feat: v4 multimodal chat input, multi-format support, and annotation detection

- Replace st.chat_input with st-multimodal-chatinput (Ctrl+V paste, drag-drop, file button)
- Extract _process_uploaded_file() shared handler (eliminates ~70 duplicated lines)
- Add XLSX (openpyxl), XLS (xlrd), DOC (olefile) parsers to file_parser.py
- Add backend/annotation_detector.py: circle detection (HoughCircles) + arrow detection (HoughLinesP clustering) + OCR correlation + LLM context formatting
- Add annotation_result field to AgentState with session persistence
- Wire annotation detection into process_input and _format_ocr_context
- Add 11 new tests: 7 annotation detector + 4 multi-format parser
- Update all docs: CLAUDE.md, README.md, CODE_GUIDE.md, ROADMAP.md

backend/file_parser.py

@@ -51,6 +51,9 @@ def parse_file(file_path: str, file_type: str = "") -> dict:
         ".webp": _parse_image,
         ".pdf":  _parse_pdf,
         ".docx": _parse_docx,
+        ".xlsx": _parse_xlsx,
+        ".xls":  _parse_xls,
+        ".doc":  _parse_doc,
     }
 
     parser = parsers.get(suffix)
@@ -72,26 +75,7 @@ def _parse_image(path: Path) -> dict:
     except Exception:
         info = "[图片: 无法读取元数据]"
 
-    # 优先 EasyOCR（Windows 兼容性更好）
-    try:
-        import easyocr
-        import numpy as np
-        reader = easyocr.Reader(["ch_sim", "en"], gpu=False, verbose=False)
-        result = reader.readtext(np.array(img))
-        lines = [text.strip() for (_, text, _) in result if text.strip()]
-        if lines:
-            return {
-                "text": f"{info}\n识别文本:\n" + "\n".join(lines),
-                "file_type": "image",
-                "method": "easyocr",
-                "error": None,
-            }
-    except ImportError:
-        pass
-    except Exception:
-        pass
-
-    # 回退 PaddleOCR
+    # 优先 PaddleOCR（精确识别）
     try:
         from paddleocr import PaddleOCR
         ocr = PaddleOCR(lang="ch")
@@ -114,6 +98,25 @@ def _parse_image(path: Path) -> dict:
     except Exception:
         pass
 
+    # 回退 EasyOCR
+    try:
+        import easyocr
+        import numpy as np
+        reader = easyocr.Reader(["ch_sim", "en"], gpu=False, verbose=False)
+        result = reader.readtext(np.array(img))
+        lines = [text.strip() for (_, text, _) in result if text.strip()]
+        if lines:
+            return {
+                "text": f"{info}\n识别文本:\n" + "\n".join(lines),
+                "file_type": "image",
+                "method": "easyocr",
+                "error": None,
+            }
+    except ImportError:
+        pass
+    except Exception:
+        pass
+
     # OCR 不可用 → 返回图片元信息 + 安装提示
     return {
         "text": f"{info}\n(如需 OCR 文字识别，请安装: pip install easyocr)",
@@ -195,6 +198,91 @@ def _parse_docx(path: Path) -> dict:
             "error": "DOCX 解析需要安装 python-docx"}
 
 
+def _parse_xlsx(path: Path) -> dict:
+    """提取 Excel .xlsx 文件中的文本。"""
+    try:
+        from openpyxl import load_workbook
+        wb = load_workbook(path, read_only=True, data_only=True)
+        parts = []
+        for name in wb.sheetnames:
+            ws = wb[name]
+            rows = []
+            for row in ws.iter_rows(values_only=True):
+                cells = [str(c) if c is not None else "" for c in row]
+                if any(c for c in cells):
+                    rows.append("\t".join(cells))
+            if rows:
+                parts.append(f"[Sheet: {name}]\n" + "\n".join(rows))
+        wb.close()
+        text = "\n\n".join(parts)
+        return {"text": text, "file_type": "xlsx", "method": "openpyxl", "error": None}
+    except ImportError:
+        pass
+    except Exception as e:
+        return {"text": "", "file_type": "xlsx", "method": "none",
+                "error": f"XLSX 解析失败: {e}"}
+    return {"text": "", "file_type": "xlsx", "method": "none",
+            "error": "XLSX 解析需要安装 openpyxl"}
+
+
+def _parse_xls(path: Path) -> dict:
+    """提取旧版 Excel .xls 文件中的文本。"""
+    try:
+        import xlrd
+        wb = xlrd.open_workbook(path)
+        parts = []
+        for name in wb.sheet_names():
+            ws = wb.sheet_by_name(name)
+            rows = []
+            for rx in range(ws.nrows):
+                cells = [str(ws.cell_value(rx, cx)) if ws.cell_value(rx, cx) != "" else ""
+                         for cx in range(ws.ncols)]
+                if any(c for c in cells):
+                    rows.append("\t".join(cells))
+            if rows:
+                parts.append(f"[Sheet: {name}]\n" + "\n".join(rows))
+        text = "\n\n".join(parts)
+        return {"text": text, "file_type": "xls", "method": "xlrd", "error": None}
+    except ImportError:
+        pass
+    except Exception as e:
+        return {"text": "", "file_type": "xls", "method": "none",
+                "error": f"XLS 解析失败: {e}"}
+    return {"text": "", "file_type": "xls", "method": "none",
+            "error": "XLS 解析需要安装 xlrd"}
+
+
+def _parse_doc(path: Path) -> dict:
+    """提取旧版 Word .doc 文件中的文本（尽力而为，二进制格式）。"""
+    try:
+        import olefile
+        ole = olefile.OleFileIO(path)
+        if not ole.exists("WordDocument"):
+            ole.close()
+            return {"text": "", "file_type": "doc", "method": "none",
+                    "error": "不是有效的 .doc 文件"}
+        raw = ole.openstream("WordDocument").read()
+        ole.close()
+        # 提取可打印 UTF-16LE 字符段
+        text = ""
+        try:
+            decoded = raw.decode("utf-16-le", errors="ignore")
+            text = "".join(c for c in decoded if c.isprintable() or c in "\n\r\t")
+        except Exception:
+            pass
+        if not text.strip():
+            return {"text": "", "file_type": "doc", "method": "olefile",
+                    "error": "无法提取文本（.doc 为二进制格式，建议转换为 .docx）"}
+        return {"text": text.strip(), "file_type": "doc", "method": "olefile", "error": None}
+    except ImportError:
+        pass
+    except Exception as e:
+        return {"text": "", "file_type": "doc", "method": "none",
+                "error": f"DOC 解析失败: {e}"}
+    return {"text": "", "file_type": "doc", "method": "none",
+            "error": "DOC 解析需要安装 olefile"}
+
+
 def _parse_text(path: Path) -> dict:
     """读取纯文本文件。"""
     try: