docvortex 0.4.3__tar.gz → 0.4.4__tar.gz

{docvortex-0.4.3/src/docvortex.egg-info → docvortex-0.4.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docvortex
-Version: 0.4.3
+Version: 0.4.4
 Summary: A fast, multi-format document parsing and conversion engine
 License-Expression: MIT
 Project-URL: Repository, https://github.com/myhloli/DocVortex
@@ -43,6 +43,7 @@ Requires-Dist: opencv-python>=4.11.0.86
 Provides-Extra: test
 Requires-Dist: pytest; extra == "test"
 Requires-Dist: pytest-cov; extra == "test"
+Requires-Dist: pytest-xdist<4,>=3.8; extra == "test"
 Requires-Dist: markdown; extra == "test"
 Requires-Dist: jsonschema<5,>=4; extra == "test"
 Provides-Extra: dev

{docvortex-0.4.3 → docvortex-0.4.4}/pyproject.toml

@@ -32,7 +32,7 @@ dependencies = [
 ]
 
 [project.optional-dependencies]
-test = ["pytest", "pytest-cov", "markdown", "jsonschema>=4,<5"]
+test = ["pytest", "pytest-cov", "pytest-xdist>=3.8,<4", "markdown", "jsonschema>=4,<5"]
 dev = ["ruff==0.14.2", "build", "twine"]
 
 [project.urls]

{docvortex-0.4.3 → docvortex-0.4.4}/src/docvortex/render/_internal/pdf/diagnostics.py

@@ -27,9 +27,8 @@ def collect_pdf_diagnostics() -> Iterator[list[Diagnostic]]:
 
 
 def report_pdf_diagnostic(code: str, message: str, page_index: int | None = None) -> None:
-    """正常布局适配使用 DEBUG，其余使用 WARNING；结构化诊断保持完整。"""
-    level = "DEBUG" if code in _DEBUG_DIAGNOSTIC_CODES else "WARNING"
-    logger.log(level, "{}: {}", code, message)
+    """PDF 渲染诊断统一使用 DEBUG 日志，结构化诊断保持完整。"""
+    logger.debug("{}: {}", code, message)
     items = _diagnostics.get()
     if items is not None:
         diagnostic = Diagnostic(code, message, page_index)

{docvortex-0.4.3 → docvortex-0.4.4}/src/docvortex/render/_internal/pdf/formula.py

@@ -356,7 +356,7 @@ def _local_name(tag: str) -> str:
 
 
 def _formula_preview(latex: str) -> str:
-    """为告警生成有界 LaTeX 摘要，避免超长公式污染日志。"""
+    """为诊断生成有界 LaTeX 摘要，避免超长公式污染日志。"""
     return latex if len(latex) <= 200 else f"{latex[:197]}..."
 
 

{docvortex-0.4.3 → docvortex-0.4.4}/src/docvortex/render/_internal/pdf/inline.py

@@ -43,7 +43,7 @@ class PdfAnchorRegistry:
         return self._names.get(normalized) if normalized else None
 
     def attach_markup(self, anchor: str | None) -> str:
-        """返回首个目标的 Paragraph anchor 标签，重复目标仅记录告警。"""
+        """返回首个目标的 Paragraph anchor 标签，重复目标仅记录诊断。"""
         normalized = (anchor or "").strip()
         name = self.resolve(normalized)
         if name is None:
@@ -85,7 +85,7 @@ class PdfInlineContext:
     next_formula_id: int = 0
 
     def location(self, page_idx: int, block_index: int | None, block_type: str) -> str:
-        """返回稳定的 page/block 告警定位文本。"""
+        """返回稳定的 page/block 诊断定位文本。"""
         return f"page_idx={page_idx}, block_index={block_index}, block_type={block_type}"
 
     def register_formula(self, image: InlineFormulaImage) -> str:

{docvortex-0.4.3 → docvortex-0.4.4}/src/docvortex/render/_internal/pdf/original.py

