print(f"""
-{status}
-""")
-        return False, old_report
-    return True, old_report
-
-
-def run_task(task: Task) -> Report:
-    start_time = datetime.now()
-
-    # Print a fancy box using rich
-    console = Console()
-    console.print(Panel("RUNNING TASK", title="Status", expand=False))
-
-    print(f"Task: {task.name}")
-
-    shell_script = get_shell_script_executing_python_file(python_file=task.task_root.joinpath("task.py").to_str(), ve_name=task.venv)
-    shell_script_root = P.tmp().joinpath(f"tmp_scripts/scheduler/{task.name}").create()
-    try:
-        if platform.system() == 'Windows':
-            shell_script = shell_script_root.joinpath("run.ps1").write_text(shell_script)
-            subprocess.run(['powershell', '-ExecutionPolicy', 'Unrestricted', shell_script], check=True)
-        elif platform.system() == 'Linux':
-            shell_script = shell_script_root.joinpath("run.sh").write_text(shell_script)
-            subprocess.run(['bash', shell_script], check=True)
-        else: res = f"Error: Unsupported platform {platform.system()}."
-        res = SUCCESS
-    except subprocess.CalledProcessError:
-        res = f"Error: The script {shell_script_root} failed to run."
-    except Exception as e:
-        res = f"Error: An unexpected error occurred while running the script {shell_script_root}: {e}"
-
-    end_time = datetime.now()
-    report = Report(name=task.name, start=start_time, end=end_time, status=res.replace('\n', '_NL_').strip().replace('=', '_eq_'))
-    report.to_path(task.report_path)
-
-    print(f"""
-✅ Task {task.name} completed with status: {res}
-""")
-    return report
+# """Task scheduler
+# """
+
+# from machineconfig.utils.utils import get_shell_script_executing_python_file
+# from machineconfig.utils.utils2 import read_ini
+# from dataclasses import dataclass
+# from datetime import datetime, timedelta
+# from typing import Optional, Any
+# from rich.console import Console
+# from rich.panel import Panel
+# import platform
+# import subprocess
+
+
+# def str2timedelta(time_str: str) -> timedelta:
+#     """Convert string to timedelta. Simple implementation for common cases."""
+#     # Handle common formats like "1h", "30m", "2d", etc.
+#     import re
+    
+#     # Parse patterns like "1h", "30m", "2d", "1w"
+#     match = re.match(r'^(\d+)([hdwm])$', time_str.lower())
+#     if match:
+#         value, unit = match.groups()
+#         value = int(value)
+        
+#         if unit == 'h':
+#             return timedelta(hours=value)
+#         elif unit == 'd':
+#             return timedelta(days=value)
+#         elif unit == 'w':
+#             return timedelta(weeks=value)
+#         elif unit == 'm':
+#             return timedelta(minutes=value)
+    
+#     # Fallback: try to parse as seconds
+#     try:
+#         return timedelta(seconds=int(time_str))
+#     except ValueError:
+#         raise ValueError(f"Cannot parse time string: {time_str}")
+
+
+# def format_table_markdown(data: list[dict[str, Any]]) -> str:
+#     """Convert list of dictionaries to markdown table format."""
+#     if not data:
+#         return ""
+    
+#     # Get all unique keys from all dictionaries
+#     all_keys = set()
+#     for row in data:
+#         all_keys.update(row.keys())
+    
+#     keys = sorted(all_keys)
+    
+#     # Create header
+#     header = "|" + "|".join(f" {key} " for key in keys) + "|"
+#     separator = "|" + "|".join(" --- " for _ in keys) + "|"
+    
+#     # Create rows
+#     rows = []
+#     for row in data:
+#         row_values = []
+#         for key in keys:
+#             value = row.get(key, "")
+#             # Convert to string and handle None values
+#             if value is None:
+#                 value = ""
+#             else:
+#                 value = str(value)
+#             row_values.append(f" {value} ")
+#         rows.append("|" + "|".join(row_values) + "|")
+    
+#     return "\n".join([header, separator] + rows)
+
+
+# SCHEDULER_DEFAULT_ROOT = PathExtended.home().joinpath("dotfiles/scripts/.scheduler")
+# SUCCESS = "success"
+# DEFAULT_CONFIG = """
+# [specs]
+# frequency = 30d
+# start = 2024-01-01 01:00
+
+# [runtime]
+# venv = ve
+# """
+
+
+# def save_ini(path: PathExtended, obj: dict[str, dict[str, Any]]) -> PathExtended:
+#     """Write a simple INI file to `path` using standard library.
+
+#     The `obj` should be a mapping of section name to a mapping of key/value pairs.
