GNU Texinfo 6.5

Short Table of Contents

Table of Contents

Texinfo

This manual is for GNU Texinfo (version 6.5, 25 August 2017), a documentation system that can produce both online information and a printed manual from a single source using semantic markup.

The first part of this master menu lists the major nodes in this Info document, including the @-command and concept indices. The rest of the menu lists all the lower-level nodes in the document.

Appendices

 — The Detailed Node Listing —

Overview of Texinfo

Writing a Texinfo File

Beginning and Ending a Texinfo File

Texinfo File Header

Document Permissions

Title and Copyright Pages

The ‘Top’ Node and Master Menu

Global Document Commands

Nodes

Menus

Chapter Structuring

Cross-references

@xref

@url, @uref{url[, text][, replacement]}

Marking Text, Words and Phrases

Indicating Definitions, Commands, etc.

Emphasizing Text

Quotations and Examples

Lists and Tables

Making a Two-column Table

@multitable: Multi-column Tables

Special Displays

Floats

Inserting Images

Footnotes

Indices

Combining Indices

Special Insertions

Special Characters: Inserting @ {} , \ #

Inserting Space

Glyphs for Text

Glyphs for Programming

Forcing and Preventing Breaks

Definition Commands

The Definition Commands

Object-Oriented Programming

Internationalization

Conditionally Visible Text

Flags: @set, @clear, conditionals, and @value

Defining New Texinfo Commands

External Macro Processors: Line Directives

Include Files

Formatting and Printing Hardcopy

Format with tex/texindex

texi2any: The Generic Translator for Texinfo

Customization Variables

Creating and Installing Info Files

Creating an Info File

Installing an Info File

Generating HTML

HTML Cross-references

Sample Texinfo Files

Using Texinfo Mode

Updating Nodes and Menus

Page Headings

Catching Mistakes

Finding Badly Referenced Nodes

Info Format Specification

Info Format General Layout

Info Format Text Constructs

Documentation is like sex: when it is good, it is very, very good; and when it is bad, it is better than nothing. —Dick Brandon

Texinfo Copying Conditions

GNU Texinfo is free software; this means that everyone is free to use it and free to redistribute it on certain conditions. Texinfo is not in the public domain; it is copyrighted and there are restrictions on its distribution, but these restrictions are designed to permit everything that a good cooperating citizen would want to do. What is not allowed is to try to prevent others from further sharing any version of Texinfo that they might get from you.

Specifically, we want to make sure that you have the right to give away copies of the programs that relate to Texinfo, that you receive source code or else can get it if you want it, that you can change these programs or use pieces of them in new free programs, and that you know you can do these things.

To make sure that everyone has such rights, we have to forbid you to deprive anyone else of these rights. For example, if you distribute copies of the Texinfo related programs, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must tell them their rights.

Also, for our own protection, we must make certain that everyone finds out that there is no warranty for the programs that relate to Texinfo. If these programs are modified by someone else and passed on, we want their recipients to know that what they have is not what we distributed, so that any problems introduced by others will not reflect on our reputation.

The precise conditions of the licenses for the programs currently being distributed that relate to Texinfo are found in the General Public Licenses that accompany them. This manual is covered by the GNU Free Documentation License (see GNU Free Documentation License).

1 Overview of Texinfo

Texinfo is a documentation system that uses a single source file to produce both online information and printed output. This means that instead of writing several different documents, one for each output format, you need only write one document.

Using Texinfo, you can create a printed document (via the TeX typesetting system) in PDF or PostScript format, including chapters, sections, cross-references, and indices. From the same Texinfo source file, you can create an HTML output file suitable for use with a web browser, you can create an Info file with special features to make browsing documentation easy, and also create a Docbook file or a transliteration to XML format.

A Texinfo source file is a plain text file containing text interspersed with @-commands (words preceded by an ‘@’) that tell the Texinfo processors what to do. Texinfo’s markup commands are almost entirely semantic; that is, they specify the intended meaning of text in the document, rather than physical formatting instructions. You can edit a Texinfo file with any text editor, but it is especially convenient to use GNU Emacs since that editor has a special mode, called Texinfo mode, that provides various Texinfo-related features. (See Texinfo Mode.)

Texinfo was devised specifically for the purpose of writing software documentation and manuals. If you want to write a good manual for your program, Texinfo has many features which we hope will make your job easier. However, it provides almost no commands for controlling the final formatting. Texinfo is not intended to be a general-purpose formatting program, so if you need to lay out a newspaper, devise a glossy magazine ad, or follow the exact formatting requirements of a publishing house, Texinfo may not be the simplest tool.

Spell “Texinfo” with a capital “T” and the other letters in lowercase. The first syllable of “Texinfo” is pronounced like “speck”, not “hex”. This odd pronunciation is derived from the pronunciation of TeX. Pronounce TeX as if the ‘X’ were the last sound in the name ‘Bach’. In the word TeX, the ‘X’ is, rather than the English letter “ex”, actually the Greek letter “chi”.

Texinfo is the official documentation format of the GNU project. More information, including manuals for GNU packages, is available at the GNU documentation web page.

1.1 Reporting Bugs

We welcome bug reports and suggestions for any aspect of the Texinfo system: programs, documentation, installation, etc. Please email them to bug-texinfo@gnu.org. You can get the latest version of Texinfo via its home page, http://www.gnu.org/software/texinfo.

For bug reports, please include enough information for the maintainers to reproduce the problem. Generally speaking, that means:

• The version number of Texinfo and the program(s) or manual(s) involved.
• The contents of any input files necessary to reproduce the bug.
• Precisely how you ran any program(s) involved.
• A description of the problem and samples of any erroneous output.
• Hardware and operating system names and versions.
• Anything else that you think would be helpful.

When in doubt whether something is needed or not, include it. It’s better to include too much than to leave out something important.

It is critical to send an actual input file that reproduces the problem. What’s not critical is to “narrow down” the example to the smallest possible input—the actual input with which you discovered the bug will suffice. (Of course, if you do do experiments, the smaller the input file, the better.)

Patches are most welcome; if possible, please make them with ‘diff -c’ (see Comparing and Merging Files) and include ChangeLog entries (see Change Log in The GNU Emacs Manual), and follow the existing coding style.

1.2 Output Formats

Here is a brief overview of the output formats currently supported by Texinfo.

Info

(Generated via makeinfo.) Info format is mostly a plain text transliteration of the Texinfo source. It adds a few control characters to provide navigational information for cross-references, indices, and so on. The Emacs Info subsystem (see Info), and the standalone info program (see GNU Info), among others, can read these files. See Info Files, and Creating and Installing Info Files.

Plain text

(Generated via makeinfo --plaintext.) This is almost the same as Info output with the navigational control characters are omitted.

HTML

(Generated via makeinfo --html.) HTML, standing for Hyper Text Markup Language, has become the most commonly used language for writing documents on the World Wide Web. Web browsers, such as Mozilla, Lynx, and Emacs-W3, can render this language online. There are many versions of HTML, both different standards and browser-specific variations. makeinfo tries to use a subset of the language that can be interpreted by any common browser, intentionally not using many newer or less widely-supported tags. Although the native output is thus rather plain, it can be customized at various levels, if desired. For details of the HTML language and much related information, see http://www.w3.org/MarkUp/. See Generating HTML.

DVI

(Generated via texi2dvi.) The DeVIce Independent binary format is output by the TeX typesetting program (http://tug.org). This is then read by a DVI ‘driver’, which knows the actual device-specific commands that can be viewed or printed, notably Dvips for translation to PostScript (see Dvips) and Xdvi for viewing on an X display (http://sourceforge.net/projects/xdvi/). See Hardcopy. (Be aware that the Texinfo language is very different from and much stricter than TeX’s usual languages: plain TeX, LaTeX, ConTeXt, etc.)

PostScript

(Generated via texi2dvi --ps.) PostScript is a page description language that became widely used around 1985 and is still used today. http://en.wikipedia.org/wiki/PostScript gives a basic description and more preferences. By default, Texinfo uses the dvips program to convert TeX’s DVI output to PostScript. See Dvips.

PDF

(Generated via texi2dvi --pdf or texi2pdf.) This format was developed by Adobe Systems for portable document interchange, based on their previous PostScript language. It can represent the exact appearance of a document, including fonts and graphics, and supporting arbitrary scaling. It is intended to be platform-independent and easily viewable, among other design goals; http://en.wikipedia.org/wiki/Portable_Document_Format and http://tug.org/TUGboat/tb22-3/tb72beebe-pdf.pdf have some background. By default, Texinfo uses the pdftex program, an extension of TeX, to output PDF; see http://tug.org/applications/pdftex. See PDF Output.

Docbook

(Generated via makeinfo --docbook.) This is an XML-based format developed some years ago, primarily for technical documentation. It therefore bears some resemblance, in broad outline, to Texinfo. See http://www.docbook.org. Various converters from Docbook to Texinfo have also been developed; see the Texinfo web pages.

XML

(Generated via makeinfo --xml.) XML is a generic syntax specification usable for any sort of content (a reference is at http://www.w3.org/XML). The makeinfo XML output, unlike all the other output formats, is a transliteration of the Texinfo source rather than processed output. That is, it translates the Texinfo markup commands into XML syntax, for further processing by XML tools. The XML contains enough information to recreate the original content, except for syntactic constructs such as Texinfo macros and conditionals. The Texinfo source distribution includes a utility script txixml2texi to do that backward transformation.

The details of the output syntax are defined in an XML DTD as usual, which is contained in a file texinfo.dtd included in the Texinfo source distribution and available via the Texinfo web pages. Texinfo XML files, and XML files in general, cannot be viewed in typical web browsers; they won’t follow the DTD reference and as a result will simply report a (misleading) syntax error.

1.3 Info Files

As mentioned above, Info format is mostly a plain text transliteration of the Texinfo source, with the addition of a few control characters to separate nodes and provide navigational information, so that Info-reading programs can operate on it.

Info files are nearly always created by processing a Texinfo source document. makeinfo, also known as texi2any, is the principal command that converts a Texinfo file into an Info file; see Generic Translator texi2any.

Generally, you enter an Info file through a node that by convention is named ‘Top’. This node normally contains just a brief summary of the file’s purpose, and a large menu through which the rest of the file is reached. From this node, you can either traverse the file systematically by going from node to node, or you can go to a specific node listed in the main menu, or you can search the index menus and then go directly to the node that has the information you want. Alternatively, with the standalone Info program, you can specify specific menu items on the command line (see Info).

If you want to read through an Info file in sequence, as if it were a printed manual, you can hit SPC repeatedly, or you get the whole file with the advanced Info command g *. (See Advanced Info commands in Info.)

The dir file in the info directory serves as the departure point for the whole Info system. From it, you can reach the ‘Top’ nodes of each of the documents in a complete Info system.

If you wish to refer to an Info file via a URI, you can use the (unofficial) syntax exemplified by the following. This works with Emacs/W3, for example:

info:emacs#Dissociated%20Press
info:///usr/info/emacs#Dissociated%20Press
info://localhost/usr/info/emacs#Dissociated%20Press

The info program itself does not follow URIs of any kind.

1.4 Printed Books

A Texinfo file can be formatted and typeset as a printed book or manual. To do this, you need TeX, a sophisticated typesetting program written by Donald Knuth of Stanford University.

A Texinfo-based book is similar to any other typeset, printed work: it can have a title page, copyright page, table of contents, and preface, as well as chapters, numbered or unnumbered sections and subsections, page headers, cross-references, footnotes, and indices.

TeX is a general purpose typesetting program. Texinfo provides a file texinfo.tex that contains information (definitions or macros) that TeX uses when it typesets a Texinfo file. (texinfo.tex tells TeX how to convert the Texinfo @-commands to TeX commands, which TeX can then process to create the typeset document.) texinfo.tex contains the specifications for printing a document. You can get the latest version of texinfo.tex from the Texinfo home page, http://www.gnu.org/software/texinfo/.

In the United States, documents are most often printed on 8.5 inch by 11 inch pages (216mm by 280mm); this is the default size. But you can also print for 7 inch by 9.25 inch pages (178mm by 235mm, the @smallbook size; or on A4 or A5 size paper (@afourpaper, @afivepaper). See @smallbook, and A4 Paper.

TeX is freely distributable. It is written in a superset of Pascal for literate programming called WEB and can be compiled either in Pascal or (by using a conversion program that comes with the TeX distribution) in C.

TeX is very powerful and has a great many features. Because a Texinfo file must be able to present information both on a character-only terminal in Info form and in a typeset book, the formatting commands that Texinfo supports are necessarily limited.

See Obtaining TeX, for information on acquiring TeX. It is not part of the Texinfo distribution.

1.5 Adding Output Formats

The output formats in the previous sections handle a wide variety of usage, but of course there is always room for more.

If you are a programmer and would like to contribute to the GNU project by implementing additional output formats for Texinfo, that would be excellent. The way to do this that would be most useful is to write a new back-end for texi2any, our reference implementation of a Texinfo parser; it creates a tree representation of the Texinfo input that you can use for the conversion. The documentation in the source file tp/Texinfo/Convert/Converter.pm is a good place to start. See Generic Translator texi2any.

Another viable approach is use the Texinfo XML output from texi2any as your input. This XML is an essentially complete representation of the input, but without the Texinfo syntax and option peculiarities, as described above.

If you still cannot resist the temptation of writing a new program that reads Texinfo source directly, let us give some more caveats: please do not underestimate the amount of work required. Texinfo is by no means a simple language to parse correctly, and remains under development, so you would be committing to an ongoing task. You are advised to check that the tests of the language that come with texi2any give correct results with your new program.

From time to time, proposals are made to generate traditional Unix man pages from Texinfo source. However, because man pages have a strict conventional format, creating a good man page requires a completely different source from that needed for the typical Texinfo applications of writing a good user tutorial and/or a good reference manual. This makes generating man pages incompatible with the Texinfo design goal of not having to document the same information in different ways for different output formats. You might as well write the man page directly.

