docling-parse 4.2.2__tar.gz → 4.2.3__tar.gz

{docling_parse-4.2.2 → docling_parse-4.2.3}/CHANGELOG.md

@@ -1,3 +1,9 @@
+## [v4.2.3](https://github.com/docling-project/docling-parse/releases/tag/v4.2.3) - 2025-08-22
+
+### Fix
+
+* Media box ([#157](https://github.com/docling-project/docling-parse/issues/157)) ([`5ded3b8`](https://github.com/docling-project/docling-parse/commit/5ded3b8f7f12936ab390f8296814e2cdddc2da43))
+
 ## [v4.2.2](https://github.com/docling-project/docling-parse/releases/tag/v4.2.2) - 2025-08-19
 
 ### Fix

{docling_parse-4.2.2/docling_parse.egg-info → docling_parse-4.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling-parse
-Version: 4.2.2
+Version: 4.2.3
 Summary: Simple package to extract text with coordinates from programmatic PDFs
 Author-email: Peter Staar <taa@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Maxim Lysak <mly@zurich.ibm.com>
 License-Expression: MIT

{docling_parse-4.2.2 → docling_parse-4.2.3/docling_parse.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling-parse
-Version: 4.2.2
+Version: 4.2.3
 Summary: Simple package to extract text with coordinates from programmatic PDFs
 Author-email: Peter Staar <taa@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Maxim Lysak <mly@zurich.ibm.com>
 License-Expression: MIT

{docling_parse-4.2.2 → docling_parse-4.2.3}/docling_parse.egg-info/SOURCES.txt

@@ -1134,6 +1134,14 @@ tests/data/cases/case_06.pdf
 tests/data/cases/case_07.pdf
 tests/data/cases/case_08.pdf
 tests/data/cases/case_09.pdf
+tests/data/groundtruth/broken_media_box_v01.pdf.page_no_1.py.json
+tests/data/groundtruth/broken_media_box_v01.pdf.page_no_1.py.json.char.txt
+tests/data/groundtruth/broken_media_box_v01.pdf.page_no_1.py.json.line.txt
+tests/data/groundtruth/broken_media_box_v01.pdf.page_no_1.py.json.word.txt
+tests/data/groundtruth/broken_media_box_v01.pdf.v1.json
+tests/data/groundtruth/broken_media_box_v01.pdf.v1.p=0.json
+tests/data/groundtruth/broken_media_box_v01.pdf.v2.json
+tests/data/groundtruth/broken_media_box_v01.pdf.v2.p=0.json
 tests/data/groundtruth/cropbox_versus_mediabox_01.pdf.page_no_1.py.json
 tests/data/groundtruth/cropbox_versus_mediabox_01.pdf.page_no_1.py.json.char.txt
 tests/data/groundtruth/cropbox_versus_mediabox_01.pdf.page_no_1.py.json.line.txt
@@ -1424,6 +1432,7 @@ tests/data/groundtruth/type3_fonts.pdf.v1.json
 tests/data/groundtruth/type3_fonts.pdf.v1.p=0.json
 tests/data/groundtruth/type3_fonts.pdf.v2.json
 tests/data/groundtruth/type3_fonts.pdf.v2.p=0.json
+tests/data/regression/broken_media_box_v01.pdf
 tests/data/regression/cropbox_versus_mediabox_01.pdf
 tests/data/regression/cropbox_versus_mediabox_02.pdf
 tests/data/regression/duplicate_bold_text_01.pdf

{docling_parse-4.2.2 → docling_parse-4.2.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "docling-parse"
-version = "4.2.2"
+version = "4.2.3"
 description = "Simple package to extract text with coordinates from programmatic PDFs"
 license = "MIT"
 keywords = ["docling", "pdf", "parser"]

{docling_parse-4.2.2 → docling_parse-4.2.3}/src/v2/pdf_resources/page_dimension.h

@@ -29,6 +29,10 @@ namespace pdflib
 
     std::pair<double, double> rotate(int angle);
     
+  private:
+
+    std::array<double, 4> normalize_page_boundaries(std::array<double, 4> bbox, std::string name);
+    
   private:
 
     bool                  initialised;
@@ -154,6 +158,33 @@ namespace pdflib
     
     return delta;
   }
+
+  std::array<double, 4> pdf_resource<PAGE_DIMENSION>::normalize_page_boundaries(std::array<double, 4> bbox, std::string name)
+  {
+    LOG_S(INFO) << __FUNCTION__;
+    
+    double llx = std::min(bbox[0], bbox[2]);
+    double lly = std::min(bbox[1], bbox[3]);
+    double urx = std::max(bbox[0], bbox[2]);
+    double ury = std::max(bbox[1], bbox[3]);
+
+    if(urx<llx)
+      {
+	LOG_S(ERROR) << "we have a malformed page-boundary for " << name << "-> llx: "<< llx << ", urx: "<< urx;
+      }
+
+    if(ury<lly)
+      {
+	LOG_S(ERROR) << "we have a malformed page-boundary for " << name << "-> lly: "<< lly << ", ury: "<< ury;
+      }    
+    
+    bbox[0] = llx;
+    bbox[1] = lly;
+    bbox[2] = urx;
+    bbox[3] = ury;
+
+    return bbox;
+  }
   
   bool pdf_resource<PAGE_DIMENSION>::init_from(nlohmann::json& data)
   {
@@ -324,6 +355,12 @@ namespace pdflib
         LOG_S(ERROR) << ss.str();
 	throw std::logic_error(ss.str());
       }
+
+    crop_bbox = normalize_page_boundaries(crop_bbox, "crop_bbox");
+    media_bbox = normalize_page_boundaries(media_bbox, "media_bbox");
+    art_bbox = normalize_page_boundaries(art_bbox, "art_bbox");
+    bleed_bbox = normalize_page_boundaries(bleed_bbox, "bleed_bbox");
+    trim_bbox = normalize_page_boundaries(trim_bbox, "trim_bbox");
   }
 
 }