BabelDOC 0.3.17__tar.gz → 0.3.18__tar.gz

{babeldoc-0.3.17 → babeldoc-0.3.18}/.github/workflows/docs.yml

@@ -17,7 +17,7 @@ jobs:
           git config user.name github-actions[bot]
           git config user.email 41898282+github-actions[bot]@users.noreply.github.com
       - name: Setup uv with Python 3.12
-        uses: astral-sh/setup-uv@0c5e2b8115b80b4c7c5ddf6ffdd634974642d182 # v5.4.1
+        uses: astral-sh/setup-uv@d4b2f3b6ecc6e67c4457f6d3e41ec42d3d0fcb86 # v5.4.2
         with:
           python-version: "3.12"
           enable-cache: true

babeldoc-0.3.18/.github/workflows/lint.yml

@@ -0,0 +1,17 @@
+name: Lint Code
+permissions:
+  contents: read
+  pull-requests: write
+on: [push]
+
+jobs:
+  lint:
+    strategy:
+      fail-fast: false
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Ruff
+        uses: astral-sh/ruff-action@v3
+      - name: AutoCorrect
+        uses: huacnlee/autocorrect-action@main

babeldoc-0.3.18/.github/workflows/pr-lint.yml

@@ -0,0 +1,46 @@
+name: Lint Code and Review Dog Report
+
+on: [pull_request]
+permissions:
+  contents: read
+  pull-requests: write
+jobs:
+  ruff:
+    name: runner / ruff
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Install Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+          
+      - name: Install ruff
+        run: pip install ruff
+        
+      - name: Install reviewdog
+        uses: reviewdog/action-setup@e04ffabe3898a0af8d0fb1af00c188831c4b5893 # v1.3.2
+        with:
+          reviewdog_version: latest
+          
+      - name: Run ruff with reviewdog
+        env:
+          REVIEWDOG_GITHUB_API_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          ruff check . --output-format=rdjson | reviewdog -f=rdjson -reporter=github-pr-review -fail-on-error
+          
+  autocorrect:
+    name: runner / autocorrect
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: AutoCorrect
+        uses: huacnlee/autocorrect-action@bf91ab3904c2908dd8e71312a8a83ed1eb632997 # v2.13.3
+      - name: Report ReviewDog
+        if: failure()
+        uses: huacnlee/autocorrect-action@bf91ab3904c2908dd8e71312a8a83ed1eb632997 # v2.13.3
+        env:
+          REVIEWDOG_GITHUB_API_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          reviewdog: true

{babeldoc-0.3.17 → babeldoc-0.3.18}/.github/workflows/publish-to-pypi.yml

@@ -35,7 +35,7 @@ jobs:
           token: ${{ secrets.GITHUB_TOKEN }}
           
       - name: Setup uv with Python 3.12
-        uses: astral-sh/setup-uv@0c5e2b8115b80b4c7c5ddf6ffdd634974642d182 # v5.4.1
+        uses: astral-sh/setup-uv@d4b2f3b6ecc6e67c4457f6d3e41ec42d3d0fcb86 # v5.4.2
         with:
           python-version: "3.12"
           enable-cache: true

{babeldoc-0.3.17 → babeldoc-0.3.18}/.github/workflows/test.yml

@@ -28,7 +28,7 @@ jobs:
           path: ~/.cache/babeldoc
           key: babeldoc-assets-${{ hashFiles('babeldoc/assets/embedding_assets_metadata.py') }}
       - name: Setup uv with Python ${{ matrix.python-version }}
-        uses: astral-sh/setup-uv@0c5e2b8115b80b4c7c5ddf6ffdd634974642d182 # v5.4.1
+        uses: astral-sh/setup-uv@d4b2f3b6ecc6e67c4457f6d3e41ec42d3d0fcb86 # v5.4.2
         with:
           python-version: ${{ matrix.python-version }}
           enable-cache: true

{babeldoc-0.3.17 → babeldoc-0.3.18}/.gitignore

@@ -12,6 +12,7 @@ web/dist
 web/dist-ssr
 web/*.local
 
+memray*
 **/*.so
 *.pdf
 *.docx

{babeldoc-0.3.17 → babeldoc-0.3.18}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: BabelDOC
-Version: 0.3.17
+Version: 0.3.18
 Summary: Yet Another Document Translator
 Project-URL: Homepage, https://github.com/funstory-ai/BabelDOC
 Project-URL: Issues, https://github.com/funstory-ai/BabelDOC/issues

babeldoc-0.3.18/babeldoc/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.3.18"

