Detailed TOC usage

The scenarios below show different features of toc

Features

Read the table of contents

Let's say you want to structure your javascript file "example.js". Single-line comments in this language start with //. You open your file and add these comments where you need them:

#!/usr/bin/env node

// ################################################################ Main section

let Section1 = "Write //, 64 hash characters and the name of section"

// ################################ Nested section

let Section2 = "Write //, 32 hash characters and the name of section"

// ################ Nested section

let Section3 = "Write //, 16 hash characters and the name of section"

// ######## Nested section

let Section4 = "Write //, 8 hash characters and the name of section"

// #### Nested section

let Section5 = "Write //, 4 hash characters and the name of section"

If you run toc example.js, the program will output the following (stdout):

// ┌───────────────────────────────────────────────────────────────┐
// │ Contents of example.js                                        │
// ├───────────────────────────────────────────────────────────────┘
// │
// ├──┐Main section
// │  └──┐Nested section
// │     └──┐Nested section
// │        └──┐Nested section
// │           └── Nested section
// │
// └───────────────────────────────────────────────────────────────

Embed the table of contents in the original file

If you want to output the toc we just saw to the file, you should run toc -f example.js and get (stderr):

Adding toc to file example.js

By opening "example.js", the file content will be:

Click to view the original `example.js` with toc

#!/usr/bin/env node

// ┌───────────────────────────────────────────────────────────────┐
// │ Contents of example.js                                        │
// ├───────────────────────────────────────────────────────────────┘
// │
// ├──┐Main section
// │  └──┐Nested section
// │     └──┐Nested section
// │        └──┐Nested section
// │           └── Nested section
// │
// └───────────────────────────────────────────────────────────────

// ################################################################ Main section

let Section1 = "Write //, 64 hash characters and the name of section"

// ################################ Nested section

let Section2 = "Write //, 32 hash characters and the name of section"

// ################ Nested section

let Section3 = "Write //, 16 hash characters and the name of section"

// ######## Nested section

let Section4 = "Write //, 8 hash characters and the name of section"

// #### Nested section

let Section5 = "Write //, 4 hash characters and the name of section"

Notice how toc recognized the shebang for node and added the table of contents after it. If you run again toc -f example.js, it recognizes that there is no need to update the toc, as no changes have been made.

However, if you add new sections to the file and run again toc -f example.js, it will update the file accordingly:

Updating toc in example.js

Click to view the modified `example.js` with toc

#!/usr/bin/env node

// ┌───────────────────────────────────────────────────────────────┐
// │ Contents of example.js                                        │
// ├───────────────────────────────────────────────────────────────┘
// │
// ├──┐Main section
// │  ├──┐New section
// │  │  └── Also New section
// │  ├──┐New section
// │  │  └── Also New section
// │  └──┐Nested section
// │     └──┐Nested section
// │        └──┐Nested section
// │           └── Nested section
// │
// └───────────────────────────────────────────────────────────────

// ################################################################ Main section

let Section_1 = "Write //, 64 hash characters and the name of section"

// ################################ New section

let Section_1_1 = "Write //, 32 hash characters and the name of section"

// ################ Also New section

let Section_1_1_1 = "Write //, 32 hash characters and the name of section"

// ################################ New section

let Section_1_1 = "Write //, 32 hash characters and the name of section"

// ################ Also New section

let Section_1_1_1 = "Write //, 32 hash characters and the name of section"

// ################################ Nested section

let Section_1_2 = "Write //, 32 hash characters and the name of section"

// ################ Nested section

let Section_1_2_3 = "Write //, 16 hash characters and the name of section"

// ######## Nested section

let Section_1_2_4 = "Write //, 8 hash characters and the name of section"

// #### Nested section

let Section_1_2_5 = "Write //, 4 hash characters and the name of section"

Process multiple files

You can alternatively write the files you want to keep updated in a simple glob list:

# ignore-this-file.txt
README.md
USAGE.md
CHANGELOG.md
toc/cli.py
toc/toc.py
# glob expansion
tests/test*.py

To keep these files up-to-date, you just need to run toc -l -f files.txt:

Skipping unchanged toc in "README.md"
Updating toc in "USAGE.md"
Adding toc to "CHANGELOG.md"
Updating toc in "toc/cli.py"
Updating toc in "toc/toc.py"
Skipping unchanged toc in "tests/test_cli.py"
SSkipping unchanged toc in  "tests/test_toc.py"

Note that more than one list can be passed in a single command, lines starting with "#" are ignored, and there is support for glob expansion.

If you feel brave enough, you can run toc * over your entire code base, as its AI¹ will:

skip directories
skip non-text files
skip non-existing files
skip non-readable files
skip non-writable files
skip files that don't have suitable "section" comments
skip files whose toc is already up-to-date
only update files whose toc can be added or updated safely²
preserve shebangs, markdown frontmatters and other declarations of edited files

Show line numbers

For very long files, it may come in handy to run toc -n example.js to see the line number of each section, similar to the page numbers in the table of contents of a book:

// ┌───────────────────────────────────────────────────────────────┐
// │ Contents of example.js                                        │
// ├───────────────────────────────────────────────────────────────┘
// │
// ├──┐Main section 19
// │  ├──┐New section 23
// │  │  └── Also New section 27
// │  ├──┐New section 31
// │  │  └── Also New section 35
// │  └──┐Nested section 39
// │     └──┐Nested section 43
// │        └──┐Nested section 47
// │           └── Nested section 51
// │
// └───────────────────────────────────────────────────────────────

Limit depth

If you want to only see headings up to a certain level, you can limit them with toc -d 2 example.js:

// ┌───────────────────────────────────────────────────────────────┐
// │ Contents of example.js                                        │
// ├───────────────────────────────────────────────────────────────┘
// │
// ├──┐Main section
// │  ├── New section
// │  ├── New section
// │  └── Nested section
// │
// └───────────────────────────────────────────────────────────────

Set a custom comment character

But how could toc recognize that // is the proper comment character for that file? Well, thanks to AI³ toc supports most programming and markup languages, from COBOL to Carbon. In case it doesn't work as expected, you can force the behavior by running toc -c "//" example.xyz.

Set a custom file extension

You can also force toc to consider an arbitrary file extension by running toc -e "html" new-file-format.html6.

Read from stdin

toc can read stdin by using - as an argument. You may want to set an extension as well: curl -s https://github.com/AlphaJack/toc | toc -e html -

Result:

<!--
// ┌───────────────────────────────────────────────────────────────┐
// │ Contents of stdin.html                                        │
// ├───────────────────────────────────────────────────────────────┘
// │
// ├──┐Search code, repositories, users, issues, pull requests...
// │  └──┐Use saved searches to filter your results more quickly
// │     └── License
// ├──┐AlphaJack/toc
// │  ├──┐About
// │  │  ├── Topics
// │  │  ├── Resources
// │  │  ├── License
// │  │  ├── Stars
// │  │  ├── Watchers
// │  │  └── Forks
// │  ├── Languages
// │  └──┐Footer
// │     └── Footer navigation
// │
// └───────────────────────────────────────────────────────────────
-->

Redirect output to another file

For testing purposes, you can run toc -o output.py input.py to choose a different output file.

Note that this option does nothing if the toc in "input.py" is already up-to-date.

The -o flag is incompatible with the -l one.

Other commands

You can run toc -h for usage info and toc -v to read the current version

Exceptional file types

Native support

For these file types, you don't need to write comments, as toc can leverage the language syntax to build the table of contents

AsciiDoc

For AsciiDoc files, just organize your sections with one or more =:

= Document Title (Level 0)

== Level 1 Section Title

= Level 0 Section Title (Part)

== Level 1 Section Title

=== Level 2 Section Title

==== Level 3 Section Title

===== Level 4 Section Title

====== Level 5 Section Title

Beancount

For Beancount files, just organize your sections with one or more *:

* Options

; comment

* Transactions
** FY2020

2020-04-20 * "Food"
 Assets:Bank                                          -20.00 EUR
 Expenses:Groceries

HTML

For HTML files, just use regular headings tags such as <h1>, <h2> and so on:

<!doctype html>
<html>
<h1>Title</h1>
<h2>Subtitle</h2>
</html>

Man pages

For man pages such as groff and mandoc, just use regular headings indicators .\TH, .\SH and .\Ss:

.TH MAN 1 "2023-09-23" "2.12.0" "Manual pager utils"

.SH NAME
man \- an interface to the system reference manuals

.SH SYNOPSIS
.\" The general command line

Markdown

For Markdown files, just organize your sections with one or more #:

# Title
   
## Section

Perl

For Perl files, use the default Pod heading style:

=pod
=encoding utf8

=head1 Title

=head2 Section

=for comment
Text

=cut

reStructuredText

For ReStructured Text files, just use the conventional heading format:

*******
Chapter
*******

=======
Section
=======

----------
Subsection
----------

Wrap around comments needed

These languages do not support single-line comments, and thus every comment should be wrapped by a multi-line comment separator

CSS

For CSS files, you have to wrap your // comments between /* and */:

/*
// ################################################################ Landscape touchscreen
*/

@media (orientation: landscape) and (hover: none) and (pointer: coarse) {

/*
// ################################ Element selectors
*/

body {background-color: blue;}

} /* end Landscape touchscreen */

OCaml

For OCaml files, you have to wrap your * comments between (* and *):

(*
* ################################################################ Unique section
*)

let () = print_endline "Hello, World!"

Integration with third-party editors

Vim and Emacs

If you place your Vim Modeline / Emacs mode as the first line, the toc will be appended after:

// vim: noai:ts=4:sw=4
#include <stdlib.h>

void main() { exit(0);}

RStudio

If you are using RStudio, you may want to end your comments with at least 4 -, = or #. This marks the comment as a foldable section:

# ################################################################ Foldable section 1 ####

print("Collapse me!")

# ################################ Foldable section 2 ####

print("Collapse me!")

Footnotes

Not even, it's just a bunch of if-else and try-excepts statements that may prevent catastrophic damage ↩
The outdated toc to be replaced is defined as the first match of a non-greedy regex ↩
No, not really, it's just a match-case statement using the file extension, defaulting to "#" ↩

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

USAGE.md

USAGE.md

Detailed TOC usage

Features

Read the table of contents

Embed the table of contents in the original file

Process multiple files

Show line numbers

Limit depth

Set a custom comment character

Set a custom file extension

Read from stdin

Redirect output to another file

Other commands

Exceptional file types

Native support

AsciiDoc

Beancount

HTML

Man pages

Markdown

Perl

reStructuredText

Wrap around comments needed

CSS

OCaml

Integration with third-party editors

Vim and Emacs

RStudio

Files

USAGE.md

Latest commit

History

USAGE.md

File metadata and controls

Detailed TOC usage

Features

Read the table of contents

Embed the table of contents in the original file

Process multiple files

Show line numbers

Limit depth

Set a custom comment character

Set a custom file extension

Read from stdin

Redirect output to another file

Other commands

Exceptional file types

Native support

AsciiDoc

Beancount

HTML

Man pages

Markdown

Perl

reStructuredText

Wrap around comments needed

CSS

OCaml

Integration with third-party editors

Vim and Emacs

RStudio

Footnotes