super-document-processor 0.1.5__tar.gz → 0.1.6__tar.gz

{super_document_processor-0.1.5 → super_document_processor-0.1.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: super-document-processor
-Version: 0.1.5
+Version: 0.1.6
 Summary: High-fidelity multi-format document parsing and chunking engine for RAG pipelines
 Author: Cinderfox
 License: MIT
@@ -27,6 +27,7 @@ Requires-Dist: python-pptx==1.0.2
 Requires-Dist: extract-msg==0.56.1
 Requires-Dist: lxml==6.1.3
 Requires-Dist: aspose-slides==26.8.0
+Requires-Dist: resvg-py==0.5.0
 Requires-Dist: python-dotenv==1.2.3
 Requires-Dist: fastapi==0.141.1
 Requires-Dist: uvicorn==0.52.4

{super_document_processor-0.1.5 → super_document_processor-0.1.6}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "super-document-processor"
-version = "0.1.5"
+version = "0.1.6"
 description = "High-fidelity multi-format document parsing and chunking engine for RAG pipelines"
 readme = "README.md"
 requires-python = ">=3.9"
@@ -44,6 +44,7 @@ dependencies = [
     "extract-msg==0.56.1",
     "lxml==6.1.3",
     "aspose-slides==26.8.0",
+    "resvg-py==0.5.0",
     "python-dotenv==1.2.3",
     "fastapi==0.141.1",
     "uvicorn==0.52.4",

{super_document_processor-0.1.5 → super_document_processor-0.1.6}/super_document_processor.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: super-document-processor
-Version: 0.1.5
+Version: 0.1.6
 Summary: High-fidelity multi-format document parsing and chunking engine for RAG pipelines
 Author: Cinderfox
 License: MIT
@@ -27,6 +27,7 @@ Requires-Dist: python-pptx==1.0.2
 Requires-Dist: extract-msg==0.56.1
 Requires-Dist: lxml==6.1.3
 Requires-Dist: aspose-slides==26.8.0
+Requires-Dist: resvg-py==0.5.0
 Requires-Dist: python-dotenv==1.2.3
 Requires-Dist: fastapi==0.141.1
 Requires-Dist: uvicorn==0.52.4

{super_document_processor-0.1.5 → super_document_processor-0.1.6}/super_document_processor.egg-info/requires.txt

@@ -4,6 +4,7 @@ python-pptx==1.0.2
 extract-msg==0.56.1
 lxml==6.1.3
 aspose-slides==26.8.0
+resvg-py==0.5.0
 python-dotenv==1.2.3
 fastapi==0.141.1
 uvicorn==0.52.4

{super_document_processor-0.1.5 → super_document_processor-0.1.6}/utils/aspose_slide_renderer.py

@@ -1,9 +1,5 @@
 import io
-import os
 import re
-import subprocess
-from pathlib import Path
-from typing import Optional
 
 def fix_svg_dimensions(svg_text: str, scale: float = 2.0) -> str:
     """
@@ -48,56 +44,57 @@ def strip_aspose_watermark_from_svg(svg_text: str) -> str:
                 break
     return svg_text
 
-def find_resvg_bin() -> Optional[Path]:
-    """Locates the standalone native resvg binary if installed."""
-    import shutil
-    for candidate in ["/usr/local/bin/resvg", "/usr/bin/resvg", "resvg", "resvg.exe"]:
-        found = shutil.which(candidate)
-        if found and os.access(found, os.X_OK):
-            return Path(found)
-    return None
+def normalize_svg_font_sizes(svg_text: str) -> str:
+    """
+    resvg does not parse 'pt' units in font-size attributes or inline CSS (e.g. font-size="18pt" or style="font-size: 18pt").
+    Converts all font-size occurrences with 'pt' to standard pixel values:
+    px = round(pt * 96 / 72, 2)
+    """
+    def _pt_attr_to_px(match):
+        num = float(match.group(1))
+        px = round(num * 96.0 / 72.0, 2)
+        return f'font-size="{px}px"'
+
+    def _pt_style_to_px(match):
+        num = float(match.group(1))
+        px = round(num * 96.0 / 72.0, 2)
+        return f'font-size: {px}px'
+
+    # Handle font-size="18pt" and font-size='18pt'
+    svg_text = re.sub(r'font-size="([0-9.]+)pt"', _pt_attr_to_px, svg_text)
+    svg_text = re.sub(r"font-size='([0-9.]+)pt'", _pt_attr_to_px, svg_text)
+    # Handle style="... font-size: 18pt; ..."
+    svg_text = re.sub(r'font-size\s*:\s*([0-9.]+)pt', _pt_style_to_px, svg_text)
+    return svg_text
 