As an alternative way to support man pages, you may find the program help2man to be useful. It generates a traditional man page from the ‘--help’ output of a program. In fact, the man pages for the programs in the Texinfo distribution are generated with this. It is GNU software written by Brendan O’Dea, available from http://www.gnu.org/software/help2man.

1.6 History

Richard M. Stallman invented the Texinfo format, wrote the initial processors, and created Edition 1.0 of this manual. Robert J. Chassell greatly revised and extended the manual, starting with Edition 1.1. Brian Fox was responsible for the standalone Texinfo distribution until version 3.8, and originally wrote the standalone makeinfo and info programs. Karl Berry has continued maintenance since Texinfo 3.8 (manual edition 2.22).

Our thanks go out to all who helped improve this work, particularly the indefatigable Eli Zaretskii and Andreas Schwab, who have provided patches beyond counting. François Pinard and David D. Zuhn, tirelessly recorded and reported mistakes and obscurities. Zack Weinberg did the impossible by implementing the macro syntax in texinfo.tex. Thanks to Melissa Weisshaus for her frequent reviews of nearly similar editions. Dozens of others have contributed patches and suggestions, they are gratefully acknowledged in the ChangeLog file. Our mistakes are our own.

Beginnings

In the 1970’s at CMU, Brian Reid developed a program and format named Scribe to mark up documents for printing. It used the @ character to introduce commands, as Texinfo does. Much more consequentially, it strove to describe document contents rather than formatting, an idea wholeheartedly adopted by Texinfo.

Meanwhile, people at MIT developed another, not too dissimilar format called Bolio. This then was converted to using TeX as its typesetting language: BoTeX. The earliest BoTeX version seems to have been 0.02 on October 31, 1984.

BoTeX could only be used as a markup language for documents to be printed, not for online documents. Richard Stallman (RMS) worked on both Bolio and BoTeX. He also developed a nifty on-line help format called Info, and then combined BoTeX and Info to create Texinfo, a mark up language for text that is intended to be read both online and as printed hard copy.

Moving forward, the original translator to create Info was written (primarily by RMS and Bob Chassell) in Emacs Lisp, namely the texinfo-format-buffer and other functions. In the early 1990s, Brian Fox reimplemented the conversion program in C, now called makeinfo.

Reimplementing in Perl

In 2012, the C makeinfo was itself replaced by a Perl implementation generically called texi2any. This version supports the same level of output customization as texi2html, an independent program originally written by Lionel Cons, later with substantial work by many others. The many additional features needed to make texi2html a replacement for makeinfo were implemented by Patrice Dumas. The first never-released version of texi2any was based on the texi2html code. That implementation, however, was abandoned in favor of the current program, which parses the Texinfo input into a tree for processing. It still supports nearly all the features of texi2html.

The new Perl program is much slower than the old C program. We hope the speed gap will close in the future, but it may not ever be entirely comparable. So why did we switch? In short, we intend and hope that the present program will be much easier than the previous C implementation of makeinfo to extend to different output styles, back-end output formats, and all other customizations. In more detail:

• HTML customization. Many GNU and other free software packages had been happily using the HTML customization features in texi2html for years. Thus, in effect two independent implementations of the Texinfo language had developed, and keeping them in sync was not simple. Adding the HTML customization possible in texi2html to a C program would have been an enormous effort.
• Unicode, and multilingual support generally, especially of east Asian languages. Although of course it’s perfectly plausible to write such support in C, in the particular case of makeinfo, it would have been tantamount to rewriting the entire program. In Perl, much of that comes essentially for free.
• Additional back-ends. The makeinfo code had become convoluted to the point where adding a new back-end was quite complex, requiring complex interactions with existing back-ends. In contrast, our Perl implementation provides a clean tree-based representation for all back-ends to work from. People have requested numerous different back-ends (LaTeX, the latest (X)HTML, …), and they will now be much more feasible to implement. Which leads to the last item:
• Making contributions easier. In general, due to the cleaner structure, the Perl program should be considerably easier than the C for anyone to read and contribute to, with the resulting obvious benefits.

See Reference Implementation, for more on the rationale for and role of texi2any.

2 Writing a Texinfo File

This chapter describes Texinfo syntax and what is required in a Texinfo file, and gives a short sample file.

2.1 General Syntactic Conventions

This section describes the general conventions used in all Texinfo documents.

• All printable ASCII characters except ‘@’, ‘{’ and ‘}’ can appear in a Texinfo file and stand for themselves. ‘@’ is the escape character which introduces commands, while ‘{’ and ‘}’ are used to surround arguments to certain commands. To put one of these special characters into the document, put an ‘@’ character in front of it, like this: ‘@@’, ‘@{’, and ‘@}’.
• In a Texinfo file, the commands you write to describe the contents of the manual are preceded by an ‘@’ character; they are called @-commands. (The ‘@’ in Texinfo has the same meaning that ‘\’ has in plain TeX.)

  Depending on what they do or what arguments¹ they take, you need to write @-commands on lines of their own, or as part of sentences. As a general rule, a command requires braces if it mingles among other text; but it does not need braces if it is on a line of its own. For more details of Texinfo command syntax, see Command Syntax.

• Whitespace following an @-command name is optional and (usually) ignored if present. The exceptions are contexts when whitespace is significant, e.g., an @example environment.
• Texinfo supports the usual quotation marks used in English and in other languages; see Inserting Quotation Marks.
• Use three hyphens in a row, ‘---’, to produce a long dash—like this (called an em dash), used for punctuation in sentences. Use two hyphens, ‘--’, to produce a medium dash (called an en dash), used primarily for numeric ranges, as in “June 25–26”. Use a single hyphen, ‘-’, to produce a standard hyphen used in compound words. For display on the screen, Info reduces three hyphens to two and two hyphens to one (not transitively!). Of course, any number of hyphens in the source remain as they are in literal contexts, such as @code and @example.
• Form feed (CTRL-l) characters in the input are handled as follows:

  PDF/DVI

  In normal text, treated as ending any open paragraph; essentially ignored between paragraphs.

  Info

  Output as-is between paragraphs (their most common use); in other contexts, they may be treated as regular spaces (and thus consolidated with surrounding whitespace).

  HTML

  Written as a numeric entity except contexts where spaces are ignored; for example, in ‘@footnote{ ^L foo}’, the form feed is ignored.

  XML

  Keep them everywhere; in attributes, escaped as ‘\f’; also, ‘\’ is escaped as ‘\\’ and newline as ‘\n’.

  Docbook

  Completely removed, as they are not allowed.

  As you can see, because of these differing requirements of the output formats, it’s not possible to use form feeds completely portably.

• Caution: Last, do not use tab characters in a Texinfo file! (Except perhaps in verbatim modes.) TeX uses variable-width fonts, which means that it is impractical at best to define a tab to work in all circumstances. Consequently, TeX treats tabs like single spaces, and that is not what they look like in the source. Furthermore, makeinfo does nothing special with tabs, and thus a tab character in your input file will usually have a different appearance in the output.

  To avoid this problem, Texinfo mode in GNU Emacs inserts multiple spaces when you press the TAB key. Also, you can run untabify in Emacs to convert tabs in a region to multiple spaces, or use the unexpand command from the shell.

2.2 Comments

You can write comments in a Texinfo file by using the @comment command, which may be abbreviated to @c. Such comments are for a person looking at the Texinfo source file. All the text on a line that follows either @comment or @c is a comment; the rest of the line does not appear in the visible output. (To be precise, the character after the @c or @comment must be something other than a dash or alphanumeric, or it will be taken as part of the command.)

Often, you can write the @comment or @c in the middle of a line, and only the text that follows after the @comment or @c command does not appear; but some commands, such as @settitle, work on a whole line. You cannot use @comment or @c within a line beginning with such a command.

In cases of nested command invocations, complicated macro definitions, etc., @c and @comment may provoke an error when processing with TeX. Therefore, you can also use the DEL character (ASCII 127 decimal, 0x7f hex, 0177 octal) as a true TeX comment character (catcode 14, in TeX internals). Everything on the line after the DEL will be ignored.

You can also have long stretches of text ignored by the Texinfo processors with the @ignore and @end ignore commands. Write each of these commands on a line of its own, starting each command at the beginning of the line. Text between these two commands does not appear in the processed output. You can use @ignore and @end ignore for writing comments. (For some caveats regarding nesting of such commands, see Conditional Nesting.)

2.3 What a Texinfo File Must Have

By convention, the name of a Texinfo file ends with one of the extensions .texinfo, .texi, .txi, or .tex.²

In order to be made into a printed manual and other output formats, a Texinfo file must begin with lines like this:

\input texinfo
@settitle name-of-manual

The contents of the file follow this beginning, and then you must end the Texinfo source with a line like this:

@bye

Here’s an explanation:

• The ‘\input texinfo’ line tells TeX to use the texinfo.tex file, which tells TeX how to translate the Texinfo @-commands into TeX typesetting commands. (Note the use of the backslash, ‘\’; this is correct for TeX.)
• The @settitle line specifies a title for the page headers (or footers) of the printed manual, and the default title and document description for the ‘<head>’ in HTML. Strictly speaking, @settitle is optional—if you don’t mind your document being titled ‘Untitled’.
• The @bye line at the end of the file on a line of its own tells the formatters that the file is ended and to stop formatting. If you leave this out, you’ll be dumped at TeX’s prompt at the end of the run.

Furthermore, you will usually provide a Texinfo file with a title page, indices, and the like, all of which are explained in this manual. But the minimum, which can be useful for short documents, is just the two lines at the beginning and the one line at the end.

2.4 A Short Sample Texinfo File

Here is a short but complete Texinfo file, so you can see how Texinfo source appears in practice. The first three parts of the file are mostly boilerplate: when writing a manual, you simply change the names as appropriate.

The complete file, without interspersed comments, is shown in Short Sample Texinfo File.

See Beginning and Ending a File, for more documentation on the commands listed here.

Header

The header tells TeX which definitions file to use, names the manual, and carries out other such housekeeping tasks.

\input texinfo
@settitle Sample Manual 1.0

Summary Description and Copyright

This segment describes the document and contains the copyright notice and copying permissions. This is done with the @copying command.

A real manual includes more text here, according to the license under which it is distributed. See GNU Sample Texts.

@copying
This is a short example of a complete Texinfo file, version 1.0.

Copyright @copyright{} 2016 Free Software Foundation, Inc.
@end copying

Titlepage, Copyright, Contents

The title and copyright segment contains the title and copyright pages for the printed manual. The segment must be enclosed between @titlepage and @end titlepage commands. The title and copyright page does not appear in the online output.

We use the @insertcopying command to include the permission text from the previous section, instead of writing it out again; it is output on the back of the title page. The @contents command generates a table of contents.

@titlepage
@title Sample Title

@c The following two commands start the copyright page.
@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage

@c Output the table of contents at the beginning.
@contents

‘Top’ Node and Master Menu

The ‘Top’ node starts off the online output; it does not appear in the printed manual. We repeat the short description from the beginning of the ‘@copying’ text, but there’s no need to repeat the copyright information, so we don’t use ‘@insertcopying’ here.

The ‘@top’ command itself helps makeinfo determine the relationships between nodes. The ‘Top’ node contains at least a top-level menu listing the chapters, and possibly a Master Menu listing all the nodes in the entire document.

@ifnottex
@node Top
@top Short Sample

This is a short sample Texinfo file.
@end ifnottex

@menu
* First Chapter::    The first chapter is the
                       only chapter in this sample.
* Index::            Complete index.
@end menu

The Body of the Document

The body segment contains all the text of the document, but not the indices or table of contents. This example illustrates a node and a chapter containing an enumerated list.

@node First Chapter
@chapter First Chapter

@cindex chapter, first

This is the first chapter.
@cindex index entry, another

Here is a numbered list.

@enumerate
@item
This is the first item.

@item
This is the second item.
@end enumerate

The End of the Document

This may contain commands for printing indices, and closes with the @bye command, which marks the end of the document.

@node Index
@unnumbered Index

@printindex cp

@bye

Some Results

Here is what the contents of the first chapter of the sample look like:

This is the first chapter.

Here is a numbered list.

1. This is the first item.
2. This is the second item.

3 Beginning and Ending a Texinfo File

This chapter expands on the minimal complete Texinfo source file previously given (see Short Sample).

Certain pieces of information must be provided at the beginning of a Texinfo file, such the title of the document and the Top node. A table of contents is also generally produced here.

Straight text outside of any command before the Top node should be avoided. Such text is treated differently in the different output formats: at the time of writing, it is visible in TeX and HTML, by default not shown in Info readers, and so on.

3.1 Sample Texinfo File Beginning

The following sample shows what is needed. The elements given here are explained in more detail in the following sections. Other commands are often included at the beginning of Texinfo files, but the ones here are the most critical.

See GNU Sample Texts, for the full texts to be used in GNU manuals.

\input texinfo
@settitle name-of-manual version

@copying
This manual is for program, version version.

Copyright @copyright{} years copyright-owner.

@quotation
Permission is granted to …
@end quotation
@end copying

@titlepage
@title name-of-manual-when-printed
@subtitle subtitle-if-any
@subtitle second-subtitle
@author author

@c  The following two commands
@c  start the copyright page.
@page
@vskip 0pt plus 1filll
@insertcopying

Published by …
@end titlepage

@c So the toc is printed at the start.
@contents

@ifnottex
@node Top
@top title

This manual is for program, version version.
@end ifnottex

@menu
* First Chapter::    Getting started …
* Second Chapter::          …
 …
* Copying::          Your rights and freedoms.
@end menu

@node First Chapter
@chapter First Chapter

@cindex first chapter
@cindex chapter, first
…

3.2 Texinfo File Header

Texinfo files start with at least two lines. These are the \input texinfo line and the @settitle line.

Also, if you want to format just part of the Texinfo file in Emacs, you must write the @settitle line between start-of-header and end-of-header lines. These start- and end-of-header lines are optional, but they do no harm, so you might as well always include them.

Any command that affects document formatting as a whole makes sense to include in the header. @synindex (see @synindex), for instance, is another command often included in the header.

Thus, the beginning of a Texinfo file looks approximately like this:

\input texinfo
@settitle Sample Manual 1.0

(See GNU Sample Texts for complete sample texts.)

