Could not detect hyperlink

[mrzz222]

hyperlink is not being detected properly even though it is present

[dosubot]

Hi @mrzz222! I'm Dosu and I’m helping the docling team.

This is a known issue affecting multiple document formats. The core problem is that while hyperlinks are detected in docling-parse, they aren't properly propagated to the final DoclingDocument output.

Format-specific issues:

• PDF: Hyperlinks are identified but only display text is exported; actual URLs are not extracted
• DOCX: URL fragments and query parameters get stripped (e.g., https://example.com/page#section?view=test becomes https://example.com/page). A fix is in progress via PR fix(docx): preserve URL fragments and query params in hyperlinks #3050.
• HTML: Best support currently, with comprehensive anchor tag handling added

Workaround for DOCX: If you're processing DOCX files, you can temporarily modify docling/backend/msword_backend.py to replace Path(c.address) with c.url to preserve full URLs.

Full hyperlink extraction support is expected with docling-parse v5, which has been under development. Could you share what document format you're working with? That would help narrow down the specific issue.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}

   

[ceberam]

Thanks @mrzz222 for opening this issue.
Could you please give more details on the issue you observed?
Could you share a document that we can use to reproduce the issue?
Also, please copy-paste the output of docling --version

[mrzz222]

Thanks @mrzz222 for opening this issue. Could you please give more details on the issue you observed? Could you share a document that we can use to reproduce the issue? Also, please copy-paste the output of docling --version

links.pdf

expected output is from each page what type of data is present for example in page 1 if it contains image output must return aspicture item else multiple types are present with text,hyperlink and image it must be identified and returned
terminal output

docling --version

/home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/requests/init.py:113: RequestsDependencyWarning: urllib3 (1.26.18) or chardet (7.0.1)/charset_normalizer (3.3.2) doesn't match a supported version!
warnings.warn(
Docling version: 2.77.0
Docling Core version: 2.69.0
Docling IBM Models version: 3.12.0
Docling Parse version: 5.5.0
Python: cpython-310 (3.10.12)
Platform: Linux-6.8.0-101-generic-x86_64-with-glibc2.35