 def render_svg_to_png_bytes(svg_string: str, scale: float = 2.0) -> bytes:
     """
-    Renders an SVG string to PNG bytes using the standalone native resvg CLI.
+    Renders an SVG string to PNG bytes using resvg_py in-memory Rust engine.
+    Ensures watermark is stripped, inch dimensions are scaled to pixels,
+    and CSS font units (pt -> px) are normalized so all text renders crisp.
     """
     clean_svg = strip_aspose_watermark_from_svg(svg_string)
     clean_svg = fix_svg_dimensions(clean_svg, scale=scale)
-    
-    resvg_bin = find_resvg_bin()
-    if not resvg_bin:
-        raise RuntimeError("Standalone native resvg binary not found in PATH or /usr/local/bin/resvg.")
-
-    import tempfile
-    with tempfile.NamedTemporaryFile(suffix=".svg", delete=False) as in_svg, \
-         tempfile.NamedTemporaryFile(suffix=".png", delete=False) as out_png:
-        in_svg.write(clean_svg.encode("utf-8"))
-        in_svg.flush()
-        in_path = in_svg.name
-        out_path = out_png.name
+    clean_svg = normalize_svg_font_sizes(clean_svg)
 
-    try:
-        print(f"⚡ [Rasterizer] Using standalone native resvg CLI at scale={scale}", flush=True)
-        cmd = [str(resvg_bin), "-z", str(scale), in_path, out_path]
-        res = subprocess.run(cmd, capture_output=True, timeout=60)
-        if res.returncode == 0 and os.path.exists(out_path):
-            with open(out_path, "rb") as f:
-                return f.read()
-        else:
-            err_msg = res.stderr.decode("utf-8", errors="ignore").strip()
-            raise RuntimeError(f"Native resvg rasterization failed: {err_msg}")
-    finally:
-        if os.path.exists(in_path):
-            os.remove(in_path)
-        if os.path.exists(out_path):
-            os.remove(out_path)
+    import resvg_py
+    return resvg_py.svg_to_bytes(svg_string=clean_svg, zoom=float(scale))
 
 def render_slide_to_png_bytes(aspose_slide, scale: float = 2.0) -> bytes:
     """
     Exports an Aspose slide to clean SVG in memory, strips the watermark, and
-    rasterizes it to high-fidelity PNG bytes.
+    rasterizes it to high-fidelity PNG bytes using resvg.
+    Falls back to Aspose native slide image rendering only if resvg is completely unavailable.
     """
-    svg_stream = io.BytesIO()
-    aspose_slide.write_as_svg(svg_stream)
-    raw_svg = svg_stream.getvalue().decode("utf-8", errors="ignore")
-    return render_svg_to_png_bytes(raw_svg, scale=scale)
+    try:
+        svg_stream = io.BytesIO()
+        aspose_slide.write_as_svg(svg_stream)
+        raw_svg = svg_stream.getvalue().decode("utf-8", errors="ignore")
+        return render_svg_to_png_bytes(raw_svg, scale=scale)
+    except Exception:
+        # Fallback to Aspose native rasterizer
+        import aspose.slides as _slides
+        img = aspose_slide.get_image(float(scale), float(scale))
+        buf = io.BytesIO()
+        img.save(buf, _slides.ImageFormat.PNG)
+        return buf.getvalue()

{super_document_processor-0.1.5 → super_document_processor-0.1.6}/main.py

@@ -701,9 +701,9 @@ def main():
             sys.exit(1)
             
         try:
-            from super_document_processor.utils.pptx_converter import convert_pptx, create_pptx_render_zip
-        except ImportError:
             from utils.pptx_converter import convert_pptx, create_pptx_render_zip
+        except ImportError:
+            from super_document_processor.utils.pptx_converter import convert_pptx, create_pptx_render_zip
             
         try:
             norm_fmt = "png" if args.format.lower() in ("png", "img", "image") else "svg"

{super_document_processor-0.1.5 → super_document_processor-0.1.6}/utils/pptx_converter.py

@@ -9,14 +9,14 @@ from typing import Optional, Union, Dict, Any, List
 from concurrent.futures import ThreadPoolExecutor
 
 try:
-    from super_document_processor.utils.aspose_slide_renderer import (
+    from utils.aspose_slide_renderer import (
         strip_aspose_watermark_from_svg,
         fix_svg_dimensions,
         render_svg_to_png_bytes,
         render_slide_to_png_bytes
     )
 except ImportError:
-    from utils.aspose_slide_renderer import (
+    from super_document_processor.utils.aspose_slide_renderer import (
         strip_aspose_watermark_from_svg,
         fix_svg_dimensions,
         render_svg_to_png_bytes,