3.2.1 The First Line of a Texinfo File

Every Texinfo file that is to be the top-level input to TeX must begin with a line that looks like this:

\input texinfo

When the file is processed by TeX, the ‘\input texinfo’ command tells TeX to load the macros needed for processing a Texinfo file. These are in a file called texinfo.tex, which should have been installed on your system along with either the TeX or Texinfo software. TeX uses the backslash, ‘\’, to mark the beginning of a command, exactly as Texinfo uses ‘@’. The texinfo.tex file causes the switch from ‘\’ to ‘@’; before the switch occurs, TeX requires ‘\’, which is why it appears at the beginning of the file.

You may optionally follow this line with a comment to tell GNU Emacs to use Texinfo mode when the file is edited:

\input texinfo   @c -*-texinfo-*-

This may be useful when Emacs doesn’t detect the file type from the file extension automatically.

3.2.2 Start of Header

A start-of-header line is a Texinfo comment that looks like this:

@c %**start of header

Write the start-of-header line on the second line of a Texinfo file. Follow the start-of-header line with an @settitle line and, optionally, with other commands that globally affect the document formatting, such as @synindex or @footnotestyle; and then by an end-of-header line (see End of Header).

The start- and end-of-header lines allow you to format only part of a Texinfo file for Info or printing. See texinfo-format commands.

The odd string of characters, ‘%**’, is to ensure that no other comment is accidentally taken for a start-of-header line. You can change it if you wish by setting the tex-start-of-header and/or tex-end-of-header Emacs variables. See Texinfo Mode Printing.

3.2.3 @setfilename: Set the Output File Name

The @setfilename line specifies the name of the output file to be generated. When present, it should be the first Texinfo command (that is, after ‘\input texinfo’). Write the @setfilename command at the beginning of a line and follow it on the same line by the Info file name.

@setfilename info-file-name

The name must be different from the name of the Texinfo file. There are two conventions for choosing the name: you can either remove the extension (such as ‘.texi’) entirely from the input file name, or (recommended) replace it with the ‘.info’ extension.

When a @setfilename line is present, the Texinfo processors ignore everything written before the @setfilename line. This is why the very first line of the file (the \input line) does not show up in the output.

If there is no @setfilename line, makeinfo uses the input file name to determine the output name: first, any of the extensions .texi, .tex, .txi or .texinfo is removed from the input file name; then, the output format specific extension is added—.html when generating HTML, .info when generating Info, etc. The \input line is still ignored in this processing, as well as leading blank lines.

When producing another output format, makeinfo will replace any final extension with the output format-specific extension (‘html’ when generating HTML, for example), or add a dot followed by the extension (‘.html’ for HTML) if the given name has no extension.

@setfilename used to be required by the Texinfo processors, and some other programs may still expect it to be present; for example, Automake (see Texinfo in GNU Automake).

Although an explicit ‘.info’ extension is preferable, some operating systems cannot handle long file names. You can run into a problem even when the file name you specify is itself short enough. This occurs because the Info formatters split a long Info file into short indirect subfiles, and name them by appending ‘-1’, ‘-2’, …, ‘-10’, ‘-11’, and so on, to the original file name. (See Tag and Split Files.) The subfile name texinfo.info-10, for example, is too long for old systems with a 14-character limit on filenames; so the Info file name for this document is texinfo rather than texinfo.info. When makeinfo is running on operating systems such as MS-DOS which impose severe limits on file names, it may remove some characters from the original file name to leave enough space for the subfile suffix, thus producing files named texin-10, gcc.i12, etc.

See also the --output option in Invoking texi2any.

3.2.4 @settitle: Set the Document Title

A Texinfo file should contain a line that looks like this:

@settitle title

Write the @settitle command at the beginning of a line and follow it on the same line by the title. Do not write anything else on the line. The @settitle command should precede everything that generates actual output. The best place for it is right after the @setfilename command (described in the previous section).

This command tells TeX the title to use in a header or footer for double-sided output, in case such headings are output. For more on headings for TeX, see Heading Generation.

In the HTML file produced by makeinfo, title serves as the document ‘<title>’. It also becomes the default document description in the ‘<head>’ part (see @documentdescription).

When the title page is used in the output, the title in the @settitle command does not affect the title as it appears on the title page. Thus, the two do not need not to match exactly. A practice we recommend is to include the version or edition number of the manual in the @settitle title; on the title page, the version number generally appears as a @subtitle so it would be omitted from the @title. See @titlepage.

3.2.5 End of Header

Follow the header lines with an end-of-header line, which is a Texinfo comment that looks like this:

@c %**end of header

See Start of Header.

3.3 Document Permissions

The copyright notice and copying permissions for a document need to appear in several places in the various Texinfo output formats. Therefore, Texinfo provides a command (@copying) to declare this text once, and another command (@insertcopying) to insert the text at appropriate points.

This section is about the license of the Texinfo document. If the document is a software manual, the software is typically under a different license—for GNU and many other free software packages, software is usually released under the GNU GPL, and manuals are released under the GNU FDL. It is helpful to state the license of the software of the manual, but giving the complete text of the software license is not necessarily required.

3.3.1 @copying: Declare Copying Permissions

The @copying command should be given very early in the document; the recommended location is right after the header material (see Texinfo File Header). It conventionally consists of a sentence or two about what the program is, identification of the documentation itself, the legal copyright line, and the copying permissions. Here is a skeletal example:

@copying
This manual is for program (version version, updated
date), which …

Copyright @copyright{} years copyright-owner.

@quotation
Permission is granted to …
@end quotation
@end copying

The @quotation has no legal significance; it’s there to improve readability in some contexts.

The text of @copying is output as a comment at the beginning of Info, HTML, XML, and Docbook output files. It is not output implicitly in plain text or TeX; it’s up to you to use @insertcopying to emit the copying information. See the next section for details.

The @copyright{} command generates a ‘c’ inside a circle when the output format supports this glyph (print and HTML always do, for instance). When the glyph is not supported in the output, it generates the three-character sequence ‘(C)’.

The copyright notice itself has the following legally-prescribed form:

Copyright © years copyright-owner.

The word ‘Copyright’ must always be written in English, even if the document is otherwise written in another language. This is due to international law.

The list of years should include all years in which a version was completed (even if it was released in a subsequent year). It is simplest for each year to be written out individually and in full, separated by commas.

The copyright owner (or owners) is whoever holds legal copyright on the work. In the case of works assigned to the FSF, the owner is ‘Free Software Foundation, Inc.’.

The copyright ‘line’ may actually be split across multiple lines, both in the source document and in the output. This often happens for documents with a long history, having many different years of publication. If you do use several lines, do not indent any of them (or anything else in the @copying block) in the source file.

See Copyright Notices in GNU Maintainer Information, for additional information. See GNU Sample Texts, for the full text to be used in GNU manuals. See GNU Free Documentation License, for the license itself under which GNU and other free manuals are distributed.

3.3.2 @insertcopying: Include Permissions Text

The @insertcopying command is simply written on a line by itself, like this:

@insertcopying

This inserts the text previously defined by @copying. To meet legal requirements, it must be used on the copyright page in the printed manual (see Copyright).

The @copying command itself causes the permissions text to appear in an Info file before the first node. The text is also copied into the beginning of each split Info output file, as is legally necessary. This location implies a human reading the manual using Info does not see this text (except when using the advanced Info command g *), but this does not matter for legal purposes, because the text is present.

Similarly, the @copying text is automatically included at the beginning of each HTML output file, as an HTML comment. Again, this text is not visible (unless the reader views the HTML source).

The permissions text defined by @copying also appears automatically at the beginning of the XML and Docbook output files.

3.4 Title and Copyright Pages

In hard copy output, the manual’s name and author are usually printed on a title page. Copyright information is usually printed on the back of the title page.

The title and copyright pages appear in printed manuals, but not in most other output formats. Because of this, it is possible to use several slightly obscure typesetting commands that are not to be used in the main text. In addition, this part of the beginning of a Texinfo file contains the text of the copying permissions that appears in the printed manual.

3.4.1 @titlepage

Start the material for the title page and following copyright page with @titlepage on a line by itself and end it with @end titlepage on a line by itself.

The @end titlepage command starts a new page and turns on page numbering (see Heading Generation). All the material that you want to appear on unnumbered pages should be put between the @titlepage and @end titlepage commands. You can force the table of contents to appear there with the @setcontentsaftertitlepage command (see Contents).

By using the @page command you can force a page break within the region delineated by the @titlepage and @end titlepage commands and thereby create more than one unnumbered page. This is how the copyright page is produced. (The @titlepage command might perhaps have been better named the @titleandadditionalpages command, but that would have been rather long!)

When you write a manual about a computer program, you should write the version of the program to which the manual applies on the title page. If the manual changes more frequently than the program or is independent of it, you should also include an edition number³ for the manual. This helps readers keep track of which manual is for which version of the program. (The ‘Top’ node should also contain this information; see The Top Node.)

Texinfo provides two main methods for creating a title page. One method uses the @titlefont, @sp, and @center commands to generate a title page in which the words on the page are centered.

The second method uses the @title, @subtitle, and @author commands to create a title page with black rules under the title and author lines and the subtitle text set flush to the right hand side of the page. With this method, you do not specify any of the actual formatting of the title page. You specify the text you want, and Texinfo does the formatting.

You may use either method, or you may combine them; see the examples in the sections below.

For sufficiently simple documents, and for the bastard title page in traditional book frontmatter, Texinfo also provides a command @shorttitlepage which takes the rest of the line as the title. The argument is typeset on a page by itself and followed by a blank page.

3.4.2 @titlefont, @center, and @sp

You can use the @titlefont, @sp, and @center commands to create a title page for a printed document. (This is the first of the two methods for creating a title page in Texinfo.)

Use the @titlefont command to select a large font suitable for the title itself. You can use @titlefont more than once if you have an especially long title.

For HTML output, each @titlefont command produces an <h1> heading, but the HTML document <title> is not affected. For that, you must put a @settitle command before the @titlefont command (see @settitle).

For example:

@titlefont{Texinfo}

Use the @center command at the beginning of a line to center the remaining text on that line. Thus,

@center @titlefont{Texinfo}

centers the title, which in this example is “Texinfo” printed in the title font.

Use the @sp command to insert vertical space. For example:

@sp 2

This inserts two blank lines on the printed page. (See @sp, for more information about the @sp command.)

A template for this method looks like this:

@titlepage
@sp 10
@center @titlefont{name-of-manual-when-printed}
@sp 2
@center subtitle-if-any
@sp 2
@center author
…
@end titlepage

The spacing of the example fits an 8.5 by 11 inch manual.

You can in fact use these commands anywhere, not just on a title page, but since they are not logical markup commands, we don’t recommend them.

3.4.3 @title, @subtitle, and @author

You can use the @title, @subtitle, and @author commands to create a title page in which the vertical and horizontal spacing is done for you automatically. This contrasts with the method described in the previous section, in which the @sp command is needed to adjust vertical spacing.

Write the @title, @subtitle, or @author commands at the beginning of a line followed by the title, subtitle, or author. The @author command may be used for a quotation in an @quotation block (see @quotation); except for that, it is an error to use any of these commands outside of @titlepage.

The @title command produces a line in which the title is set flush to the left-hand side of the page in a larger than normal font. The title is underlined with a black rule. The title must be given on a single line in the source file; it will be broken into multiple lines of output is needed.

For long titles, the @* command may be used to specify the line breaks in long titles if the automatic breaks do not suit. Such explicit line breaks are generally reflected in all output formats; if you only want to specify them for the printed output, use a conditional (see Conditionals). For example:

@title This Long Title@inlinefmt{tex,@*} Is Broken in @TeX{}

The @subtitle command sets subtitles in a normal-sized font flush to the right-hand side of the page.

The @author command sets the names of the author or authors in a middle-sized font flush to the left-hand side of the page on a line near the bottom of the title page. The names are followed by a black rule that is thinner than the rule that underlines the title.

There are two ways to use the @author command: you can write the name or names on the remaining part of the line that starts with an @author command:

@author by Jane Smith and John Doe

or you can write the names one above each other by using multiple @author commands:

@author Jane Smith
@author John Doe

A template for this method looks like this:

@titlepage
@title name-of-manual-when-printed
@subtitle subtitle-if-any
@subtitle second-subtitle
@author author
@page
…
@end titlepage

3.4.4 Copyright Page

By international treaty, the copyright notice for a book must be either on the title page or on the back of the title page. When the copyright notice is on the back of the title page, that page is customarily not numbered. Therefore, in Texinfo, the information on the copyright page should be within @titlepage and @end titlepage commands.

Use the @page command to cause a page break. To push the copyright notice and the other text on the copyright page towards the bottom of the page, use the following incantation after @page:

@vskip 0pt plus 1filll

The @vskip command inserts whitespace in the TeX output; it is ignored in all other output formats. The ‘0pt plus 1filll’ means to put in zero points of mandatory whitespace, and as much optional whitespace as needed to push the following text to the bottom of the page. Note the use of three ‘l’s in the word ‘filll’; this is correct.

To insert the copyright text itself, write @insertcopying next (see Document Permissions):

@insertcopying

Follow the copying text by the publisher, ISBN numbers, cover art credits, and other such information.

Here is an example putting all this together:

@titlepage
…
@page
@vskip 0pt plus 1filll
@insertcopying

Published by …

Cover art by …
@end titlepage

We have one more special case to consider: for plain text output, you must insert the copyright information explicitly if you want it to appear. For instance, you could have the following after the copyright page:

@ifplaintext
@insertcopying
@end ifplaintext

You could include other title-like information for the plain text output in the same place.

3.4.5 Heading Generation

Like all @end commands (see Quotations and Examples), the @end titlepage command must be written at the beginning of a line by itself, with only one space between the @end and the titlepage. It not only marks the end of the title and copyright pages, but also causes TeX to start generating page headings and page numbers.

Texinfo has two standard page heading formats, one for documents printed on one side of each sheet of paper (single-sided printing), and the other for documents printed on both sides of each sheet (double-sided printing).

