python-hwpx-automation 7.3.2__tar.gz → 7.3.3__tar.gz

{python_hwpx_automation-7.3.2/src/python_hwpx_automation.egg-info → python_hwpx_automation-7.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-hwpx-automation
-Version: 7.3.2
+Version: 7.3.3
 Summary: Task-oriented HWPX document automation for Python, with an optional MCP adapter.
 Author: Kohkyuhyun
 License-Expression: Apache-2.0
@@ -74,7 +74,7 @@ Dynamic: license-file
 
 <!-- release-state: release-approved -->
 > [!NOTE]
-> 발행이 승인된 7.3.2 후보 (release-approved; 원격 발행 대기): `python-hwpx 6.6.0 → python-hwpx-automation 7.3.2 → hwpx-plugin 2.4.0` (`5e5c23651f92785a`).
+> 발행이 승인된 7.3.3 후보 (release-approved; 원격 발행 대기): `python-hwpx 6.6.0 → python-hwpx-automation 7.3.3 → hwpx-plugin 2.4.0` (`5e5c23651f92785a`).
 > 공개 트레인: `python-hwpx 6.6.0 → python-hwpx-automation 7.3.1 → hwpx-plugin 2.4.0`.
 > 공개 좌표는 core·automation PyPI와 plugin GitHub
 > Release·marketplace·실제 marketplace 설치 관찰 후에만 승격합니다.
@@ -162,7 +162,7 @@ PDF와 쪽 PNG로 렌더합니다. 별도 환경에서 실행하므로 호출하
 고정 버전을 건드리지 않습니다.
 
 ```bash
-uvx --from "python-hwpx-automation[oracle]==7.3.2" \
+uvx --from "python-hwpx-automation[oracle]==7.3.3" \
   hwpx render-pdf in.hwpx out.pdf --png out/page --json
 ```
 

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/README.md

@@ -14,7 +14,7 @@
 
 <!-- release-state: release-approved -->
 > [!NOTE]
-> 발행이 승인된 7.3.2 후보 (release-approved; 원격 발행 대기): `python-hwpx 6.6.0 → python-hwpx-automation 7.3.2 → hwpx-plugin 2.4.0` (`5e5c23651f92785a`).
+> 발행이 승인된 7.3.3 후보 (release-approved; 원격 발행 대기): `python-hwpx 6.6.0 → python-hwpx-automation 7.3.3 → hwpx-plugin 2.4.0` (`5e5c23651f92785a`).
 > 공개 트레인: `python-hwpx 6.6.0 → python-hwpx-automation 7.3.1 → hwpx-plugin 2.4.0`.
 > 공개 좌표는 core·automation PyPI와 plugin GitHub
 > Release·marketplace·실제 marketplace 설치 관찰 후에만 승격합니다.
@@ -102,7 +102,7 @@ PDF와 쪽 PNG로 렌더합니다. 별도 환경에서 실행하므로 호출하
 고정 버전을 건드리지 않습니다.
 
 ```bash
-uvx --from "python-hwpx-automation[oracle]==7.3.2" \
+uvx --from "python-hwpx-automation[oracle]==7.3.3" \
   hwpx render-pdf in.hwpx out.pdf --png out/page --json
 ```
 

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "python-hwpx-automation"
-version = "7.3.2"
+version = "7.3.3"
 description = "Task-oriented HWPX document automation for Python, with an optional MCP adapter."
 readme = "README.md"
 authors = [{ name = "Kohkyuhyun" }]

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/core/content.py

@@ -14,8 +14,15 @@ from uuid import uuid4
 from xml.etree import ElementTree as ET
 
 from ..compat import patch_python_hwpx
-from ..storage import build_hwpx_open_safety_report
-from ..upstream import HP_NS as _HP_NS, HwpxDocument, repair_pathological_text_spacing
+from ..storage import build_hwpx_open_safety_report, open_local_document
+from ..upstream import (
+    HP_NS as _HP_NS,
+    HwpxDocument,
+    is_hwp5_source,
+    is_hwp5_target,
+    repair_pathological_text_spacing,
+)
+from .document import save_doc
 from .formatting import resolve_style_id
 from .locations import resolve_paragraph_reference
 