+#     Values are converted to strings. The parent directory is created if missing.
+#     Returns the resolved path.
+#     """
+#     import configparser
+
+#     resolved_path: PathExtended = PathExtended(path).expanduser().absolute()
+#     resolved_path.parent.mkdir(parents=True, exist_ok=True)
+
+#     config = configparser.ConfigParser()
+#     for section_name, section_values in obj.items():
+#         # Ensure all values are stringified for configparser
+#         config[section_name] = {str(k): ("" if v is None else str(v)) for k, v in section_values.items()}
+
+#     with open(resolved_path, "w", encoding="utf-8") as file:
+#         config.write(file)
+
+#     return resolved_path
+
+
+# class Register:
+#     def __init__(self, root: str):
+#         self.root = PathExtended(root).expanduser().absolute()
+
+#     def register_runtime(self, frequency_months: int = 1):
+#         start, end = self.get_report_start_end_datetimes(frequency_months=frequency_months)
+#         report = Report(name="runtime", start=start, end=end, status="success")
+#         report.to_path(self.root.joinpath("runtime.ini"))
+#         return report
+
+#     def read_runtime(self):
+#         return Report.from_path(self.root.joinpath("runtime.ini"))
+
+#     @staticmethod
+#     def format_date(date: datetime):
+#         return str(date.year)[2:] + "-" + str(date.month).zfill(2)
+#     # @staticmethod
+#     # def get_report_start_end_datetimes(frequency_months: int):
+#     #     now = datetime.now()
+#     #     import numpy as np
+#     #     chunks = np.arange(start=1, stop=12, step=frequency_months)
+#     #     chunk_now_start_month: int = chunks[chunks <= now.month][-1]
+#     #     chunk_now_start = datetime(year=now.year, month=chunk_now_start_month, day=1)
+#     #     from dateutil.relativedelta import relativedelta
+#     #     previous_chunk_start = chunk_now_start - relativedelta(months=frequency_months)
+#     #     return previous_chunk_start, chunk_now_start
+
+
+# @dataclass
+# class Report:
+#     name: str
+#     start: datetime
+#     end: datetime
+#     status: str
+
+#     @classmethod
+#     def from_path(cls, path: PathExtended, return_default_if_not_found: bool=False):
+#         if not path.exists():
+#             if return_default_if_not_found:
+#                 return Report(name=path.parent.name, start=datetime(year=2000, month=1, day=1), end=datetime(year=2000, month=1, day=1), status="NA")
+#             else:
+#                 raise ValueError(f"Could not find report at {path}")
+#         ini = read_ini(path)['report']
+#         return cls(
+#             name=ini["name"],
+#             start=datetime.fromisoformat(ini["start"]),
+#             end=datetime.fromisoformat(ini["end"]),
+#             status=ini["status"],
+#         )
+#     def to_path(self, path: PathExtended):
+#         save_ini(path=path, obj={
+#             'report': {
+#                 'name': self.name,
+#                 'start': self.start.isoformat(),
+#                 'end': self.end.isoformat(),
+#                 'status': str(self.status),
+#             }
+#         })
+
+
+# @dataclass
+# class Task:
+#     name: str
+#     task_root: PathExtended
+#     frequency: timedelta
+#     start: datetime
+#     venv: str
+#     @property
+#     def report_path(self):
+#         return self.task_root.joinpath("report.ini")
+
+
+# def read_task_from_dir(path: PathExtended):
+#     tasks_config = read_ini(path.joinpath("config.ini"))
+#     task = Task(name=path.name,
+#                 task_root=path,
+#                 frequency=str2timedelta(tasks_config["specs"]["frequency"]),
+#                 start=datetime.fromisoformat(tasks_config["specs"]["start"]),
+#                 venv=tasks_config["runtime"]["venv"],
+#                 # output_dir=PathExtended(a_task_section["output_dir"]).expanduser().absolute(),
+#                 )
+#     return task
+
+
+# def main(root: Optional[str] = None, ignore_conditions: bool=True):
+#     if root is None: root_resolved = SCHEDULER_DEFAULT_ROOT
+#     else: root_resolved = PathExtended(root).expanduser().absolute()
+#     # Replace crocodile List usage with pathlib iteration
+#     from pathlib import Path
+#     # Find all `task.py` files under root and use their parent directories
+#     tasks_dirs = list({PathExtended(p.parent) for p in Path(str(root_resolved)).rglob("task.py")})
+
+#     # Print a fancy box using rich
+#     console = Console()
