camelot-py 2.0 (pdfium, без Ghostscript/OpenCV), v1.31

2026-06-17 21:31:37 +04:00
parent e613716178
commit 6187e4dee8
3 changed files with 21 additions and 16 deletions
@@ -124,22 +124,27 @@ def _parse_doc(file_bytes: bytes) -> dict:
 def _parse_pdf(file_bytes: bytes) -> dict:
    result = {"elements": [], "errors": []}

-    # Попробовать tabula (Java) — точнее таблицы
-    tables_from_tabula = False
+    # Попробовать camelot (pdfium, без системных зависимостей)
+    tables_from_camelot = False
    try:
-        import tabula
-        tables = tabula.read_pdf(io.BytesIO(file_bytes), pages="all", multiple_tables=True)
-        for tbl in tables:
-            rows = [list(tbl.columns)] + tbl.values.tolist()
-            result["elements"].append({
-                "type": "table",
-                "rows": rows,
-            })
-        tables_from_tabula = len(tables) > 0
+        import camelot
+        for flavor in ("lattice", "stream"):
+            try:
+                ctables = camelot.read_pdf(io.BytesIO(file_bytes), pages="all", flavor=flavor)
+                for ct in ctables:
+                    rows = [list(ct.df.columns)] + ct.df.values.tolist()
+                    result["elements"].append({
+                        "type": "table",
+                        "rows": rows,
+                        "page": ct.page,
+                    })
+                    tables_from_camelot = True
+            except Exception:
+                pass
    except ImportError:
-        result["errors"].append("tabula not installed, using pdfplumber")
+        result["errors"].append("camelot not installed, using pdfplumber")
    except Exception as e:
-        result["errors"].append(f"tabula error: {e}")
+        result["errors"].append(f"camelot error: {e}")

    # pdfplumber: текст + таблицы (если tabula не сработал)
    try:
@@ -158,7 +163,7 @@ def _parse_pdf(file_bytes: bytes) -> dict:
                                "text": line,
                                "page": pn,
                            })
-                if not tables_from_tabula:
+                if not tables_from_camelot:
                    tables = page.extract_tables()
                    for tbl in tables:
                        if tbl:
@@ -59,7 +59,7 @@
 <body>
  <div class="topbar">
    <img src="{{ url_for('static', filename='nubes-logo.svg') }}" alt="Nubes">
-    <span class="title">Сверка договоров <span style="font-weight:400;color:var(--muted);font-size:12px;">v1.30</span></span>
+    <span class="title">Сверка договоров <span style="font-weight:400;color:var(--muted);font-size:12px;">v1.31</span></span>
  </div>

  <div class="content">