MarkLogic含空格与特殊字符的URI创建失败求助

创建了一个MarkLogic转换（transform），目的是在数据入库时转换URL编码的[]和空格字符，但处理后的URI被判定无效，触发500错误REST-INVALIDPARAM。

相关代码 #

XQuery转换代码 #

xquery version "1.0-ml";

module namespace space = "http://marklogic.com/rest-api/transform/space-to-space";

declare function space:transform(
    $context    as map:map,
    $params     as map:map,
    $content    as document-node()
  ) as document-node()
{

    let $puts := (
        xdmp:log($params),
        xdmp:log($context),
        map:put($context, "uri", fn:replace(map:get($context, "uri"), "%5B+", "[")),
        map:put($context, "uri", fn:replace(map:get($context, "uri"), "%5D+", "]")),
        map:put($context, "uri", fn:replace(map:get($context, "uri"), "%20+", " ")),
        xdmp:log($context)
    )
    
    return $content
    
};

Python测试代码 #

def upload_document(self, inputContent, uri, fileType, database, collection):
        if fileType == 'XML':
            headers = {'Content-type': 'application/xml'}
            fileBytes = str.encode(inputContent)
        elif fileType == 'TXT':
            headers = {'Content-type': 'text/*'}
            fileBytes = str.encode(inputContent)
        else:
            headers = {'Content-type': 'application/octet-stream'}
            fileBytes = inputContent

        endpoint = ML_DOCUMENTS_ENDPOINT
        params = {}

        if uri is not None:
            encodedUri = urllib.parse.quote(uri)
            endpoint = endpoint + "?uri=" + encodedUri

        if database is not None:
            params['database'] = database

        if collection is not None:
            params['collection'] = collection

        params['transform'] = 'space-to-space'

        req = PreparedRequest()
        req.prepare_url(endpoint, params)

        response = requests.put(req.url, data=fileBytes, headers=headers, auth=HTTPDigestAuth(ML_USER_NAME, ML_PASSWORD))
        print('upload_document result: ' + str(response.status_code))

        if response.status_code == 400:
            print(response.text)

日志内容 #

1. 2023-02-13 16:59:00.067 Info: {}
2. 2023-02-13 16:59:00.067 Info:
   {"input-type":"application/octet-stream",
   "uri":"/Judgment/26856/supportingfiles/[TEST] 57_image1.PNG", "output-type":"application/octet-stream"}
3. 2023-02-13 16:59:00.067 Info:
   {"input-type":"application/octet-stream",
   "uri":"/Judgment/26856/supportingfiles/[TEST] 57_image1.PNG", "output type":"application/octet-stream"}
4. 2023-02-13 16:59:00.653 Info: Status 500: REST-INVALIDPARAM: (err:FOER0000)
   Invalid parameter: invalid uri:
   /Judgment/26856/supportingfiles/[TEST] 57_image1.PNG

问题分析 #

1. Transform代码逻辑无效：REST API在调用transform前已经自动解码了uri参数，所以fn:replace匹配不到%5B、%20这类编码字符，导致context中的uri没有被修改（日志显示处理前后uri一致）。
2. URI合法性问题：MarkLogic的REST API不允许文档URI包含空格，因此解码后的带空格URI触发REST-INVALIDPARAM错误。

解决方案 #

方案1：调整URI格式，替换空格为合法字符 #

如果不需要保留空格，修改transform将空格替换为下划线（MarkLogic允许[]在文档URI中）：

xquery version "1.0-ml";

module namespace space = "http://marklogic.com/rest-api/transform/space-to-space";

declare function space:transform(
    $context    as map:map,
    $params     as map:map,
    $content    as document-node()
  ) as document-node()
{
    let $originalUri := map:get($context, "uri")
    -- 替换空格为下划线，保留[]
    let $newUri := fn:replace($originalUri, "\s", "_")
    let $_ := map:put($context, "uri", $newUri)
    let $_ := xdmp:log(xdmp:concat("Updated URI: ", $newUri))
    return $content
};

方案2：优化Python代码的URI处理 #

如果需要保留空格（确认MarkLogic版本支持），让prepare_url自动处理URI编码，移除transform中无效的替换逻辑：

def upload_document(self, inputContent, uri, fileType, database, collection):
    if fileType == 'XML':
        headers = {'Content-type': 'application/xml'}
        fileBytes = str.encode(inputContent)
    elif fileType == 'TXT':
        headers = {'Content-type': 'text/*'}
        fileBytes = str.encode(inputContent)
    else:
        headers = {'Content-type': 'application/octet-stream'}
        fileBytes = inputContent

    endpoint = ML_DOCUMENTS_ENDPOINT
    params = {}

    if uri is not None:
        params['uri'] = uri  # 不手动编码，交给prepare_url自动处理

    if database is not None:
        params['database'] = database

    if collection is not None:
        params['collection'] = collection

    params['transform'] = 'space-to-space'  # 仅用于日志或其他非URI修改的处理

    req = PreparedRequest()
    req.prepare_url(endpoint, params)

    response = requests.put(req.url, data=fileBytes, headers=headers, auth=HTTPDigestAuth(ML_USER_NAME, ML_PASSWORD))
    print('upload_document result: ' + str(response.status_code))

    if response.status_code == 400:
        print(response.text)

方案3：直接用XQuery插入文档（绕过REST API限制） #

如果必须保留空格和[]作为URI，可直接调用XQuery的xdmp:document-insert插入，避免REST API的URI验证：

xquery version "1.0-ml";
declare variable $uri as xs:string external;
declare variable $content as document-node() external;

xdmp:document-insert($uri, $content);

通过Python调用MarkLogic的XQuery执行端点运行这段代码即可。

内容的提问来源于stack exchange，提问作者Eugene