Add generic production context helpers

src/production_analytics/context/__init__.py

@@ -0,0 +1,62 @@
+"""Deterministic ERP context helpers, independent of database and timeseries access."""
+
+import math
+import re
+
+__all__ = ["build_enlyze_production_order", "extract_nominal_width_m"]
+
+# Capture malformed numeric tokens too, so a valid-looking suffix cannot become width.
+_NUMBER_TOKEN = r"[+-]?(?:[0-9]+(?:[.,][0-9]+)*|inf(?:inity)?|nan)(?:[eE][+-]?[0-9]+)?"
+_DIMENSIONS = re.compile(
+    rf"(?<![\w.,+\-/])(?P<width>{_NUMBER_TOKEN})\s*x\s*"
+    rf"(?P<length>{_NUMBER_TOKEN})\s*m(?![\w/²³^])",
+    re.IGNORECASE,
+)
+_PLAIN_NUMBER = re.compile(r"[0-9]+(?:[.,][0-9]+)?")
+
+
+def build_enlyze_production_order(erp_production_order: str, format_template: str) -> str:
+    """Format a single ERP order using explicit caller-supplied configuration.
+
+    The template must contain exactly one literal {production_order} placeholder
+    and no other braces. Outer order whitespace is stripped; the order
+    must otherwise contain ASCII digits only (leading zeros are preserved).
+    This is not a parser for ENLYZE identifiers, combined or otherwise.
+    """
+    order = erp_production_order.strip()
+    if not order or not order.isascii() or not order.isdigit():
+        raise ValueError("ERP production order must be a non-empty ASCII digit string")
+    placeholder = "{production_order}"
+    literal = format_template.replace(placeholder, "")
+    if format_template.count(placeholder) != 1 or "{" in literal or "}" in literal:
+        raise ValueError("Format template must contain exactly one {production_order} placeholder")
+    return format_template.replace(placeholder, order)
+
+
+def extract_nominal_width_m(article_description: str | None) -> float | None:
+    """Read one unambiguous positive '<width> x <length> m' pair from ERP text.
+
+    Decimal comma/point and variable whitespace are supported. Multiple pairs,
+    dimension chains, signed/scientific notation and invalid dimensions fail
+    closed.
+    """
+    if article_description is None:
+        return None
+    matches = list(_DIMENSIONS.finditer(article_description))
+    if len(matches) != 1:
+        return None
+    match = matches[0]
+    # Do not mistake the tail of a three-dimensional expression for a width pair.
+    if re.search(r"[x×]\s*$", article_description[:match.start()], re.IGNORECASE):
+        return None
+    if re.match(r"\s*[x×]", article_description[match.end():], re.IGNORECASE):
+        return None
+    values = []
+    for token in (match["width"], match["length"]):
+        if _PLAIN_NUMBER.fullmatch(token) is None:
+            return None
+        value = float(token.replace(",", "."))
+        if not math.isfinite(value) or value <= 0:
+            return None
+        values.append(value)
+    return values[0]