The International Journal of Interactive Multimedia and Artificial Intelligence (IJIMAI) is a quarterly journal which provides an interdisciplinary forum in which scientists and professionals can share their research results and report new advances on artificial intelligence tools, theory, methodologies, systems, architectures integrating multiple technologies, problems including demonstrations of effectiveness, or tools that use AI with interactive multimedia techniques.

This template allows researchers to write and send papers to IJIMAI, directly written in Typst! IJIMAI is the first journal ever to accept contributions authored in Typst.

Quick start

To use it in the web application, find the ijimai template when creating a project. To learn about creating a project, see documentation.

To use it with the CLI, run typst init @preview/ijimai:1.0.0 in a terminal.

The project example and structure can also be found in the template directory.

The template requires providing configuration, bibliography, and authors’ photos. Additionally, the project must include 5 required headings and a first paragraph (at least one word):

#import "@preview/ijimai:1.0.0": *
#show: ijimai.with(
  config: toml("paper.toml"),
  bibliography: "bibliography.yaml", // or .bib file
  read: path => read-raw(path),
)

= Introduction
Typst is a new markup-based typesetting system for the sciences. It is designed
to be an alternative both to advanced tools like LaTeX and simpler tools like
Word and Google Docs.

= CRediT authorship contribution statement
// Content for this section is generated automatically (this comment can be
// removed).

= Data statement
= Declaration of conflicts of interest
= Acknowledgment

All the needed information is written in a TOML configuration file.

If you do not have the Unit OT font, you can download it from the repository. Once done, upload the UnitOT-Regular.otf and UnitOT-LightItalic.otf files to your project directory in the Typst web application. If you are using Typst locally, install font files in your operating system so that they are available to any program, alternatively use --font-path (paired with --ignore-system-fonts) flag and pass the directory with the downloaded fonts (see typst compile --help).

Documentation

This section describes the usage/features of the template with an accent on common issues.

The template includes a lot of automations that have different degrees of potentially causing some issues (due to human error or nature of automations’ implementation). This is why this documentation tries to thoroughly go through all the special features and be transparent about how they work (or might not work).

[!NOTE] If you find any discrepancy in the behavior described below and the actual behavior when using the template, please, open an issue in the repository with the problematic example attached. If you are getting compilation errors, please, make sure to read the documentation carefully and fully before creating an issue.

Heading names

The template uses titleize package to format all headings in title case. This means that manually recreating title case is unnecessary. Look inside the template or at the snippet in the Quick start for proffered (but not enforced) way of writing heading names.

Required sections

Each paper must contain next sections (in the order of listing):

• Introduction
• CRediT Authorship Contribution Statement
• Data Statement
• Declaration of Conflicts of Interest
• Acknowledgment

The title case used for headings is not necessary in the source file. Refer to Heading names above for details.

The “Introduction” section is the very first section in a paper, while the other 4 sections are the very last ones. Additionally, the “Acknowledgment” section allows for next alternative spellings (but prefer the one above):

• Acknowledgments
• Acknowledgement
• Acknowledgements

To see description for the (last 4) sections, refer to the documentation test.

Introduction section and first paragraph

The template treats the very first paragraph (below Introduction heading) specially. The first word is being heavily styled. It is uppercased, its color and weight are modified, and on top of that a dropcap function from the droplet package is applied to it. However, in order for dropcap to work correctly (span multiple lines down), it needs access to the whole paragraph.

For this reason, first-paragraph function exists. But unlike previous versions, it is no longer required. One reason being interchangeability: to use this or another template simply change the template function, while keeping the document content completely unchanged. Another reason, is a newly developed algorithm (par show rule) that can automatically detect and modify only the very first paragraph. Due to shear complexity of the algorithm, there are some caveats laid out below.

The par show rule’s first paragraph detection is based on the concept of location proximity, examples of which can be found in https://github.com/typst/typst/issues/2953. The anchor, that help find the paragraph is the “Introduction” heading element. It can be safely used as an anchor due to being a required heading (see Required sections). Because spacing between the heading and the paragraph (as well as heading/paragraph font size) is controlled by the template, this approach becomes more reliable, enabling the use of a fixed value for proximity detection.

