Source code for energy_data_retrieval.pv_catalogue.inhenergy

"""Offline staging of reviewed INHENERGY SI technical-table columns.

Input is Poppler ``pdftotext -layout`` output from a separately checksum-verified
PDF. Only the reviewed model sequence is accepted. Shared scalar cells may repeat
across models; partially merged cells and wrapped rows require manual review.
No network, model-name rating inference, current summation or canonical upsert.
"""

from __future__ import annotations

import math
import re
import unicodedata


[docs] def read_inhenergy_layout(text: str, *, models: list[str]) -> list[dict[str, object]]: """Extract exact power/current columns, refusing layout or identity drift.""" text = unicodedata.normalize("NFKC", text) if "INHENERGY CO., LTD." not in text or "GRID-CONNECTED INVERTER" not in text: raise ValueError("Not the reviewed INHENERGY technical-table format") if not models or len(set(models)) != len(models): raise ValueError("Reviewed model identities must be nonempty and unique") headers = [line for line in text.splitlines() if "Technical Data" in line] if len(headers) != 1: raise ValueError("Expected exactly one technical table") observed = re.findall(r"SI-\s*\d+(?:\.\d+)?K(?:-[ST]2)?", headers[0]) observed = [re.sub(r"\s+", "", model) for model in observed] if observed != models: raise ValueError("Model sequence differs from reviewed exact identities") def row(label: str) -> str: found = [line.split(label, 1)[1].strip() for line in text.splitlines() if label in line] if len(found) != 1 or not found[0]: raise ValueError(f"Missing, duplicate or wrapped row: {label}") return found[0] def quantities(label: str, unit: str) -> list[float]: value = row(label) pattern = re.compile(rf"(\d+(?:\.\d+)?)\s*({unit})", re.IGNORECASE) matches = list(pattern.finditer(value)) if pattern.sub("", value).strip() or len(matches) not in {1, len(models)}: raise ValueError(f"Unresolved units or partially merged cells: {label}") values = [float(m[1]) * (1000 if m[2].lower() == "kw" else 1) for m in matches] if not all(math.isfinite(v) and v > 0 for v in values): raise ValueError(f"Nonpositive/nonfinite physical quantity: {label}") return values * len(models) if len(values) == 1 else values columns = { "pac_rated_w": quantities("Rated output power", "kW|W"), "output_current_max_a": quantities("Max. output current", "A"), "vdc_max_v": quantities("Max. PV input voltage", "V"), "max_output_power_w": quantities("Max. output power", "kW|W"), "max_pv_access_power_w": quantities("Max. PV access power", "kW|W"), "max_pv_input_power_w": quantities("Max. PV input power", "kW|W"), "rated_output_current_a": quantities("Rated output current", "A"), } shared = { "operating_voltage_as_reported": row("Operation voltage range"), "ac_voltage_as_reported": row("Rated output voltage"), "frequency_as_reported": row("Rated output frequency"), "table_mppt_current_cells_as_reported": row("MPPT max."), "dimensions_as_reported": row("Dimensions (L/W/H) in mm"), "table_weight_cells_as_reported": row("Weight"), "table_cooling_cells_as_reported": row("Heat dissipation mode"), "night_power_as_reported": row("Self-Consumption night"), "communications_as_reported": row("Communication interface"), "warranty_as_reported": row("Warranty"), } return [ {"model": model, **{key: values[i] for key, values in columns.items()}, **shared} for i, model in enumerate(models) ]