@@ -131,17 +131,19 @@ class OriginalPdfRenderer(_PdfRenderer):
             )
             if reasons:
                 plan.exception_count += 1
-                self._warning("pdf_layout_font_exception", f"reason={','.join(reasons)}, {detail}", item.page_idx, item.block)
+                self._diagnostic(
+                    "pdf_layout_font_exception", f"reason={','.join(reasons)}, {detail}", item.page_idx, item.block
+                )
             if item.geometry_conflict:
-                self._warning("pdf_title_geometry_conflict", detail, item.page_idx, item.block)
+                self._diagnostic("pdf_title_geometry_conflict", detail, item.page_idx, item.block)
             elif item.clearance_unavailable:
-                self._warning("pdf_title_clearance_unavailable", detail, item.page_idx, item.block)
+                self._diagnostic("pdf_title_clearance_unavailable", detail, item.page_idx, item.block)
             ox0, oy0, ox1, oy1 = item.original_rect
             dx0, dy0, dx1, dy1 = item.draw_rect
             expanded = dx0 < ox0 - 0.001 or dy0 < oy0 - 0.001 or dx1 > ox1 + 0.001 or dy1 > oy1 + 0.001
             if expanded:
                 plan.expanded_count += 1
-                self._warning("pdf_title_layout_expanded", detail, item.page_idx, item.block)
+                self._diagnostic("pdf_title_layout_expanded", detail, item.page_idx, item.block)
             plan.titles.append(
                 {
                     "page_idx": item.page_idx,
@@ -190,7 +192,9 @@ class OriginalPdfRenderer(_PdfRenderer):
                         )
                     )
                 return prepared
-            self._warning("pdf_layout_approximate", "Missing child bbox; fitting the group inside its parent", page_idx, block)
+            self._diagnostic(
+                "pdf_layout_approximate", "Missing child bbox; fitting the group inside its parent", page_idx, block
+            )
         assert block.bbox is not None
         x0, y0, x1, y1 = block.bbox
         width, height = (x1 - x0) * page_width, (y1 - y0) * page_height
@@ -226,14 +230,14 @@ class OriginalPdfRenderer(_PdfRenderer):
                 if prepared is not None:
                     return [self._prepared_image_flowable(prepared, block)]
             if isinstance(block, (TableBodyBlock, ChartBodyBlock)):
-                self._warning(
+                self._diagnostic(
                     "pdf_layout_visual_fallback", "Region image unavailable; rendering structured content", page_idx, block
                 )
                 if "<table" in block.content.lower():
                     try:
                         return list(self._html_tables(block.content, page_idx=page_idx, block=block))
                     except PdfTableError as exc:
-                        self._warning("pdf_table_fallback", str(exc), page_idx, block)
+                        self._diagnostic("pdf_table_fallback", str(exc), page_idx, block)
                 if block.content.strip():
                     return [self._preformatted(_plain_html_text(block.content), page_idx, block, self.styles.spatial_table)]
             return [self._placeholder("image unavailable", page_idx=page_idx, block=block, width=self.available_width)]
@@ -281,7 +285,7 @@ class OriginalPdfRenderer(_PdfRenderer):
         url: str | None = None,
     ) -> Paragraph:
         """用可测量段落展示缺失内容，避免固定高度的占位表格产生文字外溢。"""
-        self._warning("pdf_content_placeholder", label, page_idx, block)
+        self._diagnostic("pdf_content_placeholder", label, page_idx, block)
         spans = [TextSpan(type="text", content=label[:320] or "content unavailable")]
         if url:
             spans = [HyperlinkSpan(type="hyperlink", url=url, content=spans)]
@@ -322,11 +326,11 @@ class OriginalPdfRenderer(_PdfRenderer):
         """只绘制已经确认的测量结果，正文不重新排版，标题使用独立的安全绘制框。"""
         fit = item.fit
         if fit.small_text:
-            self._warning(
+            self._diagnostic(
                 "pdf_layout_small_text", f"Block fitted below 6 pt (scale={fit.scale:.4f})", item.page_idx, item.block
             )
         if fit.scale < 0.999:
-            self._warning("pdf_layout_scaled", f"Block fitted at scale={fit.scale:.4f}", item.page_idx, item.block)
+            self._diagnostic("pdf_layout_scaled", f"Block fitted at scale={fit.scale:.4f}", item.page_idx, item.block)
         rect = item.draw_rect or item.original_rect
         canvas.saveState()
         try:

{docvortex-0.4.3 → docvortex-0.4.4}/src/docvortex/render/_internal/pdf/renderer.py