@@ -500,6 +507,15 @@ def copy_document_file(source: str, destination: str = None) -> str:
         destination = f"{stem}_copy.{ext}"
     source_path = Path(source)
     destination_path = Path(destination)
+    if is_hwp5_source(source_path) != is_hwp5_target(destination_path):
+        # The destination extension picks the format, so a copy between .hwp
+        # and .hwpx is written through the document model, not byte-copied.
+        document = open_local_document(source_path, role="copy source")
+        try:
+            save_doc(document, str(destination_path))
+        finally:
+            document.close()
+        return str(destination_path)
     if source_path.suffix.lower() != ".hwpx" and destination_path.suffix.lower() != ".hwpx":
         shutil.copy2(source_path, destination_path)
         return str(destination_path)

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/core/document.py

@@ -5,7 +5,7 @@ from pathlib import Path
 from typing import Any
 
 from ..configuration import env_value
-from ..storage import LocalDocumentStorage, require_hwpx_editor_open_safe
+from ..storage import LocalDocumentStorage, open_local_document
 from ..upstream import HwpxDocument, blank_document_template_bytes, open_document
 
 
@@ -18,8 +18,7 @@ def _local_storage() -> LocalDocumentStorage:
 def open_doc(path: str) -> HwpxDocument:
     if not os.path.exists(path):
         raise FileNotFoundError(f"파일을 찾을 수 없습니다: {path}")
-    require_hwpx_editor_open_safe(Path(path), role="local HWPX open")
-    return open_document(path)
+    return open_local_document(Path(path), role="local HWPX open")
 
 
 def save_doc(doc: HwpxDocument, path: str, *, quality: Any = None) -> dict[str, Any]:

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/core/transactions.py

@@ -12,8 +12,12 @@ from pathlib import Path
 from typing import Any, Iterable
 
 from .. import quality as quality_contract
-from ..storage import build_hwpx_verification_report, require_hwpx_editor_open_safe
-from ..upstream import HwpxDocument, open_document
+from ..storage import (
+    build_hwpx_verification_report,
+    require_hwpx_editor_open_safe,
+    require_native_hwp5,
+)
+from ..upstream import HwpxDocument, hwp5_document_bytes, is_hwp5_target, open_document
 
 _MAX_SUMMARY_ITEMS = 20
 _MAX_SNIPPET = 80
@@ -341,9 +345,13 @@ def save_dry_run(
     document: HwpxDocument, target: str | Path, *, quality: Any = None
 ) -> dict[str, Any]:
     target_path = Path(target)
+    hwp_target = is_hwp5_target(target_path)
+    if hwp_target:
+        require_native_hwp5()
+    # The gate checks HWPX parts, so an .hwp target is staged as HWPX.
     fd, tmp_name = tempfile.mkstemp(
         prefix=f".{target_path.stem}.dry-run.",
-        suffix=target_path.suffix or ".hwpx",
+        suffix=".hwpx" if hwp_target else (target_path.suffix or ".hwpx"),
         dir=str(target_path.parent),
     )
     tmp_path = Path(tmp_name)
@@ -366,6 +374,10 @@ def save_dry_run(
                 "dry-run HWPX failed open-safety verification: "
                 + verification["openSafety"]["summary"]
             )
+        if hwp_target:
+            # Raises Hwp5Error (hwp5-write-unsupported) as the real save would.
+            hwp5_document_bytes(document)
+            verification["format"] = "hwp"
         diff = semantic_diff(target_path, tmp_path) if target_path.exists() else None
         verification["filePath"] = str(target_path)
         verification["dryRunTempDeleted"] = True

