feat(file-summary): 实现文件处理技能链路

review_agent/file_summary/skills/__init__.py

@@ -0,0 +1 @@
+

review_agent/file_summary/skills/archive_extract.py

@@ -0,0 +1,26 @@
+from __future__ import annotations
+
+from pathlib import Path
+
+from review_agent.models import FileSummaryBatchAttachment
+
+from ..paths import resolve_storage_path
+from ..services.archive import ARCHIVE_EXTENSIONS, extract_archive
+from .base import BaseSkill, SkillResult, WorkflowContext
+
+
+class ArchiveExtractSkill(BaseSkill):
+    name = "archive_extract"
+
+    def run(self, context: WorkflowContext) -> SkillResult:
+        extracted_count = 0
+        target_dir = Path(context.batch.work_dir or "")
+        if not target_dir:
+            return SkillResult(success=True, data={"extracted_count": 0})
+
+        for binding in FileSummaryBatchAttachment.objects.filter(batch=context.batch):
+            path = resolve_storage_path(binding.attachment.storage_path)
+            if path.suffix.lower().lstrip(".") not in ARCHIVE_EXTENSIONS:
+                continue
+            extracted_count += len(extract_archive(path, target_dir))
+        return SkillResult(success=True, data={"extracted_count": extracted_count})

review_agent/file_summary/skills/base.py

@@ -0,0 +1,24 @@
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+
+from review_agent.models import FileSummaryBatch
+
+
+@dataclass(frozen=True)
+class WorkflowContext:
+    batch: FileSummaryBatch
+
+
+@dataclass
+class SkillResult:
+    success: bool
+    data: dict = field(default_factory=dict)
+    message: str = ""
+
+
+class BaseSkill:
+    name = ""
+
+    def run(self, context: WorkflowContext) -> SkillResult:
+        raise NotImplementedError

review_agent/file_summary/skills/document_page_count.py

@@ -0,0 +1,64 @@
+from __future__ import annotations
+
+from review_agent.models import FileSummaryItem
+
+from ..services.page_count import SUPPORTED_EXTENSIONS, count_document_pages
+from .base import BaseSkill, SkillResult, WorkflowContext
+
+
+class DocumentPageCountSkill(BaseSkill):
+    name = "document_page_count"
+
+    def run(self, context: WorkflowContext) -> SkillResult:
+        success_files = failed_files = unsupported_files = uncertain_files = total_pages = 0
+        for item in context.batch.items.order_by("file_index"):
+            if item.file_type not in SUPPORTED_EXTENSIONS:
+                item.statistics_status = FileSummaryItem.StatisticsStatus.UNSUPPORTED
+                unsupported_files += 1
+                item.save(update_fields=["statistics_status", "updated_at"])
+                continue
+
+            result = None
+            for attempt in range(1, 4):
+                result = count_document_pages(item.storage_path)
+                item.retry_count = attempt - 1
+                if result.status != "failed":
+                    break
+            item.statistics_status = result.status
+            item.page_count = result.page_count
+            item.error_message = result.error_message
+            item.save(
+                update_fields=[
+                    "statistics_status",
+                    "page_count",
+                    "retry_count",
+                    "error_message",
+                    "updated_at",
+                ]
+            )
+
+            if result.status == FileSummaryItem.StatisticsStatus.SUCCESS:
+                success_files += 1
+                total_pages += result.page_count or 0
+            elif result.status == FileSummaryItem.StatisticsStatus.UNCERTAIN:
+                uncertain_files += 1
+            elif result.status == FileSummaryItem.StatisticsStatus.UNSUPPORTED:
+                unsupported_files += 1
+            else:
+                failed_files += 1
+
+        context.batch.success_files = success_files
+        context.batch.failed_files = failed_files
+        context.batch.unsupported_files = unsupported_files
+        context.batch.uncertain_files = uncertain_files
+        context.batch.total_pages = total_pages
+        context.batch.save(
+            update_fields=[
+                "success_files",
+                "failed_files",
+                "unsupported_files",
+                "uncertain_files",
+                "total_pages",
+            ]
+        )
+        return SkillResult(success=True)

review_agent/file_summary/skills/file_inventory.py

@@ -0,0 +1,21 @@
+from __future__ import annotations
+
+from pathlib import Path
+
+from review_agent.models import FileSummaryBatchAttachment
+
+from ..paths import resolve_storage_path
+from ..services.inventory import scan_files_to_items
+from .base import BaseSkill, SkillResult, WorkflowContext
+
+
+class FileInventorySkill(BaseSkill):
+    name = "file_inventory"
+
+    def run(self, context: WorkflowContext) -> SkillResult:
+        roots = [
+            resolve_storage_path(binding.attachment.storage_path)
+            for binding in FileSummaryBatchAttachment.objects.filter(batch=context.batch)
+        ]
+        items = scan_files_to_items(batch=context.batch, roots=roots)
+        return SkillResult(success=True, data={"total_files": len(items)})

review_agent/file_summary/skills/product_detect.py

@@ -0,0 +1,12 @@
+from __future__ import annotations
+
+from ..services.product_detect import detect_product_name
+from .base import BaseSkill, SkillResult, WorkflowContext
+
+
+class ProductDetectSkill(BaseSkill):
+    name = "product_detect"
+
+    def run(self, context: WorkflowContext) -> SkillResult:
+        product_name = detect_product_name(context.batch)
+        return SkillResult(success=True, data={"product_name": product_name})

review_agent/file_summary/skills/registry.py

@@ -0,0 +1,22 @@
+from __future__ import annotations
+
+from .base import BaseSkill, SkillResult, WorkflowContext
+
+
+class SkillRegistry:
+    def __init__(self):
+        self._skills: dict[str, BaseSkill] = {}
+
+    def register(self, skill: BaseSkill) -> None:
+        if not skill.name:
+            raise ValueError("Skill 必须声明 name。")
+        self._skills[skill.name] = skill
+
+    def get(self, name: str) -> BaseSkill:
+        try:
+            return self._skills[name]
+        except KeyError as exc:
+            raise KeyError(f"Skill 未注册：{name}") from exc
+
+    def execute(self, name: str, context: WorkflowContext) -> SkillResult:
+        return self.get(name).run(context)