Configuration manual for PageSeeder

Publication types

This feature requires PageSeeder v5.96 or later.


A “publication” is a hierarchical collection of  PSML component documents joined by “embed” or “transclude-type” XRefs.

The core of the feature is a process that seamlessly resolves the publication component into a single document for presentation or numbering. The following documentation explains the options for processing a publication.

For additional background, we recommend completing this tutorial. The challenges of document numbering are widely recognized and the sample files of the tutorial help to explain how PageSeeder addresses this issue.

System Concepts

Following are some concepts that are fundamental to PageSeeder publications.

Publication ID

Any PSML document with embed or transclude XRefs can become a publication by simply adding a publication ID via the document properties. A listing of all documents with a publication ID is visible through the Group Publications block. The document with the ID is known as the Publication Root.

Note: Future PageSeeder releases may have tools to manage publication IDs but at present it is the responsibility of users to ensure IDs are unique. To prevent name clashes, some implementations may need to manage IDs externally or use group-level prefixes.

Components / Shared components

Component documents are what the publication assembles into a single document for processing. In this model, multiple publications can reference (share) the same, individual component but presentation and numbering depend on which publication is current and the publication config used to create it.

embed and transclude

These are the only two XRef types processable in a publication.

References document type

Although it is not mandatory, the references document type is the logical format for the root document of a publication.

Publication config

Stores all configuration information for a given Publication Type in the following folder:


Current publication

The publication that is controlling the display of the component document currently in the document view. To see what other publications the component is included in, click in the box at the top of the Table of Contents.

Table of Contents (ToC)

In the PageSeeder interface, a publication ToC lists the high-level headings of every component and the complete headings from whatever component document is open in the Document view.

Publication headings

The <heading> element is a fundamental part of PSML and in a publication, it is the primary structural and navigational object for publications.

Publication paragraphs

In publications, all the properties of a <heading> element are available to the <para> element, with one important distinction – <para> elements do not display in the ToC.

Numbered vs Unnumbered

Both <heading> and <para> elements support an attribute of @numbered that is available through the number icon in the editing interface. Any object where this attribute has a value of "true" will be subject to processing by the publication-config.

Block Labels

The publication can also process objects wrapped in a <block> element and a value in the @label attribute that corresponds to a declaration in the publication config. Use cases for block labels include captions on figures or tables.

The location of the <block> element in the publication-config depends on the numbering requirements. It must be below any other counter needed for the object. If a figure caption needs the chapter number as a prefix – say Chapter 4: Figure 7 –  and the heading for chapter is in the config at the first level then the <block> element for figure should be declared in the config with a <scheme> element where the value of the @level attribute is at least “2”.

@level and @indent

In PSML, the <heading> element has a @level attribute and the <para> element has an @indent attribute. For the purposes of the publication configuration, unless the documentation is explicit, please consider any references to “level” as also meaning “indent”.

Microsoft Word

To reliably generate correctly numbered docx files, it is important to calibrate the Word template and the publication config.

Xref Config

These are project-level configurable templates that automatically format links such as references to numbered paragraphs, footnotes or citations.

<xref display="template" title="{prefix}">


Publications add a new dimension to PageSeeder document versions, especially documents that are part of multiple publications.

Best Practice

Getting the best results out of publications is easier under the following circumstances:

  1. Start every document with a heading level 1 then let the references document and publication process calculate the final level of heading. This is easier than trying to anticipate and adjust the position of the heading in the hierarchy
  2. Do not put any paragraphs above the level 1 heading in a document.
  3. To include single fragments in a publication, use XRef-type of transclude, not “embed”
  4. Whenever using an XRef-type of “embed” the link cannot be one-way, it must reverse.


The in-built “best practice” schema will validate for these issues.

Publication management

Following are the details for some routine tasks:

Create a publication

To create a publication requires a publication ID. These can consist of letters, numbers, hyphens or underscores up to 250 characters long. This value must be unique for the PageSeeder server (similar to Document ID).

To create a publication from an existing set of documents follow these steps:

  1. Navigate to the root document and click the Properties tab.
  2. Enter an ID of your choice under Publication ID (usually less than 20 characters).

  3. Unless users have specific instructions to choose a Publication type, ignore this field and let the system assign the default type.

  4. Click on Save.


Now click the View tab on the horizontal menu, then open the Table of contents block to the left of the document. This will display all the heading across the publication files – see below.


Delete a publication

To delete a publication, clear the value from the publication ID field on the Properties page.


This will also delete all the versions associated with the publication.

Change the publication root

To change the root document of a publication, navigate to the new root document and enter the publication ID on the Properties page. To do this the user must also have contributor access to the old root document.

Publication configuration

Each publication has a type that specifies the levels and numbering. If no value has been selected through the Publication Type field on the Properties page of the root document, then the type will be default.

