fixed the tests

Signed-off-by: Peter Staar <taa@zurich.ibm.com>
DS4SD · Nov 7, 2024 · 5c82ff9 · 5c82ff9
1 parent b154d4f
commit 5c82ff9
Show file tree

Hide file tree

Showing 5 changed files with 17 additions and 4,427 deletions.
diff --git a/docling/backend/html_backend.py b/docling/backend/html_backend.py
@@ -114,15 +114,19 @@ def walk(self, element, doc):
                         _log.info(f" -> error treating child: {exc}")
                         raise exc
 
+            else:
+                _log.debug(f"ignoring element of type {type(element)}")
+
+            """
             elif isinstance(element, Tag):
                 try:
                     self.analyse_element(element, 0, doc)
                 except Exception as exc:
                     _log.info(f" -> error treating elem: {exc}")
                     raise exc
-            else:
-                _log.debug(f"ignoring element of type {type(element)}")
+            """
 
+
         except Exception as exc:
             _log.debug(f"error walking element: {type(element)}")
             pass

diff --git a/docling/cli/main.py b/docling/cli/main.py
@@ -77,6 +77,7 @@ def export_documents(
     export_md: bool,
     export_txt: bool,
     export_doctags: bool,
+    export_itxt: bool,
 ):
 
     success_count = 0
@@ -115,6 +116,12 @@ def export_documents(
                     _log.info(f"writing Doc Tags output to {fname}")
                     fp.write(conv_res.document.export_to_document_tokens())
 
+            # Export Indented Text format:
+            if export_itxt:
+                fname = output_dir / f"{doc_filename}.itxt"
+                with fname.open("w") as fp:
+                    _log.info(f"writing Doc Tags output to {fname}")
+                    fp.write(conv_res.document._export_to_indented_text())                    
         else:
             _log.warning(f"Document {conv_res.input.file} failed to convert.")
             failure_count += 1
@@ -200,6 +207,7 @@ def convert(
     export_md = OutputFormat.MARKDOWN in to_formats
     export_txt = OutputFormat.TEXT in to_formats
     export_doctags = OutputFormat.DOCTAGS in to_formats
+    export_itxt = OutputFormat.INDENTED_TEXT in to_formats
 
     match ocr_engine:
         case OcrEngine.EASYOCR:
@@ -243,6 +251,7 @@ def convert(
         export_md=export_md,
         export_txt=export_txt,
         export_doctags=export_doctags,
+        export_itxt=export_itxt,
     )
 
     end_time = time.time() - start_time

diff --git a/docling/datamodel/base_models.py b/docling/datamodel/base_models.py
@@ -39,6 +39,7 @@ class OutputFormat(str, Enum):
     JSON = "json"
     TEXT = "text"
     DOCTAGS = "doctags"
+    INDENTED_TEXT = "itxt"
 
 
 FormatToExtensions: Dict[InputFormat, List[str]] = {