Files
site/WESP_REL/app/lab/etl/agrostar_pdf_import.py
T
2026-07-17 12:57:18 +03:00

89 lines
2.8 KiB
Python

"""AgroStar PDF (печатная форма анализа) → AgrostarParseResult."""
from __future__ import annotations
import re
from typing import Any
from app.lab.etl.agrostar_tabular import (
parse_tabular_triplet_lines,
rows_to_label_values,
sample_from_agro_fields,
tabular_rows_to_agro_fields,
)
from app.lab.etl.agrostar_xml_import import AgrostarParseResult
def _pdf_text(data: bytes) -> str:
try:
import fitz # pymupdf
except ImportError as exc:
raise RuntimeError("Для PDF нужен pymupdf (pip install pymupdf)") from exc
doc = fitz.open(stream=data, filetype="pdf")
parts = [doc[i].get_text() for i in range(doc.page_count)]
doc.close()
return "\n".join(parts)
def _line_after(lines: list[str], marker: str) -> str:
for i, line in enumerate(lines):
if line == marker and i + 1 < len(lines):
return lines[i + 1]
return ""
def _is_agrostar_pdf(text: str) -> bool:
t = text or ""
return "ИНФОРМАЦИЯ ОБ ОБРАЗЦЕ" in t or "РЕЗУЛЬТАТЫ АНАЛИЗА" in t
def parse_agrostar_pdf(data: bytes) -> AgrostarParseResult:
result = AgrostarParseResult(lab_name="АгроСтар")
try:
text = _pdf_text(data)
except RuntimeError as exc:
result.errors.append(str(exc))
return result
if not _is_agrostar_pdf(text):
result.errors.append("Не похоже на отчёт AgroStar (PDF)")
return result
lines = [ln.strip() for ln in text.splitlines() if ln.strip()]
sample_no = _line_after(lines, "Образец №")
farm_name = _line_after(lines, "Клиент")
date_printed = _line_after(lines, "Дата анализа")
desc = _line_after(lines, "Описание")
label_parts: list[str] = []
if "Образец" in lines:
idx = lines.index("Образец")
if idx + 1 < len(lines) and lines[idx + 1] != "Описание":
label_parts.append(lines[idx + 1])
if idx + 2 < len(lines) and lines[idx + 2] not in ("Описание", "Клиент"):
label_parts.append(lines[idx + 2])
triplets = parse_tabular_triplet_lines(lines)
agro_fields = tabular_rows_to_agro_fields(rows_to_label_values(triplets))
if not agro_fields:
result.errors.append("В PDF не найдены показатели анализа")
return result
sample = sample_from_agro_fields(
sample_no=sample_no,
farm_name=farm_name,
date_printed=date_printed,
desc_1=desc or " ".join(label_parts),
agro_fields=agro_fields,
)
result.samples.append(sample)
return result
def looks_like_agrostar_pdf(data: bytes) -> bool:
try:
text = _pdf_text(data)
except RuntimeError:
return False
return _is_agrostar_pdf(text)