static-sitemap-cli

static-sitemap-cli

Simple CLI to pre-generate XML sitemaps for static sites locally.

Built in 10 minutes. :stuck_out_tongue_winking_eye:

Install

npm i -g static-sitemap-cli

Usage

Syntax: static-sitemap-cli <BASEURL> [options]

At its simplest, just go to your dist folder and run:

static-sitemap-cli https://example.com > sitemap.xml

Or, because it's quite a mouthful:

sscli https://example.com > sitemap.xml

Where sscli is just an alias of static-sitemap-cli. CLI by default outputs to stdout - so that you can pipe it to do other cool stuff. CLI also allows you to pipe in BASEURL via stdin.

Arguments

Argument	Description
BASEURL	Base URL that is prefixed to all location entries. For example: https://example.com/

Options

Option	Long	Description
-h	--help	show CLI help
-V	--version	show CLI version
-r	--root	[default: current dir] root directory to start from
-m	--match	[default: **/*.html,!404.html] list of globs to match
-p	--priority	glob-priority pair (eg: foo/*.html=0.1)
-c	--changefreq	glob-changefreq pair (eg: foo/*.html=daily)
-n	--no-clean	disable clean URLs
-l	--slash	add trailing slash to all URLs
-t	--text	output as .TXT instead
-s	--save	save output to XML and TXT files directly
-o	--output-dir	specify the output dir; used together with --save; defaults to root working directory
-v	--verbose	be more verbose
N/A	--follow-noindex	removes pages with noindex meta tag from sitemap (up to 5x slower due to reading and parsing every HTML file)

Clean URLs

Whether or not to include the .html extension. By default, something like:

rootDir/foo/index.html becomes https://example.com/foo.

rootDir/foo/bar/foobar.html becomes https://example.com/foo/bar/foobar.

Pass -n option to disable this behavior.

Trailing Slashes

Controls whether or not URLs should include trailing slashes. For example:

rootDir/bar/index.html becomes https://example.com/bar/.

For obvious reasons, this cannot be used together with -n.

Ignore Some Files

The -m flag allows multiple entries to be input. By default it's set to the following globs: **/*.html and !404.html. You can change the glob pattern matches to suit your use-case, like:

sscli https://example.com -m '**/*.html' -m '!404.html' -m '!**/ignore/**' -m '!this/other/specific/file.html'

Glob-* Pairs

The -p and -c flags allow multiple entries and accept glob-* pairs as input. A glob-* pair is input as <glob-pattern>=<value>, where = is used as the separator. For example, a glob-frequency pair should be input as events/**/*.html=daily.

Latter entries will override the former. So for example in

sscli https://example.com -c '**/*=weekly' -c 'events/**=daily'

all URL entries will contain <changefreq>weekly</changefreq> while pages that match event/** will contain <changefreq>daily</changefreq>.

Output as Text

Sitemaps can be formatted as a simple text file as well, where each line contains exactly one URL. Pass the option -t to do so. In this case, --priority and --changefreq are redundant and ignored.

Examples

Create sitemap for dist folder

static-sitemap-cli https://example.com -r dist > dist/sitemap.xml

OR

sscli https://example.com -r dist > dist/sitemap.xml

Ignore a bunch of files

sscli https://example.com -m '**/*.html' '!404.html' '!**/ignore/**' '!this/other/specific/file.html' > sm.xml

Set priority of certain pages

By default, the optional <priority> label (protocol reference) is excluded, so every pages' default is 0.5. To change the relative priority of certain pages:

sscli https://example.com -p '**/{foo,bar}/**=0.1' '**/important/**=0.9' > sm.xml

Set changefreq of all pages to weekly, and some to daily

sscli https://example.com -c '**/*=weekly' -c 'events/**=daily' > sm.xml

Pipe in the base URL

echo https://example.com | sscli > sm.xml

Save XML and TXT files into a specified location directly

sscli https://example.com -r 'src' -s -o 'dist'

To-do

Add tests! :sweat_smile:

Tests

Run npm run test.

Implementation Notes

To slash or not to slash

First of all, search engines treat trailing slashes the same only for root URLs.

1. https://example.com
2. https://example.com/
3. https://example.com/about
4. https://example.com/about/

(1) and (2) are root URLs and are treated exactly the same; while (3) and (4) are different and are treated as 2 unique addresses. This can be verified through devtools - where you'll notice there aren't 301 redirects when (1) or (2) are entered into the URL address bar.

Internally, browsers append the slash when a root URL is entered, but hides the slash when displayed in the URL address bar - for vanity purposes.

To synchronise with browser behaviour, this commit adds the trailing slash for all root URLs, even if the --slash flag is unused.

Is this important? Not really - most of the time; but if you're using Google AMP, then yes, the trailing slash on all root URLs is important. Why? Because of how AMP Cache stores the root URL always with the trailing slash - so you can use your sitemap to perform cache-busting operations.

Changelog

v1.4.0 - 2020-07-09:

• Add noindex meta tag detection feature per #9. (Thanks @davwheat!)

v1.3.3 - 2020-07-07:

• Update dependencies.

v1.3.2 - 2020-02-22:

• Update the changelog.
• Update dependencies.

v1.3.1 - 2020-02-22:

• Fixes #7 typo in README re --changefreq alias. (Thanks @joshtaylor!)

v1.3.0 - 2020-01-10:

• --save now outputs BOTH sitemap.xml and sitemap.txt formats.
• Update dependencies.

v1.2.0 - 2019-09-26:

• Always add trailing slash to root urls. (ref: implementation notes)

v1.1.0 - 2019-08-18:

• BREAKING: Trailing slash alias -s renamed to -l. Sorry. :cry:
• Add feature save directly to file <rootDir>/sitemap.xml instead of stdout.

v1.0.1 - 2019-08-16:

• Bugfix - empty line at EOF in text mode.

v1.0.0 - 2019-08-15:

• BREAKING: --ignore is deprecated. Use --match instead.
• BREAKING: Glob-* pairs are no longer comma-seperated. Use = instead.
• BREAKING: Logic for multiple glob-* pairs changed. Later pairs override the earlier ones now.
• Major refactor of original codebase; discontinued usage of globby and sitemap in favour of fast-glob, micromatch, and js2xmlparser.
• Resulting code should be much easier to reason with and maintain now.
• Add feature to output as text (one URL per line).
• Add verbose mode to see some console feedback.
• And finally, add tests with ~95% coverage.

v0.2.0 - 2019-07-31:

• Allow BASEURL to be piped in also.
• Refactor some dependencies.

v0.1.1 - 2019-07-27:

• Bugfix: properly check rootDir before replacing.
• Add new alias sscli because the original is quite a mouthful.

v0.1.0 - 2019-07-26:

• Initial release.
• Built in 10 minutes. :stuck_out_tongue_winking_eye: