jarvis-ai-assistant 0.7.16__py3-none-any.whl → 1.0.2__py3-none-any.whl

jarvis/jarvis_sec/checkers/rust_checker.py

@@ -19,11 +19,12 @@ from __future__ import annotations
 
 import re
 from pathlib import Path
-from typing import List, Sequence, Tuple
+from typing import List
+from typing import Sequence
+from typing import Tuple
 
 from ..types import Issue
 
-
 # ---------------------------
 # 规则库（正则表达式）
 # ---------------------------
@@ -38,11 +39,16 @@ RE_ASSUME_INIT = re.compile(r"\bassume_init\s*\(")
 RE_UNWRAP = re.compile(r"\bunwrap\s*\(", re.IGNORECASE)
 RE_EXPECT = re.compile(r"\bexpect\s*\(", re.IGNORECASE)
 RE_EXTERN_C = re.compile(r'extern\s+"C"')
-RE_UNSAFE_IMPL = re.compile(r"\bunsafe\s+impl\s+(?:Send|Sync)\b|\bimpl\s+unsafe\s+(?:Send|Sync)\b", re.IGNORECASE)
+RE_UNSAFE_IMPL = re.compile(
+    r"\bunsafe\s+impl\s+(?:Send|Sync)\b|\bimpl\s+unsafe\s+(?:Send|Sync)\b",
+    re.IGNORECASE,
+)
 
 # 结果忽略/下划线绑定（可能忽略错误）
 RE_LET_UNDERSCORE = re.compile(r"\blet\s+_+\s*=\s*.+;")
-RE_MATCH_IGNORE_ERR = re.compile(r"\.ok\s*\(\s*\)|\.ok\?\s*;|\._?\s*=\s*.+\.err\(\s*\)", re.IGNORECASE)  # 粗略
+RE_MATCH_IGNORE_ERR = re.compile(
+    r"\.ok\s*\(\s*\)|\.ok\?\s*;|\._?\s*=\s*.+\.err\(\s*\)", re.IGNORECASE
+)  # 粗略
 
 # 类型转换相关
 RE_AS_CAST = re.compile(r"\b\w+\s+as\s+[A-Za-z_]\w*", re.IGNORECASE)
@@ -90,6 +96,7 @@ RE_ZEROED = re.compile(r"\bzeroed\s*\(")
 # 公共工具
 # ---------------------------
 
+
 def _safe_line(lines: Sequence[str], idx: int) -> str:
     if 1 <= idx <= len(lines):
         return lines[idx - 1]
@@ -101,7 +108,9 @@ def _strip_line(s: str, max_len: int = 200) -> str:
     return s if len(s) <= max_len else s[: max_len - 3] + "..."
 
 
-def _window(lines: Sequence[str], center: int, before: int = 3, after: int = 3) -> List[Tuple[int, str]]:
+def _window(
+    lines: Sequence[str], center: int, before: int = 3, after: int = 3
+) -> List[Tuple[int, str]]:
     start = max(1, center - before)
     end = min(len(lines), center + after)
     return [(i, _safe_line(lines, i)) for i in range(start, end + 1)]
@@ -118,9 +127,9 @@ def _remove_comments_preserve_strings(text: str) -> str:
     n = len(text)
     in_sl_comment = False  # //
     in_bl_comment = False  # /* */
-    in_string = False      # "
-    in_char = False        # '
-    in_raw_string = False   # r"..." 或 r#"..."#
+    in_string = False  # "
+    in_char = False  # '
+    in_raw_string = False  # r"..." 或 r#"..."#
     raw_string_hash_count = 0  # 原始字符串的 # 数量
     escape = False
 
