BabelDOC 0.4.0__tar.gz → 0.4.2__tar.gz

{babeldoc-0.4.0 → babeldoc-0.4.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: BabelDOC
-Version: 0.4.0
+Version: 0.4.2
 Summary: Yet Another Document Translator
 Project-URL: Homepage, https://github.com/funstory-ai/BabelDOC
 Project-URL: Issues, https://github.com/funstory-ai/BabelDOC/issues
@@ -17,7 +17,6 @@ Requires-Dist: charset-normalizer>=2.0.0
 Requires-Dist: configargparse>=1.7
 Requires-Dist: cryptography>=36.0.0
 Requires-Dist: freetype-py>=2.5.1
-Requires-Dist: hdbscan>=0.8.40
 Requires-Dist: httpx[socks]>=0.27.0
 Requires-Dist: huggingface-hub>=0.27.0
 Requires-Dist: hyperscan>=0.7.13
@@ -38,7 +37,6 @@ Requires-Dist: rapidocr-onnxruntime>=1.4.4
 Requires-Dist: rich>=13.9.4
 Requires-Dist: rtree>=1.4.0
 Requires-Dist: scikit-image>=0.25.2
-Requires-Dist: scikit-learn<1.8.0
 Requires-Dist: tenacity>=9.0.0
 Requires-Dist: tiktoken>=0.9.0
 Requires-Dist: toml>=0.10.2

babeldoc-0.4.2/babeldoc/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.4.2"

{babeldoc-0.4.0 → babeldoc-0.4.2}/babeldoc/const.py

@@ -3,7 +3,7 @@ import shutil
 import subprocess
 from pathlib import Path
 
-__version__ = "0.4.0"
+__version__ = "0.4.2"
 
 CACHE_FOLDER = Path.home() / ".cache" / "babeldoc"
 

{babeldoc-0.4.0 → babeldoc-0.4.2}/babeldoc/format/pdf/document_il/midend/styles_and_formulas.py

@@ -19,15 +19,13 @@ from babeldoc.format.pdf.document_il.utils.formular_helper import (
     is_formulas_middle_char,
 )
 from babeldoc.format.pdf.document_il.utils.formular_helper import is_formulas_start_char
+from babeldoc.format.pdf.document_il.utils.formular_helper import update_formula_data
 from babeldoc.format.pdf.document_il.utils.layout_helper import LEFT_BRACKET
 from babeldoc.format.pdf.document_il.utils.layout_helper import RIGHT_BRACKET
 from babeldoc.format.pdf.document_il.utils.layout_helper import build_layout_index
 from babeldoc.format.pdf.document_il.utils.layout_helper import (
     calculate_y_iou_for_boxes,
 )
-from babeldoc.format.pdf.document_il.utils.layout_helper import (
-    formular_height_ignore_char,
-)
 from babeldoc.format.pdf.document_il.utils.layout_helper import is_bullet_point
 from babeldoc.format.pdf.document_il.utils.layout_helper import is_same_style
 from babeldoc.format.pdf.translation_config import TranslationConfig
@@ -40,6 +38,9 @@ class StylesAndFormulas:
         self.translation_config = translation_config
         self.font_mapper = FontMapper(translation_config)
 
+    def update_formula_data(self, formula: PdfFormula):
+        update_formula_data(formula)
+
     def process(self, document: Document):
         with self.translation_config.progress_monitor.stage_start(
             self.stage_name,
@@ -50,6 +51,12 @@ class StylesAndFormulas:
                 self.process_page(page)
                 pbar.advance()
 
+    def update_all_formula_data(self, page: Page):
+        for para in page.pdf_paragraph:
+            for comp in para.pdf_paragraph_composition:
+                if comp.pdf_formula:
+                    self.update_formula_data(comp.pdf_formula)
+
     def process_page(self, page: Page):
         """处理页面，包括公式识别和偏移量计算"""
         build_layout_index(page)
@@ -60,6 +67,7 @@ class StylesAndFormulas:
         # self.process_page_offsets(page)
         self.process_translatable_formulas(page)
         self.process_page_offsets(page)
+        self.update_all_formula_data(page)
         self.process_page_styles(page)
 
         # clean up to save memory
@@ -590,31 +598,6 @@ class StylesAndFormulas:
             self.update_line_data(new_line)
             return PdfParagraphComposition(pdf_line=new_line)
 
-    def update_formula_data(self, formula: PdfFormula):
-        min_x = min(char.visual_bbox.box.x for char in formula.pdf_character)
-        max_x = max(char.visual_bbox.box.x2 for char in formula.pdf_character)
-        if not all(map(formular_height_ignore_char, formula.pdf_character)):
-            min_y = min(
-                char.visual_bbox.box.y
-                for char in formula.pdf_character
-                if not formular_height_ignore_char(char)
-            )
-            max_y = max(
-                char.visual_bbox.box.y2
-                for char in formula.pdf_character
-                if not formular_height_ignore_char(char)
-            )
-        else:
-            min_y = min(char.visual_bbox.box.y for char in formula.pdf_character)
-            max_y = max(char.visual_bbox.box.y2 for char in formula.pdf_character)
-        formula.box = Box(min_x, min_y, max_x, max_y)
-        if not formula.y_offset:
-            formula.y_offset = 0
-        if not formula.x_offset:
-            formula.x_offset = 0
-        if not formula.x_advance:
-            formula.x_advance = 0
-
     def is_translatable_formula(self, formula: PdfFormula) -> bool:
         """判断公式是否只包含需要正常翻译的字符（数字、空格和英文逗号）"""
         text = "".join(char.char_unicode for char in formula.pdf_character)