In full generality, you can control the headings in different ways:

• The conventional way is to write a @setchapternewpage command before the title page commands, if required, and then have the @end titlepage command start generating page headings in the manner desired.

  Most documents are formatted with the standard single-sided or double-sided headings, (sometimes) using @setchapternewpage odd for double-sided printing and (almost always) no @setchapternewpage command for single-sided printing (see @setchapternewpage).

• Alternatively, you can use the @headings command to prevent page headings from being generated or to start them for either single or double-sided printing. Write a @headings command immediately after the @end titlepage command. To turn off headings, write @headings off. See @headings.
• Or, you may specify your own page heading and footing format. See Headings.

3.5 Generating a Table of Contents

The @chapter, @section, and other structuring commands (see Chapter Structuring) supply the information to make up a table of contents, but they do not cause an actual table to appear in the manual. To do this, you must use the @contents and/or @summarycontents command(s).

@contents

Generates a table of contents in a printed manual, including all chapters, sections, subsections, etc., as well as appendices and unnumbered chapters. Headings generated by @majorheading, @chapheading, and the other @…heading commands do not appear in the table of contents (see Structuring Command Types).

@shortcontents

@summarycontents

(@summarycontents is a synonym for @shortcontents.)

Generates a short or summary table of contents that lists only the chapters, appendices, and unnumbered chapters. Sections, subsections and subsubsections are omitted. Only a long manual needs a short table of contents in addition to the full table of contents.

Both contents commands should be written on a line by themselves, and placed near the beginning of the file, after the @end titlepage (see @titlepage), before any sectioning command. The contents commands automatically generate a chapter-like heading at the top of the first table of contents page, so don’t include any sectioning command such as @unnumbered before them.

Since an Info file uses menus instead of tables of contents, the Info formatting commands ignore the contents commands. But the contents are included in plain text output (generated by makeinfo --plaintext) and in other output formats, such as HTML.

When makeinfo writes a short table of contents while producing HTML output, the links in the short table of contents point to corresponding entries in the full table of contents rather than the text of the document. The links in the full table of contents point to the main text of the document.

3.6 The ‘Top’ Node and Master Menu

The ‘Top’ node is the node in which a reader enters an Info manual. As such, it should begin with a brief description of the manual (including the version number), and end with a master menu for the whole manual. Of course you should include any other general information you feel a reader would find helpful.

It is conventional and desirable to write a @top sectioning command line containing the title of the document immediately after the @node Top line (see @top Command).

The contents of the ‘Top’ node should appear only in the online output; none of it should appear in printed output, so enclose it between @ifnottex and @end ifnottex commands. (TeX does not print either an @node line or a menu; they appear only in Info; strictly speaking, you are not required to enclose these parts between @ifnottex and @end ifnottex, but it is simplest to do so. See Conditionally Visible Text.)

3.6.1 Top Node Example

Here is an example of a Top node.

@ifnottex
@node Top
@top Sample Title

This is the text of the top node.
@end ifnottex

Additional general information.

@menu
* First Chapter::
* Second Chapter::
…
* Index::

@end menu

3.6.2 Parts of a Master Menu

A master menu is the main menu. It is customary to include a detailed menu listing all the nodes in the document in this menu.

Like any other menu, a master menu is enclosed in @menu and @end menu and does not appear in the printed output.

Generally, a master menu is divided into parts.

• The first part contains the major nodes in the Texinfo file: the nodes for the chapters, chapter-like sections, and the appendices.
• The second part contains nodes for the indices.
• The third and subsequent parts contain a listing of the other, lower-level nodes, often ordered by chapter. This way, rather than go through an intermediary menu, an inquirer can go directly to a particular node when searching for specific information. These menu items are not required; add them if you think they are a convenience. If you do use them, put @detailmenu before the first one, and @end detailmenu after the last; otherwise, makeinfo will get confused.

Each section in the menu can be introduced by a descriptive line. So long as the line does not begin with an asterisk, it will not be treated as a menu entry. (See Writing a Menu, for more information.)

For example, the master menu for this manual looks like the following (but has many more entries):

@menu
* Copying Conditions::  Your rights.
* Overview::            Texinfo in brief.
…

* Command and Variable Index::
* General Index::

@detailmenu
--- The Detailed Node Listing ---

Overview of Texinfo

* Reporting Bugs:: …
…

Beginning a Texinfo File

* Sample Beginning:: …
…
@end detailmenu
@end menu

3.7 Global Document Commands

Besides the basic commands mentioned in the previous sections, here are additional commands which affect the document as a whole. They are generally all given before the Top node, if they are given at all.

3.7.1 @documentdescription: Summary Text

When producing HTML output for a document, makeinfo writes a ‘<meta>’ element in the ‘<head>’ to give some idea of the content of the document. By default, this description is the title of the document, taken from the @settitle command (see @settitle). To change this, use the @documentdescription environment, as in:

@documentdescription
descriptive text.
@end documentdescription

This will produce the following output in the ‘<head>’ of the HTML:

<meta name=description content="descriptive text.">

@documentdescription must be specified before the first node of the document.

3.7.2 @setchapternewpage: Blank Pages Before Chapters

In an officially bound book, text is usually printed on both sides of the paper, chapters start on right-hand pages, and right-hand pages have odd numbers. But in short reports, text often is printed only on one side of the paper. Also in short reports, chapters sometimes do not start on new pages, but are printed on the same page as the end of the preceding chapter, after a small amount of vertical whitespace.

You can use the @setchapternewpage command with various arguments to specify how TeX should start chapters and whether it should format headers for printing on one or both sides of the paper (single-sided or double-sided printing).

Write the @setchapternewpage command at the beginning of a line followed by its argument.

For example, you would write the following to cause each chapter to start on a fresh odd-numbered page:

@setchapternewpage odd

You can specify one of three alternatives with the @setchapternewpage command:

@setchapternewpage off

Cause TeX to typeset a new chapter on the same page as the last chapter, after skipping some vertical whitespace. Also, cause TeX to format page headers for single-sided printing.

@setchapternewpage on

Cause TeX to start new chapters on new pages and to format page headers for single-sided printing. This is the form most often used for short reports or personal printing. This is the default.

@setchapternewpage odd

Cause TeX to start new chapters on new, odd-numbered pages (right-handed pages) and to typeset for double-sided printing. This is the form most often used for books and manuals.

Texinfo does not have a @setchapternewpage even command, because there is no printing tradition of starting chapters or books on an even-numbered page.

If you don’t like the default headers that @setchapternewpage sets, you can explicit control them with the @headings command. See @headings.

At the beginning of a manual or book, pages are not numbered—for example, the title and copyright pages of a book are not numbered. By convention, table of contents and frontmatter pages are numbered with roman numerals and not in sequence with the rest of the document.

The @setchapternewpage has no effect in output formats that do not have pages, such as Info and HTML.

We recommend not including any @setchapternewpage command in your document source at all, since such desired pagination is not intrinsic to the document. For a particular hard copy run, if you don’t want the default output (no blank pages, same headers on all pages) use the --texinfo option to texi2dvi to specify the output you want.

3.7.3 The @headings Command

The @headings command is rarely used. It specifies what kind of page headings and footings to print on each page. Usually, this is controlled by the @setchapternewpage command. You need the @headings command only if the @setchapternewpage command does not do what you want, or if you want to turn off predefined page headings prior to defining your own. Write a @headings command immediately after the @end titlepage command.

You can use @headings as follows:

@headings off

Turn off printing of page headings.

@headings single

Turn on page headings appropriate for single-sided printing.

@headings double

Turn on page headings appropriate for double-sided printing.

@headings singleafter

@headings doubleafter

Turn on single or double headings, respectively, after the current page is output.

@headings on

Turn on page headings: single if ‘@setchapternewpage on’, double otherwise.

For example, suppose you write @setchapternewpage off before the @titlepage command to tell TeX to start a new chapter on the same page as the end of the last chapter. This command also causes TeX to typeset page headers for single-sided printing. To cause TeX to typeset for double sided printing, write @headings double after the @end titlepage command.

You can stop TeX from generating any page headings at all by writing @headings off on a line of its own immediately after the line containing the @end titlepage command, like this:

@end titlepage
@headings off

The @headings off command overrides the @end titlepage command, which would otherwise cause TeX to print page headings.

You can also specify your own style of page heading and footing. See Page Headings, for more information.

3.7.4 @paragraphindent: Controlling Paragraph Indentation

The Texinfo processors may insert whitespace at the beginning of the first line of each paragraph, thereby indenting that paragraph. You can use the @paragraphindent command to specify this indentation. Write a @paragraphindent command at the beginning of a line followed by either ‘asis’ or a number:

@paragraphindent indent

The indentation is according to the value of indent:

asis

Do not change the existing indentation (not implemented in TeX).

none

0

Omit all indentation.

n

Indent by n space characters in Info output, by n ems in TeX.

The default value of indent is 3. @paragraphindent is ignored for HTML output.

It is best to write the @paragraphindent command before the end-of-header line at the beginning of a Texinfo file, so the region formatting commands indent paragraphs as specified. See Start of Header.

3.7.5 @firstparagraphindent: Indenting After Headings

As you can see in the present manual, the first paragraph in any section is not indented by default. Typographically, indentation is a paragraph separator, which means that it is unnecessary when a new section begins. This indentation is controlled with the @firstparagraphindent command:

@firstparagraphindent word

The first paragraph after a heading is indented according to the value of word:

none

Prevents the first paragraph from being indented (default). This option is ignored by makeinfo if @paragraphindent asis is in effect.

insert

Include normal paragraph indentation. This respects the paragraph indentation set by a @paragraphindent command (see @paragraphindent).

@firstparagraphindent is ignored for HTML and Docbook output.

It is best to write the @firstparagraphindent command before the end-of-header line at the beginning of a Texinfo file, so the region formatting commands indent paragraphs as specified. See Start of Header.

3.7.6 @exampleindent: Environment Indenting

The Texinfo processors indent each line of @example and similar environments. You can use the @exampleindent command to specify this indentation. Write an @exampleindent command at the beginning of a line followed by either ‘asis’ or a number:

@exampleindent indent

The indentation is according to the value of indent:

asis

Do not change the existing indentation (not implemented in TeX).

0

Omit all indentation.

n

Indent environments by n space characters in Info output, by n ems in TeX.

The default value of indent is 5 spaces in Info, and 0.4in in TeX, which is somewhat less. (The reduction is to help TeX fit more characters onto physical lines.)

It is best to write the @exampleindent command before the end-of-header line at the beginning of a Texinfo file, so the region formatting commands indent paragraphs as specified. See Start of Header.

3.8 Ending a Texinfo File

The end of a Texinfo file should include commands to create indices (see Printing Indices & Menus), and the @bye command to mark the last line to be processed. For example:

@node Index
@unnumbered Index

@printindex cp

@bye

An @bye command terminates Texinfo processing. None of the formatters process anything following @bye; any such text is completely ignored. The @bye command should be on a line by itself.

Thus, if you wish, you may follow the @bye line with arbitrary notes. Also, you may follow the @bye line with a local variables list for Emacs, most typically a ‘compile-command’ (see Using the Local Variables List).

4 Nodes

A node is a region of text that begins at a @node command, and continues until the next @node command. To specify a node, write a @node command at the beginning of a line, and follow it with the name of the node. Each node contains the discussion of one topic. Info readers display one node at a time, and provide commands for the user to move to related nodes. The HTML output can be similarly navigated.

Nodes are used as the targets of cross-references. Cross-references, such as the one at the end of this sentence, are made with @xref and related commands; see Cross References. Cross-references can be sprinkled throughout the text, and provide a way to represent links that do not fit a hierarchical structure.

Normally, you put a node command immediately before each chapter structuring command—for example, an @section or @subsection line. (See Chapter Structuring.). You must do this even if you do not intend to format the file for Info. This is because TeX uses both @node names and chapter-structuring names in the output for cross-references. The only time you are likely to use the chapter structuring commands without also using nodes is if you are writing a document that contains no cross references and will only be printed, not transformed into Info, HTML, or other formats.

4.1 Texinfo Document Structure

Nodes can contain menus, which contain the names of child nodes within the parent node; for example, a node corresponding to a chapter would have a menu of the sections in that chapter. The menus allow the user to move to the child nodes in a natural way in the online output.

In addition, nodes contain node pointers that name other nodes. The ‘Next’ and ‘Previous’ pointers form nodes at the same sectioning level into a chain. As you might imagine, the ‘Next’ pointer links to the next node, and the ‘Previous’ pointer links to the previous node. Thus, for example, all the nodes that are at the level of sections within a chapter are linked together, and the order in this chain is the same as the order of the children in the menu of the parent chapter. Each child node records the parent node name as its ‘Up’ pointer.

The Info and HTML output from makeinfo for each node includes links to the ‘Next’, ‘Previous’, and ‘Up’ nodes. The HTML also uses the accesskey attribute with the values ‘n’, ‘p’, and ‘u’ respectively. This allows people using web browsers to follow the navigation using (typically) M-letter, e.g., M-n for the ‘Next’ node, from anywhere within the node. Node pointers and menus provide structure for Info files just as chapters, sections, subsections, and the like provide structure for printed books. The two structures are theoretically distinct; in practice, however, the tree structure of printed books is essentially always used for the node and menu structure also, as this leads to a document which is easiest to follow. See Texinfo Document Structure.

Typically, the sectioning structure and the node structure are completely parallel, with one node for each chapter, section, etc., and with the nodes following the same hierarchical arrangement as the sectioning. Thus, if a node is at the logical level of a chapter, its child nodes are at the level of sections; similarly, the child nodes of sections are at the level of subsections.

Although it is technically possible to create Texinfo documents with only one structure or the other, or for the two structures not to be parallel, or for either the sectioning or node structure to be abnormally formed, etc., this is not at all recommended. To the best of our knowledge, all the Texinfo manuals currently in general use do follow the conventional parallel structure.

4.2 Choosing Node Names

The name of a node identifies the node. For all the details of node names, see Node Line Requirements).

Here are some suggestions for node names:

• Try to pick node names that are informative but short.

  In the Info file, the file name, node name, and pointer names are all inserted on one line, which may run into the right edge of the window. (This does not cause a problem with Info, but is ugly.)

• Try to pick node names that differ from each other near the beginnings of their names. This way, it is easy to use automatic name completion in Info.
• Conventionally, node names are capitalized in the same way as section and chapter titles. In this manual, initial and significant words are capitalized; others are not. In other manuals, just initial words and proper nouns are capitalized. Either way is fine; we recommend just being consistent.

Because node names are used in cross-references, it is not desirable to casually change them once published. Such name changes invalidate references from other manuals, from mail archives, and so on. See HTML Xref Link Preservation.

The pointers from a given node enable you to reach other nodes and consist simply of the names of those nodes. The pointers are usually not specified explicitly, as makeinfo can determine them (see makeinfo Pointer Creation).

Normally, a node’s ‘Up’ pointer contains the name of the node whose menu mentions that node. The node’s ‘Next’ pointer contains the name of the node that follows the present node in that menu and its ‘Previous’ pointer contains the name of the node that precedes it in that menu. When a node’s ‘Previous’ node is the same as its ‘Up’ node, both pointers name the same node.

Usually, the first node of a Texinfo file is the ‘Top’ node, and its ‘Up’ pointer points to the dir file, which contains the main menu for all of Info.

4.3 Writing an @node Line

The easiest way to write an @node line is to write @node at the beginning of a line and then the name of the node, like this:

@node node-name

After you have inserted an @node line, you should immediately write an @-command for the chapter or section and insert its name. Next (and this is important!), put in several index entries. Usually, you will find at least two and often as many as four or five ways of referring to the node in the index. Use them all. This will make it much easier for people to find the node.

If you wish, you can ignore @node lines altogether in your first draft and then use the texinfo-insert-node-lines command to create @node lines for you. However, we do not recommend this practice. It is better to name the node itself at the same time that you write a segment so you can easily make cross-references. Useful cross-references are an especially important feature of a good Texinfo manual.

Even when you explicitly specify all pointers, you cannot write the nodes in the Texinfo source file in an arbitrary order! Because formatters must process the file sequentially, irrespective of node pointers, you must write the nodes in the order you wish them to appear in the output. For Info format one can imagine that the order may not matter, but it matters for the other formats.

You may optionally follow the node name argument to @node with up to three optional arguments on the rest of the same line, separating the arguments with commas. These are the names of the ‘Next’, ‘Previous’, and ‘Up’ pointers, in that order. We recommend omitting them if your Texinfo document is hierarchically organized, as virtually all are (see makeinfo Pointer Creation).

Any spaces before or after each name on the @node line are ignored.

The template for a fully-written-out node line with ‘Next’, ‘Previous’, and ‘Up’ pointers looks like this:

@node node-name, next, previous, up

The node-name argument must be present, but the others are optional. If you wish to specify some but not others, just insert commas as needed, as in: ‘@node mynode,,,uppernode’. However, we recommend leaving off all the pointers and letting makeinfo determine them.

If you are using GNU Emacs, you can use the update node commands provided by Texinfo mode to insert the names of the pointers; or (recommended), you can leave the pointers out of the Texinfo file and let makeinfo insert node pointers into the Info file it creates. (See Texinfo Mode, and makeinfo Pointer Creation.)

Alternatively, you can insert the ‘Next’, ‘Previous’, and ‘Up’ pointers yourself. If you do this, you may find it helpful to use the Texinfo mode keyboard command C-c C-c n. This command inserts ‘@node’ and a comment line listing the names of the pointers in their proper order. The comment line helps you keep track of which arguments are for which pointers. This comment line is especially useful if you are not familiar with Texinfo.

4.4 @node Line Requirements

Names used with @node have several requirements:

• All the node names in a single Texinfo file must be unique.

  This means, for example, that if you end every chapter with a summary, you must name each summary node differently. You cannot just call them all “Summary”. You may, however, duplicate the titles of chapters, sections, and the like. Thus you can end each chapter with a section called “Summary”, so long as the node names for those sections are all different.

• Node names can contain @-commands. The output is generally the natural result of the command; for example, using @TeX{} in a node name results in the TeX logo being output, as it would be in normal text. Cross-references should use @TeX{} just as the node name does.

  For Info and HTML output, especially, it is necessary to expand commands to some sequence of plain characters; for instance, @TeX{} expands to the three letters ‘TeX’ in the Info node name. However, cross-references to the node should not take the “shortcut” of using ‘TeX’; stick to the actual node name, commands and all.

  Some commands do not make sense in node names; for instance, environments (e.g., @quotation), commands that read a whole line as their argument (e.g., @sp), and plenty of others.

  For the complete list of commands that are allowed, and their expansion for HTML identifiers and file names, see HTML Xref Command Expansion. The expansions for Info are generally given with main the description of the command.

  Prior to the Texinfo 5 release in 2013, this feature was supported in an ad hoc way (the --commands-in-node-names option to makeinfo). Now it is part of the language.

• Unfortunately, you cannot reliably use periods, commas, or colons within a node name; these can confuse the Info reader. Also, a node name may not start with a left parenthesis preceding a right parenthesis, as in (not)allowed, since this syntax is used to specify an external manual. (Perhaps these limitations will be removed some day.)

  makeinfo warns about such problematic usage in node names, menu items, and cross-references. If you don’t want to see the warnings, you can set the customization variable INFO_SPECIAL_CHARS_WARNING to ‘0’ (see Other Customization Variables).

  Also, if you insist on using these characters in node names (accepting the resulting substandard Info output), in order not to confuse the Texinfo processors you must still escape those characters, by using either special insertions (see Inserting a Comma) or @asis (see @asis). For example:

  @node foo@asis{::}bar
  

  As an example of avoiding the special characters, the following is a section title in this manual:

  @section @code{@@unnumbered}, @code{@@appendix}: ...
  

  But the corresponding node name lacks the commas and the subtitle:

  @node @unnumbered @appendix
  

• Case is significant in node names.
• Spaces before and after names on the ‘@node’ line are ignored. Multiple whitespace characters “inside” a name are collapsed to a single space. For example:

  @node foo bar
  @node  foo bar,
  @node foo bar ,
  @node foo  bar,
  @node  foo  bar ,
  

  all define the same node, namely ‘foo bar’. In menu entries, this is the name that should be used: no leading or trailing spaces, and a single internal space. (For cross-references, the node name used in the Texinfo sources is automatically normalized in this way.)

• The next/previous/up pointers on @node lines must be the names of nodes. (It’s recommended to leave out these explicit node pointer names, which automatically avoids any problem here; see makeinfo Pointer Creation.)

4.5 The First Node

The first node of a Texinfo file is the Top node, except in an included file (see Include Files). The Top node should contain a short summary, copying permissions, and a master menu. See The Top Node, for more information on the Top node contents and examples.

Here is a description of the node pointers to be used in the Top node:

• The Top node (which must be named ‘top’ or ‘Top’) should have as its ‘Up’ node the name of a node in another file, where there is a menu that leads to this file. Specify the file name in parentheses.

  Usually, all Info files are available through a single virtual Info tree, constructed from multiple directories. In this case, use ‘(dir)’ as the parent of the Top node; this specifies the top-level node in the dir file, which contains the main menu for the Info system as a whole. (Each directory with Info files is intended to contain a file named dir.)

  That’s fine for Info, but for HTML output, one might well want the Up link from the Top node to go somewhere other than ‘dir.html’. For example, for GNU the natural place would be http://www.gnu.org/manual/ (a web page collecting links to most GNU manuals), better specified as just /manual/ if the manual will be installed on www.gnu.org. This can be specified with the TOP_NODE_UP_URL customization variable (see HTML Customization Variables), as in

  $ makeinfo --html -c TOP_NODE_UP_URL=/manual/ ...
  

  All links to (dir) will be replaced by the given url.

• The ‘Prev’ node of the Top node is usually either omitted or also set to (dir). Either is fine.
• The ‘Next’ node of the Top node should be the first chapter in your document.

See Installing an Info File, for more information about installing an Info file in the info directory.

It is usually best to leave the pointers off entirely and let the tools implicitly define them, with this simple result:

@node Top

4.6 The @top Sectioning Command

The @top command is a special sectioning command that you should only use after a ‘@node Top’ line at the beginning of a Texinfo file. The @top command tells the makeinfo formatter which node is to be used as the root of the node tree.

It produces the same sort of output as @unnumbered (see @unnumbered @appendix).

The @top node is conventionally wrapped in an @ifnottex conditional so that it will not appear in TeX output (see Conditionals). Thus, in practice, a Top node usually looks like this:

@ifnottex
@node Top
@top your-manual-title

very-high-level-summary
@end ifnottex

@top is ignored when raising or lowering sections. That is, it is never lowered and nothing can be raised to it (see Raise/lower sections).

4.7 Node and Menu Illustration

Here is a diagram that illustrates a Texinfo file with three chapters, each of which contains two sections.

The “root” is at the top of the diagram and the “leaves” are at the bottom. This is how such a diagram is drawn conventionally; it illustrates an upside-down tree. For this reason, the root node is called the ‘Top’ node, and ‘Up’ node pointers carry you closer to the root.

                         Top
                          |
        -------------------------------------
       |                  |                  |
    Chapter 1          Chapter 2          Chapter 3
       |                  |                  |
    --------           --------           --------
   |        |         |        |         |        |
Section  Section   Section  Section   Section  Section
  1.1      1.2       2.1      2.2       3.1      3.2

Using explicit pointers (not recommended, but for shown for purposes of the example), the fully-written command to start Chapter 2 would be this:

@node     Chapter 2,  Chapter 3, Chapter 1, Top
@comment  node-name,  next,      previous,  up

This @node line says that the name of this node is “Chapter 2”, the name of the ‘Next’ node is “Chapter 3”, the name of the ‘Previous’ node is “Chapter 1”, and the name of the ‘Up’ node is “Top”. You can (and should) omit writing out these node names if your document is hierarchically organized (see makeinfo Pointer Creation), but the pointer relationships still obtain.

Note: ‘Next’ and ‘Previous’ refer to nodes at the same hierarchical level in the manual, not necessarily to the next node within the Texinfo file. In the Texinfo file, the subsequent node may be at a lower level—a section-level node most often follows a chapter-level node, for example. (The ‘Top’ node contains the exception to this rule. Since the ‘Top’ node is the only node at that level, ‘Next’ refers to the first following node, which is almost always a chapter or chapter-level node.)

To go to Sections 2.1 and 2.2 using Info, you need a menu inside Chapter 2. (See Menus.) You would write the menu just before the beginning of Section 2.1, like this:

   @menu
   * Sect. 2.1::    Description of this section.
   * Sect. 2.2::    Description.
   @end menu

Using explicit pointers, the node for Sect. 2.1 is written like this:

@node     Sect. 2.1, Sect. 2.2, Chapter 2, Chapter 2
@comment  node-name, next,      previous,  up

In Info format, the ‘Next’ and ‘Previous’ pointers of a node usually lead to other nodes at the same level—from chapter to chapter or from section to section (sometimes, as shown, the ‘Previous’ pointer points up); an ‘Up’ pointer usually leads to a node at the level above (closer to the ‘Top’ node); and a ‘Menu’ leads to nodes at a level below (closer to ‘leaves’). (A cross-reference can point to a node at any level; see Cross References.)

A @node command and a chapter structuring command are conventionally used together, in that order, often followed by indexing commands. (As shown in the example above, you may follow the @node line with a comment line, e.g., to show which pointer is which if explicit pointers are used.) The Texinfo processors use this construct to determine the relationships between nodes and sectioning commands.

Here is the beginning of the chapter in this manual called “Ending a Texinfo File”. This shows an @node line followed by an @chapter line, and then by indexing lines.

@node Ending a File
@chapter Ending a Texinfo File
@cindex Ending a Texinfo file
@cindex Texinfo file ending
@cindex File ending

An earlier version of the manual used explicit node pointers. Here is the beginning of the same chapter for that case. This shows an @node line followed by a comment line, a @chapter line, and then by indexing lines.

@node    Ending a File, Structuring, Beginning a File, Top
@comment node-name,     next,        previous,         up
@chapter Ending a Texinfo File
@cindex Ending a Texinfo file
…

4.8 makeinfo Pointer Creation

The makeinfo program can automatically determine node pointers for a hierarchically organized document. This implicit node pointer creation feature in makeinfo relieves you from the need to update menus and pointers manually or with Texinfo mode commands. (See Updating Nodes and Menus.) We highly recommend taking advantage of this.

To do so, write your @node lines with just the name of the node:

@node My Node

You do not need to write out the ‘Next’, ‘Previous’, and ‘Up’ pointers.

Then, you must write a sectioning command, such as @chapter or @section, on the line immediately following each truncated @node line (except that comment lines may intervene). This is where it normally goes.

Also, you must write the name of each node (except for the ‘Top’ node) in a menu that is one or more hierarchical levels above the node’s level.

Finally, you must follow the ‘Top’ @node line with a line beginning with @top to mark the top-level node in the file. See @top Command.

If you use a detailed menu in your master menu (see Master Menu Parts), mark it with the @detailmenu … @end detailmenu environment, or makeinfo will get confused, typically about the last and/or first node in the document.

In most cases, you will want to take advantage of this feature and not redundantly specify node pointers that the programs can determine. However, Texinfo documents are not required to be organized hierarchically or in fact to contain sectioning commands at all (for example, if you never intend the document to be printed), so node pointers may still be specified explicitly, in full generality.

4.9 Menus

Menus contain pointers to subordinate nodes. In online output, you use menus to go to such nodes. Menus have no effect in printed manuals and do not appear in them.

4.9.1 Writing a Menu

A menu consists of a @menu command on a line by itself, followed by menu entry lines or menu comment lines, and then followed by an @end menu command on a line by itself.

A menu looks like this:

@menu
Larger Units of Text

* Files::                       All about handling files.
* Multiples: Buffers.           Multiple buffers; editing
                                 several files at once.
@end menu

In a menu, every line that begins with an ‘* ’ is a menu entry. (Note the space after the asterisk.)

