From 8c60dfa0e66dc279020b96777597f6df9cce5741 Mon Sep 17 00:00:00 2001
From: Maksym Lysak <mly@zurich.ibm.com>
Date: Mon, 21 Oct 2024 16:42:18 +0200
Subject: [PATCH] Fixed example run_md, added origin info to md_backend

Signed-off-by: Maksym Lysak <mly@zurich.ibm.com>
---
 docling/backend/md_backend.py | 18 +++++++++++++++++-
 docs/examples/run_md.py       |  9 +--------
 2 files changed, 18 insertions(+), 9 deletions(-)

diff --git a/docling/backend/md_backend.py b/docling/backend/md_backend.py
index e59dd51e..5e5f7496 100644
--- a/docling/backend/md_backend.py
+++ b/docling/backend/md_backend.py
@@ -10,6 +10,7 @@ import marko.inline
 from docling_core.types.doc import (
     DocItemLabel,
     DoclingDocument,
+    DocumentOrigin,
     GroupLabel,
     TableCell,
     TableData,
@@ -235,7 +236,22 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
 
     def convert(self) -> DoclingDocument:
         _log.debug("converting Markdown...")
-        doc = DoclingDocument(name="Test")
+
+        fname = ""
+        if isinstance(self.path_or_stream, Path):
+            fname = self.path_or_stream.name
+
+        origin = DocumentOrigin(
+            filename=fname,
+            mimetype="text/markdown",
+            binary_hash=self.document_hash,
+        )
+        if len(fname) > 0:
+            docname = Path(fname).stem
+        else:
+            docname = "stream"
+
+        doc = DoclingDocument(name=docname, origin=origin)
 
         if self.is_valid():
             # Parse the markdown into an abstract syntax tree (AST)
diff --git a/docs/examples/run_md.py b/docs/examples/run_md.py
index d6fb9b11..46be97e2 100644
--- a/docs/examples/run_md.py
+++ b/docs/examples/run_md.py
@@ -13,14 +13,7 @@ _log = logging.getLogger(__name__)
 
 
 def main():
-    input_paths = [
-        Path("README.md"),
-        Path("scratch_a/2203.01017v2.md"),
-        Path("scratch_a/2206.01062.md"),
-        Path("scratch_a/2305.03393v1.md"),
-        Path("scratch_a/redp5110.md"),
-        Path("scratch_a/redp5695.md"),
-    ]
+    input_paths = [Path("README.md")]
 
     for path in input_paths:
         in_doc = InputDocument(