@@ -159,7 +168,7 @@ def _remove_comments_preserve_strings(text: str) -> str:
                 # 检查是否有足够的 # 来结束原始字符串
                 hash_count = 0
                 j = i - 1
-                while j >= 0 and text[j] == '#':
+                while j >= 0 and text[j] == "#":
                     hash_count += 1
                     j -= 1
                 if hash_count == raw_string_hash_count:
@@ -238,7 +247,7 @@ def _remove_comments_preserve_strings(text: str) -> str:
             # 计算 # 的数量
             raw_string_hash_count = 1
             j = i + 1
-            while j < n and text[j] == '#':
+            while j < n and text[j] == "#":
                 raw_string_hash_count += 1
                 j += 1
             if j < n and text[j] == '"':
@@ -252,7 +261,7 @@ def _remove_comments_preserve_strings(text: str) -> str:
             # 字节原始字符串：br#"
             raw_string_hash_count = 1
             j = i + 2
-            while j < n and text[j] == '#':
+            while j < n and text[j] == "#":
                 raw_string_hash_count += 1
                 j += 1
             if j < n and text[j] == '"':
@@ -303,21 +312,21 @@ def _mask_strings_preserve_len(text: str) -> str:
     in_raw_string = False
     raw_string_hash_count = 0
     escape = False
-    
+
     i = 0
     n = len(text)
-    
+
     while i < n:
         ch = text[i]
         nxt = text[i + 1] if i + 1 < n else ""
         nxt2 = text[i + 2] if i + 2 < n else ""
-        
+
         if in_raw_string:
             if ch == '"':
                 # 检查是否有足够的 # 来结束原始字符串
                 hash_count = 0
                 j = i - 1
-                while j >= 0 and text[j] == '#':
+                while j >= 0 and text[j] == "#":
                     hash_count += 1
                     j -= 1
                 if hash_count == raw_string_hash_count:
@@ -338,7 +347,7 @@ def _mask_strings_preserve_len(text: str) -> str:
                     res.append(" ")
                 i += 1
             continue
-        
+
         if in_string:
             if escape:
                 # 保留转义反斜杠为两字符（反斜杠+空格），以不破坏列对齐过多
@@ -356,7 +365,7 @@ def _mask_strings_preserve_len(text: str) -> str:
                 res.append(" ")
             i += 1
             continue
-            
+
         if in_char:
             if escape:
                 res.append(" ")
@@ -373,7 +382,7 @@ def _mask_strings_preserve_len(text: str) -> str:
                 res.append(" ")
             i += 1
             continue
-        
+
         # 检测原始字符串开始：r"..." 或 r#"..."# 或 br"..." 或 b"..."（字节字符串）
         if ch == "r" and nxt == '"':
             # 简单原始字符串：r"
@@ -396,7 +405,7 @@ def _mask_strings_preserve_len(text: str) -> str:
             # 带 # 的原始字符串：r#"
             raw_string_hash_count = 1
             j = i + 1
-            while j < n and text[j] == '#':
+            while j < n and text[j] == "#":
                 raw_string_hash_count += 1
                 j += 1
             if j < n and text[j] == '"':
@@ -410,7 +419,7 @@ def _mask_strings_preserve_len(text: str) -> str:
             # 字节原始字符串：br#"
             raw_string_hash_count = 1
             j = i + 2
-            while j < n and text[j] == '#':
+            while j < n and text[j] == "#":
                 raw_string_hash_count += 1
                 j += 1
             if j < n and text[j] == '"':
@@ -422,7 +431,7 @@ def _mask_strings_preserve_len(text: str) -> str:
                     res.append(text[k])
                 i = j + 1
                 continue
-        
+
         # 处理 b"..."（字节字符串，不是原始字符串，需要在原始字符串检测之后）
         if ch == "b" and nxt == '"' and nxt2 != "r":
             in_string = True
@@ -440,14 +449,16 @@ def _mask_strings_preserve_len(text: str) -> str:
             res.append("'")
             i += 1
             continue
-        
+
         res.append(ch)
         i += 1
-    
+
     return "".join(res)
 
 
-def _has_safety_comment_around(lines: Sequence[str], line_no: int, radius: int = 5) -> bool:
+def _has_safety_comment_around(
+    lines: Sequence[str], line_no: int, radius: int = 5
+) -> bool:
     """
     Rust 社区约定在 unsafe 附近写 SAFETY: 注释说明前置条件。
     如存在，适当降低置信度。
@@ -489,6 +500,7 @@ def _severity_from_confidence(conf: float) -> str:
 # 规则实现
 # ---------------------------
 