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/errors.py

@@ -73,6 +73,7 @@ DEFAULT_ERROR_SUGGESTIONS: dict[str, str] = {
     "EMPTY_MATCH": "찾을 텍스트나 앵커 조건을 실제 문서 내용에 맞게 수정하세요.",
     "HWP_CONVERSION_FAILED": "원본 HWP가 손상되지 않았는지 확인하고 변환 로그의 원인을 기준으로 다시 시도하세요.",
     "HWP_TEXT_EXTRACT_FAILED": "HWP 파일을 다른 뷰어에서 열어 유효성을 확인한 뒤 텍스트 추출을 다시 실행하세요.",
+    "HWPX_PACKAGE_REQUIRED": "copy_document로 .hwpx 사본을 만들어 이 도구를 쓰고, .hwp가 필요하면 결과를 copy_document로 .hwp 경로에 복사하세요. search_and_replace·set_table_cell_text 같은 문서 편집 도구는 .hwp를 그대로 편집합니다.",
     "IDEMPOTENT_REPLAY": "이미 적용된 planId입니다. 최신 preview/plan을 새로 만든 뒤 apply를 다시 호출하세요.",
     "MEMO_NOT_FOUND": "list_memos 또는 문서 리소스에서 memoId를 다시 확인한 뒤 호출하세요.",
     "MISSING_NODE": "대상 섹션, 문단, 표, 셀이 아직 존재하는지 조회한 뒤 새 앵커로 다시 실행하세요.",
@@ -81,7 +82,10 @@ DEFAULT_ERROR_SUGGESTIONS: dict[str, str] = {
     "PIPELINE_ERROR": "details.pipelineCode와 hint를 확인한 뒤 preview부터 다시 생성하세요.",
     "PLAN_RECORD_MISSING": "preview_edit 또는 plan_edit으로 planId를 새로 발급받은 뒤 apply를 다시 호출하세요.",
     "PREVIEW_REQUIRED": "preview_edit으로 변경 내용을 확인한 뒤 반환된 planId로 apply_edit을 호출하세요.",
-    "READ_ONLY_HWP_DOCUMENT": "HWP 바이너리는 직접 저장할 수 없습니다. 먼저 HWPX로 변환한 뒤 편집하세요.",
+    "HWP_DOCUMENT_UNREADABLE": "한/글에서 문서를 열어 다시 저장한 뒤 사용하세요. details.hwp5Code가 원인입니다.",
+    "HWP_ENCRYPTED_DOCUMENT": "한/글에서 문서 암호·배포용 보호를 풀고 다시 저장한 뒤 사용하세요. DRM 문서는 열 수 없습니다.",
+    "HWP_WRITE_UNSUPPORTED": "HWP 5.0으로 쓸 수 없는 내용이 있습니다. 출력 경로를 .hwpx로 지정해 저장하세요.",
+    "READ_ONLY_HWP_DOCUMENT": "python-hwpx 6.6.0 이상을 설치하면 .hwp 문서를 그대로 열고 편집·저장할 수 있습니다.",
     "RENDER_PREVIEW_INVALID_MODE": "mode는 'pages' 또는 'long'으로 지정한 뒤 다시 실행하세요.",
     "RENDER_PREVIEW_INVALID_SCREENSHOT_MODE": "screenshot은 'auto', 'require', 'off' 중 하나로 지정하세요.",
     "RENDER_PREVIEW_UNAVAILABLE": "matching python-hwpx 버전을 설치하고 MCP 서버를 새로 시작한 뒤 다시 실행하세요.",
@@ -175,3 +179,52 @@ def build_error_payload(
         suggestion=suggestion
         or suggestion_for_error(code, details=details, hint=hint),
     ).model_dump(exclude_none=True, by_alias=True)
+
+
+class HwpDocumentError(RuntimeError):
+    """An ``.hwp`` request the installed engine cannot carry out."""
+
+    def __init__(
+        self, message: str, *, code: str, details: Optional[Dict[str, Any]] = None
+    ) -> None:
+        super().__init__(message)
+        self.code = code
+        self.message = message
+        self.details = details
+
+    def to_payload(self) -> Dict[str, Any]:
+        return build_error_payload(
+            code=self.code, message=self.message, details=self.details
+        )
+
+
+_HWP5_ENCRYPTED_CODES = frozenset({"hwp5-password", "hwp5-distribution", "hwp5-drm"})
+
+
+def hwp5_error_payload(exc: BaseException) -> Optional[Dict[str, Any]]:
+    """The MCP error for python-hwpx's ``Hwp5Error``; None for any other exception.
+
+    Matched by its ``hwp5-*`` code so this module needs no core import.
+    """
+
+    hwp5_code = getattr(exc, "code", None)
+    if not isinstance(hwp5_code, str) or not hwp5_code.startswith("hwp5-"):
+        return None
+    details = {"hwp5Code": hwp5_code}
+    if hwp5_code == "hwp5-write-unsupported":
+        return build_error_payload(
+            code="HWP_WRITE_UNSUPPORTED",
+            message="문서에 HWP 5.0(.hwp)으로 쓸 수 없는 내용이 있어 아무것도 저장하지 않았습니다.",
+            details=details,
+        )
+    if hwp5_code in _HWP5_ENCRYPTED_CODES:
+        return build_error_payload(
+            code="HWP_ENCRYPTED_DOCUMENT",
+            message="암호·배포용·DRM으로 보호된 HWP 문서는 열 수 없습니다.",
+            details=details,
+        )
+    return build_error_payload(
+        code="HWP_DOCUMENT_UNREADABLE",
+        message="HWP 5.0(.hwp) 문서를 읽을 수 없습니다.",
+        details=details,
+    )

