Trixy7/trixy_core/utils/template_formatter.py

# -*- coding: utf-8 -*-
"""
Template-Formatter mit Platzhalter-System und Ausdruecken.

Verarbeitungs-Pipeline:
    1. {placeholder}-Variablen aufloesen (date, sys, env, ctx, app)
    2. (:math)-Ausdruecke auswerten
    3. (condition ? true : false) Logik-Ausdruecke auswerten
    4. <tag>-basierte String-/Zahlenformatierung

Unterstuetzte Namespaces fuer {placeholder}:
    - {date.now.*}        Aktuelles Datum/Uhrzeit
    - {date.tomorrow.*}   Morgen
    - {date.yesterday.*}  Gestern
    - {sys.*}             Systeminformationen
    - {env.*}             Umgebungsvariablen
    - {ctx.*}             Werte aus dem uebergebenen Kontext-Dict
    - {app.*}             Werte aus der Server-Applikation (magische Konstanten)

Mathematische Ausdruecke:
    (:10 + 5)             → 15
    (:2 ^ 10)             → 1024
    <calc>3 * 4</calc>    → 12

Logische Ausdruecke:
    (5 > 3 ? ja : nein)   → ja
    (x == 1 ? eins :: x == 2 ? zwei : andere)

String-Operationen:
    <upper>text</upper>      → TEXT
    <lower>TEXT</lower>      → text
    <trim>  text  </trim>    → text
    <len>hello</len>         → 5
    <repeat=3>ab</repeat>    → ababab
    <substr=1,3>hello</substr> → ell
    <strpos=ll>hello</strpos>  → 2
    <replace=alt,neu>alt</replace> → neu

Zahlenformatierung:
    <number>1234567</number>     → 1.234.567
    <int>3.7</int>               → 4
    <decimal=2>3.1</decimal>     → 3.10
    <size>1073741824</size>      → 1 GB
"""

from __future__ import annotations

import locale
import os
import platform
import re
import socket
from datetime import datetime, timedelta
from typing import Any


# Wochentage und Monate auf Deutsch
_WEEKDAY_NAMES = [
    "Montag", "Dienstag", "Mittwoch", "Donnerstag",
    "Freitag", "Samstag", "Sonntag",
]

_MONTH_NAMES = [
    "", "Januar", "Februar", "Maerz", "April", "Mai", "Juni",
    "Juli", "August", "September", "Oktober", "November", "Dezember",
]

# Pattern fuer {placeholder} (keine verschachtelten Klammern)
_PLACEHOLDER_RE = re.compile(r"\{([^{}]+)\}")

# Pattern fuer (:math-expression)
_MATH_INLINE_RE = re.compile(r"\(:([^)]+)\)")

# Pattern fuer <calc>...</calc> und <math>...</math>
_MATH_TAG_RE = re.compile(r"<(calc|math)>(.*?)</\1>", re.DOTALL | re.IGNORECASE)

# Pattern fuer (condition ? true : false) — muss ? enthalten
_LOGIC_RE = re.compile(r"\(([^()]*?\?[^()]*?)\)")

# String-Tags
_UPPER_RE = re.compile(r"<upper>(.*?)</upper>", re.DOTALL | re.IGNORECASE)
_LOWER_RE = re.compile(r"<lower>(.*?)</lower>", re.DOTALL | re.IGNORECASE)
_TRIM_RE = re.compile(r"<trim>(.*?)</trim>", re.DOTALL | re.IGNORECASE)
_LEN_RE = re.compile(r"<len>(.*?)</len>", re.DOTALL | re.IGNORECASE)
_REPEAT_RE = re.compile(r"<repeat=(\d+)>(.*?)</repeat>", re.DOTALL | re.IGNORECASE)
_SUBSTR_RE = re.compile(
    r"<substr=(-?\d+)(?:,(-?\d+))?>(.*?)</substr>", re.DOTALL | re.IGNORECASE
)
_STRPOS_RE = re.compile(r"<strpos=(.+?)>(.*?)</strpos>", re.DOTALL | re.IGNORECASE)
_REPLACE_RE = re.compile(r"<replace=(.+?)>(.*?)</replace>", re.DOTALL | re.IGNORECASE)