+#     console.print(Panel("TASK SCHEDULER INITIALIZED", title="Status", expand=False))
+
+#     print(f"📁 Root directory resolved: {root_resolved}")
+
+#     tasks: list[Task] = []
+#     for a_dir in tasks_dirs:
+#         tasks.append(read_task_from_dir(a_dir))
+
+#     from machineconfig.utils.utils import choose_multiple_options
+    
+#     # Create data for tasks display
+#     task_data = [Report.from_path(path=a_task.report_path).__dict__ for a_task in tasks]
+#     task_display = format_table_markdown(task_data)
+#     tasks_chosen_raw = choose_multiple_options(task_display.splitlines(), "📋 Choose tasks to run")
+#     tasks_chosen = [tasks[int(a_task_chosen.split("|")[1])] for a_task_chosen in tasks_chosen_raw]
+
+#     print(f"""
+# 🎯 Selected Tasks:
+# {task_display}
+# """)
+
+#     result: list[Report] = []
+#     for a_task in tasks_chosen:
+#         if not ignore_conditions:
+#             answer, report = should_task_run(a_task)
+#         else:
+#             answer, report = True, None
+#         if answer: report = run_task(a_task)
+#         else:
+#             assert report is not None
+#         result.append(report)
+
+#     result_data = [r.__dict__ for r in result]
+#     result_display = format_table_markdown(result_data)
+#     print(f"""
+# ✅ Task Execution Results:
+# {result_display}
+# """)
+#     return ""
+
+
+# def should_task_run(task: Task, tolerance_mins: int = 1440) -> tuple[bool, Optional[Report]]:
+#     if not task.report_path.exists():
+#         print(f"""
+# ⚠️  Task {task.name} has no record of being run before. Running now...
+# """)
+#         return True, None
+#     old_report = Report.from_path(task.report_path)
+#     time_since_execution = datetime.now() - old_report.end
+#     if time_since_execution > task.frequency:
+#         print(f"""
+# ⚠️  Task {task.name} has not been run for {time_since_execution}. It is meant to run every {task.frequency}. Running now if time is okay...
+# """)
+#     elif old_report.status != SUCCESS:
+#         print(f"""
+# ⚠️  Task {task.name} last run failed. Running now if time is okay...
+# """)
+#     else:
+#         print(f"""
+# ✅ Task `{task.name}` was run successfully {time_since_execution} ago. Skipping...
+# """)
+#         return False, old_report
+
+#     suitable_run_time = task.start.time()
+#     time_now = datetime.now().time()
+#     min_diff = abs(suitable_run_time.hour - time_now.hour) * 60 + abs(suitable_run_time.minute - time_now.minute)
+#     if not min_diff < tolerance_mins:
+#         status = f"⌚ Time now is not suitable for running task {task.name} (Ideally, it should be run at {suitable_run_time})"
+#         print(f"""
+# {status}
+# """)
+#         return False, old_report
+#     return True, old_report
+
+
+# def run_task(task: Task) -> Report:
+#     start_time = datetime.now()
+
+#     # Print a fancy box using rich
+#     console = Console()
+#     console.print(Panel("RUNNING TASK", title="Status", expand=False))
+
+#     print(f"Task: {task.name}")
+
+#     shell_script = get_shell_script_executing_python_file(python_file=str(task.task_root.joinpath("task.py")), ve_name=task.venv)
+#     shell_script_root = PathExtended.tmp().joinpath(f"tmp_scripts/scheduler/{task.name}")
+#     shell_script_root.mkdir(parents=True, exist_ok=True)
+#     try:
+#         if platform.system() == 'Windows':
+#             shell_script = shell_script_root.joinpath("run.ps1").write_text(shell_script)
+#             subprocess.run(['powershell', '-ExecutionPolicy', 'Unrestricted', shell_script], check=True)
+#         elif platform.system() in ['Linux', 'Darwin']:
+#             shell_script = shell_script_root.joinpath("run.sh").write_text(shell_script)
+#             subprocess.run(['bash', shell_script], check=True)
+#         else: res = f"Error: Unsupported platform {platform.system()}."
+#         res = SUCCESS
+#     except subprocess.CalledProcessError:
+#         res = f"Error: The script {shell_script_root} failed to run."
+#     except Exception as e:
+#         res = f"Error: An unexpected error occurred while running the script {shell_script_root}: {e}"
+
+#     end_time = datetime.now()
+#     report = Report(name=task.name, start=start_time, end=end_time, status=res.replace('\n', '_NL_').strip().replace('=', '_eq_'))
+#     report.to_path(task.report_path)
+
+#     print(f"""
+# ✅ Task {task.name} completed with status: {res}
+# """)
+#     return report

