camelot для таблиц PDF (lattice+stream), fallback на pdfplumber, v1.27

site/parser.py

@@ -123,7 +123,15 @@ def _parse_doc(file_bytes: bytes) -> dict:
 
 def _parse_pdf(file_bytes: bytes) -> dict:
     result = {"elements": [], "errors": []}
+    import tempfile, os as _os
+
+    # Сохранить PDF во временный файл (camelot требует путь)
+    tmp_path = None
     try:
+        with tempfile.NamedTemporaryFile(suffix=".pdf", delete=False) as tmp:
+            tmp.write(file_bytes)
+            tmp_path = tmp.name
+
         import pdfplumber
 
         with pdfplumber.open(io.BytesIO(file_bytes)) as pdf:
@@ -142,18 +150,48 @@ def _parse_pdf(file_bytes: bytes) -> dict:
                                 "page": pn,
                             })
 
-                # Таблицы на странице
-                tables = page.extract_tables()
-                for tbl in tables:
-                    if tbl:
+        # Таблицы — camelot (точнее), иначе pdfplumber
+        tables_found = False
+        try:
+            import camelot
+            # Сначала lattice (таблицы с рамками)
+            for flavor in ("lattice", "stream"):
+                try:
+                    ctables = camelot.read_pdf(tmp_path, pages="all", flavor=flavor)
+                    for ct in ctables:
+                        rows = [list(ct.df.columns)] + ct.df.values.tolist()
                         result["elements"].append({
                             "type": "table",
-                            "rows": tbl,
-                            "page": pn,
+                            "rows": rows,
+                            "page": ct.page,
                         })
+                        tables_found = True
+                except Exception:
+                    pass
+        except ImportError:
+            result["errors"].append("camelot not installed, using pdfplumber")
+        except Exception as e:
+            result["errors"].append(f"camelot error: {e}")
+
+        # Fallback: pdfplumber таблицы если camelot ничего не нашёл
+        if not tables_found:
+            with pdfplumber.open(io.BytesIO(file_bytes)) as pdf:
+                for page in pdf.pages:
+                    pn = page.page_number
+                    tables = page.extract_tables()
+                    for tbl in tables:
+                        if tbl:
+                            result["elements"].append({
+                                "type": "table",
+                                "rows": tbl,
+                                "page": pn,
+                            })
 
     except Exception as e:
         result["errors"].append(f"pdf parse error: {e}")
+    finally:
+        if tmp_path and _os.path.exists(tmp_path):
+            _os.unlink(tmp_path)
 
     return result