+
 def _rule_unsafe(lines: Sequence[str], relpath: str) -> List[Issue]:
     issues: List[Issue] = []
     for idx, s in enumerate(lines, start=1):
@@ -809,13 +821,18 @@ def _rule_unsafe_mem_ops(lines: Sequence[str], relpath: str) -> List[Issue]:
     """
     issues: List[Issue] = []
     for idx, s in enumerate(lines, start=1):
-        if not (RE_COPY_NONOVERLAPPING.search(s) or RE_COPY.search(s) or RE_WRITE.search(s) or RE_READ.search(s)):
+        if not (
+            RE_COPY_NONOVERLAPPING.search(s)
+            or RE_COPY.search(s)
+            or RE_WRITE.search(s)
+            or RE_READ.search(s)
+        ):
             continue
         # 检查是否在 unsafe 块中
         window_text = " ".join(t for _, t in _window(lines, idx, before=5, after=5))
         if "unsafe" not in window_text.lower():
             continue  # 这些函数必须在 unsafe 块中使用
-        
+
         conf = 0.8
         if _has_safety_comment_around(lines, idx):
             conf -= 0.1
@@ -938,7 +955,7 @@ def _rule_refcell_borrow(lines: Sequence[str], relpath: str) -> List[Issue]:
     """
     issues: List[Issue] = []
     refcell_vars: set[str] = set()
-    
+
     # 收集 RefCell 变量
     for idx, s in enumerate(lines, start=1):
         if RE_REFCELL.search(s):
@@ -946,7 +963,7 @@ def _rule_refcell_borrow(lines: Sequence[str], relpath: str) -> List[Issue]:
             m = re.search(r"\bRefCell\s*<[^>]+>\s*([A-Za-z_]\w*)", s, re.IGNORECASE)
             if m:
                 refcell_vars.add(m.group(1))
-    
+
     # 检测 borrow/borrow_mut 的使用
     for idx, s in enumerate(lines, start=1):
         for var in refcell_vars:
@@ -1043,6 +1060,7 @@ def _rule_uninit_zeroed(lines: Sequence[str], relpath: str) -> List[Issue]:
 # 对外主入口
 # ---------------------------
 
+
 def analyze_rust_text(relpath: str, text: str) -> List[Issue]:
     """
     基于提供的文本进行 Rust 启发式分析。
@@ -1105,4 +1123,4 @@ def analyze_rust_files(base_path: str, relative_paths: List[str]) -> List[Issue]
         p = Path(f)
         if p.suffix.lower() == ".rs":
             out.extend(analyze_rust_file(base, p))
-    return out
+    return out

jarvis/jarvis_sec/cli.py

@@ -20,10 +20,18 @@ from pathlib import Path
 from typing import Optional
 
 import typer
-from jarvis.jarvis_utils.utils import init_env
+
+from jarvis.jarvis_sec.report import aggregate_issues
+from jarvis.jarvis_sec.report import format_csv_report
+from jarvis.jarvis_utils.output import PrettyOutput
+
 # removed: set_config import（避免全局覆盖模型组配置）
-from jarvis.jarvis_sec.workflow import run_with_agent, direct_scan, format_markdown_report as format_markdown_report_workflow
-from jarvis.jarvis_sec.report import format_csv_report, aggregate_issues
+from jarvis.jarvis_sec.workflow import direct_scan
+from jarvis.jarvis_sec.workflow import (
+    format_markdown_report as format_markdown_report_workflow,
+)
+from jarvis.jarvis_sec.workflow import run_with_agent
+from jarvis.jarvis_utils.utils import init_env
 
 app = typer.Typer(
     add_completion=False,
@@ -32,27 +40,38 @@ app = typer.Typer(
 )
 