# Zahlen-Tags
_NUMBER_RE = re.compile(r"<number(?:=([^>]*))?>(.*?)</number>", re.DOTALL | re.IGNORECASE)
_INT_RE = re.compile(r"<int(?:=([^>]*))?>(.*?)</int>", re.DOTALL | re.IGNORECASE)
_DECIMAL_RE = re.compile(
    r"<decimal(?:=([^>]*))?>(.*?)</decimal>", re.DOTALL | re.IGNORECASE
)
_SIZE_RE = re.compile(r"<size>(.*?)</size>", re.DOTALL | re.IGNORECASE)


# =====================================================================
# Haupt-API
# =====================================================================

def format_template(
    template: str,
    context: dict[str, Any] | None = None,
    now: datetime | None = None,
    application: Any | None = None,
) -> str:
    """
    Verarbeitet ein Template mit der vollstaendigen Pipeline.

    Args:
        template: Text mit Platzhaltern und Ausdruecken.
        context: Optionaler Kontext fuer {ctx.*}-Variablen.
        now: Optionaler Zeitpunkt (default: datetime.now()).
        application: Optionale Referenz zur Applikation fuer {app.*}.

    Returns:
        Der formatierte Text.
    """
    if not template:
        return template

    dt_now = now or datetime.now()
    ctx = context or {}

    text = template

    # Phase 1: {placeholder}-Variablen aufloesen
    text = _resolve_placeholders(text, dt_now, ctx, application)

    # Phase 2+3: Math und Logik iterativ aufloesen (max 32 Durchlaeufe)
    for _ in range(32):
        prev = text
        text = _process_math(text)
        text = _process_logic(text)
        if text == prev:
            break

    # Phase 4: String-Operationen
    text = _process_string_ops(text)

    # Phase 5: Zahlenformatierung
    text = _process_number_format(text)

    return text


# =====================================================================
# Phase 1: Platzhalter-Auflosung
# =====================================================================

def _resolve_placeholders(
    text: str, now: datetime, ctx: dict[str, Any], app: Any | None
) -> str:
    """Ersetzt alle {placeholder}-Variablen."""
    if "{" not in text:
        return text

    def _replace(match: re.Match) -> str:
        key = match.group(1).strip()
        try:
            value = _resolve(key, now, ctx, app)
            if value is not None:
                return str(value)
        except Exception:
            pass
        return match.group(0)

    return _PLACEHOLDER_RE.sub(_replace, text)


def _resolve(key: str, now: datetime, ctx: dict[str, Any], app: Any | None) -> Any:
    """Loest einen einzelnen Platzhalter-Schluessel auf."""
    parts = key.split(".")

    if len(parts) < 2:
        return None

    namespace = parts[0].lower()

    if namespace == "date":
        return _resolve_date(parts[1:], now)
    if namespace == "sys":
        return _resolve_sys(parts[1:])
    if namespace == "env":
        return _resolve_env(parts[1:])
    if namespace == "ctx":
        return _resolve_ctx(parts[1:], ctx)
    if namespace == "app" and app is not None:
        return _resolve_app(parts[1:], app)

    return None


# --- Datum/Uhrzeit ---

def _resolve_date(parts: list[str], now: datetime) -> Any:
    """Loest date.*-Platzhalter auf."""
    if not parts:
        return None

    base_key = parts[0].lower()
    if base_key == "now":
        dt = now
    elif base_key == "tomorrow":
        dt = now + timedelta(days=1)
    elif base_key == "yesterday":
        dt = now - timedelta(days=1)
    else:
        return None

    if len(parts) < 2:
        return dt.strftime("%Y-%m-%d %H:%M:%S")

    field = parts[1].lower()
    return _date_field(dt, field)