machineconfig/utils/utils.py

@@ -2,7 +2,7 @@
 Utils
 """
 
-from crocodile.file_management import P
+from machineconfig.utils.path_reduced import P as PathExtended
 # import crocodile.environment as env
 import machineconfig
 from machineconfig.utils.options import check_tool_exists, choose_cloud_interactively, choose_multiple_options, choose_one_option, choose_ssh_host, display_options
@@ -26,12 +26,12 @@ _ = sanitize_path
 _ = match_file_name
 
 
-LIBRARY_ROOT = P(machineconfig.__file__).resolve().parent  # .replace(P.home().to_str().lower(), P.home().str)
+LIBRARY_ROOT = PathExtended(machineconfig.__file__).resolve().parent  # .replace(str(PathExtended.home()).lower(), PathExtended.home().str)
 REPO_ROOT = LIBRARY_ROOT.parent.parent
-CONFIG_PATH = P.home().joinpath(".config/machineconfig")
+CONFIG_PATH = PathExtended.home().joinpath(".config/machineconfig")
 INSTALL_VERSION_ROOT = CONFIG_PATH.joinpath("cli_tools_installers/versions")
-INSTALL_TMP_DIR = P.home().joinpath("tmp_results", "tmp_installers")
-DEFAULTS_PATH = P.home().joinpath("dotfiles/machineconfig/defaults.ini")
+INSTALL_TMP_DIR = PathExtended.home().joinpath("tmp_results", "tmp_installers")
+DEFAULTS_PATH = PathExtended.home().joinpath("dotfiles/machineconfig/defaults.ini")
 
 
 # def get_latest_version(url: str) -> None:
@@ -51,7 +51,7 @@ DEFAULTS_PATH = P.home().joinpath("dotfiles/machineconfig/defaults.ini")
 
 
 def check_dotfiles_version_is_beyond(commit_dtm: str, update: bool=False):
-    dotfiles_path = str(P.home().joinpath("dotfiles"))
+    dotfiles_path = str(PathExtended.home().joinpath("dotfiles"))
     from git import Repo
     repo = Repo(path=dotfiles_path)
     last_commit = repo.head.commit
@@ -66,7 +66,7 @@ def check_dotfiles_version_is_beyond(commit_dtm: str, update: bool=False):
         main(cloud=None, path=dotfiles_path)
     return res
 
-def wait_for_jobs_to_finish(root: P, pattern: str, wait_for_n_jobs: int, max_wait_minutes: float) -> bool:
+def wait_for_jobs_to_finish(root: PathExtended, pattern: str, wait_for_n_jobs: int, max_wait_minutes: float) -> bool:
     wait_finished: bool=False
     import time
     t0 = time.time()

machineconfig/utils/utils2.py

@@ -0,0 +1,42 @@
+
+
+from pathlib import Path
+from typing import Optional, Any
+
+def randstr(length: int = 10, lower: bool = True, upper: bool = True, digits: bool = True, punctuation: bool = False, safe: bool = False, noun: bool = False) -> str:
+    if safe:
+        import secrets
+        return secrets.token_urlsafe(length)  # interannly, it uses: random.SystemRandom or os.urandom which is hardware-based, not pseudo
+    if noun:
+        import randomname
+        return randomname.get_name()
+    import string
+    import random
+    population = (string.ascii_lowercase if lower else "") + (string.ascii_uppercase if upper else "") + (string.digits if digits else "") + (string.punctuation if punctuation else "")
+    return ''.join(random.choices(population, k=length))
+
+
+def read_ini(path: 'Path', encoding: Optional[str] = None):
+    if not Path(path).exists() or Path(path).is_dir(): raise FileNotFoundError(f"File not found or is a directory: {path}")
+    import configparser
+    res = configparser.ConfigParser()
+    res.read(filenames=[str(path)], encoding=encoding)
+    return res
+def read_json(path: 'Path', r: bool = False, **kwargs: Any) -> Any:  # return could be list or dict etc
+    import json
+    try:
+        mydict = json.loads(Path(path).read_text(encoding='utf-8'), **kwargs)
+    except Exception:
+        import pyjson5
+        mydict = pyjson5.loads(Path(path).read_text(encoding='utf-8'), **kwargs)  # file has C-style comments.
+    _ = r
+    return mydict
+def read_toml(path: 'Path'):
+    import tomli
+    return tomli.loads(path.read_text(encoding='utf-8'))
+
+def pprint(obj: dict[Any, Any], title: str) -> None:
+    from rich import inspect
+    inspect(type("TempStruct", (object,), obj)(), value=False, title=title, docs=False, dunder=False, sort=False)
+def get_repr(obj: dict[Any, Any], sep: str = "\n", justify: int = 15, quotes: bool = False):
+    return sep.join([f"{key:>{justify}} = {repr(val) if quotes else val}" for key, val in obj.items()])