PacketStation/server/packetstation/text.py

"""Input cleaning and limits: the old build's cleanText, cleanDeep and validation regexes, now on the server.

Every string that enters the database goes through `clean_text` with the field's limit. Control characters
(except tab, LF and CR), DEL, the zero-width and bidi override characters are removed, the text is NFC
normalised, trimmed and cut to the limit.
"""

from __future__ import annotations

import re
import unicodedata
from typing import Any

# C0 controls except tab (09), LF (0A) and CR (0D); DEL; zero-width and bidi marks; isolates.
_STRIP = re.compile("[\u0000-\u0008\u000b\u000c\u000e-\u001f\u007f​-‏‪-‮⁦-⁩]")

RX_HEAT = re.compile(r"^[A-Za-z0-9][A-Za-z0-9 ./#_-]{0,39}$")
RX_STAMP = re.compile(r"^[A-Za-z0-9-]{1,10}$")
RX_EMP = re.compile(r"^[A-Z0-9-]{1,20}$")
RX_PROJECT = re.compile(r"^[A-Z0-9][A-Z0-9-]{0,19}$")
RX_LOT = re.compile(r"^[A-Z0-9]{1,3}$")

# Character limits per field. Reasons for IT edits and send-backs need at least REASON_MIN.
LIMITS: dict[str, int] = {
    "title": 200,
    "customer": 120,
    "drawing": 60,
    "rev": 10,
    "wps": 40,
    "op": 40,
    "notes": 2000,
    "reason": 500,
    "note": 500,
    "heat": 40,
    "stamp": 10,
    "location": 120,
    "part_id": 60,
    "description": 240,
    "material": 120,
    "length": 40,
    "project_number": 20,
    "project_name": 120,
    "po": 60,
    "lot": 3,
    "unit_name": 120,
    "doc_title": 160,
    "ecn": 40,
    "name": 160,
    "order_pin": 40,
    "file_name": 160,
    "job_note": 200,
    "filler": 40,
    "preheat": 40,
    "weld_text": 40,
    "doc_note": 300,
    "qc_doc_title": 140,
    "qc_doc_rev": 12,
    "qc_doc_note": 600,
    "ncr": 40,
    "itp_field": 120,
}
REASON_MIN = 8
MAX_BODY_BYTES = 200_000


def clean_text(value: Any, max_len: int = 2000) -> str:
    """The old cleanText: strip controls and invisible marks, NFC, trim, cut. None becomes ''."""
    if value is None:
        return ""
    text = value if isinstance(value, str) else str(value)
    text = _STRIP.sub("", text)
    text = unicodedata.normalize("NFC", text).strip()
    return text[:max_len]


def clean_opt(value: Any, max_len: int) -> str | None:
    """clean_text, but an empty result is None (for nullable columns)."""
    text = clean_text(value, max_len)
    return text or None


def clean_deep(value: Any, max_len: int = 4000, depth: int = 0) -> Any:
    """Clean every string inside a JSON-like value. Depth is capped at 12 like the old build."""
    if depth > 12:
        return value
    if isinstance(value, str):
        return clean_text(value, max_len)
    if isinstance(value, list):
        return [clean_deep(x, max_len, depth + 1) for x in value]
    if isinstance(value, dict):
        return {str(k)[:120]: clean_deep(v, max_len, depth + 1) for k, v in value.items()}
    return value


def limit(field: str) -> int:
    return LIMITS.get(field, 2000)


def reason_ok(value: Any) -> str | None:
    """A cleaned reason of at least REASON_MIN characters, else None."""
    text = clean_text(value, LIMITS["reason"])
    return text if len(text) >= REASON_MIN else None