If a paragraph is modified through a par show rule, the new value will create a uniquely new paragraph, causing the show rule to be applied to it as well (despite it being the result of modifications). To detect that the first paragraph styling was already applied (or not applied at all, more on this later), a state (counter) is used. This in turn means context must be utilized inside of the show rule. At present, Typst’s context + state behavior can quickly get out of hand if not handled with care. Still, because of context requiring at least 2 consecutive compilations to converge, some behavior can become impossible to achieve, or it can be very finicky. And the first paragraph show rule indeed behaves a bit strangely (but not for the user). For example, due to some strange context behavior it is impossible (?) to detect whether the first letter uses uppercase or lowercase. Having said that, the par show rule in its current implementation behaves as excepted, passing all the tests.

By default, the first paragraph should be added normally, without using the first-paragraph function. However, in the case when the The "first-paragraph" function must be used exactly once error appears in the presence of the first paragraph, as a fallback method, first-paragraph function must be used. Because of the aforementioned use of state, explicitly using this function increases the counter, automatically disabling the internal automatic use of this function (i.e., implicit use of the function). By using the function explicitly, any issue with paragraph parsing is eliminated.

If this or another related error is still present, completely disabling the show rule might help, especially if the document uses its own par show rule (which might break because of the template also overriding this element’s behavior). To disable it, set auto-first-paragraph parameter of the ijimai function to false (set to true by default).

CRediT section and roles

The “CRediT Authorship Contribution Statement” section needs to have a specific content with a specific formatting. To further remove the manual labor and human error, this section is entirely automated, which is why it must be left empty in the source file (only the heading is added).

The section’s content is entirely generated from the configuration file. Each author must have the credit key the value for which is an array of strings. Each string element represent a single contributor role, which are defined in ANSI/NISO Z39.104-2022. The configuration must use specific (simplified) aliases to the official roles, which are as follows:

#(
  conceptualization: [Conceptualization],
  data-curation: [Data curation],
  formal-analysis: [Formal analysis],
  funding-acquisition: [Funding acquisition],
  investigation: [Investigation],
  methodology: [Methodology],
  project-administration: [Project administration],
  resources: [Resources],
  software: [Software],
  supervision: [Supervision],
  validation: [Validation],
  visualization: [Visualization],
  writing-original-draft: [Writing -- original draft],
  writing-review-editing: [Writing -- review & editing],
)

An example of a valid list of roles for an author:

credit = ["conceptualization", "methodology", "resources", "validation"]

Each author should have at least 1 role specified. If an invalid name is provided, the template will give a corresponding error with a list of allowed string values (roles).

Even though this automation is achieved by creating the section content from within the heading show rule, both Typst’s outline (not used in the template) and PDF’s Document Outline are not affected. Which means neither of them will contain content of the section.

Short author list

The short author list is a short piece of text used in the citation box at the bottom of the first page.

By default, the template will prioritize the paper.short-author-list value in the configuration TOML file, if present. If the short-author-list key is absent, the template will try to generate its value automatically. Generally speaking, no action from authors is needed, however one can face some issues.

It will succeed if all authors have exactly two words in their name value: first name and last name (i.e., surname). It will use the first letter (grapheme cluster, to be exact) from the first name (then put . ), and the whole last name (for each author). Then these pieces of text are joined together with , between each but last pair, last 2 names will have and between them. Even though this is automated, make sure to double check the output in the citation box to avoid any mistakes.

For example, for authors "John Doe" and "Jane Smith", it will generate “J. Doe” and “J. Smith”, and the result will be “J. Doe and J. Smith”.

It will fail with an error message Failed to generate short author list if it fails to split the name string into exactly 2 parts. This can happen either

• due to human error of having more than 1 whitespace in the string,
• or due to the name naturally having more than 2 parts (or less than 2).

This can happen when an author has prefix(es) and/or other additional parts in their name, as is the case with “Johannes Diderik van der Waals”. See https://en.wikipedia.org/wiki/Van_(Dutch) and https://academia.stackexchange.com/a/46500 for some context and more examples.