 
-
-
 @app.command("agent", help="Agent模式（单Agent逐条子任务分析）")
 def agent(
-    path: str = typer.Option(".", "--path", "-p", help="待分析的根目录（默认当前目录）"),
-
+    path: str = typer.Option(
+        ".", "--path", "-p", help="待分析的根目录（默认当前目录）"
+    ),
     llm_group: Optional[str] = typer.Option(
-        None, "--llm-group", "-g", help="使用的模型组（仅对本次运行生效，不修改全局配置）"
+        None,
+        "--llm-group",
+        "-g",
+        help="使用的模型组（仅对本次运行生效，不修改全局配置）",
     ),
     output: Optional[str] = typer.Option(
-        "report.md", "--output", "-o", help="最终报告输出路径（默认 ./report.md）。如果后缀为 .csv，则输出 CSV 格式；否则输出 Markdown 格式"
+        "report.md",
+        "--output",
+        "-o",
+        help="最终报告输出路径（默认 ./report.md）。如果后缀为 .csv，则输出 CSV 格式；否则输出 Markdown 格式",
     ),
-
     cluster_limit: int = typer.Option(
-        50, "--cluster-limit", "-c", help="聚类每批最多处理的告警数（按文件分批聚类，默认50）"
+        50,
+        "--cluster-limit",
+        "-c",
+        help="聚类每批最多处理的告警数（按文件分批聚类，默认50）",
     ),
     enable_verification: bool = typer.Option(
-        True, "--enable-verification/--no-verification", help="是否启用二次验证（默认开启）"
+        True,
+        "--enable-verification/--no-verification",
+        help="是否启用二次验证（默认开启）",
     ),
     force_save_memory: bool = typer.Option(
-        False, "--force-save-memory/--no-force-save-memory", help="强制使用记忆（默认关闭）"
+        False,
+        "--force-save-memory/--no-force-save-memory",
+        help="强制使用记忆（默认关闭）",
     ),
 ) -> None:
     # 初始化环境，确保平台/模型等全局配置就绪（避免 NoneType 平台）
@@ -64,7 +83,6 @@ def agent(
 
     # 若指定了模型组：仅对本次运行生效，透传给 Agent；不修改全局配置（无需 set_config）
 
-
     text: Optional[str] = None
     try:
         text = run_with_agent(
@@ -77,19 +95,23 @@ def agent(
         )
     except Exception as e:
         try:
-            typer.secho(f"[jarvis_sec] Agent 分析过程出错，将回退到直扫基线（fast）：{e}", fg=typer.colors.YELLOW, err=True)
+            PrettyOutput.auto_print(
+                f"⚠️ [jarvis_sec] Agent 分析过程出错，将回退到直扫基线（fast）：{e}"
+            )
         except Exception:
             pass
         text = None
 
     if not text or not str(text).strip():
         try:
-            typer.secho("[jarvis_sec] Agent 无输出，回退到直扫基线（fast）。", fg=typer.colors.YELLOW, err=True)
+            PrettyOutput.auto_print(
+                "⚠️ [jarvis_sec] Agent 无输出，回退到直扫基线（fast）。"
+            )
         except Exception:
             pass
         result = direct_scan(path)
         # 根据输出文件后缀选择格式
-        if output and output.lower().endswith('.csv'):
+        if output and output.lower().endswith(".csv"):
             # 使用 report.py 中的函数来格式化 CSV
             report_json = aggregate_issues(
                 result.get("issues", []),
@@ -119,15 +141,15 @@ def agent(
             p.parent.mkdir(parents=True, exist_ok=True)
             p.write_text(md_text, encoding="utf-8")
             try:
-                typer.secho(f"[jarvis_sec] Markdown 报告已写入: {p}", fg=typer.colors.GREEN)
+                PrettyOutput.auto_print(f"✅ [jarvis_sec] Markdown 报告已写入: {p}")
             except Exception:
                 pass
         except Exception as e:
             try:
-                typer.secho(f"[jarvis_sec] 写入Markdown报告失败: {e}", fg=typer.colors.RED, err=True)
+                PrettyOutput.auto_print(f"❌ [jarvis_sec] 写入Markdown报告失败: {e}")
             except Exception:
                 pass
-    typer.echo(text)
+    PrettyOutput.auto_print(text)
 
 
 def main() -> int:
@@ -136,4 +158,4 @@ def main() -> int:
 
 
 if __name__ == "__main__":
-    sys.exit(main())
+    sys.exit(main())