after running below output in terminal
python3 docling_d.py

/home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/requests/init.py:113: RequestsDependencyWarning: urllib3 (1.26.18) or chardet (7.0.1)/charset_normalizer (3.3.2) doesn't match a supported version!
warnings.warn(
[INFO] 2026-03-11 09:53:57,479 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-03-11 09:53:57,490 [RapidOCR] download_file.py:60: File exists and is valid: /home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/rapidocr/models/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-03-11 09:53:57,491 [RapidOCR] main.py:53: Using /home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/rapidocr/models/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-03-11 09:53:57,624 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-03-11 09:53:57,626 [RapidOCR] download_file.py:60: File exists and is valid: /home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/rapidocr/models/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-03-11 09:53:57,627 [RapidOCR] main.py:53: Using /home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/rapidocr/models/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-03-11 09:53:57,670 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-03-11 09:53:57,689 [RapidOCR] download_file.py:60: File exists and is valid: /home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/rapidocr/models/ch_PP-OCRv4_rec_infer.onnx
[INFO] 2026-03-11 09:53:57,689 [RapidOCR] main.py:53: Using /home/mohammad/Desktop/chroma_upload/pad/lib/python3.10/site-packages/rapidocr/models/ch_PP-OCRv4_rec_infer.onnx
schema_name='DoclingDocument' version='1.9.0' name='links' origin=DocumentOrigin(mimetype='application/pdf', binary_hash=4575220423892836957, filename='links.pdf', uri=None) furniture=GroupItem(self_ref='#/furniture', parent=None, children=[], content_layer=<ContentLayer.FURNITURE: 'furniture'>, meta=None, name='root', label=<GroupLabel.UNSPECIFIED: 'unspecified'>) body=GroupItem(self_ref='#/body', parent=None, children=[RefItem(cref='#/texts/0'), RefItem(cref='#/texts/1'), RefItem(cref='#/texts/2'), RefItem(cref='#/texts/3'), RefItem(cref='#/texts/4'), RefItem(cref='#/texts/5'), RefItem(cref='#/texts/6'), RefItem(cref='#/texts/7'), RefItem(cref='#/texts/8'), RefItem(cref='#/texts/9'), RefItem(cref='#/texts/10'), RefItem(cref='#/texts/11'), RefItem(cref='#/texts/12')], content_layer=<ContentLayer.BODY: 'body'>, meta=None, name='root', label=<GroupLabel.UNSPECIFIED: 'unspecified'>) groups=[] texts=[SectionHeaderItem(self_ref='#/texts/0', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.SECTION_HEADER: 'section_header'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=76.69292, t=785.8815294726563, r=197.83292, b=763.5615294726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 12))], source=[], comments=[], orig='Links in PDF', text='Links in PDF', formatting=None, hyperlink=None, level=1), TextItem(self_ref='#/texts/1', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=61.69191, t=726.6163294726563, r=533.7070599999994, b=673.5463294726562, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 425))], source=[], comments=[], orig='There are two types of links in PDF: a link to a specific position in the same PDF document, and a link to an external document. The internal-destination property of fo:basic-link indicates to link to a position in the same document. Its destination may be either an ID or a page number. The external-destination property indicates to link to an external document. Its destination may be either an external file or a website.', text='There are two types of links in PDF: a link to a specific position in the same PDF document, and a link to an external document. The internal-destination property of fo:basic-link indicates to link to a position in the same document. Its destination may be either an ID or a page number. The external-destination property indicates to link to an external document. Its destination may be either an external file or a website.', formatting=None, hyperlink=None), SectionHeaderItem(self_ref='#/texts/2', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.SECTION_HEADER: 'section_header'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=56.69594, t=650.9462194726563, r=245.39594, b=637.5542194726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 34))], source=[], comments=[], orig='Linking to an internal destination', text='Linking to an internal destination', formatting=None, hyperlink=None, level=1), TextItem(self_ref='#/texts/3', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=56.69595, t=627.6163394726563, r=527.0459499999995, b=616.5463394726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 118))], source=[], comments=[], orig='Linking to an ID ： Specify internal-destination="Link-01" on the link; specify id="Link-01" on the link's destination.', text='Linking to an ID ： Specify internal-destination="Link-01" on the link; specify id="Link-01" on the link's destination.', formatting=None, hyperlink=None), TextItem(self_ref='#/texts/4', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=56.69596, t=603.6163394726563, r=302.10596000000004, b=592.5463394726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 60))], source=[], comments=[], orig='Linking to a page number (page 2) ： internal-destination="2"', text='Linking to a page number (page 2) ： internal-destination="2"', formatting=None, hyperlink=None), TextItem(self_ref='#/texts/5', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=56.69597, t=579.6163394726563, r=495.48595999999964, b=568.5463394726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 106))], source=[], comments=[], orig='Linking to a page number (page 2) and setting the display ratio (200%) ： internal-destination="2#zoom=200"', text='Linking to a page number (page 2) and setting the display ratio (200%) ： internal-destination="2#zoom=200"', formatting=None, hyperlink=None), SectionHeaderItem(self_ref='#/texts/6', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.SECTION_HEADER: 'section_header'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=56.69398999999997, t=549.9462194726563, r=248.06998999999996, b=536.5542194726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 34))], source=[], comments=[], orig='Linking to an external destination', text='Linking to an external destination', formatting=None, hyperlink=None, level=1), TextItem(self_ref='#/texts/7', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=56.69001, t=526.6163394726563, r=480.65000999999955, b=515.5463394726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 105))], source=[], comments=[], orig='Linking to an external file (attachment-sample-1.pdf). ： external-destination="./attachment-sample-1.pdf"', text='Linking to an external file (attachment-sample-1.pdf). ： external-destination="./attachment-sample-1.pdf"', formatting=None, hyperlink=None), TextItem(self_ref='#/texts/8', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=56.69004, t=502.61633947265625, r=507.4600499999997, b=491.5463394726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 107))], source=[], comments=[], orig='Linking to a website (https://www.antennahouse.com/) ： external-destination="https://www.antennahouse.com/"', text='Linking to a website (https://www.antennahouse.com/) ： external-destination="https://www.antennahouse.com/"', formatting=None, hyperlink=None), TextItem(self_ref='#/texts/9', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.FURNITURE: 'furniture'>, meta=None, label=<DocItemLabel.PAGE_FOOTER: 'page_footer'>, prov=[ProvenanceItem(page_no=1, bbox=BoundingBox(l=472.82091, t=47.44734947265624, r=538.5809100000001, b=38.59134947265625, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 19))], source=[], comments=[], orig='Antenna House, Inc.', text='Antenna House, Inc.', formatting=None, hyperlink=None), SectionHeaderItem(self_ref='#/texts/10', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.SECTION_HEADER: 'section_header'>, prov=[ProvenanceItem(page_no=2, bbox=BoundingBox(l=56.69292, t=812.1528794726563, r=199.20792, b=795.5478794726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 19))], source=[], comments=[], orig='About Antenna House', text='About Antenna House', formatting=None, hyperlink=None, level=1), TextItem(self_ref='#/texts/11', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=2, bbox=BoundingBox(l=66.68792, t=787.1163394726563, r=296.27794, b=776.0463394726563, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 55))], source=[], comments=[], orig='See also https://www.antennahouse.com/for more details.', text='See also https://www.antennahouse.com/for more details.', formatting=None, hyperlink=None), TextItem(self_ref='#/texts/12', parent=RefItem(cref='#/body'), children=[], content_layer=<ContentLayer.BODY: 'body'>, meta=None, label=<DocItemLabel.TEXT: 'text'>, prov=[ProvenanceItem(page_no=2, bbox=BoundingBox(l=472.82091, t=47.44734947265624, r=538.5809100000001, b=38.59134947265625, coord_origin=<CoordOrigin.BOTTOMLEFT: 'BOTTOMLEFT'>), charspan=(0, 19))], source=[], comments=[], orig='Antenna House, Inc.', text='Antenna House, Inc.', formatting=None, hyperlink=None)] pictures=[] tables=[] key_value_items=[] form_items=[] pages={1: PageItem(size=Size(width=595.275634765625, height=841.8897094726562), image=None, page_no=1), 2: PageItem(size=Size(width=595.275634765625, height=841.8897094726562), image=None, page_no=2)}

PDF Content Summary

Page 1 -> sectionheaderitem, text
Page 2 -> sectionheaderitem, text

[hussainarslan]

@ceberam I looked into it and i found the problem to be that docling-parse extracts PDF hyperlink annotations into SegmentedPdfPage.hyperlinks, but the PDF pipeline never matches them to text clusters during page assembly. All TextItem.hyperlink fields remain None.

I have just made a pr ( #3131 ) addressing this issue. I have gone with the following fix strategy:
Spatially match hyperlink annotation rects against cluster bounding boxes during page assembly, aggregate coverage per URI for multi-rect links, and thread the resolved hyperlink through the reading order model to the final DoclingDocument.