{babeldoc-0.4.0 → babeldoc-0.4.2}/babeldoc/format/pdf/document_il/midend/typesetting.py

@@ -15,6 +15,7 @@ from babeldoc.format.pdf.document_il import PdfParagraphComposition
 from babeldoc.format.pdf.document_il import PdfStyle
 from babeldoc.format.pdf.document_il import il_version_1
 from babeldoc.format.pdf.document_il.utils.fontmap import FontMapper
+from babeldoc.format.pdf.document_il.utils.formular_helper import update_formula_data
 from babeldoc.format.pdf.translation_config import TranslationConfig
 from babeldoc.format.pdf.translation_config import WatermarkOutputMode
 
@@ -481,6 +482,7 @@ class TypesettingUnit:
                 y_offset=self.formular.y_offset * scale,
                 x_advance=self.formular.x_advance * scale,
             )
+            update_formula_data(new_formula)
             return TypesettingUnit(formular=new_formula)
 
         elif self.unicode:
@@ -819,7 +821,7 @@ class Typesetting:
             # 更新 x 坐标
             current_x = relocated_unit.box.x2
             if prev_x > current_x:
-                logger.warning(f"坐标回绕！！！TypesettingUnit: {unit}, ")
+                logger.warning(f"坐标回绕！！！TypesettingUnit: {unit.box}, ")
 
             last_unit = relocated_unit
 

{babeldoc-0.4.0 → babeldoc-0.4.2}/babeldoc/format/pdf/document_il/utils/formular_helper.py

@@ -3,8 +3,13 @@ import functools
 import re
 import unicodedata
 
+from babeldoc.format.pdf.document_il.il_version_1 import Box
 from babeldoc.format.pdf.document_il.il_version_1 import Page
+from babeldoc.format.pdf.document_il.il_version_1 import PdfFormula
 from babeldoc.format.pdf.document_il.utils.fontmap import FontMapper
+from babeldoc.format.pdf.document_il.utils.layout_helper import (
+    formular_height_ignore_char,
+)
 from babeldoc.format.pdf.translation_config import TranslationConfig
 
 
@@ -301,3 +306,29 @@ def is_formulas_font(font_name: str, formular_font_pattern: str | None) -> bool:
         return True
 
     return False
+
+
+def update_formula_data(formula: PdfFormula):
+    min_x = min(char.visual_bbox.box.x for char in formula.pdf_character)
+    max_x = max(char.visual_bbox.box.x2 for char in formula.pdf_character)
+    if not all(map(formular_height_ignore_char, formula.pdf_character)):
+        min_y = min(
+            char.visual_bbox.box.y
+            for char in formula.pdf_character
+            if not formular_height_ignore_char(char)
+        )
+        max_y = max(
+            char.visual_bbox.box.y2
+            for char in formula.pdf_character
+            if not formular_height_ignore_char(char)
+        )
+    else:
+        min_y = min(char.visual_bbox.box.y for char in formula.pdf_character)
+        max_y = max(char.visual_bbox.box.y2 for char in formula.pdf_character)
+    formula.box = Box(min_x, min_y, max_x, max_y)
+    if not formula.y_offset:
+        formula.y_offset = 0
+    if not formula.x_offset:
+        formula.x_offset = 0
+    if not formula.x_advance:
+        formula.x_advance = 0

{babeldoc-0.4.0 → babeldoc-0.4.2}/babeldoc/format/pdf/high_level.py

@@ -640,7 +640,6 @@ def migrate_toc(
     old_doc = Document(translation_config.input_file)
     if not old_doc:
         return
-
     try:
         fix_filter(old_doc)
         fix_null_xref(old_doc)
@@ -653,6 +652,15 @@ def migrate_toc(
         logger.info("No TOC found in the original PDF, skipping migration.")
         return
 
+    if translation_config.only_include_translated_page:
+        total_page = set(range(0, len(old_doc)))
+
+        pages_to_translate = {
+            i for i in len(old_doc) if translation_config.should_translate_page(i + 1)
+        }
+
+        should_removed_page = list(total_page - pages_to_translate)
+
     files = {
         translate_result.dual_pdf_path,
         # translate_result.mono_pdf_path,

{babeldoc-0.4.0 → babeldoc-0.4.2}/babeldoc/main.py

@@ -21,7 +21,7 @@ from babeldoc.translator.translator import OpenAITranslator
 from babeldoc.translator.translator import set_translate_rate_limiter
 
 logger = logging.getLogger(__name__)
-__version__ = "0.4.0"
+__version__ = "0.4.2"
 
 
 def create_parser():

{babeldoc-0.4.0 → babeldoc-0.4.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "BabelDOC"
-version = "0.4.0"
+version = "0.4.2"
 description = "Yet Another Document Translator"
 license = "AGPL-3.0"
 readme = "README.md"
@@ -47,9 +47,7 @@ dependencies = [
     "rapidocr-onnxruntime>=1.4.4",
     "pyzstd>=0.17.0",
     "hyperscan>=0.7.13",
-    "rtree>=1.4.0",
-    "hdbscan>=0.8.40",
-    "scikit-learn<1.8.0"
+    "rtree>=1.4.0"
 ]
 
 [project.optional-dependencies]
@@ -159,7 +157,7 @@ pythonpath = [".", "src"]
 testpaths = ["tests"]
 
 [bumpver]
-current_version = "0.4.0"
+current_version = "0.4.2"
 version_pattern = "MAJOR.MINOR.PATCH[.PYTAGNUM]"
 
 [bumpver.file_patterns]