feat: add memory export filter;

core/config.py

@@ -39,6 +39,7 @@ FULL_TEXT_SEGMENT_ID = -1
 ## 规则集ID列表，需与rules.xlsx中的sheet名称保持一致!!!
 ALL_RULESET_IDS = [
     "通用",
+    "通用 (新)",
     "借款",
     "担保",
     "财务口",

core/memory.py

@@ -13,6 +13,7 @@ from utils.http_util import upload_file
 from utils.doc_util import DocBase
 from core.config import META_KEY, FILE_SUFFIX, use_lufa
 from core.finding_preprocessor import FindingPreprocessor
+from core.tools.memory_export_filter import MemoryExportFilter
 
 logger = logging.getLogger(__name__)
 
@@ -84,6 +85,7 @@ class MemoryStore:
         self.merge_facts: List[Dict[str, Any]] = []
         self.findings: Dict[str, List[Finding]] = {}
         self._finding_preprocessor = FindingPreprocessor()
+        self._export_filter = MemoryExportFilter()
         self._load()
 
     # ---------------------- facts ----------------------
@@ -333,7 +335,10 @@ class MemoryStore:
                 ("suggestion", "建议"),
             ]
 
-            grouped_items = list(self.findings.items())
+            grouped_items = [
+                (key, self._export_filter.filter_findings(values))
+                for key, values in self.findings.items()
+            ]
             if grouped_items:
                 first_key, first_values = grouped_items[0]
                 ws_first = wb.active
@@ -456,7 +461,9 @@ class MemoryStore:
         target_key = self._normalize_finding_key(finding_key)
 
         with self._lock:
-            target_findings = list(self._get_findings_bucket(target_key))
+            target_findings = self._export_filter.filter_findings(
+                self._get_findings_bucket(target_key)
+            )
             comments: List[Dict[str, Any]] = []
             for idx, f in enumerate(target_findings, start=1):
                 segment_id = int(f.segment_id or 0)

core/tools/memory_export_filter.py

+from __future__ import annotations
+
+from collections.abc import Iterable, Mapping
+from typing import Any, List, Optional, TypeVar
+
+
+FindingType = TypeVar("FindingType")
+
+
+class MemoryExportFilter:
+    """Filter findings from an export without changing the memory store."""
+
+    DEFAULT_EXCLUDED_TITLES = frozenset({"通用审查"})
+
+    def __init__(self, excluded_titles: Optional[Iterable[str]] = None) -> None:
+        titles = (
+            self.DEFAULT_EXCLUDED_TITLES
+            if excluded_titles is None
+            else excluded_titles
+        )
+        self._excluded_titles = {
+            str(title).strip() for title in titles if str(title).strip()
+        }
+
+    def filter_findings(self, findings: Iterable[FindingType]) -> List[FindingType]:
+        return [finding for finding in findings if self.should_export(finding)]
+
+    def should_export(self, finding: Any) -> bool:
+        rule_title = self._get_rule_title(finding).strip()
+        return rule_title not in self._excluded_titles
+
+    @staticmethod
+    def _get_rule_title(finding: Any) -> str:
+        if isinstance(finding, Mapping):
+            rule_title = finding.get("rule_title") or ""
+        else:
+            rule_title = getattr(finding, "rule_title", "")
+        return str(rule_title)