To define a custom publication-type needs a publication-config.xml file in the . To create this file, do the following:

  1. In the Developer perspective, go to Document config under the Dev tab on the horizontal menu.

  2. Select Create publication type and enter a name using only letter, number and underscore characters.

  3. Select the create icon next to publication-config.xml.

  4. Modify the publication (see options below).

  5. Save the publication config file.

  6. Navigate to the publication root and click the Properties tab.

  7. In the Publication type field, select the new type.

  8. Save the changes.


Select reload after any edits. This will clear the cache and regenerate the tables of contents with the new settings.


Below is the default publication-config.xml used by PageSeeder.

  <toc title-collapse="always" />
  <levels xref-relative-to="heading"
      para-adjust="numbering" />
      <scheme level="1" type="decimal"   format="[1.]" />
      <scheme level="2" type="decimal"   format="[1.][2]" />
      <scheme level="3" type="decimal"   format="[1.][2.][3]" />
      <scheme level="4" type="decimal"   format="[1.][2.][3.][4]" />
      <scheme level="5" type="decimal"   format="[1.][2.][3.][4.][5]" />
      <scheme level="6" type="loweralpha"  format="[(6)]"
            element="para" />
      <scheme level="7" type="lowerroman"  format="[(7)]"
            element="para" />
      <scheme level="8" type="upperroman"  format="[(8)]"
            element="para" />
      <scheme level="9" type="upperalpha"  format="[9]"
            element="para" />

<toc> element

This element is optional and can have the following optional attribute:

AttributeDescriptionDefault value

As a default, PageSeeder documents have the same value in the title and the first heading of the document. This option avoids that content appearing twice in the ToC. Allowed values are:

  • always – if the first heading in a document is above all other headings, then always collapse it;
  • auto – if the first heading in a document is above all other headings and it matches the document title, then collapse it;
  • never – never collapse the headings.

<levels> element

When processing the publication components into a single instance, this element controls the resolution of the levels. This can affect are the numbering and presentation of headings, paragraphs and XRefs. This element is optional and can have the following optional attributes:

AttributeDescriptionDefault value

Calculates the link title by processing the heading levels of the target document. This is important when the title includes a generated number, for example “see 2.4.a – Operations and Maintenance”. Allowed values are:

  • heading – this option treats the XRefs target as a child of current heading in the source document. With this value, resolving an XRef below a heading 3 in the source document, will promote the headings and numbered paragraphs in the target by three levels.
  • document – this option does not acknowledge where the XRef exists in the content of the source document. It only adjusts the heading and paragraph levels in line with the document setting in the ToC. If the source document is in the ToC with an increase of 1 level, the resolution of the target will promote all heading and numbered paragraphs by one increment.

Determines how paragraph indents are calculated when the documents in the publication are consolidated. Allowed values are:

  • heading: indents under a heading 3 in the source document, will increase the @level by 3  in the target
  • document: relative to containing document, i.e. paragraphs take the level of their @indent.

Indicates whether to adjust the @level attribute on the <heading> element. Allowed values are:

  • numbering: adjust for numbering only.
  • content: adjust for numbering and in document content.

Indicates whether to adjust the @indent attribute on the <para> element. Allowed values are:

  • numbering: adjust for numbering only.
  • content: adjust for numbering and in document content.

<numbering> element

The <numbering> element applies numbering schemes to PSML document-types or documents that have specific document labels. Without this element, there will be no generation of numbers in the publication. The <numbering> element supports the following optional attributes:

AttributeDescriptionDefault value

Non-consecutive, or “skipped” heading levels are when a <heading level="3"> follows <heading level="1">. Processing options include the following:

  • 1 – the publication will make the missing levels a “1” – e.g. 2.1.3
  • 0 – the publication will make the missing levels a “0” – e.g. 2.0.3
  • strip – the publication will remove any reference to missing levels – e.g. 2.3

Numbering only applies to documents with the specified document-label and the document-label must exist on both the source and target documents.

<scheme> element

Each @level attribute on this element represents the intersection of content and the properties of the publication. All @level references are explicit. There is no way to imply headings or paragraph levels, so content with no declaration in the scheme becomes part of the “skipped-levels” processing. (see <numbering> element above).

This element can have the following attributes:

AttributeDescriptionDefault value

numbering level (required).


Each level format (required).

Defined by one or more patterns [x(level)x] where:

  • (level) – defines the level of numbering from1 to 9;
  • x – is a string of static content to decorate the numbering. For example, “Figure –“ before each image number or “Chapter ” before each heading 1.

The value of a @level attribute can only be once per hierarchy. To use more than once requires each instance have a different block-label.

For example, a specification can process the level 2 numbering differently in the Appendix than the body, but one of the numbering schemes must have a block-label for disambiguation.


