使用Python向Apache Jena Fuseki上传大RDF文件遇表单过大错误

我按官方文档搭建了独立运行的Apache Jena Fuseki服务器，localhost:3030访问正常。Python代码能上传小体积.ttl格式RDF文件，但大文件上传失败，服务器日志报错：

Caused by: java.lang.IllegalStateException: form too large > 20000000
at org.eclipse.jetty.server.FormFields.checkMaxLength(FormFields.java:318) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.server.FormFields.parse(FormFields.java:307) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.server.FormFields.parse(FormFields.java:39) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.io.content.ContentSourceCompletableFuture.parse(ContentSourceCompletableFuture.java:104) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.server.handler.ContextHandler$ScopedContext.run(ContextHandler.java:1212) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.server.handler.ContextRequest$OnContextDemand.run(ContextRequest.java:74) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.util.thread.SerializedInvoker$Link.run(SerializedInvoker.java:191) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.server.internal.HttpConnection$DemandContentCallback.succeeded(HttpConnection.java:679) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.io.FillInterest.fillable(FillInterest.java:99) ~[fuseki-server.jar:5.0.0]
at org.eclipse.jetty.io.SelectableChannelEndPoint$1.run(SelectableChannelEndPoint.java:53) ~[fuseki-server.jar:5.0.0]

最初采用SPARQL INSERT DATA方式上传，代码如下：

input_location = "C:/......../Added_Triples.ttl"

with open(input_location, 'r') as f:
    content = f.read()

rdf_string_no_prefixes = "\n".join(line for line in content.split("\n") if not line.startswith("@prefix"))

update_query = """ 
    PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
    PREFIX CSRO: <http://www.semanticweb.org/aagr657/ontologies/2023/9/CraneSpaceRepresentationOntology#>
    PREFIX LinkOnt: <http://purl.org/ConstructLinkOnt/LinkOnt#>
    PREFIX bot: <https://w3id.org/bot#>
    PREFIX expr: <https://w3id.org/express#>
    PREFIX geo: <http://www.opengis.net/ont/geosparql#>
    PREFIX geom: <http://rdf.bg/geometry.ttl#>
    PREFIX ifc: <https://standards.buildingsmart.org/IFC/DEV/IFC2X3/TC1/OWL>
    PREFIX inst: <https://www.ugent.be/myAwesomeFirstBIMProject#>
    PREFIX owl: <http://www.w3.org/2002/07/owl#>
    PREFIX sf: <http://www.opengis.net/ont/sf#>
    PREFIX omg: <https://w3id.org/omg#>
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
    PREFIX lbd: <https://linkedbuildingdata.org/LBD#>
    PREFIX props: <http://lbd.arch.rwth-aachen.de/props#>
    PREFIX unit: <http://qudt.org/vocab/unit/&gt;
    PREFIX IFC4-PSD: <https://www.linkedbuildingdata.net/IFC4-PSD#>
    PREFIX smls: <https://w3id.org/def/smls-owl#>
    PREFIX fog: <https://w3id.org/fog#>
    PREFIX cc: <http://creativecommons.org/ns#>
    PREFIX dce: <http://purl.org/dc/elements/1.1/&gt;
    PREFIX express: <https://w3id.org/express#>
    PREFIX list: <https://w3id.org/list#>
    PREFIX vann: <http://purl.org/vocab/vann/&gt;
    PREFIX expr: <https://w3id.org/express#>
    PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
    PREFIX : <https://standards.buildingsmart.org/IFC/DEV/IFC2x3/TC1/OWL#>

    INSERT DATA {
        %s
    }
    """ % (rdf_string_no_prefixes)
sparql = SPARQLWrapper("http://localhost:3030/your-dataset/update")
sparql.setMethod(POST)
sparql.setQuery(update_query)

sparql.query()

尝试修改jetty.xml解决但未找到该文件，临时分块上传耗时过长，不想采用。之后换requests.post方式上传，代码如下：

import requests
file_location = "C:/.........../Added_Triples.ttl"
sparql_endpoint = "http://localhost:3030/construction_dataset_2/update"

headers = {'Content-Type': 'text/turtle;charset=utf-8'}
data = open(file_location, 'r').read()
response = requests.post(sparql_endpoint, headers=headers, data=data)

出现错误：

Exception has occurred: ConnectionError
('Connection aborted.', ConnectionAbortedError(10053, 'An established connection was aborted by the software in your host machine', None, 10053, None))
ConnectionAbortedError: [WinError 10053] An established connection was aborted by the software in your host machine

服务器日志显示：

11:40:03 INFO Fuseki :: [23] 415 Unsupported Media Type (0 ms)
11:40:03 INFO Fuseki :: [24] POST http://localhost:3030/construction_dataset_2/update
11:40:03 INFO Fuseki :: [24] 415 Unsupported Media Type (0 ms)
11:42:17 INFO Fuseki :: [25] POST http://localhost:3030/construction_dataset_2/update

需要无需分块、可一次性上传大RDF文件的解决方案。

1. 调整Jetty表单大小限制（适配SPARQL Update方式） #

Fuseki 5.x使用内嵌Jetty服务器，默认表单大小限制为20MB，可通过启动参数直接调整，无需寻找jetty.xml：

• Windows环境：修改fuseki.bat中的启动命令，添加Jetty参数：

  java -Dorg.eclipse.jetty.server.FormFields.maxSize=100000000 -jar fuseki-server.jar --update /your-dataset
  

• Linux/macOS环境：修改fuseki.sh中的启动命令，添加Jetty参数：

  java -Dorg.eclipse.jetty.server.FormFields.maxSize=100000000 -jar fuseki-server.jar --update /your-dataset
  

  其中100000000代表100MB，可根据实际文件大小调整为对应数值（如500MB写500000000）。

2. 使用Fuseki原生文件上传端点（推荐大文件） #

Fuseki提供专门的文件上传接口，比SPARQL INSERT DATA更高效，且原生支持大文件：

Python代码示例 #

import requests

file_path = "C:/......../Added_Triples.ttl"
# 上传到默认图，若需指定命名图，在URL后追加?graph=<你的命名图URI>
upload_endpoint = "http://localhost:3030/construction_dataset_2/data"

with open(file_path, 'rb') as f:
    # 采用multipart/form-data格式上传文件
    files = {'file': ('Added_Triples.ttl', f, 'text/turtle')}
    # 若需要覆盖目标图现有数据，添加参数?overwrite=true
    response = requests.post(upload_endpoint, files=files)

print(response.status_code)
print(response.text)

3. 修复直接POST Turtle数据的错误 #

之前的415错误是因为/update端点仅接受SPARQL Update请求，而非直接的Turtle数据。若要直接提交Turtle内容，需使用数据集的/data端点：

正确代码示例 #

import requests

file_path = "C:/......../Added_Triples.ttl"
data_endpoint = "http://localhost:3030/construction_dataset_2/data"

headers = {'Content-Type': 'text/turtle; charset=utf-8'}
# PUT会替换目标图所有数据，POST会追加数据；指定命名图加?graph=<URI>
with open(file_path, 'r', encoding='utf-8') as f:
    response = requests.post(data_endpoint, headers=headers, data=f)

print(response.status_code)
print(response.text)

注意：若文件过大，仍需先按方案1调整Jetty的请求大小限制，可同时添加两个参数确保覆盖：

java -Dorg.eclipse.jetty.server.FormFields.maxSize=100000000 -Dorg.eclipse.jetty.server.Request.maxFormContentSize=100000000 -jar fuseki-server.jar --update /your-dataset

内容的提问来源于stack exchange，提问作者AJAY AGRAWAL