def _date_field(dt: datetime, field: str) -> Any:
    """Gibt ein einzelnes Datumsfeld zurueck."""
    mapping = {
        "year": dt.year,
        "month": dt.month,
        "month_name": _MONTH_NAMES[dt.month],
        "day": dt.day,
        "hour": dt.hour,
        "minute": dt.minute,
        "second": dt.second,
        "weekday": _WEEKDAY_NAMES[dt.weekday()],
        "weekday_num": dt.isoweekday(),
        "date": dt.strftime("%Y-%m-%d"),
        "time": dt.strftime("%H:%M:%S"),
        "iso": dt.isoformat(),
        "timestamp": int(dt.timestamp()),
        "week": dt.isocalendar()[1],
    }
    return mapping.get(field)


# --- System ---

def _resolve_sys(parts: list[str]) -> Any:
    """Loest sys.*-Platzhalter auf."""
    if not parts:
        return None

    key = ".".join(parts).lower()
    mapping = {
        "hostname": socket.gethostname,
        "os.name": lambda: platform.system(),
        "os.version": lambda: platform.version(),
        "os.release": lambda: platform.release(),
        "os.platform": lambda: platform.platform(),
        "os.arch": lambda: platform.machine(),
        "python.version": lambda: platform.python_version(),
        "user": lambda: os.getenv("USER") or os.getenv("USERNAME") or "unknown",
        "pid": lambda: os.getpid(),
        "cpu_count": lambda: os.cpu_count(),
    }

    resolver = mapping.get(key)
    if resolver:
        return resolver()
    return None


# --- Umgebungsvariablen ---

def _resolve_env(parts: list[str]) -> Any:
    """Loest env.*-Platzhalter auf."""
    if not parts:
        return None
    var_name = ".".join(parts)
    return os.getenv(var_name, "")


# --- Kontext ---

def _resolve_ctx(parts: list[str], ctx: dict[str, Any]) -> Any:
    """Loest ctx.*-Platzhalter auf (verschachtelte Dict-Keys)."""
    if not parts or not ctx:
        return None

    current: Any = ctx
    for part in parts:
        if isinstance(current, dict):
            current = current.get(part)
        else:
            return None
        if current is None:
            return None
    return current


# --- Applikation (Magische Klassenkonstanten) ---

def _resolve_app(parts: list[str], app: Any) -> Any:
    """
    Loest app.*-Platzhalter gegen die Applikation auf.

    Navigiert ueber Properties und Attribute des Applikationsobjekts.
    Beispiele:
        {app.satellites.count}          → Anzahl registrierter Satellites
        {app.satellites.connected}      → Anzahl verbundener Satellites
        {app.scheduler.job_count}       → Anzahl Scheduler-Jobs
        {app.scheduler.running}         → Ob Scheduler laeuft
        {app.plugins.count}             → Anzahl geladener Plugins
        {app.config.network.command_port} → Config-Wert
        {app.uptime}                    → Laufzeit in Sekunden
    """
    if not parts:
        return None

    current: Any = app

    for part in parts:
        if current is None:
            return None

        # Spezielle Zugriffe fuer bekannte Trixy-Typen
        result = _resolve_app_special(current, part)
        if result is not _UNRESOLVED:
            current = result
            continue

        # Generischer Attribut-Zugriff
        if hasattr(current, part):
            current = getattr(current, part)
        elif hasattr(current, f"_{part}"):
            current = getattr(current, f"_{part}")
        elif isinstance(current, dict):
            current = current.get(part)
        else:
            return None

    # Callables nicht zurueckgeben (Properties sind ok, Methoden nicht)
    if callable(current) and not isinstance(current, (int, float, str, bool, list, dict)):
        try:
            current = current()
        except Exception:
            return None

    return current


# Sentinel fuer nicht-aufgeloeste Spezialzugriffe
_UNRESOLVED = object()


def _resolve_app_special(obj: Any, key: str) -> Any:
    """
    Behandelt spezielle Zugriffe auf bekannte Trixy-Objekte.