The expression of numbering (optional). Allowed values are:

  • decimal
  • upperalpha
  • loweralpha
  • upperroman
  • lowerroman.

The PSML element that each numbering level applies to (optional). Allowed values are:

  • heading
  • para
  • any


DocX does not support "any".


Spawns a separate numbering stream when this value corresponds to the value of the @label attribute on a <block> element.

Designed to support use cases such as numbering figures and tables separate to paragraphs or headings (optional). For example, counting tables consecutively, say, 1 to 100 instead of Table 1.1, Table 3.1 and so on.

The block label matches the @level attribute on the editing view of the source document and PageSeeder manages the properties after the publication processing. There is no need for the attribute value to consider the post-processing value.


  <toc title-collapse="always" />
  <levels xref-relative-to="heading"
      para-adjust="content" />
      <scheme level="1" type="decimal"  format="[1.]" />
      <scheme level="2" type="decimal"  format="[1.][2]" />
      <scheme level="3" type="decimal"  format="[1.][2.][3]" />
      <scheme level="4" type="decimal"   format="[1.][2.][3.][4]" />
      <scheme level="5" type="decimal"    format="[1.][2.][3.][4.][5]" />
      <scheme level="6" type="loweralpha"  format="[(6)]" />
      <scheme level="7" type="lowerroman"   format="[(7)]" />
      <scheme level="8" type="upperroman"    format="[(8)]" />
      <scheme level="9" type="upperalpha" format="[9]" />
  <numbering skipped-levels="strip" document-label="detail">
      <scheme level="1" type="decimal"     format="[1.]" />
      <scheme level="2" type="decimal"    format="[1.][2.]" />
  <scheme level="2" type="decimal"  format="[Table 1-][2]"  block-label="table-caption" />
  <scheme level="2" type="upperalpha" format="[Fig 1-][2]"  block-label="figure-caption" />
      <scheme level="3" type="decimal"   format="[1.][2.][3.]"  element="any"/>
      <scheme level="4" type="decimal"       format="[1.][2.][3.][4.]" element="any"/>
      <scheme level="5" type="loweralpha"     format="[(5)]"     element="para"/>
      <scheme level="6" type="lowerroman"      format="[(6)]"     element="para" />
      <scheme level="7" type="upperalpha"     format="[(7)]" />
      <scheme level="8" type="upperroman"      format="[(8)]" />
      <scheme level="9" type="loweralpha"      format="[9]" />

docx support

It is important to understand the consequences of the setting in the publication-config.xml file and their effect on Microsoft Word or applications that can process docx file format such as Adobe InDesign.

In many cases, getting the desired results from a publication may require coordinating the publication config and the docx template in the context of a third-party application.

The intention of the publication properties is not necessarily to transfer these settings to docx or other formats. In many cases, the intention of these properties is to represent the same logic so that the people editing have a representation that is closer to their final format. This may be something as simple as image numbers, so references are the same in the editing view as they will be in the final format. For example, [Fig 1-2] instead image-456.jpg.

To calibrate the publication with a docx template or an InDesign import may require some deep understanding of non-PageSeeder technologies. In addition to the publication-config, users may need to configure the Default Numbering in the PageSeeder word-export-template.doc. See How to configure auto-numbering for additional information.

As many people will know already, applications like Word and InDesign are powerful, but complex. The good news is that PageSeeder publications can support very demanding numbering and done properly, users can expect the numbering in their Word documents to be correct every single time they export. Unfortunately, this needs Word skills, not just PageSeeder.

Contact Allette Systems to discuss your requirements.

Versioning a publication

All the component documents in a publication can be versioned by taking the following these steps:

  1. Navigate to the publication root document and click the Version tab.
  2. Click on Create publication version and enter the new version number.
  3. Click Submit.

Refresh the version page and mouse over the identifier for the new version. It should display something like Publication ID: [mypub1] in the pop-up box.

To create a publication version for a single document, do the following:

  1. Navigate to a document in the publication and select the Version tab.

  2. Select Create version and enter an identifier for the version.

  3. Select a Publication from the drop-down menu.

  4. Submit.

In the change history of a fragment, publication versions have the publication ID in [square brackets] after the version identifier.

Exporting a publication

To export a publication version, follow these steps:

  1. Select Developer perspective from the cube icon (top left).
  2. Navigate to the publication root (alternatively, where a publication is the content of a single folder, navigate to the location).
  3. For the root document, select the Export icon from the action block the upper left of the Document view. (for the Folder view, select Export from icon to the right of the folder)
  4. In the root document, select Export as PSML and enter an appropriate number under Maximum depth of forward xrefs.
  5. If applicable, select Version to export.
  6.  Select the appropriate value under Publication ID for version. The publication ID must correspond to the publication version or compare version. A publication ID with no version or compare version will export the original uploaded version.
  7.  Click Export.

Created on , last edited on