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/handlers/agent_document.py

@@ -9,6 +9,8 @@ from typing import Any
 
 
 from ..document_state import document_state_payload
+from ..errors import HwpDocumentError
+from ..storage import require_hwpx_package
 from ..agent_document import (
     AgentChildLimit,
     AgentCommandList,
@@ -100,6 +102,7 @@ def apply_document_commands(
     try:
         resolved_input = resolve_path(filename)
         resolved_output = resolve_path(output)
+        require_hwpx_package(resolved_input, output=resolved_output)
         quality_contract.assert_write_capability()
         store = ScopedIdempotencyStore(
             RUNTIME_SERVICES.idempotency_cache,
@@ -132,6 +135,8 @@ def apply_document_commands(
         if Path(state_path).exists():
             payload.update(document_state_payload(state_path))
         return payload
+    except HwpDocumentError:
+        raise
     except Exception as exc:
         return agent_batch_error_payload(
             exc,

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/handlers/read_export.py

@@ -35,6 +35,7 @@ from ..office.compliance import DEFAULT_POLICY, mask_pii
 from ..runtime_services import RUNTIME_SERVICES
 from ..upstream import (
     HP_NS,
+    hwp5_conversion_report,
     open_document,
 )
 from ..utils.helpers import default_max_chars, resolve_path, truncate_response
@@ -777,16 +778,19 @@ def get_document_info(filename: str) -> dict:
     path = resolve_path(filename)
     doc = open_doc(path)
     file_size = Path(path).stat().st_size
-    return _with_document_state(
-        {
-            "filename": filename,
-            "sections": len(doc.sections),
-            "paragraphs": _paragraph_count(doc),
-            "tables": _table_count(doc),
-            "file_size": str(file_size),
-        },
-        path,
-    )
+    info: dict[str, Any] = {
+        "filename": filename,
+        "sections": len(doc.sections),
+        "paragraphs": _paragraph_count(doc),
+        "tables": _table_count(doc),
+        "file_size": str(file_size),
+    }
+    conversion = hwp5_conversion_report(doc)
+    if conversion is not None:
+        # An .hwp is edited through the document model; this is what it left out.
+        info["format"] = "hwp"
+        info["hwpConversion"] = conversion
+    return _with_document_state(info, path)
 
 
 def get_document_text(

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/identity.json

@@ -7,9 +7,9 @@
     "candidate": {
       "pythonHwpx": "6.6.0",
       "canonicalDistribution": "python-hwpx-automation",
-      "canonicalAutomation": "7.3.2",
+      "canonicalAutomation": "7.3.3",
       "compatibilityDistribution": "hwpx-mcp-server",
-      "compatibility": "7.3.2",
+      "compatibility": "7.3.3",
       "plugin": "2.4.0",
       "contractHash": "5e5c23651f92785a"
     },

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/mixed_form.py

@@ -41,7 +41,7 @@ from pydantic import (
 from .execution_lock import PUBLIC_MUTATION_LOCK
 from .office.agent._batch_publication import GuardedSavePipeline
 from .office.rendering import resolve_hancom_backend
-from .storage import build_hwpx_open_safety_report
+from .storage import build_hwpx_open_safety_report, require_hwpx_package
 from .utils.helpers import resolve_path
 from .workspace import (
     WorkspaceMissingParentGuard,
@@ -1374,6 +1374,7 @@ def analyze_mixed_form_plan(
 
     request = _authorize_public_plan(_payload(plan, apply=False))
     source = Path(request["source"])
+    require_hwpx_package(source, output=request["output"])
     source_snapshot = _capture_path_snapshot(source)
     if source_snapshot.data is None:
         raise FileNotFoundError(source)
@@ -1494,6 +1495,7 @@ def _apply_canonical_mixed_form_plan_locked(
 
     compiled = validate_mixed_form_plan(compiled).to_dict()
     source, output = _authorize_compiled_plan(compiled)
+    require_hwpx_package(source, output=output)
     source_snapshot = _capture_path_snapshot(source)
     if source_snapshot.data is None:  # pragma: no cover - compiled plan invariant
         raise FileNotFoundError(source)

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/office/rendering/oracle.py

@@ -149,6 +149,16 @@ def _complete_package(path: str) -> bool:
         return False
 
 
+def _unchanged(path: str, before: os.stat_result) -> bool:
+    """True when ``path`` still has the size and modification time in ``before``."""
+
+    try:
+        now = os.stat(path)
+    except OSError:
+        return False
+    return (now.st_size, now.st_mtime_ns) == (before.st_size, before.st_mtime_ns)
+
+
 def _promote_file(staged: str, target: str) -> None:
     """Move a finished file to ``target`` without leaving a partial ``target``.
 
@@ -442,6 +452,8 @@ class WindowsComOracle(RenderBackend):
         by a complete package: a failed refresh leaves it as it was. Returns
         True when the file was re-saved. Only an ``.hwpx`` file is refreshed
         (Hancom saves it as HWPX), and a link is followed to the file it names.
+        A document that changed while Hancom ran (size or modification time)
+        is not overwritten: the refresh returns False and the change stays.
         """
 
         src = os.path.realpath(hwpx_path)
@@ -459,6 +471,7 @@ class WindowsComOracle(RenderBackend):
             try:
                 os.makedirs(os.path.dirname(job_src))
                 os.makedirs(os.path.dirname(job_out))
+                before = os.stat(src)
                 shutil.copyfile(src, job_src)
             except OSError:
                 return False
@@ -467,6 +480,8 @@ class WindowsComOracle(RenderBackend):
             entry = entries[0] if entries else None
             if not (isinstance(entry, dict) and entry.get("saved") and _complete_package(job_out)):
                 return False
+            if not _unchanged(src, before):
+                return False
             try:
                 _promote_file(job_out, src)
             except OSError:

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/ops_services/context.py

@@ -4,13 +4,16 @@ from __future__ import annotations
 
 import hashlib
 import logging
+from io import BytesIO
 from pathlib import Path
+from zipfile import ZipFile
 from typing import Any, Callable, Dict, List, Optional, Tuple
 