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/const.py

@@ -3,7 +3,7 @@ import shutil
 import subprocess
 from pathlib import Path
 
-__version__ = "0.3.17"
+__version__ = "0.3.18"
 
 CACHE_FOLDER = Path.home() / ".cache" / "babeldoc"
 

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/document_il/backend/pdf_creater.py

@@ -31,11 +31,11 @@ def _subset_fonts_process(pdf_path, output_path):
         pdf = pymupdf.open(pdf_path)
         pdf.subset_fonts(fallback=False)
         pdf.save(output_path)
-        # 返回0表示成功
+        # 返回 0 表示成功
         os._exit(0)
     except Exception as e:
         logger.error(f"Error in font subsetting subprocess: {e}")
-        # 返回1表示失败
+        # 返回 1 表示失败
         os._exit(1)
 
 
@@ -69,11 +69,11 @@ def _save_pdf_clean_process(
             deflate_fonts=deflate_fonts,
             linear=linear,
         )
-        # 返回0表示成功
+        # 返回 0 表示成功
         os._exit(0)
     except Exception as e:
         logger.error(f"Error in save PDF with clean=True subprocess: {e}")
-        # 返回1表示失败
+        # 返回 1 表示失败
         os._exit(1)
 
 

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/document_il/frontend/il_creater.py

@@ -479,6 +479,7 @@ class ILCreater:
         self.current_page_font_name_id_map = {}
         self.current_page_font_char_bounding_box_map = {}
         self.mupdf_font_map: dict[int, pymupdf.Font] = {}
+        self.graphic_state_pool = {}
 
     def on_finish(self):
         self.progress.__exit__(None, None, None)
@@ -797,6 +798,21 @@ class ILCreater:
             f"{arg} {op}" for op, arg in gs.passthrough_instruction
         )
 
+        # 可能会影响部分 graphic state 准确度。不过 BabelDOC 仅使用 passthrough_per_char_instruction
+        # 所以应该是没啥影响
+        # 但是池化 graphic state 后可以减少内存占用
+        if (
+            graphic_state.passthrough_per_char_instruction
+            not in self.graphic_state_pool
+        ):
+            self.graphic_state_pool[graphic_state.passthrough_per_char_instruction] = (
+                graphic_state
+            )
+        else:
+            graphic_state = self.graphic_state_pool[
+                graphic_state.passthrough_per_char_instruction
+            ]
+
         return graphic_state
 
     def on_lt_char(self, char: LTChar):

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/document_il/midend/paragraph_finder.py

@@ -237,7 +237,7 @@ class ParagraphFinder:
             is_small_char = char_area < median_char_area * 0.1
 
             # 检查是否需要开始新段落
-            # 如果字符面积小于中位数面积的10%且当前段落已有字符，则跳过新段落检测
+            # 如果字符面积小于中位数面积的 10% 且当前段落已有字符，则跳过新段落检测
             if not (is_small_char and current_line_chars) and (
                 current_layout is None
                 or char_layout.id != current_layout.id

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/document_il/midend/styles_and_formulas.py

@@ -395,7 +395,7 @@ class StylesAndFormulas:
                             right_line = comp.pdf_line
                             break
 
-                # 计算与左右文本的y轴交集
+                # 计算与左右文本的 y 轴交集
                 left_intersection = 0
                 right_intersection = 0
 
@@ -413,7 +413,7 @@ class StylesAndFormulas:
                     if intersection_end > intersection_start:
                         right_intersection = intersection_end - intersection_start
 
-                # 如果有两个文本段落，将交集较小的设为none
+                # 如果有两个文本段落，将交集较小的设为 none
                 if left_line and right_line:
                     if left_intersection < right_intersection:
                         left_line = None

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/document_il/midend/typesetting.py