@@ -332,7 +332,7 @@ class _PdfRenderer:
                 flowable.spaceAfter = 7
                 return [flowable]
             except PdfFormulaError as exc:
-                self._warning("pdf_formula_fallback", str(exc), page_idx, block)
+                self._diagnostic("pdf_formula_fallback", str(exc), page_idx, block)
         if _has_image_payload(block):
             image = self._try_prepared_block_image(block, page_idx)
             if image is not None:
@@ -433,7 +433,7 @@ class _PdfRenderer:
                         rendered.extend(self._html_tables(content, page_idx=page_idx, block=child))
                         continue
                     except PdfTableError as exc:
-                        self._warning("pdf_table_fallback", str(exc), page_idx, child)
+                        self._diagnostic("pdf_table_fallback", str(exc), page_idx, child)
                     if _has_image_payload(child):
                         image = self._try_prepared_block_image(child, page_idx)
                         if image is not None:
@@ -473,7 +473,7 @@ class _PdfRenderer:
                     try:
                         rendered.extend(self._html_tables(content, page_idx=page_idx, block=child))
                     except PdfTableError as exc:
-                        self._warning("pdf_table_fallback", str(exc), page_idx, child)
+                        self._diagnostic("pdf_table_fallback", str(exc), page_idx, child)
                         if not image_succeeded:
                             rendered.append(self._preformatted(_plain_html_text(content), page_idx, child, self.styles.body))
                 elif content and not image_succeeded:
@@ -573,7 +573,7 @@ class _PdfRenderer:
             try:
                 prepared = prepare_html_image(source, self.asset_resolver)
             except PdfAssetError as exc:
-                self._warning("pdf_image_unavailable", str(exc), page_idx, block)
+                self._diagnostic("pdf_image_unavailable", str(exc), page_idx, block)
                 return self._placeholder(
                     f"image unavailable: {alt_text}",
                     block=block,
@@ -616,11 +616,11 @@ class _PdfRenderer:
         )
 
     def _try_prepared_block_image(self, block: ImagePayloadBlock, page_idx: int) -> PreparedImage | None:
-        """尝试离线准备图片并把所有素材错误降级为 warning。"""
+        """尝试离线准备图片，并把素材错误统一记录为 DEBUG 诊断。"""
         try:
             return prepare_block_image(block, self.asset_resolver)
         except PdfAssetError as exc:
-            self._warning("pdf_image_unavailable", str(exc), page_idx, block)
+            self._diagnostic("pdf_image_unavailable", str(exc), page_idx, block)
             return None
 
     def _prepared_image_flowable(
@@ -659,7 +659,7 @@ class _PdfRenderer:
         url: str | None = None,
     ) -> Table:
         """创建带浅色边框、可选远程链接和定位文本的稳定占位框。"""
-        self._warning("pdf_content_placeholder", label, page_idx, block)
+        self._diagnostic("pdf_content_placeholder", label, page_idx, block)
         normalized = re.sub(r"\s+", " ", label).strip()[:_MAX_PLACEHOLDER_TEXT] or "content unavailable"
         markup = render_plain_text_markup(normalized)
         if url:
@@ -688,11 +688,11 @@ class _PdfRenderer:
 
     @staticmethod
     def _location(page_idx: int, block: BlockBase) -> str:
-        """返回 PDF 告警与占位使用的稳定 page/block 定位。"""
+        """返回 PDF 诊断与占位使用的稳定 page/block 定位。"""
         return f"page_idx={page_idx}, block_index={block.index}, block_type={block.type}"
 
-    def _warning(self, code: str, message: str, page_idx: int, block: BlockBase) -> None:
-        """用统一位置格式同时报告日志与结构化 PDF 诊断。"""
+    def _diagnostic(self, code: str, message: str, page_idx: int, block: BlockBase) -> None:
+        """用统一位置格式同时报告 DEBUG 日志与结构化 PDF 诊断。"""
         report_pdf_diagnostic(code, f"{message} ({self._location(page_idx, block)})", page_idx)
 
 

{docvortex-0.4.3 → docvortex-0.4.4}/src/docvortex/version.py

@@ -1,3 +1,3 @@
-__version__ = "0.4.3"
+__version__ = "0.4.4"
 
 __all__ = ["__version__"]

{docvortex-0.4.3 → docvortex-0.4.4/src/docvortex.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docvortex
-Version: 0.4.3
+Version: 0.4.4
 Summary: A fast, multi-format document parsing and conversion engine
 License-Expression: MIT
 Project-URL: Repository, https://github.com/myhloli/DocVortex
@@ -43,6 +43,7 @@ Requires-Dist: opencv-python>=4.11.0.86
 Provides-Extra: test
 Requires-Dist: pytest; extra == "test"
 Requires-Dist: pytest-cov; extra == "test"
+Requires-Dist: pytest-xdist<4,>=3.8; extra == "test"
 Requires-Dist: markdown; extra == "test"
 Requires-Dist: jsonschema<5,>=4; extra == "test"
 Provides-Extra: dev