+from ..errors import HwpDocumentError, hwp5_error_payload
 from ..hwp_support import HwpBinaryError, extract_hwp_text
 from ..core.locator import RegisteredHandle
 from ..core.context import default_session_lifecycle_policy
-from ..storage import DocumentStorage, LocalDocumentStorage
+from ..storage import DocumentStorage, LocalDocumentStorage, require_hwpx_package
 from ..workspace import (
     WorkspacePathError,
 )
@@ -18,6 +21,8 @@ from ..upstream import (
     HwpxDocument,
     HwpxOxmlParagraph,
     HwpxOxmlTable,
+    hwpx_view_bytes,
+    native_hwp5_supported,
 )
 
 logger = logging.getLogger("hwpx_automation.hwpx_ops")
@@ -158,12 +163,50 @@ class DocumentContext:
     def _relative_path(self, path: Path) -> str:
         return self.storage.relative_path(path)
 
+    def _is_legacy_hwp(self, resolved: Path) -> bool:
+        """An ``.hwp`` the installed core cannot open; only its preview text is readable."""
+
+        return resolved.suffix.lower() == ".hwp" and not native_hwp5_supported()
+
+    def _text_source(self, resolved: Path) -> Path | ZipFile:
+        """What the text extractor reads: the HWPX file, or an ``.hwp``'s HWPX model."""
+
+        if resolved.suffix.lower() != ".hwp":
+            return resolved
+        try:
+            return ZipFile(BytesIO(hwpx_view_bytes(resolved)))
+        except Exception as exc:
+            hwp_error = self._hwp_error(exc)
+            if hwp_error is None:
+                raise
+            raise hwp_error from exc
+
+    def _hwp_error(self, exc: BaseException) -> Optional[RuntimeError]:
+        """The operation error for an ``.hwp`` refusal, or None for any other failure."""
+
+        if isinstance(exc, HwpDocumentError):
+            return self._new_error(exc.code, exc.message, details=exc.details)
+        payload = hwp5_error_payload(exc)
+        if payload is None:
+            return None
+        return self._new_error(
+            payload["code"], payload["message"], details=payload.get("details")
+        )
+
+    def _require_hwpx_package(self, source: Path, output: Optional[str] = None) -> None:
+        """Refuse ``.hwp`` for editors that patch HWPX package bytes."""
+
+        try:
+            require_hwpx_package(source, output=output)
+        except HwpDocumentError as exc:
+            raise self._new_error(exc.code, exc.message, details=exc.details) from exc
+
     def _open_document(self, path: str) -> Tuple[HwpxDocument, Path]:
         resolved = self._resolve_path(path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._is_legacy_hwp(resolved):
             raise self._new_error(
                 "READ_ONLY_HWP_DOCUMENT",
-                "HWP 파일은 편집이 불가합니다. 먼저 convert_hwp_to_hwpx 도구로 HWPX 변환 후 편집하세요.",
+                "설치된 python-hwpx는 HWP 5.0(.hwp) 문서를 열거나 저장하지 못합니다.",
             )
         try:
             document, resolved = self.storage.open_document(path)
@@ -186,6 +229,9 @@ class DocumentContext:
                 details={"requestedName": Path(path).name},
             ) from exc
         except Exception as exc:  # pragma: no cover - delegated to backend
+            hwp_error = self._hwp_error(exc)
+            if hwp_error is not None:
+                raise hwp_error from exc
             raise self._new_error(
                 "DOCUMENT_OPEN_FAILED",
                 f"failed to open '{path}': {exc}",

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/ops_services/form_fields.py

@@ -299,6 +299,7 @@ class FormFieldService:
             ) from exc
 
         source_path = self._context._resolve_path(path)
+        self._context._require_hwpx_package(source_path, output)
         output_precondition = None
         if isinstance(self._context.storage, LocalDocumentStorage):
             output_precondition = (
@@ -468,6 +469,7 @@ class FormFieldService:
                 "installed python-hwpx does not provide hwpx.body_patch.apply_body_ops",
             ) from exc
         source_path = self._context._resolve_path(path)
+        self._context._require_hwpx_package(source_path, output)
         output_precondition = None
         if isinstance(self._context.storage, LocalDocumentStorage):
             output_precondition = (
@@ -633,6 +635,7 @@ class FormFieldService:
             ) from exc
 
         blank = self._context._resolve_path(path)
+        self._context._require_hwpx_package(blank, output)
         md = self._context._resolve_path(review_md)
         output_precondition = None
         if isinstance(self._context.storage, LocalDocumentStorage):