@@ -221,8 +221,8 @@ class TypesettingUnit:
                 # 连接号
                 "～",  # 全角波浪号
                 "-",  # 连字符减号
-                "–",  # 短破折号(EN DASH)
-                "—",  # 长破折号(EM DASH)
+                "–",  # 短破折号 (EN DASH)
+                "—",  # 长破折号 (EM DASH)
                 # 间隔号
                 "·",  # 中间点
                 "・",  # 片假名中间点
@@ -767,7 +767,7 @@ class Typesetting:
         line_spacing = 1.7  # 初始行距为 1.7
         min_scale = 0.1  # 最小缩放因子
         min_line_spacing = 1.4  # 最小行距
-        expand_space_flag = 0  # 0: 未扩展, 1: 已向下扩展, 2: 已向右扩展
+        expand_space_flag = 0  # 0: 未扩展，1: 已向下扩展，2: 已向右扩展
 
         while scale >= min_scale:
             # 尝试布局排版单元

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/docvision/rpc_doclayout.py

@@ -43,10 +43,10 @@ def encode_image(image) -> bytes:
 
 
 @retry(
-    stop=stop_after_attempt(3),  # 最多重试3次
+    stop=stop_after_attempt(3),  # 最多重试 3 次
     wait=wait_exponential(
         multiplier=1, min=1, max=10
-    ),  # 指数退避策略，初始1秒，最大10秒
+    ),  # 指数退避策略，初始 1 秒，最大 10 秒
     retry=retry_if_exception_type((httpx.HTTPError, Exception)),  # 针对哪些异常重试
     before_sleep=lambda retry_state: logger.warning(
         f"Request failed, retrying in {retry_state.next_action.sleep} seconds... "

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/high_level.py

@@ -586,6 +586,7 @@ def _do_translate_single(
     logger.debug(f"finish parse il from {temp_pdf_path}")
     docs = il_creater.create_il()
     logger.debug(f"finish create il from {temp_pdf_path}")
+    del il_creater
     if translation_config.debug:
         xml_converter.write_json(
             docs,
@@ -657,6 +658,7 @@ def _do_translate_single(
         il_translator = ILTranslator(translate_engine, translation_config)
 
     il_translator.translate(docs)
+    del il_translator
     logger.debug(f"finish ILTranslator from {temp_pdf_path}")
     if translation_config.debug:
         xml_converter.write_json(

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/main.py

@@ -23,7 +23,7 @@ from babeldoc.translation_config import TranslationConfig
 from babeldoc.translation_config import WatermarkOutputMode
 
 logger = logging.getLogger(__name__)
-__version__ = "0.3.17"
+__version__ = "0.3.18"
 
 
 def create_parser():

{babeldoc-0.3.17 → babeldoc-0.3.18}/babeldoc/translation_config.py

@@ -60,7 +60,7 @@ class TranslationConfig:
         enhance_compatibility: bool = False,
         report_interval: float = 0.1,
         min_text_length: int = 5,
-        use_side_by_side_dual: bool = True,  # Deprecated: 是否使用拼版式双语 PDF（并排显示原文和译文） 向下兼容选项，已停用。
+        use_side_by_side_dual: bool = True,  # Deprecated: 是否使用拼版式双语 PDF（并排显示原文和译文）向下兼容选项，已停用。
         use_alternating_pages_dual: bool = False,
         watermark_output_mode: WatermarkOutputMode = WatermarkOutputMode.Watermarked,
         # Add split-related parameters

{babeldoc-0.3.17 → babeldoc-0.3.18}/docs/CONTRIBUTING.md

@@ -181,3 +181,9 @@ trang babeldoc/document_il/il_version_1.rnc babeldoc/document_il/il_version_1.xs
 # Generate Python classes from XSD
 xsdata generate babeldoc/document_il/il_version_1.xsd --package babeldoc.document_il
 ```
+
+##### Profile memory usage
+
+```bash
+uv run memray run --native --aggregate babeldoc/main.py -c yadt.toml
+```

{babeldoc-0.3.17 → babeldoc-0.3.18}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "BabelDOC"
-version = "0.3.17"
+version = "0.3.18"
 description = "Yet Another Document Translator"
 license = "AGPL-3.0"
 readme = "README.md"
@@ -140,7 +140,8 @@ dev = [
     "pre-commit>=4.1.0",
     "pygments>=2.19.1",
     "ruff>=0.9.2",
-    "pytest>=8.3.4"
+    "pytest>=8.3.4",
+    "memray>=1.17.1",
 ]
 
 [tool.pytest.ini_options]
@@ -148,7 +149,7 @@ pythonpath = [".", "src"]
 testpaths = ["tests"]
 
 [bumpver]
-current_version = "0.3.17"
+current_version = "0.3.18"
 version_pattern = "MAJOR.MINOR.PATCH[.PYTAGNUM]"
 
 [bumpver.file_patterns]