If one or more authors are affected by this, the template will give the error mentioned above and the document will not compile. To fix this, paper.short-author-list must be added with correct value to the configuration file.

Keywords

The paper.keywords key contains a list of keywords for the a paper. However, the list must be sorted in lexographical order. This manual task is also automated, an therefore is not a concern for the author. This is outlined to remove potential confusion in case initial order and sorted order are different.

Figure caption

The figure caption (figure.caption.where(kind: image)) must always include a period at the end. While the table caption (figure.caption.where(kind: table)) must omit it. This requirement is automatically handled by the template and is extensively tested.

Although the implementation handles every element/type separately, which makes it prone to error of not handling all possible (or rather practical) use cases. If the <content type> was not handled properly error is encountered, please open an issue with the error message included. This is a bug (and a shortcoming of the way the implementation was designed).

API reference

ijimai

/// The template function.
///
/// ```typ
/// #show: ijimai.with(
///   config: toml("paper.toml"),
///   bibliography: "bibliography.yaml",
///   read: path => read-raw(path),
/// )
/// ```
///
/// - config (dictionary): Configuration that defines paper and author metadata.
/// - photos (str, array): Path to directory with authors' photos or an array
///     of raw photos data (bytes of images).
/// - bibliography (str, bytes): Path to the bibliography file or raw content of
///     the file (bytes).
/// - read (function): A special function that allows passing path strings to
///     the other parameters. Should be set to `path => read-raw(path)`.
/// - auto-first-paragraph (bool): Whether to automatically detect and style the
///     first paragraph. Normally this parameter should be ignored.
/// - body (content): Content passed to the template function when it is
///     applied.
#let ijimai(
  config: none,
  photos: "photos/",
  bibliography: none,
  read: none,
  auto-first-paragraph: true,
  body,
)

The config parameter is a dictionary with various required metadata that can be found in ./template/paper.toml. Technically, the configuration can be stored in any form, but it’s better to use TOML, to remove possibility of a human error when converting to a different file format, etc.

The photos parameter can be, e.g., "author/" or ("path/to/photo1.jpg", "path/photo2.png").map(read-raw).

The bibliography parameter can be, e.g., "bibliography.yaml" or read-raw("bibliography.yaml").

The read parameter makes use of a special path treatment when creating an unnamed function.

For the auto-first-paragraph parameter, see Introduction section and first paragraph section for more details.

read-raw

/// An alias to `read.with(encoding: none)`. Intended to be used with
/// `ijimai.read` (also `ijimai.photos`, or `ijimai.bibliography`).
#let read-raw

first-paragraph

/// Function that styles the first paragraph of the Introduction section.
///
/// - first-word (str, content): First word that will be styled.
/// - body (str, content): The rest of the first paragraph that is needed to
///     properly style the first word.
#let first-paragraph(first-word, body)

See Introduction section and first paragraph section for more details.

no-indent

/// Remove indentation for a specific paragraph. This is useful when using
/// "where" paragraph right after an equation that must not be indented.
///
/// - body: (content): Paragraph that needs to be unindented.
#let no-indent(body)

See ./template/paper.typ for a use case example.

Eq

/// Add "Equation" supplement before the equation reference.
///
/// Intended to be used only for equations that are referenced immediately at
/// the start of a sentence, as per IJIMAI requirements. In any other
///
/// Named with a capital letter to avoid shadowing `math.eq` that is directly
/// accessible in the math mode. As a consequence, it also reminds that it must
/// be used at the start of a sentence.
/// situations, reference equations as usual (directly).
///
/// ```typ
/// Usage: #Eq[@pythagorean], #Eq(<pythagorean>), #Eq(ref(<pythagorean>))
/// ```
///
/// - reference: (ref, label): equation's label or reference to it.
#let Eq(reference)

See ./template/paper.typ for a use case example.

Examples

Simple

2 authors

Special issue

In case of doubts

Please, bear in mind this template is continuously been bettered. In case of doubts, please, send an email to alberto.corbi@unir.net.