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/ops_services/package_validation.py

@@ -135,7 +135,7 @@ class PackageValidationService:
         resolved = self._context._resolve_path(path)
         patterns = [re.compile(pat) for pat in (forbid_patterns or [])]
         warnings: List[Dict[str, Any]] = []
-        with create_text_extractor(resolved) as extractor:
+        with create_text_extractor(self._context._text_source(resolved)) as extractor:
             for paragraph in extractor.iter_document_paragraphs():
                 text = paragraph.text()
                 if max_line_len is not None and len(text) > max_line_len:

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/ops_services/planning.py

@@ -39,7 +39,7 @@ class PlanningService:
     def _ensure_planner_document(self, doc_id: str, path: str) -> None:
         resolved = self._context._resolve_path(path)
         paragraphs: List[str] = []
-        with create_text_extractor(resolved) as extractor:
+        with create_text_extractor(self._context._text_source(resolved)) as extractor:
             for paragraph in extractor.iter_document_paragraphs():
                 paragraphs.append(paragraph.text(preserve_breaks=True))
         self._manager.register_document(doc_id, "\n".join(paragraphs))

{python_hwpx_automation-7.3.2 → python_hwpx_automation-7.3.3}/src/hwpx_automation/ops_services/read_query.py

@@ -18,6 +18,7 @@ from ..upstream import (
     AnnotationOptions,
     create_object_finder,
     create_text_extractor,
+    hwp5_conversion_report,
 )
 
 from .context import DocumentContext
@@ -45,7 +46,7 @@ class ReadQueryService:
     def get_paragraphs_by_handle(self, handle_id: str) -> Dict[str, Any]:
         handle = self._context.get_registered_handle(handle_id)
         resolved = self._context._resolve_path(handle.path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             paragraphs, _, _ = self._context._read_only_hwp_paragraphs(handle.path)
             hwp_serialized = [
                 ParagraphResourceEntry(paragraphIndex=index, text=text)
@@ -57,7 +58,7 @@ class ReadQueryService:
             return model.model_dump(by_alias=True)
 
         serialized: List[ParagraphResourceEntry] = []
-        with create_text_extractor(resolved) as extractor:
+        with create_text_extractor(self._context._text_source(resolved)) as extractor:
             for paragraph in extractor.iter_document_paragraphs():
                 serialized.append(
                     ParagraphResourceEntry(
@@ -73,7 +74,7 @@ class ReadQueryService:
     def get_tables_by_handle(self, handle_id: str) -> Dict[str, Any]:
         handle = self._context.get_registered_handle(handle_id)
         resolved = self._context._resolve_path(handle.path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             model = DocumentTablesResource(handleId=handle.handle_id, tables=[])
             return model.model_dump(by_alias=True)
 
@@ -92,7 +93,7 @@ class ReadQueryService:
 
     def open_info(self, path: str) -> Dict[str, Any]:
         resolved = self._context._resolve_path(path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             paragraphs, _, source = self._context._read_only_hwp_paragraphs(path)
             stat = resolved.stat()
             meta = {
@@ -123,6 +124,10 @@ class ReadQueryService:
             "size": stat.st_size,
             "modified": datetime.fromtimestamp(stat.st_mtime).isoformat(),
         }
+        conversion = hwp5_conversion_report(document)
+        if conversion is not None:
+            meta["format"] = "hwp"
+            meta["hwpConversion"] = conversion
         return {
             "meta": meta,
             "sectionCount": section_count,
@@ -169,7 +174,7 @@ class ReadQueryService:
         with_footnotes: bool = False,
     ) -> Dict[str, Any]:
         resolved = self._context._resolve_path(path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             hwp_paragraphs, _, _ = self._context._read_only_hwp_paragraphs(path)
             effective_limit = (
                 self._context.paging_limit if limit is None else max(1, limit)
@@ -195,7 +200,7 @@ class ReadQueryService:
         paragraphs: List[str] = []
         next_offset: Optional[int] = None
         start = max(0, offset)
-        with create_text_extractor(resolved) as extractor:
+        with create_text_extractor(self._context._text_source(resolved)) as extractor:
             paragraph_iter = extractor.iter_document_paragraphs()
             sentinel = object()
 