{docvortex-0.4.3 → docvortex-0.4.4}/src/docvortex.egg-info/requires.txt

@@ -33,5 +33,6 @@ twine
 [test]
 pytest
 pytest-cov
+pytest-xdist<4,>=3.8
 markdown
 jsonschema<5,>=4

docvortex-0.4.4/tests/unittest/test_pdf_diagnostic_levels.py

@@ -0,0 +1,48 @@
+"""验证 PDF 渲染诊断统一使用 DEBUG 且结构化结果不变。"""
+
+from __future__ import annotations
+
+import pytest
+from loguru import logger
+
+from docvortex.render._internal.pdf.diagnostics import collect_pdf_diagnostics, report_pdf_diagnostic
+from docvortex.result import Diagnostic
+
+
+_PDF_RENDER_DIAGNOSTIC_CODES = (
+    "pdf_layout_font_exception",
+    "pdf_layout_small_text",
+    "pdf_layout_scaled",
+    "pdf_title_geometry_conflict",
+    "pdf_title_clearance_unavailable",
+    "pdf_title_layout_expanded",
+    "pdf_layout_approximate",
+    "pdf_layout_visual_fallback",
+    "pdf_layout_reflow_fallback",
+    "pdf_formula_fallback",
+    "pdf_table_fallback",
+    "pdf_image_unavailable",
+    "pdf_content_placeholder",
+    "pdf_duplicate_anchor",
+    "pdf_unmatched_link",
+    "unknown_diagnostic",
+)
+
+
+@pytest.mark.parametrize("code", _PDF_RENDER_DIAGNOSTIC_CODES)
+def test_all_pdf_render_diagnostics_use_debug(code: str) -> None:
+    """所有 PDF 渲染诊断均为 DEBUG，重复日志不改变结构化诊断去重结果。"""
+    records: list[str] = []
+    warnings: list[str] = []
+    debug_sink = logger.add(records.append, level="DEBUG", format="{level.name}|{message}")
+    warning_sink = logger.add(warnings.append, level="WARNING", format="{level.name}|{message}")
+    try:
+        with collect_pdf_diagnostics() as items:
+            report_pdf_diagnostic(code, "example", 4)
+            report_pdf_diagnostic(code, "example", 4)
+        assert items == [Diagnostic(code, "example", 4)]
+        assert [str(message).strip() for message in records] == [f"DEBUG|{code}: example"] * 2
+        assert warnings == []
+    finally:
+        logger.remove(debug_sink)
+        logger.remove(warning_sink)

docvortex-0.4.3/tests/unittest/test_pdf_diagnostic_levels.py

@@ -1,39 +0,0 @@
-"""验证 PDF 日志分级不改变结构化诊断及其去重行为。"""
-
-from __future__ import annotations
-
-import pytest
-from loguru import logger
-
-from docvortex.render._internal.pdf.diagnostics import collect_pdf_diagnostics, report_pdf_diagnostic
-from docvortex.result import Diagnostic
-
-
-@pytest.mark.parametrize(
-    "code,level",
-    [
-        ("pdf_title_layout_expanded", "DEBUG"),
-        ("pdf_layout_scaled", "DEBUG"),
-        ("pdf_layout_small_text", "WARNING"),
-        ("pdf_title_geometry_conflict", "WARNING"),
-        ("pdf_image_unavailable", "WARNING"),
-        ("pdf_formula_fallback", "WARNING"),
-        ("unknown_diagnostic", "WARNING"),
-    ],
-)
-def test_log_level_preserves_structured_diagnostics(code: str, level: str) -> None:
-    """按真实日志级别捕获输出，重复诊断仍只在结构化列表保存一次。"""
-    records: list[str] = []
-    warnings: list[str] = []
-    debug_sink = logger.add(records.append, level="DEBUG", format="{level.name}|{message}")
-    warning_sink = logger.add(warnings.append, level="WARNING", format="{level.name}|{message}")
-    try:
-        with collect_pdf_diagnostics() as items:
-            report_pdf_diagnostic(code, "example", 4)
-            report_pdf_diagnostic(code, "example", 4)
-        assert items == [Diagnostic(code, "example", 4)]
-        assert [str(message).strip() for message in records] == [f"{level}|{code}: example"] * 2
-        assert len(warnings) == (2 if level == "WARNING" else 0)
-    finally:
-        logger.remove(debug_sink)
-        logger.remove(warning_sink)