GetDocumentAbstract API¶

OBSOLETE -" This API is retained for backward compatibility only. Use GetDocumentAbstract1 for new integrations. GetDocumentAbstract1 returns a consistent XML structure for both success and error cases; this API does not.

Returns the full-text abstract (indexed text content) of a specified version of a document.

Endpoint¶

/srv.asmx/GetDocumentAbstract

Methods¶

• GET /srv.asmx/GetDocumentAbstract?AuthenticationTicket=...&Path=...&VersionNumber=...

• POST /srv.asmx/GetDocumentAbstract (form data)

• SOAP Action: http://tempuri.org/GetDocumentAbstract

Parameters¶

Version Number Format¶

infoRouter uses a large-integer version numbering scheme where version numbers pack a major, a minor and a revision into one integer as major * 1000000 + minor * 1000 + revision, so the first version is 1000000 and the second is 1000001 - not 2000000, which would be major version 2. Pass 0 to get the published version's abstract, or the latest version's when the document has never been published.

Response¶

Warning: The success and error responses use different XML structures -" this inconsistency is the primary reason this API is considered obsolete. See GetDocumentAbstract1 for a consistent response format.

Success Response¶

On success, the abstract text is returned inside an <abstract> child element:

<response success="true" error="">

  <abstract AppliedById="3" AppliedBy="SYSTEM" DateApplied="2024-03-15T14:32:07.000Z">This document covers the Q1 2024 financial results including

revenue figures, expense breakdowns, and year-over-year comparisons...</abstract>

</response>

Who wrote the abstract¶

The element carries the author, in the same three attributes GetPropertySets uses on its <Log> node and GetDocumentSummary uses on its <Value>.

An abstract is always produced by the server - from the email parser or the content index, never typed - so in practice this is the system account. It is read from the stored row rather than assumed, which is why all three are absent on an abstract written before the author was recorded, and on a - where there is none.

Error Response¶

On any error, a flat <response> element is returned without the <abstract> child:

<response success="false" error="[900] Authentication failed" />

This structural difference between success (has <abstract> child) and error (flat element with only attributes) is what the obsolete warning refers to.

Required Permissions¶

The calling user must have at least read access to the document. The abstract is extracted from the full-text search index; if the document has not been indexed, the returned abstract may be empty.

Example¶

GET Request¶

GET /srv.asmx/GetDocumentAbstract

  ?AuthenticationTicket=3f2504e0-4f89-11d3-9a0c-0305e82c3301

  &Path=/Finance/Reports/Q1-2024-Report.pdf

  &VersionNumber=0

HTTP/1.1

POST Request¶

POST /srv.asmx/GetDocumentAbstract HTTP/1.1

Content-Type: application/x-www-form-urlencoded

AuthenticationTicket=3f2504e0-4f89-11d3-9a0c-0305e82c3301

&Path=/Finance/Reports/Q1-2024-Report.pdf

&VersionNumber=0

SOAP Request¶

<soap:Envelope xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/"

               xmlns:tns="http://tempuri.org/">

  <soap:Body>

    <tns:GetDocumentAbstract>

      <tns:AuthenticationTicket>3f2504e0-4f89-11d3-9a0c-0305e82c3301</tns:AuthenticationTicket>

      <tns:Path>/Finance/Reports/Q1-2024-Report.pdf</tns:Path>

      <tns:VersionNumber>0</tns:VersionNumber>

    </tns:GetDocumentAbstract>

  </soap:Body>

</soap:Envelope>

JavaScript¶

Every call answers XML with HTTP 200, success or not, so success is the thing to branch on and errorCode is the number to report.

async function call(action, params) {
  const response = await fetch(`/srv.asmx/${action}?${new URLSearchParams(params)}`);
  const root = new DOMParser()
    .parseFromString(await response.text(), 'text/xml')
    .documentElement;

  if (root.getAttribute('success') !== 'true') {
    throw new Error(`${root.getAttribute('errorCode')}: ${root.getAttribute('error')}`);
  }
  return root;
}

Reads the indexed text of one document version, in an <abstract> element.

const root = await call('GetDocumentAbstract', {
  authenticationTicket: ticket,
  Path: '/Finance/Reports/Q1.pdf',
  VersionNumber: 0            // 0 means the published version
});

const element = root.querySelector('abstract');
console.log(element.textContent,
            element.getAttribute('AppliedBy'),
            element.getAttribute('DateApplied'));

The only difference between the two is the element name. GetDocumentAbstract writes the text into <abstract>, which is what it has always done; GetDocumentAbstract1 writes the identical text, with the identical attributes, into <Value> - the element the rest of the API uses for a single returned value. Same parameters, same content, same author attributes. Prefer the numbered one in new code and keep reading <abstract> in anything already written against it.

Notes¶

• This API is obsolete. Use GetDocumentAbstract1 for all new integrations.

• The abstract text is sourced from the full-text search index. If the document has not been indexed, the abstract may be empty.

• VersionNumber=0 retrieves the abstract for the published version of the document, or for its latest version when it has never been published.

• Version numbers between 1 and 999,999 are rejected. Use 0 for the published version, or a number a document actually carries - the first version is 1000000 and the second 1000001.

• Both full infoRouter paths and short document ID paths (~D{id} or ~D{id}.ext) are accepted for the Path parameter.

Related APIs¶

• GetDocumentAbstract1 - Preferred replacement with consistent XML response structure

• GetDocument - Get full document metadata and properties

• GetDocumentVersions - Get the version history list for a document

Error Codes¶

The errorCode values this operation returns, checked against a running server:

A call with no ticket signs in as the anonymous user, so a document in a library flagged as anonymous can be read without authenticating.