A line that does not start with an ‘* ’ may also appear in a menu. Such a line is not a menu entry but rather a menu comment line that appears in the Info file. In the example above, the line ‘Larger Units of Text’ is such a menu comment line; the two lines starting with ‘* ’ are menu entries.

Technically, menus can carry you to any node, regardless of the structure of the document; even to nodes in a different Info file. However, we do not recommend making use of this, because it is hard for readers to follow. Also, the makeinfo implicit pointer creation feature (see makeinfo Pointer Creation) and GNU Emacs Texinfo mode updating commands work only to create menus of subordinate nodes in a hierarchically structured document. It is much better to use cross-references to refer to arbitrary nodes.

makeinfo can automatically generate menus in nodes for Info and HTML output, based on the chapter structure of the document. To specify that you want it to do this, place the line ‘@validatemenus off’ near the beginning of the document.

In Info, a user selects a node with the m (Info-menu) command. The menu entry name is what the user types after the m command. In the HTML output from makeinfo, the accesskey attribute is used with the values ‘1’…‘9’ for the first nine entries. This allows people using web browsers to follow the first menu entries using (typically) M-digit, e.g., M-1 for the first entry.

4.9.2 A Menu Example

A menu looks like this in Texinfo:

@menu
* menu entry name: Node name.   A short description.
* Node name::                   This form is preferred.
@end menu

This produces:

* menu:

* menu entry name: Node name.   A short description.
* Node name::                   This form is preferred.

Here is an example as you might see it in a Texinfo file:

@menu
Larger Units of Text

* Files::                       All about handling files.
* Multiples: Buffers.           Multiple buffers; editing
                                 several files at once.
@end menu

This produces:

* menu:
Larger Units of Text

* Files::                       All about handling files.
* Multiples: Buffers.           Multiple buffers; editing
                                 several files at once.

In this example, the menu has two entries. ‘Files’ is both a menu entry name and the name of the node referred to by that name. ‘Multiples’ is the menu entry name; it refers to the node named ‘Buffers’. The line ‘Larger Units of Text’ is a comment; it appears in the menu, but is not an entry.

Since no file name is specified with either ‘Files’ or ‘Buffers’, they must be the names of nodes in the same Info file (see Referring to Other Info Files).

4.9.3 Menu Location

There may be at most one menu in a node. A menu is conventionally located at the end of a node, without any regular text or additional commands between the @end menu and the beginning of the next node.

This convention is useful, since a reader who uses the menu could easily miss any such text. Also, any such post-menu text will be considered part of the menu in Info output (which has no marker for the end of a menu). Thus, a line beginning with ‘* ’ will likely be incorrectly handled.

It’s usually best if a node with a menu does not contain much text. If you find yourself with a lot of text before a menu, we generally recommend moving all but a couple of paragraphs into a new subnode. Otherwise, it is easy for readers to miss the menu.

4.9.4 The Parts of a Menu

A menu entry has three parts, only the second of which is required:

1. The menu entry name (optional).
2. The name of the node (required).
3. A description of the item (optional).

The template for a generic menu entry looks like this (but see the next section for one more possibility):

* menu-entry-name: node-name.   description

Follow the menu entry name with a single colon, and follow the node name with tab, comma, newline, or the two characters period and space (‘. ’).

The third part of a menu entry is a descriptive phrase or sentence. Menu entry names and node names are often short; the description explains to the reader what the node is about. A useful description complements the node name rather than repeats it. The description, which is optional, can spread over multiple lines; if it does, some authors prefer to indent the second line while others prefer to align it with the first (and all others). It’s up to you. An empty line, or the next menu entry, ends a description.

Space characters in a menu are preserved as-is in the Info output; this allows you to format the menu as you wish. Unfortunately you must type node names without any extra spaces or some versions of some Info readers will not find the node (see Node Line Requirements).

makeinfo warns when the text of a menu item (and node names and cross-references) contains a problematic construct that will interfere with its parsing in Info. If you don’t want to see the warnings, you can set the customization variable INFO_SPECIAL_CHARS_WARNING to ‘0’ (see Other Customization Variables).

4.9.5 Less Cluttered Menu Entry

When the menu entry name and node name are the same, you can write the name immediately after the asterisk and space at the beginning of the line and follow the name with two colons.

For example, write

* Name::                        description

instead of

* Name: Name.                   description

We recommend using the node name for the menu entry name whenever possible, since it reduces visual clutter in the menu.

4.9.6 Referring to Other Info Files

You can create a menu entry that enables a reader in Info to go to a node in another Info file by writing the file name in parentheses just before the node name. Some examples:

@menu
* first-entry-name:(filename)nodename.     description
* (filename)second-node::                  description
@end menu

For example, to refer directly to the ‘Outlining’ and ‘Rebinding’ nodes in the Emacs Manual, you could write a menu like this:

@menu
* Outlining: (emacs)Outline Mode. The major mode for
                                   editing outlines.
* (emacs)Rebinding::              How to redefine the
                                   meaning of a key.
@end menu

If you do not list the node name, but only name the file, then Info presumes that you are referring to the ‘Top’ node. Examples:

* Info: (info).         Documentation browsing system.
* (emacs)::             The extensible, self-documenting
                         text editor.

The GNU Emacs Texinfo mode menu updating commands only work with nodes within the current buffer, so you cannot use them to create menus that refer to other files. You must write such menus by hand.

5 Chapter Structuring

Texinfo’s chapter structuring commands divide a document into a hierarchy of chapters, sections, subsections, and subsubsections. These commands generate large headings in the text, like the one above. They also provide information for generating the table of contents (see Generating a Table of Contents).

Normally you put a @node command immediately before each chapter structuring command. See Nodes.

5.1 Tree Structure of Sections

A Texinfo file is usually structured like a book with chapters, sections, subsections, and the like. This structure can be visualized as a tree (or rather as an upside-down tree) with the root at the top and the levels corresponding to chapters, sections, subsection, and subsubsections.

Here is a diagram that shows a Texinfo file with three chapters, each with two sections.

                         Top
                          |
        -------------------------------------
       |                  |                  |
    Chapter 1          Chapter 2          Chapter 3
       |                  |                  |
    --------           --------           --------
   |        |         |        |         |        |
Section  Section   Section  Section   Section  Section
  1.1      1.2       2.1      2.2       3.1      3.2

In a Texinfo file that has this structure, the beginning of Chapter 2 would be written like this:

@node    Chapter 2
@chapter Chapter 2

For purposes of example, here is how it would be written with explicit node pointers:

@node    Chapter 2,  Chapter 3, Chapter 1, Top
@chapter Chapter 2

The chapter structuring commands are described in the sections that follow; the @node command is described in the previous chapter (see Nodes).

5.2 Structuring Command Types

The chapter structuring commands fall into four groups or series, each of which contains structuring commands corresponding to the hierarchical levels of chapters, sections, subsections, and subsubsections.

The four groups of commands are the @chapter series, the @unnumbered series, the @appendix series, and the @heading series. Each command produces a title with a different appearance in the body of the document. Some of the commands list their titles in the tables of contents, while others do not. Here are the details:

• The @chapter and @appendix series of commands produce numbered or lettered entries both in the body of a document and in its table of contents.
• The @unnumbered series of commands produce unnumbered entries both in the body of a document and in its table of contents. The @top command, which has a special use, is a member of this series (see @top Command). An @unnumbered section is a normal part of the document structure.
• The @heading series of commands produce simple unnumbered headings that do not appear in a table of contents, are not associated with nodes, and cannot be cross-referenced. These heading commands never start a new page.

When a @setchapternewpage command says to do so, the @chapter, @unnumbered, and @appendix commands start new pages in the printed manual; the @heading commands do not. See @setchapternewpage.

Here is a summary:

5.3 @chapter: Chapter Structuring

@chapter identifies a chapter in the document–the highest level of the normal document structuring hierarchy. Write the command at the beginning of a line and follow it on the same line by the title of the chapter. The chapter is numbered automatically, starting from 1.

For example, the present chapter in this manual is entitled “@chapter: Chapter Structuring”; the @chapter line looks like this:

@chapter @code{@@chapter}: Chapter Structuring

In TeX, the @chapter command produces a chapter heading in the document.

In Info and plain text output, the @chapter command causes the title to appear on a line by itself, with a line of asterisks inserted underneath. So, the above example produces the following output:

5 Chapter Structuring
*********************

In HTML, the @chapter command produces an <h2>-level header by default (controlled by the CHAPTER_HEADER_LEVEL customization variable, see Other Customization Variables).

In the XML and Docbook output, a <chapter> element is produced that includes all the following sections, up to the next chapter.

5.4 @unnumbered, @appendix: Chapters with Other Labeling

Use the @unnumbered command to start a chapter-level element that appears without chapter numbers of any kind. Use the @appendix command to start an appendix that is labeled by letter (‘A’, ‘B’, …) instead of by number; appendices are also at the chapter level of structuring.

Write an @appendix or @unnumbered command at the beginning of a line and follow it on the same line by the title, just as with @chapter.

Texinfo also provides a command @centerchap, which is analogous to @unnumbered, but centers its argument in the printed and HTML outputs. This kind of stylistic choice is not usually offered by Texinfo. It may be suitable for short documents.

With @unnumbered, if the name of the associated node is one of these English words (case-insensitive):

Acknowledgements  Colophon  Dedication  Preface

then the Docbook output uses corresponding special tags (<preface>, etc.) instead of the default <chapter>. The argument to @unnumbered itself can be anything, and is output as the following <title> text as usual.

5.5 @majorheading, @chapheading: Chapter-level Headings

The @majorheading and @chapheading commands produce chapter-like headings in the body of a document.

However, neither command produces an entry in the table of contents, and neither command causes TeX to start a new page in a printed manual.

In TeX, a @majorheading command generates a larger vertical whitespace before the heading than a @chapheading command but is otherwise the same.

In Info and plain text, the @majorheading and @chapheading commands produce the same output as @chapter: the title is printed on a line by itself with a line of asterisks underneath. Similarly for HTML. The only difference is the lack of numbering and the lack of any association with nodes. See @chapter.

5.6 @section: Sections Below Chapters

An @section command identifies a section within a chapter unit, whether created with @chapter, @unnumbered, or @appendix, following the numbering scheme of the chapter-level command. Thus, within a @chapter chapter numbered ‘1’, the sections are numbered ‘1.1’, ‘1.2’, etc.; within an @appendix “chapter” labeled ‘A’, the sections are numbered ‘A.1’, ‘A.2’, etc.; within an @unnumbered chapter, the section gets no number. The output is underlined with ‘=’ in Info and plain text.

To make a section, write the @section command at the beginning of a line and follow it on the same line by the section title. For example,

@section This is a section

might produce the following in Info:

5.7 This is a section
=====================

Section titles are listed in the table of contents.

The TeX, HTML, Docbook, and XML output is all analogous to the chapter-level output, just “one level down”; see @chapter.

5.7 @unnumberedsec, @appendixsec, @heading

The @unnumberedsec, @appendixsec, and @heading commands are, respectively, the unnumbered, appendix-like, and heading-like equivalents of the @section command (see the previous section).

@unnumberedsec and @appendixsec do not need to be used in ordinary circumstances, because @section may also be used within @unnumbered and @appendix chapters; again, see the previous section.

@unnumberedsec

The @unnumberedsec command may be used within an unnumbered chapter or within a regular chapter or appendix to produce an unnumbered section.

@appendixsec

@appendixsection

@appendixsection is a longer spelling of the @appendixsec command; the two are synonymous.

Conventionally, the @appendixsec or @appendixsection command is used only within appendices.

@heading

You may use the @heading command (almost) anywhere for a section-style heading that will not appear in the table of contents. The @heading-series commands can appear inside most environments, for example, though pathological and useless locations such as inside @titlepage, as an argument to another command, etc., are not allowed.

5.8 @subsection: Subsections Below Sections

Subsections are to sections as sections are to chapters; see @section. In Info and plain text, subsection titles are underlined with ‘-’. For example,

@subsection This is a subsection

might produce

1.2.3 This is a subsection
--------------------------

Subsection titles are listed in the table of contents.

The TeX, HTML, Docbook, and XML output is all analogous to the chapter-level output, just “two levels down”; see @chapter.

5.9 The @subsection-like Commands

The @unnumberedsubsec, @appendixsubsec, and @subheading commands are, respectively, the unnumbered, appendix-like, and heading-like equivalents of the @subsection command. (See @subsection.)

@unnumberedsubsec and @appendixsubsec do not need to be used in ordinary circumstances, because @subsection may also be used within sections of @unnumbered and @appendix chapters (see @section).

An @subheading command produces a heading like that of a subsection except that it is not numbered and does not appear in the table of contents. Similarly, an @unnumberedsubsec command produces an unnumbered heading like that of a subsection and an @appendixsubsec command produces a subsection-like heading labeled with a letter and numbers; both of these commands produce headings that appear in the table of contents. In Info and plain text, the @subsection-like commands generate a title underlined with hyphens.

5.10 @subsection and Other Subsub Commands

The fourth and lowest level sectioning commands in Texinfo are the ‘subsub’ commands. They are:

@subsubsection

Subsubsections are to subsections as subsections are to sections. (See @subsection.) Subsubsection titles appear in the table of contents.

@unnumberedsubsubsec

Unnumbered subsubsection titles appear in the table of contents, but lack numbers. Otherwise, unnumbered subsubsections are the same as subsubsections.

@appendixsubsubsec

Conventionally, appendix commands are used only for appendices and are lettered and numbered appropriately. They also appear in the table of contents.

@subsubheading

The @subsubheading command may be used anywhere that you want a small heading that will not appear in the table of contents.

As with subsections, @unnumberedsubsubsec and @appendixsubsubsec do not need to be used in ordinary circumstances, because @subsubsection may also be used within subsections of @unnumbered and @appendix chapters (see @section).

In Info, ‘subsub’ titles are underlined with periods. For example,

@subsubsection This is a subsubsection

might produce

1.2.3.4 This is a subsubsection
...............................

The TeX, HTML, Docbook, and XML output is all analogous to the chapter-level output, just “three levels down”; see @chapter.