    Gibt _UNRESOLVED zurueck wenn nicht anwendbar.
    """
    cls_name = type(obj).__name__

    # SatelliteManager
    if cls_name == "SatelliteManager":
        if key == "count":
            return len(obj) if hasattr(obj, "__len__") else 0
        if key == "connected":
            return len(obj.get_connected()) if hasattr(obj, "get_connected") else 0
        if key == "disconnected":
            return len(obj.get_disconnected()) if hasattr(obj, "get_disconnected") else 0
        # Zugriff auf einzelnen Satellite per ID
        if hasattr(obj, "get"):
            sat = obj.get(key)
            if sat is not None:
                return sat

    # Satellite
    if cls_name == "Satellite":
        if key == "connected":
            return obj.is_connected if hasattr(obj, "is_connected") else False
        if key == "room":
            return obj.room_id if hasattr(obj, "room_id") else ""
        if key == "alias":
            return obj.alias if hasattr(obj, "alias") else ""

    # SchedulerService / Scheduler
    if cls_name in ("SchedulerService", "Scheduler"):
        if key == "running":
            return obj.is_running if hasattr(obj, "is_running") else False

    # ServiceContainer
    if cls_name == "ServiceContainer":
        if key == "count":
            return len(obj.services) if hasattr(obj, "services") else 0
        # Zugriff auf einzelnen Service
        if hasattr(obj, "get_service"):
            svc = obj.get_service(key)
            if svc is not None:
                return svc

    # PluginManager
    if cls_name == "PluginManager":
        if key == "count":
            return obj.plugin_count if hasattr(obj, "plugin_count") else 0
        if key == "loaded":
            return obj.loaded_plugins if hasattr(obj, "loaded_plugins") else []

    # ConfigManager — config-Werte auslesen
    if cls_name == "ConfigManager":
        return _UNRESOLVED  # Standard-Attributzugriff nutzen

    return _UNRESOLVED


# =====================================================================
# Phase 2: Mathematische Ausdruecke
# =====================================================================

def _process_math(text: str) -> str:
    """Wertet alle (:expr) und <calc>/<math> Ausdruecke aus."""
    from trixy_core.utils.math_eval import eval_math

    # (:expression)
    def _inline_replace(match: re.Match) -> str:
        result = eval_math(match.group(1))
        return result if result is not None else match.group(0)

    text = _MATH_INLINE_RE.sub(_inline_replace, text)

    # <calc>expression</calc> / <math>expression</math>
    def _tag_replace(match: re.Match) -> str:
        result = eval_math(match.group(2))
        return result if result is not None else match.group(0)

    text = _MATH_TAG_RE.sub(_tag_replace, text)

    return text


# =====================================================================
# Phase 3: Logische Ausdruecke
# =====================================================================

def _process_logic(text: str) -> str:
    """Wertet alle (cond ? true : false) Ausdruecke aus."""
    from trixy_core.utils.logic_eval import eval_logic

    def _replace(match: re.Match) -> str:
        result = eval_logic(match.group(1))
        return result if result is not None else match.group(0)

    return _LOGIC_RE.sub(_replace, text)


# =====================================================================
# Phase 4: String-Operationen
# =====================================================================

def _process_string_ops(text: str) -> str:
    """Verarbeitet alle String-Tags."""
    # Reihenfolge wie in PrintText: substr/len/strpos zuerst, dann extra

    # <substr=START,LENGTH>text</substr>
    def _substr(match: re.Match) -> str:
        start = int(match.group(1))
        length = int(match.group(2)) if match.group(2) else None
        content = match.group(3)
        if length is not None:
            return content[start:start + length]
        return content[start:]

    text = _SUBSTR_RE.sub(_substr, text)

    # <len>text</len>
    text = _LEN_RE.sub(lambda m: str(len(m.group(1))), text)

    # <strpos=search>text</strpos>
    def _strpos(match: re.Match) -> str:
        needle = match.group(1)
        haystack = match.group(2)
        return str(haystack.find(needle))

    text = _STRPOS_RE.sub(_strpos, text)