@@ -239,7 +244,7 @@ class ReadQueryService:
             unique_indexes.add(int(index))
 
         resolved = self._context._resolve_path(path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             paragraphs, _, _ = self._context._read_only_hwp_paragraphs(path)
             hwp_collected = {
                 idx: paragraphs[idx] for idx in unique_indexes if idx < len(paragraphs)
@@ -266,7 +271,7 @@ class ReadQueryService:
             )
 
         collected: Dict[int, str] = {}
-        with create_text_extractor(resolved) as extractor:
+        with create_text_extractor(self._context._text_source(resolved)) as extractor:
             for paragraph in extractor.iter_document_paragraphs():
                 para_index = paragraph.index
                 if para_index in unique_indexes and para_index not in collected:
@@ -292,7 +297,7 @@ class ReadQueryService:
 
     def text_extract_report(self, path: str, mode: str = "plain") -> Dict[str, Any]:
         resolved = self._context._resolve_path(path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             paragraphs, _, source = self._context._read_only_hwp_paragraphs(path)
             return {
                 "content": "\n".join(paragraphs)
@@ -307,7 +312,7 @@ class ReadQueryService:
                 endnote="inline",
                 control="placeholder",
             )
-        with create_text_extractor(resolved) as extractor:
+        with create_text_extractor(self._context._text_source(resolved)) as extractor:
             content = extractor.extract_text(
                 annotations=annotations,
                 include_nested=True,
@@ -322,11 +327,11 @@ class ReadQueryService:
         lock_keywords: Optional[Sequence[str]] = None,
     ) -> Dict[str, Any]:
         resolved = self._context._resolve_path(path)
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             paragraphs, _, source = self._context._read_only_hwp_paragraphs(path)
         else:
             paragraphs = []
-            with create_text_extractor(resolved) as extractor:
+            with create_text_extractor(self._context._text_source(resolved)) as extractor:
                 for paragraph in extractor.iter_document_paragraphs():
                     paragraphs.append(paragraph.text(preserve_breaks=True))
             source = "hwpx.text_extractor"
@@ -474,7 +479,7 @@ class ReadQueryService:
             return snippet
 
         pattern = re.compile(query) if is_regex else None
-        if resolved.suffix.lower() == ".hwp":
+        if self._context._is_legacy_hwp(resolved):
             paragraphs, _, _ = self._context._read_only_hwp_paragraphs(path)
             for para_index, text in enumerate(paragraphs):
                 if is_regex:
@@ -512,7 +517,7 @@ class ReadQueryService:
                         start = found + len(query)
             return {"matches": matches}
 
-        with create_text_extractor(resolved) as extractor:
+        with create_text_extractor(self._context._text_source(resolved)) as extractor:
             for paragraph in extractor.iter_document_paragraphs():
                 text = paragraph.text()
                 if is_regex:
@@ -558,7 +563,7 @@ class ReadQueryService:
         max_results: int = 200,
     ) -> Dict[str, Any]:
         resolved = self._context._resolve_path(path)
-        finder = create_object_finder(resolved)
+        finder = create_object_finder(self._context._text_source(resolved))
         objects = []
         for found in finder.iter(tag=tag_name, limit=max_results):
             element = found.element
@@ -582,7 +587,7 @@ class ReadQueryService:
         max_results: int = 200,
     ) -> Dict[str, Any]:
         resolved = self._context._resolve_path(path)
-        finder = create_object_finder(resolved)
+        finder = create_object_finder(self._context._text_source(resolved))
         tag_filter = None if element_type in {None, "", "*"} else element_type
         attr_matcher: Any = value if value is not None else (lambda _: True)
         objects = []