Preparing Texts

This section speaks broadly to preparing different document types for ingestion into Manifold. For best practices on authoring in the various formats with attention paid to individual elements common to scholarly writing, consults the Writing for the Web section. For a walk-through on ingesting material into Manifold as a text, see the Customizing section.

Once you have created a project, you may add any number of prepared texts into it. Manifold supports all of the following file types for text ingestion. Click on a type below for preparation suggestions specific to that format.

Type Extension Available Embeds
EPUB .epub
  • Images
  • Audio
  • Video
  • iFrames
Markdown .md
  • Images
  • Audio
  • Video
  • iFrames
HTML .htm or .html
  • Images
  • Audio
  • Video
  • iFrames
Google Docs url
  • Images
Word .docx
  • Images

Some of these formats allow for different media to be embedded with in them as components of the source text instead of as ancillary resources.

Instructions on how to encode embeds available to EPUBs as well as Markdown and HTML files will be described in the HTML section.

Thoughts on how embedded media in Google and Word docs will be treated in their respective sections.

Note. In the table above, Images refers to static renderings as well as animated GIFs.

There are two means to add a new text into Manifold: where a text is created from a single document and where a single text is created from a collection of different documents. These are, respectively, the Document and Manifest ingestion strategies described in this section.


The document ingestion strategy allows a user to create a new text in Manifold from a single document source.

Document Strategy

If the source document’s headers are well defined, Manifold will recognize them and include them in the Contents dropdown. If there are no headers or if the headers that are present do not include any semantic markup, the Contents dropdown will show as empty.

See the sections specific to the individual formats for detailed instructions on how to format headers to render as expected in the Contents dropdown.

Unstructured Document

The title of the document will automatically populate the Title field in the backend and display in the reader’s title bar. The title bar alternates between the title of the text and the title of the active text section (e.g., book title versus chapter title) for structured documents. For unstructured documents, the system will assume the title supplied during ingest to be both the title of the text and the title of the active text section. If a backend user adjusts the Title field after ingest, the title that was initially supplied will remain as the text section title, and the newly input title will serve as the text title.

The following table scopes out the source of the text title for each of the available ingestible file types when uploaded as a single document:

Source File Text Title Source
EPUBa The dc:title element in the content.opf file
Markdown Metadata Header
HTML Metadata Header
Google Doc Title of the document
Word Paragraph styled as Title

aEPUBs can only be ingested using the Document strategy but are themselves self-contained structured documents. As such the system will be able to ascertain the title of the individual text sections based off the metadata provided within the EPUB.


A Manifest upload can be composed of a collection of one file type or a mixture of them—excepting EPUBs. Thus, it is possible to create a single Manifold text from one or more Google Docs, Word files (saved in the DOCX format), Markdown files, or HTML files: different file types and sources combined to form one whole.

Manifest Strategy

When the various source documents have been prepared according to the requirements of their format, users should compress (or zip) all of the documents—along with the required YAML file described below—into a single archive, which can then be uploaded into Manifold. The structure of YAML file will be represented on the reader’s contents dropdown, and its content will inform what is displayed on the title bar. See the YAML section following for details.

Note. When creating a zip archive to upload into Manifold, it is better to individually select all of the files and then compress them than to simply select an entire folder on which to run the compression. The latter method can result in corrupted pathways that may cause ingestion errors.


In order for Manifold to properly ingest multiple source documents at once, the system requires a YAML file (extension .yml) to be included along with them in the compressed archive that will be loaded into the system.

In cases where there are no source documents to upload (e.g., if you are creating a text from multiple Google Docs) then only the YAML file needs to be loaded. When uploaded alone, the YAML file does not need to be compressed or zipped.

The YAML file can be named in accordance with the user's preference, so long as the extension remains .yml. Thus manifest.yml and contents.yml will both work in the system.

The YAML file is a means for users to tell the system how to organize the materials that are being ingested into the system and how they relate to one another.

If any of the fields require characters that are not letters, excepting source_path, enclose your input within single quotes as shown in the example below.

When creating or modifying a YAML file, be sure to use a plain text editor like Notepad (PC), Text Edit (Mac), Sublime Text (download), or (download).

Indentations matter in YAML, so prepare your file carefully, using two spaces for each indent. The example below can be copied and pasted for easy modification.

  title: 'Title for the Text: As a Whole'
  language: en
  rights: 'This text is licensed under a CC BY-NC-ND 4.0 license.'
  description: 'This text considers a, b, and c.'
    - name: Rowan Ono
    - name: Max Duo
    - name: Ida Davis
    - name: Danielle Idra
  - label: Title of First Section
    source_path: chap_1.html
    start_section: true
      - label: Title of Second Section
          - label: Title of Next Section
  - label: Title of Another Section
    source_path: chapt_2.html
  - label: Title of One More Section

Meta Block

  • The meta block allows users to describe the text that is being created as a whole.
  • The title field in the meta block will be understood by Manifold as being the title of the text as whole.
  • Authors (creators) and contributors cannot be assigned to specific sections within the text being created; instead include all authors and contributors in the meta block.
  • Other than creators and contributors, fields can only have one input, as shown here.
  • The language abbreviation comes from the ISO 639-1 standard.

TOC Block

  • The TOC block allows users to name, order, and locate the source files for the system.
  • The input for the label serves as the title of the text section and will appear in both the contents dropdown in the reader as well as on the title bar.
  • The source_path tells Manifold where to find the file: it’s location in the compressed archive or from a remote host. If there are no folders within the archive being compressed then the location is simply the name of the file. If the file is nested in a folder, it would be {folder-name}\{filename.extension} or, for example part1\
  • The start_section describes what part of the text the Manifold reader will open to. The field can only be used once and can be associated with any section. Thus a user can have the reader open to chapter 1 instead of, say, a preface.
  • The children tag allows users to nest content. This nesting will also be rendered as expected on the Contents dropdown. It is important that nested content be spaced as is shown in the sample above, with each level of nesting described by an indent of two spaces). Manifold can render nesting up to three levels in the contents dropdown. See the Nested Text Sections following to learn how advanced users can make text headings navigable in the Contents dropdown.

Nested Text Sections

Advanced users who want to include headings as navigable items in the Contents dropdown can do so for any HTML or Markdown texts in the Manifest. This will require that each heading designated to appear in the Contents dropdown include an id attribute that matches a hashed source pathway in the YAML file. See the preparing HTML section for details. The instructions and coding in that section are appropriate for both HTML and Markdown texts.

Note. It is not presently possible to map specific headings from individual Word or Google docs into the reader's Contents dropdown.

Document Metadata

Each source document being uploaded to Manifold can include the following metadata. When ingesting single documents all of the metadata can be included in the upload. When ingesting through the Manifest strategy, the source document should only include title metadata, with the rest included in the YAML file.

  • Title
  • Date
  • Language
  • Rights
  • Description
  • Creator
  • Contributor

There can be multiple creators (authors) and contributors for each upload. The system will automatically attempt to match existing records of the same name. If none is found, then the system will create a new record.

Following are related themes that link to other pertinent sections concerning how you work with texts in the system: