BabelDOC 0.3.7__tar.gz → 0.3.9__tar.gz

{babeldoc-0.3.7 → babeldoc-0.3.9}/.github/ISSUE_TEMPLATE/bug_report.yaml

@@ -69,7 +69,7 @@ body:
     validations:
       required: false
 
-  - type: file
+  - type: textarea
     id: pdf
     attributes:
       label: Original PDF File

{babeldoc-0.3.7 → babeldoc-0.3.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: BabelDOC
-Version: 0.3.7
+Version: 0.3.9
 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-0.3.9/babeldoc/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.3.9"

{babeldoc-0.3.7 → babeldoc-0.3.9}/babeldoc/const.py

@@ -3,7 +3,7 @@ import shutil
 import subprocess
 from pathlib import Path
 
-__version__ = "0.3.7"
+__version__ = "0.3.9"
 
 CACHE_FOLDER = Path.home() / ".cache" / "babeldoc"
 

{babeldoc-0.3.7 → babeldoc-0.3.9}/babeldoc/converter.py

@@ -1,4 +1,3 @@
-import base64
 import logging
 import re
 import unicodedata
@@ -93,13 +92,7 @@ class PDFConverterEx(PDFConverter):
         textwidth = font.char_width(cid)
         textdisp = font.char_disp(cid)
 
-        font_name = font.fontname
-        if isinstance(font_name, bytes):
-            try:
-                font_name = font_name.decode("utf-8")
-            except UnicodeDecodeError:
-                font_name = "BASE64:" + base64.b64encode(font_name).decode("utf-8")
-        font_id = self.il_creater.current_page_font_name_id_map.get(font_name, None)
+        font_id = self.il_creater.current_page_font_name_id_map.get(font.xobj_id, None)
 
         item = AWLTChar(
             matrix,

{babeldoc-0.3.7 → babeldoc-0.3.9}/babeldoc/document_il/frontend/il_creater.py

@@ -350,6 +350,8 @@ class ILCreater:
                 _, encoding = self.mupdf.xref_get_key(xref_id, "Encoding")
                 if encoding == "/Identity-H" or encoding == "/Identity-V":
                     encoding_length = 2
+                if encoding == "/WinAnsiEncoding":
+                    encoding_length = 1
                 else:
                     _, to_unicode_id = self.mupdf.xref_get_key(xref_id, "ToUnicode")
                     to_unicode_bytes = self.mupdf.xref_stream(
@@ -442,7 +444,7 @@ class ILCreater:
                 )
         except Exception:
             pass
-        self.current_page_font_name_id_map[font_name] = font_id
+        self.current_page_font_name_id_map[xref_id] = font_id
         if self.xobj_id in self.xobj_map:
             self.xobj_map[self.xobj_id].pdf_font.append(il_font_metadata)
         else:

{babeldoc-0.3.7 → babeldoc-0.3.9}/babeldoc/document_il/midend/paragraph_finder.py

@@ -29,7 +29,7 @@ class ParagraphFinder:
 
     # 定义项目符号的正则表达式模式
     BULLET_POINT_PATTERN = re.compile(
-        r"[■•⚫⬤◆◇○●◦‣⁃▪▫∗†‡¹²³⁴⁵⁶⁷⁸⁹⁰₁₂₃₄₅₆₇₈₉₀ᵃᵇᶜᵈᵉᶠᵍʰⁱʲᵏˡᵐⁿᵒᵖᵍʳˢᵗᵘᵛʷˣʸᶻ†‡§¶※⁑⁂⁕⁎⁜⁑❧☙⁋‖‽]"
+        r"[■•⚫⬤◆◇○●◦‣⁃▪▫∗†‡¹²³⁴⁵⁶⁷⁸⁹⁰₁₂₃₄₅₆₇₈₉₀ᵃᵇᶜᵈᵉᶠᵍʰⁱʲᵏˡᵐⁿᵒᵖᵍʳˢᵗᵘᵛʷˣʸᶻ†‡§¶※⁑⁂⁕⁎⁜⁑❧☙⁋‖‽]"
     )
 
     def __init__(self, translation_config: TranslationConfig):

{babeldoc-0.3.7 → babeldoc-0.3.9}/babeldoc/document_il/midend/typesetting.py

@@ -111,6 +111,7 @@ class TypesettingUnit:
             r"\uAB30-\uAB6F"  # Latin Extended E
             r"\u0250-\u02A0"  # IPA Extensions
             r"\u0400-\u04FF"  # Cyrillic
+            r"'"
             r"]+$",
             unicode,
         ):

{babeldoc-0.3.7 → babeldoc-0.3.9}/babeldoc/main.py

@@ -23,7 +23,7 @@ from babeldoc.translation_config import TranslationConfig
 from babeldoc.translation_config import WatermarkOutputMode
 
 logger = logging.getLogger(__name__)
-__version__ = "0.3.7"
+__version__ = "0.3.9"
 
 
 def create_parser():

{babeldoc-0.3.7 → babeldoc-0.3.9}/babeldoc/pdfinterp.py

@@ -104,6 +104,7 @@ class PDFPageInterpreterEx(PDFPageInterpreter):
                         objid = spec.objid
                     spec = dict_value(spec)
                     font = self.rsrcmgr.get_font(objid, spec)
+                    font.xobj_id = objid
                     self.il_creater.on_page_resource_font(font, objid, fontid)
                     self.fontmap[fontid] = font
                     self.fontmap[fontid].descent = 0  # hack fix descent

{babeldoc-0.3.7 → babeldoc-0.3.9}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "BabelDOC"
-version = "0.3.7"
+version = "0.3.9"
 description = "Yet Another Document Translator"
 license = "AGPL-3.0"
 readme = "README.md"
@@ -148,7 +148,7 @@ pythonpath = [".", "src"]
 testpaths = ["tests"]
 
 [bumpver]
-current_version = "0.3.7"
+current_version = "0.3.9"
 version_pattern = "MAJOR.MINOR.PATCH[.PYTAGNUM]"
 
 [bumpver.file_patterns]

{babeldoc-0.3.7 → babeldoc-0.3.9}/uv.lock

@@ -48,7 +48,7 @@ wheels = [
 
 [[package]]
 name = "babeldoc"
-version = "0.3.7"
+version = "0.3.9"
 source = { editable = "." }
 dependencies = [
     { name = "bitstring" },

babeldoc-0.3.7/babeldoc/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.3.7"