SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2408.09869 · 2024

Docling Technical Report

arXiv · PDF · Open in the Atlas

Code that ran

6 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 6 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2408.09869.json · badge: /api/ran/2408.09869.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
DS4SD/docling-core canonical (classifier label)
classifier basis: census
5 ran · 1 no recorded run
DS4SD/docling canonical (classifier label)
classifier basis: census
1 ran · 0 no recorded run
FunctionStatusWhere it lives
docling_document_to_legacy Ran
ran; our contract draft was wrong, not the code
DS4SD/docling-core/docling_core/utils/legacy.py
code served (permissive licence) · get_code("7b2fb2c0eebc6858")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned _NullLegacyDocument. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from typing import Any

class MockDocument:
    """Mock document that mimics a docling document structure."""
    def __init__(self):
        self.main_text = []
        self.content = "Sample document content"
        self.metadata = {}

def docling_document_to_legacy(doc):
    """Function to convert docling document to legacy format."""
    class _NullLegacyDocument:
        main_text: list[object] = []

        def _resolve_ref(self, item):
            return item

        def export_to_markdown(self, *args, **kwargs):
            return ''

        def export_to_document_tokens(self, *args, **kwargs):
            return ''
    
    return _NullLegacyDocument()


# Create a mock document object to pass to the function
mock_doc = MockDocument()

ARGS = {
    'doc': mock_doc
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type _NullLegacyDocument — a _NullLegacyDocument output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a _NullLegacyDocument output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
ensure_unique_list Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
DS4SD/docling-core/docling_core/utils/validators.py
code served (permissive licence) · get_code("a52e30781c6c4cfc")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

# Construct arguments for ensure_unique_list
# The function expects a list and deduplicates it while preserving order

ARGS = {
    "values": [1, 2, 3, 2, 1, 4, 3, 5]
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_input_rejection_cause Ran
ran; our contract draft was wrong, not the code
DS4SD/docling/docling/datamodel/document.py
code served (permissive licence) · get_code("41f5effd5983f81d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ValueError. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from typing import Optional, NamedTuple

class RejectionInfo(NamedTuple):
    original_error: Optional[BaseException]

class InputDocument:
    def __init__(self, rejection: Optional[RejectionInfo] = None):
        self._rejection = rejection

# Test case 1: No rejection (None)
in_doc_valid = InputDocument(rejection=None)

# Test case 2: With rejection containing an original error
test_error = ValueError("Test validation error")
rejection_info = RejectionInfo(original_error=test_error)
in_doc_rejected = InputDocument(rejection=rejection_info)

# Use the rejected document as the main argument
ARGS = {
    "in_doc": in_doc_rejected
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ValueError — a ValueError output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ValueError output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
resolve_remote_filename Ran
ran; our contract draft was wrong, not the code
DS4SD/docling-core/docling_core/utils/file.py
code served (permissive licence) · get_code("8ff59a7c7a42ae57")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

from pathlib import Path
from pydantic import AnyHttpUrl

# Create a valid AnyHttpUrl instance
http_url = AnyHttpUrl("https://example.com/path/to/document.pdf")

# Create response headers dict
response_headers = {
    'Content-Disposition': 'attachment; filename="myfile.txt"',
    'Content-Type': 'text/plain'
}

# Create fallback filename
fallback_filename = 'file'

ARGS = {
    'http_url': http_url,
    'response_headers': response_headers,
    'fallback_filename': fallback_filename
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pydantic — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
validate_datetime Ran
ran; our contract draft was wrong, not the code
DS4SD/docling-core/docling_core/utils/validators.py
code served (permissive licence) · get_code("36b91c9280a3d9fb")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned datetime. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

from datetime import datetime

def handler(v):
    """Simple handler that returns the value."""
    return v

ARGS = {
    "v": datetime(2024, 1, 15, 10, 30, 45),
    "handler": handler
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type datetime — a datetime output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a datetime output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
validate_unique_list Ran
ran; our contract draft was wrong, not the code
DS4SD/docling-core/docling_core/utils/validators.py
code served (permissive licence) · get_code("be7da761ed51ee1f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from collections.abc import Hashable
from typing import Any, TypeVar
from pydantic_core import PydanticCustomError

torch.manual_seed(0)

# Test case 1: Valid unique list of integers
valid_unique_list = [1, 2, 3, 4, 5]

ARGS = {
    "v": valid_unique_list
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pydantic_core, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
generate_json_schema Not yet run
unverified (harvested, no recorded run)
DS4SD/docling-core/docling_core/utils/generate_jsonschema.py
code served (permissive licence) · get_code("652c46e5dea51af6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 7 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

This technical report introduces Docling, an easy to use, self-contained, MIT-licensed open-source package for PDF document conversion. It is powered by state-of-the-art specialized AI models for layout analysis (DocLayNet) and table structure recognition (TableFormer), and runs efficiently on commodity hardware in a small resource budget. The code interface allows for easy extensibility and addition of new features and models.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2408.09869")
get_code_for_paper("2408.09869")
have("2408.09869")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2408.09869.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2408.09869.svg)](https://syntology.ai/paper/2408.09869)

Connect an agent — have() is free.