    # <upper>text</upper>
    text = _UPPER_RE.sub(lambda m: m.group(1).upper(), text)

    # <lower>text</lower>
    text = _LOWER_RE.sub(lambda m: m.group(1).lower(), text)

    # <trim>text</trim>
    text = _TRIM_RE.sub(lambda m: m.group(1).strip(), text)

    # <repeat=N>text</repeat>
    def _repeat(match: re.Match) -> str:
        n = min(int(match.group(1)), 1000)
        return match.group(2) * n

    text = _REPEAT_RE.sub(_repeat, text)

    # <replace=old,new>text</replace>
    def _replace_tag(match: re.Match) -> str:
        params = match.group(1)
        content = match.group(2)
        # Komma splitten, aber \, escapen
        parts = re.split(r"(?<!\\),", params, maxsplit=1)
        if len(parts) < 2:
            return content
        old = parts[0].replace("\\,", ",")
        new = parts[1].replace("\\,", ",")
        return content.replace(old, new)

    text = _REPLACE_RE.sub(_replace_tag, text)

    return text


# =====================================================================
# Phase 5: Zahlenformatierung
# =====================================================================

def _process_number_format(text: str) -> str:
    """Verarbeitet Zahlen-Tags."""

    # <number=FORMAT>value</number>
    def _fmt_number(match: re.Match) -> str:
        fmt = match.group(1) or ""
        raw = match.group(2).strip()
        try:
            num = float(raw)
        except ValueError:
            return raw

        if not fmt:
            # Standard: Tausender-Trennung mit Punkt
            if num == int(num):
                return f"{int(num):,}".replace(",", ".")
            return f"{num:,.2f}".replace(",", "X").replace(".", ",").replace("X", ".")

        # N0-N9: Anzahl Dezimalstellen mit Tausender-Trennung
        m = re.match(r"[Nn](\d)", fmt)
        if m:
            decimals = int(m.group(1))
            formatted = f"{num:,.{decimals}f}"
            return formatted.replace(",", "X").replace(".", ",").replace("X", ".")

        return raw

    text = _NUMBER_RE.sub(_fmt_number, text)

    # <int>value</int>
    def _fmt_int(match: re.Match) -> str:
        raw = match.group(2).strip()
        try:
            return str(round(float(raw)))
        except ValueError:
            return raw

    text = _INT_RE.sub(_fmt_int, text)

    # <decimal=N>value</decimal>
    def _fmt_decimal(match: re.Match) -> str:
        fmt = match.group(1) or "2"
        raw = match.group(2).strip()
        try:
            num = float(raw)
            decimals = int(fmt) if fmt.isdigit() else 2
            return f"{num:.{decimals}f}"
        except ValueError:
            return raw

    text = _DECIMAL_RE.sub(_fmt_decimal, text)

    # <size>bytes</size> → menschenlesbare Groesse
    def _fmt_size(match: re.Match) -> str:
        raw = match.group(1).strip()
        try:
            size = float(raw)
        except ValueError:
            return raw
        return _format_byte_size(size)

    text = _SIZE_RE.sub(_fmt_size, text)

    return text


def _format_byte_size(size: float) -> str:
    """Formatiert Bytes als menschenlesbare Groesse."""
    units = ["Bytes", "KB", "MB", "GB", "TB", "PB", "EB"]

    unit_index = 0
    value = abs(size)

    while value >= 1000 and unit_index < len(units) - 1:
        value /= 1024
        unit_index += 1

    sign = "-" if size < 0 else ""

    if unit_index == 0:
        return f"{sign}{int(value)} {units[0]}"

    if value < 10:
        return f"{sign}{value:.2f} {units[unit_index]}"
    if value < 100:
        return f"{sign}{value:.1f} {units[unit_index]}"
    return f"{sign}{value:.0f} {units[unit_index]}"