5.11 @part: Groups of Chapters

The final sectioning command is @part, to mark a part of a manual, that is, a group of chapters or (rarely) appendices. This behaves quite differently from the other sectioning commands, to fit with the way such “parts” are conventionally used in books.

No @node command is associated with @part. Just write the command on a line by itself, including the part title, at the place in the document you want to mark off as starting that part. For example:

@part Part I:@* The beginning

As can be inferred from this example, no automatic numbering or labeling of the @part text is done. The text is taken as-is.

Because parts are not associated with nodes, no general text can follow the @part line. To produce the intended output, it must be followed by a chapter-level command (including its node). Thus, to continue the example:

@part Part I:@* The beginning

@node Introduction
@chapter Introduction
...

In the TeX output, the @part text is included in both the normal and short tables of contents (see Contents), without a page number (since that is the normal convention). In addition, a “part page” is output in the body of the document, with just the @part text. In the example above, the @* causes a line break on the part page (but is replaced with a space in the tables of contents). This part page is always forced to be on an odd (right-hand) page, regardless of the chapter pagination (see @setchapternewpage).

In the HTML output, the @part text is similarly included in the tables of contents, and a heading is included in the main document text, as part of the following chapter or appendix node.

In the XML and Docbook output, the <part> element includes all the following chapters, up to the next <part>. A <part> containing chapters is also closed at an appendix.

In the Info and plain text output, @part has no effect.

@part is ignored when raising or lowering sections (see next section). That is, it is never lowered and nothing can be raised to it.

5.12 Raise/lower Sections: @raisesections and @lowersections

The @raisesections and @lowersections commands implicitly raise and lower the hierarchical level of following chapters, sections and the other sectioning commands (excluding parts).

That is, the @raisesections command changes sections to chapters, subsections to sections, and so on. Conversely, the @lowersections command changes chapters to sections, sections to subsections, and so on. Thus, a @lowersections command cancels a @raisesections command, and vice versa.

You can use @lowersections to include text written as an outer or standalone Texinfo file in another Texinfo file as an inner, included file (see Include Files). Typical usage looks like this:

@lowersections
@include somefile.texi
@raisesections

(Without the @raisesections, all the subsequent sections in the main file would also be lowered.)

If the included file being lowered has a @top node, you’ll need to conditionalize its inclusion with a flag (see @set @value).

As a practical matter, you generally only want to raise or lower large chunks, usually in external files as shown above. The final result has to have menus that take the raising and lowering into account, so you cannot just arbitrarily sprinkle @raisesections and @lowersections commands throughout the document.

Repeated use of the commands continues to raise or lower the hierarchical level a step at a time. An attempt to raise above ‘chapter’ reproduces chapter commands; an attempt to lower below ‘subsubsection’ reproduces subsubsection commands. Also, lowered subsubsections and raised chapters will not work with makeinfo’s feature of implicitly determining node pointers, since the menu structure cannot be represented correctly.

Write each @raisesections and @lowersections command on a line of its own.

6 Cross-references

Cross-references are used to refer the reader to other parts of the same or different Texinfo files.

6.1 What References Are For

Often, but not always, a printed document should be designed so that it can be read sequentially. People tire of flipping back and forth to find information that should be presented to them as they need it.

However, in any document, some information will be too detailed for the current context, or incidental to it; use cross-references to provide access to such information. Also, an online help system or a reference manual is not like a novel; few read such documents in sequence from beginning to end. Instead, people look up what they need. For this reason, such creations should contain many cross references to help readers find other information that they may not have read.

In a printed manual, a cross-reference results in a page reference, unless it is to another manual altogether, in which case the cross-reference names that manual. In Info, a cross-reference results in an entry that you can follow using the Info ‘f’ command. (See Following cross-references in Info.) In HTML, a cross-reference results in an hyperlink.

The various cross-reference commands use nodes (or anchors, see @anchor) to define cross-reference locations. TeX needs nodes to define cross-reference locations. When TeX generates a DVI file, it records each node’s page number and uses the page numbers in making references. Thus, even if you are writing a manual that will only be printed, and not used online, you must nonetheless write @node lines in order to name the places to which you make cross-references.

6.2 Different Cross-reference Commands

There are three different cross-reference commands:

@xref

Used to start a sentence in the printed manual and in HTML with ‘See …’ or an Info cross-reference saying ‘*Note name: node.’.

@ref

Used within or, more often, at the end of a sentence; produces just the reference in the printed manual and in HTML without the preceding ‘See’ (same as @xref for Info).

@pxref

Used within parentheses, at the end of a sentence, or otherwise before punctuation, to make a reference. Its output starts with a lowercase ‘see’ in the printed manual and in HTML, and a lowercase ‘*note’ in Info. (‘p’ is for ‘parenthesis’.)

Additionally, there are commands to produce references to documents outside the Texinfo system. The @cite command is used to make references to books and manuals. @url produces a URL, for example a reference to a page on the World Wide Web. @inforef is used to make a reference to an Info file for which there is no printed manual.

6.3 Parts of a Cross-reference

A cross-reference command requires only one argument, which is the name of the node to which it refers. Here is a simple example:

@xref{Node name}.

In Info output, this produces

*Note Node name::.

In a printed manual, the output is

See Section nnn [Node name], page ppp.

A cross-reference command may contain up to four additional arguments. By using these arguments, you can provide a cross-reference name, a topic description or section title for the printed output, the name of a different manual file, and the name of a different printed manual. To refer to another manual as a whole, the manual file and/or the name of the printed manual are the only required arguments (see Referring to a Manual as a Whole).

Here is an example of a full five-part cross-reference:

@xref{Node name, Online Label, Printed Label,
info-file-name, A Printed Manual}, for details.

which produces

*Note Online Label: (info-file-name)Node name,
for details.

in Info and

See section “Printed Label” in A Printed Manual, for details.

in a printed book.

The five possible arguments for a cross-reference are:

1. The node or anchor name (required, except for reference to whole manuals). This is the location to which the cross-reference takes you. In a printed document, the location of the node provides the page reference only for references within the same document. Use @node to define the node (see Writing a Node), or @anchor (see @anchor).

   Write a node name in a cross-reference in exactly the same way as in the @node line, including the same capitalization; otherwise, the formatters may not find the reference.

2. A label for online output. It is usually omitted; then the topic description (third argument) is used if it was specified; if that was omitted as well, the node name is used.
3. A label for printed output. Often, this is the title or topic of the section. This is used as the name of the reference in the printed manual. If omitted, the node name is used.
4. The name of the manual file in which the reference is located, if it is different from the current file. This name is used both for Info and HTML.
5. The name of a printed manual from a different Texinfo file.

The template for a full five argument cross-reference looks like this:

@xref{node-name, online-label, printed-label,
info-file-name, printed-manual-title}

Whitespace before and after the commas separating these arguments is ignored. To include a comma in one of the arguments, use @comma{} (see Inserting a Comma).

When processing with TeX, a comma is automatically inserted after the page number for cross-references to within the same manual, unless the closing brace of the argument is followed by non-whitespace (such as a comma or period). This gives you the choice of whether to have a comma there in Info or HTML output. For example,

@xref{Another Section} for more information

produces ‘See Another Section, page ppp, for more information’ in the printed output, and ‘*Note Another Section:: for more information’ in the Info output.

If an unwanted comma is added, follow the argument with a command such as ‘@:’. For example, ‘@xref{Hurricanes}@: --- for the details’ produces

See Hurricanes, page ppp — for the details

instead of ‘See Hurricanes, page ppp, — for the details’.

Cross-references with one, two, three, four, and five arguments are described separately following the description of @xref.

makeinfo warns when the text of a cross-reference (and node names and menu items) contains a problematic construct that will interfere with its parsing in Info. If you don’t want to see the warnings, you can set the customization variable INFO_SPECIAL_CHARS_WARNING to ‘0’ (see Other Customization Variables).

6.4 @xref

The @xref command generates a cross-reference for the beginning of a sentence.

6.4.1 @xref with One Argument

The simplest form of @xref takes one argument, the name of another node in the same Texinfo file.

For example,

@xref{Tropical Storms}.

produces

*Note Tropical Storms::.

in Info and

See Section 3.1 [Tropical Storms], page 24.

in a printed manual.

6.4.2 @xref with Two Arguments

With two arguments, the second is used as a label for the online output.

The template is like this:

@xref{node-name, online-label}.

For example,

@xref{Electrical Effects, Lightning}.

produces:

*Note Lightning: Electrical Effects.

in Info and

See Section 5.2 [Electrical Effects], page 57.

in a printed manual, where the node name is printed.

The second argument to cross-references must observe some of the restrictions for node names (see Node Line Requirements). The most common issue is that colons cannot be used, since that interferes with the parsing of the Info file.

6.4.3 @xref with Three Arguments

A third argument replaces the node name in the TeX output. The third argument should be the name of the section in the printed output, or else state the topic discussed by that section.

The template is like this:

@xref{node-name, online-label, printed-label}.

For example,

@xref{Electrical Effects, Lightning, Thunder and Lightning},
for details.

produces

*Note Lightning: Electrical Effects, for details.

in Info and

See Section 5.2 [Thunder and Lightning], page 57, for details.

in a printed manual.

If a third argument is given and the second one is empty, then the third argument serves for both. (Note how two commas, side by side, mark the empty second argument.)

@xref{Electrical Effects, , Thunder and Lightning},
for details.

produces

*Note Thunder and Lightning: Electrical Effects, for details.

in Info and

See Section 5.2 [Thunder and Lightning], page 57, for details.

in a printed manual.

The third argument to cross-references must observe some of the restrictions for node names (see Node Line Requirements). The most common issue is that colons cannot be used, since that interferes with the parsing of the Info file.

As a practical matter, it is often best to write cross-references with just the first argument if the node name and the section title are the same (or nearly so), and with the first and third arguments only if the node name and title are different.

Texinfo offers a setting to use the section title instead of node names by default in cross-references (an explicitly specified third argument still takes precedence):

@xrefautomaticsectiontitle on

Typically this line would be given near the beginning of the document and used for the whole manual. But you can turn it off if you want (@xrefautomaticsectiontitle off), for example, if you’re including some other sub-document that doesn’t have suitable section names.

6.4.4 @xref with Four and Five Arguments

In a cross-reference, a fourth argument specifies the name of another Info file, different from the file in which the reference appears, and a fifth argument specifies its title as a printed manual.

The full template is:

@xref{node-name, online-label, printed-label,
info-file-name, printed-manual-title}.

For example,

@xref{Electrical Effects, Lightning, Thunder and Lightning,
weather, An Introduction to Meteorology}.

produces this output in Info:

*Note Lightning: (weather)Electrical Effects.

As you can see, the name of the Info file is enclosed in parentheses and precedes the name of the node.

In a printed manual, the reference looks like this:

See section “Thunder and Lightning” in An Introduction to Meteorology.

The title of the printed manual is typeset like @cite; and the reference lacks a page number since TeX cannot know to which page a reference refers when that reference is to another manual.

Next case: often, you will leave out the second argument when you use the long version of @xref. In this case, the third argument, the topic description, will be used as the cross-reference name in Info. For example,

@xref{Electrical Effects, , Thunder and Lightning,
weather, An Introduction to Meteorology}.

produces

*Note Thunder and Lightning: (weather)Electrical Effects.

in Info and

See section “Thunder and Lightning” in An Introduction to Meteorology.

in a printed manual.

Next case: If the node name and the section title are the same in the other manual, you may also leave out the section title. In this case, the node name is used in both instances. For example,

@xref{Electrical Effects,,,
weather, An Introduction to Meteorology}.

produces

*Note (weather)Electrical Effects::.

in Info and

See section “Electrical Effects” in An Introduction to Meteorology.

in a printed manual.

A very unusual case: you may want to refer to another manual file that is within a single printed manual—when multiple Texinfo files are incorporated into the same TeX run but can create separate Info or HTML output files. In this case, you need to specify only the fourth argument, and not the fifth.

Finally, it’s also allowed to leave out all the arguments except the fourth and fifth, to refer to another manual as a whole. See the next section.

6.5 Referring to a Manual as a Whole

Ordinarily, you must always name a node in a cross-reference. However, it’s not unusual to want to refer to another manual as a whole, rather than a particular section within it. In this case, giving any section name is an unnecessary distraction.

So, with cross-references to other manuals (see Four and Five Arguments), if the first argument is either ‘Top’ (capitalized just that way) or omitted entirely, and the third argument is omitted, the printed output includes no node or section name. (The Info output includes ‘Top’ if it was given.) For example,

@xref{Top,,, make, The GNU Make Manual}.

produces

*Note (make)Top::.

and

See The GNU Make Manual.

Info readers will go to the Top node of the manual whether or not the ‘Top’ node is explicitly specified.

It’s also possible (and is historical practice) to refer to a whole manual by specifying the ‘Top’ node and an appropriate entry for the third argument to the @xref command. Using this idiom, to make a cross-reference to The GNU Make Manual, you would write:

@xref{Top,, Overview, make, The GNU Make Manual}.

which produces

*Note Overview: (make)Top.

in Info and

See section “Overview” in The GNU Make Manual.

in a printed manual.

In this example, ‘Top’ is the name of the first node, and ‘Overview’ is the name of the first section of the manual. There is no widely-used convention for naming the first section in a printed manual, this is just what the Make manual happens to use. This arbitrariness of the first name is a principal reason why omitting the third argument in whole-manual cross-references is preferable.

6.6 @ref

@ref is nearly the same as @xref except that it does not generate a ‘See’ in the printed output, just the reference itself. This makes it useful as the last part of a sentence.

For example,

For more information, @pxref{This}, and @ref{That}.

produces in Info:

For more information, *note This::, and *note That::.

and in printed output:

For more information, see Section 1.1 [This], page 1, and Section 1.2 [That], page 2.

The @ref command can tempt writers to express themselves in a manner that is suitable for a printed manual but looks awkward in the Info format. Bear in mind that your audience could be using both the printed and the Info format. For example:

Sea surges are described in @ref{Hurricanes}.

looks ok in the printed output:

Sea surges are described in Section 6.7 [Hurricanes], page 72.

but is awkward to read in Info, “note” being a verb:

Sea surges are described in *note Hurricanes::.

6.7 @pxref

The parenthetical reference command, @pxref, is nearly the same as @xref, but it is best used at the end of a sentence or before a closing parenthesis. The command differs from @xref in that TeX typesets the reference for the printed manual with a lowercase ‘see’ rather than an uppercase ‘See’.

With one argument, a parenthetical cross-reference looks like this:

… storms cause flooding (@pxref{Hurricanes}) …

which produces

… storms cause flooding (*note Hurricanes::) …

in Info and

… storms cause flooding (see Section 6.7 [Hurricanes], page 72) …

in a printed manual.

With two arguments, a parenthetical cross-reference has this template:

… (@pxref{node-name, cross-reference-name}) …

which produces

… (*note cross-reference-name: node-name.) …

in Info and

… (see Section nnn [node-name], page ppp) …

in a printed manual.

@pxref can be used with up to five arguments, just like @xref (see @xref).

In past versions of Texinfo, it was not allowed to write punctuation after a @pxref, so it could be used only before a right parenthesis. This is no longer the case, so now it can be used (for example) at the end of a sentence, where a lowercase “see” works best. For instance:

… For more information, @pxref{More}.

which outputs (in Info):

… For more information, *note More::.

As a matter of style, @pxref is best used at the ends of sentences. Although it technically works in the middle of a sentence, that location breaks up the flow of reading.

6.8 @anchor: Defining Arbitrary Cross-reference Targets

An anchor is a position in your document, labelled so that cross-references can refer to it, just as they can to nodes. You create an anchor with the @anchor command, and give the label as a normal brace-delimited argument. For example:

This marks the @anchor{x-spot}spot.
…
@xref{x-spot,,the spot}.

produces:

This marks the spot.
…
See [the spot], page 1.

As you can see, the @anchor command itself produces no output. This example defines an anchor ‘x-spot’ just before the word ‘spot’. You can refer to it later with an @xref or other cross reference command, as shown (see Cross References).

It is best to put @anchor commands just before the position you wish to refer to; that way, the reader’s eye is led on to the correct text when they jump to the anchor. You can put the @anchor command on a line by itself if that helps readability of the source. Whitespace (including newlines) is ignored after @anchor.

Anchor names and node names may not conflict. Anchors and nodes are given similar treatment in some ways; for example, the goto-node command takes either an anchor name or a node name as an argument. (See Go to node in Info.)

Also like node names, anchor names cannot include some characters (see Node Line Requirements).

Because of this duality, when you delete or rename a node, it is usually a good idea to define an @anchor with the old name. That way, any links to the old node, whether from other Texinfo manuals or general web pages, keep working. You can also do this with the RENAMED_NODES_FILE feature of makeinfo (see HTML Xref Link Preservation). Both methods keep links on the web working; the only substantive difference is that defining anchors also makes the old node names available when reading the document in Info.

6.9 @inforef: Cross-references to Info-only Material

@inforef is used for making cross-references to Info documents—even from a printed manual. This might be because you want to refer to conditional @ifinfo text (see Conditionals), or because printed output is not available (perhaps because there is no Texinfo source), among other possibilities.

The command takes either two or three arguments, in the following order:

1. The node name.
2. The cross-reference name (optional).
3. The Info file name.

The template is:

@inforef{node-name, cross-reference-name, info-file-name}

For example,

@inforef{Advanced, Advanced Info commands, info},
for more information.

produces (in Info):

*Note Advanced Info commands: (info)Advanced,
for more information.

and (in the printed output):

See Info file info, node ‘Advanced’, for more information.

(This particular example is not realistic, since the Info manual is written in Texinfo, so all formats are available. In fact, we don’t know of any extant Info-only manuals.)

The converse of @inforef is @cite, which is used to refer to printed works for which no Info form exists. See @cite.

6.10 @url, @uref{url[, text][, replacement]}

@uref produces a reference to a uniform resource locator (url). It takes one mandatory argument, the url, and two optional arguments which control the text that is displayed. In HTML and PDF output, @uref produces a link you can follow. (To merely indicate a url without creating a link people can follow, use @indicateurl, see @indicateurl.)

@url is a synonym for @uref. (Originally, @url had the meaning of @indicateurl, but in practice it was almost always misused. So we’ve changed the meaning.)

The second argument, if specified, is the text to display (the default is the url itself); in Info, DVI, and PDF output, but not in HTML output, the url is output in addition to this text.

The third argument, if specified, is the text to display, but in this case the url is not output in any format. This is useful when the text is already sufficiently referential, as in a man page. Also, if the third argument is given, the second argument is ignored.

6.10.1 @url Examples

First, here is an example of the simplest form of @url, with just one argument. The given url is both the target and the visible text of the link:

The official GNU ftp site is @uref{http://ftp.gnu.org/gnu}.

produces:

The official GNU ftp site is http://ftp.gnu.org/gnu.

Two-argument form of @url

Here is an example of the two-argument form:

The official @uref{http://ftp.gnu.org/gnu, GNU ftp site}
holds programs and texts.

which produces:

The official GNU ftp site
holds programs and texts.

that is, the Info (and TeX, etc.) output is this:

The official GNU ftp site (http://ftp.gnu.org/gnu)
holds programs and texts.

while the HTML output is this:

The official <a href="http://ftp.gnu.org/gnu">GNU ftp site</a>
holds programs and texts.

Three-argument form of @url

Finally, an example of the three-argument form:

The @uref{/man.cgi/1/ls,,ls} program …

which, except for HTML, produces:

The ls program …

but with HTML:

The <a href="/man.cgi/1/ls">ls</a> program …

By the way, some people prefer to display urls in the unambiguous format:

<URL:http://host/path>

You can use this form in the input file if you wish. We feel it’s not necessary to include the ‘<URL:’ and ‘>’ in the output, since to be useful any software that tries to detect urls in text already has to detect them without the ‘<URL:’.

6.10.2 URL Line Breaking

TeX allows line breaking within urls at only a few characters (which are special in urls): ‘&’, ‘.’, ‘#’, ‘?’, and ‘/’ (but not between two ‘/’ characters). A tiny amount of stretchable space is also inserted around these characters to help with line breaking.

For HTML output, modern browsers will also do line breaking within displayed urls. If you need to allow breaks at other characters you can insert @/ as needed (see Line Breaks).

By default, in TeX any such breaks at special characters will occur after the character. Some people prefer such breaks to happen before the special character. This can be controlled with the @urefbreakstyle command (this command has effect only in TeX):

@urefbreakstyle how

where the argument how is one of these words:

‘after’

(the default) Potentially break after the special characters.

‘before’

Potentially break before the special characters.

‘none’

Do not consider breaking at the special characters at all; any potential breaks must be manually inserted.

6.10.3 @url PDF Output Format

If the ultimate purpose of a PDF is only to be viewed online, perhaps similar to HTML in some inchoate way, you may not want the urls to be included in the visible text (just as urls are not visible to readers of web pages). Texinfo provides a PDF-specific option for this, which must be used inside @tex:

@tex
\global\urefurlonlylinktrue
@end tex

The result is that @url{http://www.gnu.org, GNU} has the visible output of just ‘GNU’, with a link target of http://www.gnu.org. Ordinarily, the visible output would include both the label and the url: ‘GNU (http://www.gnu.org)’.

This option only has effect when the PDF output is produced with the pdfTeX program, not with other ways of getting from Texinfo to PDF (e.g., TeX to DVI to PDF). Consequently, it is ok to specify this option unconditionally within @tex, as shown above. It is ignored when DVI is being produced.

6.10.4 PDF Colors

By default, urls and cross-reference links are printed in black in PDF output. Very occasionally, however, you may want to highlight such “live” links with a different color, as is commonly done on web pages. Texinfo provides a PDF-specific option for specifying these colors, which must be used inside @tex:

@tex
\global\def\linkcolor{1 0 0}  % red
\global\def\urlcolor{0 1 0}   % green
@end tex

\urlcolor changes the color of @url output (both the actual url and any textual label), while \linkcolor changes the color for cross-references to nodes, etc. They are independent.

The three given values must be numbers between 0 and 1, specifying the amount of red, green, and blue respectively.

These definitions only have an effect when the PDF output is produced with the pdfTeX program, not with other ways of getting from Texinfo to PDF (e.g., TeX to DVI to PDF). Consequently, it is ok to specify this option unconditionally within @tex, as shown above. It is ignored when DVI is being produced.

We do not recommend colorizing just for fun; unless you have a specific reason to use colors, best to skip it.

6.11 @cite{reference}

Use the @cite command for the name of a book that lacks a companion Info file. The command produces italics in the printed manual, and quotation marks in the Info file.

If a book is written in Texinfo, it is better to use a cross-reference command since a reader can easily follow such a reference in Info. See @xref.

7 Marking Text, Words and Phrases

In Texinfo, you can mark words and phrases in a variety of ways. The Texinfo formatters use this information to determine how to highlight the text. You can specify, for example, whether a word or phrase is a defining occurrence, a metasyntactic variable, or a symbol used in a program. Also, you can emphasize text, in several different ways.

7.1 Indicating Definitions, Commands, etc.

Texinfo has commands for indicating just what kind of object a piece of text refers to. For example, email addresses are marked by @email; that way, the result can be a live link to send email when the output format supports it. If the email address was simply marked as “print in a typewriter font”, that would not be possible.

7.1.1 Highlighting Commands are Useful

The commands serve a variety of purposes:

@code{sample-code}

Indicate text that is a literal example of a piece of a program. See @code.

@kbd{keyboard-characters}

Indicate keyboard input. See @kbd.

@key{key-name}

Indicate the conventional name for a key on a keyboard. See @key.

@samp{text}

Indicate text that is a literal example of a sequence of characters. See @samp.

@verb{text}

Write a verbatim sequence of characters. See @verb.

@var{metasyntactic-variable}

Indicate a metasyntactic variable. See @var.

@env{environment-variable}

Indicate an environment variable. See @env.

@file{file-name}

Indicate the name of a file. See @file.

@command{command-name}

Indicate the name of a command. See @command.

@option{option}

Indicate a command-line option. See @option.

@dfn{term}

Indicate the introductory or defining use of a term. See @dfn.

@cite{reference}

Indicate the name of a book. See @cite.

@abbr{abbreviation}

Indicate an abbreviation, such as ‘Comput.’.

@acronym{acronym}

Indicate an acronym. See @acronym.

@indicateurl{uniform-resource-locator}

Indicate an example (that is, nonfunctional) uniform resource locator. See @indicateurl. (Use @url (see @url) for live urls.)

@email{email-address[, displayed-text]}

Indicate an electronic mail address. See @email.

7.1.2 @code{sample-code}

Use the @code command to indicate text that is a piece of a program and which consists of entire syntactic tokens. Enclose the text in braces.

Thus, you should use @code for an expression in a program, for the name of a variable or function used in a program, or for a keyword in a programming language.

Use @code for command names in languages that resemble programming languages, such as Texinfo. For example, @code and @samp are produced by writing ‘@code{@@code}’ and ‘@code{@@samp}’ in the Texinfo source, respectively.

It is incorrect to alter the case of a word inside a @code command when it appears at the beginning of a sentence. Most computer languages are case sensitive. In C, for example, Printf is different from the identifier printf, and most likely is a misspelling of it. Even in languages which are not case sensitive, it is confusing to a human reader to see identifiers spelled in different ways. Pick one spelling and always use that. If you do not want to start a sentence with a command name written all in lowercase, you should rearrange the sentence.

In the Info output, @code results in single quotation marks around the text. In other formats, @code argument is typeset in a typewriter (monospace) font. For example,

The function returns @code{nil}.

produces this:

The function returns nil.

Here are some cases for which it is preferable not to use @code:

• For shell command names, such as ls (use @command).
• For environment variables, such as TEXINPUTS (use @env).
• For shell options, such as ‘-c’, when such options stand alone (use @option).
• An entire shell command often looks better if written using @samp rather than @code. In this case, the rule is to choose the more pleasing format.
• For a string of characters shorter than a syntactic token. For example, if you are writing about ‘goto-ch’, which is just a part of the name for the goto-char Emacs Lisp function, you should use @samp.
• In general, when writing about the characters used in a token; for example, do not use @code when you are explaining what letters or printable symbols can be used in the names of functions. (Use @samp.) Also, you should not use @code to mark text that is considered input to programs unless the input is written in a language that is like a programming language. For example, you should not use @code for the keystroke commands of GNU Emacs (use @kbd instead) although you may use @code for the names of the Emacs Lisp functions that the keystroke commands invoke.

By default, TeX will consider breaking lines at ‘-’ and ‘_’ characters within @code and related commands. This can be controlled with @allowcodebreaks (see @allowcodebreaks). The HTML output attempts to respect this for ‘-’, but ultimately it is up to the browser’s behavior. For Info, it seems better never to make such breaks.

For Info, the quotes are omitted in the output of the @code command and related commands (e.g., @kbd, @command), in typewriter-like contexts such as the @example environment (see @example) and @code itself, etc.

To control which quoting characters are implicitly inserted by Texinfo processors in the output of ‘@code’, etc., see the OPEN_QUOTE_SYMBOL and CLOSE_QUOTE_SYMBOL customization variables (see Other Customization Variables). This is separate from how actual quotation characters in the input document are handled (see Inserting Quote Characters).

7.1.3 @kbd{keyboard-characters}

Use the @kbd command for characters of input to be typed by users. For example, to refer to the characters M-a, write:

@kbd{M-a}

and to refer to the characters M-x shell, write:

@kbd{M-x shell}

By default, the @kbd command produces a different font (slanted typewriter instead of normal typewriter), so users can distinguish the characters that they are supposed to type from those that the computer outputs.

Since the usage of @kbd varies from manual to manual, you can control the font switching with the @kbdinputstyle command. This command has no effect on Info output. Write this command at the beginning of a line with a single word as an argument, one of the following: