Finally a tool for typography nerds.
JoliTypo is a tool fixing Microtypography glitches inside your HTML content.
use JoliTypo\Fixer;
$fixer = new Fixer(array('Ellipsis', 'Dash', 'SmartQuotes', 'CurlyQuote', 'Hyphen'));
$fixedContent = $fixer->fix('<p>"Tell me Mr. Anderson... what good is a phone call... if you\'re unable to speak?" -- Agent Smith, <em>Matrix</em>.</p>');
<p>“Tell me Mr. Ander­son… what good is a phone call… if you’re unable to speak?”—Agent Smith, <em>Matrix</em>.</p>
“Tell me Mr. Anderson… what good is a phone call… if you’re unable to speak?”—Agent Smith, Matrix.
It's designed to be:
- language agnostic (you can fix
fr_FR
,fr_CA
,en_US
... You tell JoliTypo what to fix); - easy to integrate into modern PHP projects (composer and autoload);
- robust (make use of
\DOMDocument
instead of parsing HTML with dummy regexp); - smart enough to avoid Javascript, Code, CSS processing... (configurable protected tags list);
- fully tested;
- fully open and usable in any project (MIT License).
You can try it with the online demo!
Just tell the Fixer class which Fixer you want to run on your content and then, call fix()
:
use JoliTypo\Fixer;
$fixer = new Fixer(array("SmartQuotes", "FrenchNoBreakSpace"));
$fixer->setLocale('fr_FR');
$fixedContent = $fixer->fix('<p>Je suis "très content" de t\'avoir invité sur <a href="https://jolicode.com/">Jolicode.com</a> !</p>');
For your ease of use, you can find ready to use list of Fixer for your language here. Micro-typography is nothing like a standard or a law, what really matters is consistency, so feel free to use your own lists.
Please be advised that JoliTypo works best on HTML content; it will also work on plain text, but will be less smart about
smart quotes. When fixing a complete HTML document, potential <head>
, <html>
and <body>
tags may be removed.
To fix non HTML content, use the fixString()
method:
use JoliTypo\Fixer;
$fixer = new Fixer(array("Trademark", "SmartQuotes"));
$fixedContent = $fixer->fixString('Here is a "protip(c)"!'); // Here is a “protip©”!
You can run a standalone version of JoliTypo by downloading the PHAR version
Run jolitypo --help
to know how to configure the Fixer.
Requirements are handled by Composer (libxml and mbstring are required).
composer require jolicode/jolitypo
Usage outside composer is also possible, just add the src/
directory to any PSR-0 compatible autoloader.
- (Built-in) Symfony Bundle
- (Built-in) Twig extension
- (Built-in) CLI
- Wordpress plugin
- Drupal module
- Joomla plugin
- MODX Extra
Replaces the simple dash -
by a ndash –
between numbers (dates ranges...) and the double --
by a mdash —
.
Replaces the letter x between numbers (12 x 123
) by a times entity (×
, the real mathematical symbol).
Replaces the three dots ...
by an ellipsis …
.
Converts dumb quotes " "
to all kinds of smart style quotation marks (“ ”
, « »
, „ “
...). Handles a good variety of locales,
like English, Arabic, French, Italian, Spanish, Irish, German...
See the code for more details, and do not forget to specify a locale on the Fixer instance.
This Fixer replaces legacy EnglishQuotes
, FrenchQuotes
and GermanQuotes
.
Replaces some classic spaces by non-breaking spaces following the French typographic code.
No break space are placed before :
, thin no break space before ;
, !
and ?
.
Removes space before ,
and makes sure there is only one space after.
Makes use of org_heigl/hyphenator
, a tool enabling word-hyphenation in PHP.
This Hyphenator uses the pattern-files from OpenOffice which are based on the pattern-files created for TeX.
There are only some locales available for this fixer: af_ZA, ca, da_DK, de_AT, de_CH, de_DE, en_GB, en_UK, et_EE, fr, hr_HR, hu_HU, it_IT, lt_LT, nb_NO, nn_NO, nl_NL, pl_PL, pt_BR, ro_RO, ru_RU, sk_SK, sl_SI, sr, zu_ZA.
You can read more about this fixer on the official github repository.
This Fixer requires a Locale to be set on the Fixer with $fixer->setLocale('fr_FR');
. Default to en_GB
.
Proper hyphenation is mandatory in justified text and you should avoid word breaking in titles with this line of CSS: hyphens:none;
.
⚠ Be aware that the current screen readers are unable to spell correctly the words containing ­
tags. The Hyphen filter should therefore be used with caution or you might reduce your website's accessibility.
Replaces straight quotes '
with curly ones ’
.
There is one exception to consider: foot and inch marks (minutes and second marks). Purists use prime ′
, this fixer uses straight quotes for compatibility.
Read more about Curly quotes.
Handles trademark symbol ™
, a registered trademark symbol ®
, and a copyright symbol ©
. This fixer replaces
commonly used approximations: (r)
, (c)
and (TM)
. A non-breaking space is put between numbers and copyright symbols too.
Adds a non-breaking space between a numeral and its unit. Like this: 12_h
, 42_฿
or 88_%
. It was named Numeric
before release 1.0.2, but BC is kept for now.
It is really easy to make your own Fixers, feel free to extend the provided ones if they do not fit your typographic rules.
$fixer = new Fixer(array('Ellipsis', 'Dimension', 'Unit', 'Dash', 'SmartQuotes', 'NoSpaceBeforeComma', 'CurlyQuote', 'Hyphen', 'Trademark'));
$fixer->setLocale('en_GB');
Those rules apply for most of the recommendations of "Abrégé du code typographique à l'usage de la presse", ISBN: 9782351130667.
$fixer = new Fixer(array('Ellipsis', 'Dimension', 'Unit', 'Dash', 'SmartQuotes', 'FrenchNoBreakSpace', 'NoSpaceBeforeComma', 'CurlyQuote', 'Hyphen', 'Trademark'));
$fixer->setLocale('fr_FR');
Mostly the same as fr_FR, but the space before punctuation points is not mandatory.
$fixer = new Fixer(array('Ellipsis', 'Dimension', 'Unit', 'Dash', 'SmartQuotes', 'NoSpaceBeforeComma', 'CurlyQuote', 'Hyphen', 'Trademark'));
$fixer->setLocale('fr_CA');
Mostly the same as en_GB, according to Typefacts and Wikipedia.
$fixer = new Fixer(array('Ellipsis', 'Dimension', 'Unit', 'Dash', 'SmartQuotes', 'NoSpaceBeforeComma', 'CurlyQuote', 'Hyphen', 'Trademark'));
$fixer->setLocale('de_DE');
More to come (contributions welcome!).
$fixer = new Fixer(array('Ellipsis', 'Dimension', 'Dash', 'SmartQuotes', 'CurlyQuote', 'Hyphen'));
$fixedContent = $fixer->fix("<p>Some user contributed HTML which does not use proper glyphs.</p>");
$fixer->setRules(array('CurlyQuote'));
$fixedContent = $fixer->fix("<p>I'm only replacing single quotes.</p>");
$fixer->setRules(array('Hyphen'));
$fixer->setLocale('en_GB'); // I tell which locale to use for Hyphenation and SmartQuotes
$fixedContent = $fixer->fix("<p>Very long words like Antidisestablishmentarianism.</p>");
If you want to add your own Fixer to the list, you have to implement JoliTypo\FixerInterface
.
Then just give JoliTypo their fully qualified name, or even instance:
// by FQN
$fixer = new Fixer(array('Ellipsis', 'Acme\\YourOwn\\TypoFixer'));
$fixedContent = $fixer->fix("<p>Content fixed by the 2 fixers.</p>");
// or instances, or both
$fixer = new Fixer(array('Ellipsis', 'Acme\\YourOwn\\TypoFixer', new Acme\\YourOwn\\PonyFixer("Some parameter")));
$fixedContent = $fixer->fix("<p>Content fixed by the 3 fixers.</p>");
Protected tags is a list of HTML tag names that the DOM parser must avoid. Nothing in those tags will be fixed.
$fixer = new Fixer(array('Ellipsis'));
$fixer->setProtectedTags(array('pre', 'a'));
$fixedContent = $fixer->fix("<p>Fixed...</p> <pre>Not fixed...</pre> <p>Fixed... <a>Not Fixed...</a>.</p>");
- Write tests;
- A Fixer is run on a piece of text, no HTML to deal with;
- Implement
JoliTypo\FixerInterface
; - Send your Pull request.
- You MUST write code in english;
- you MUST follow PSR2 and Symfony coding standard (run
composer cs
on your branch); - you MUST run the tests (run
composer test
); - you MUST comply to the MIT license;
- you SHOULD write documentation.
If you add a new Fixer, please provide sources and references about the typographic rule you want to fix.
- Windows XP : Thin No-Break Space can't be used, all other spaces are ignored, but they do not look bad (normal space).
- Mac OS Snow Leopard : no no-break space, half no-break space, ems and en-dash but doesn't look bad (normal space).
BUT if you use a font (@font-face
maybe) that contains all those glyphs, there will be no issues.
There is a known issue preventing JoliTypo to work correctly with APC versions older than 3.1.11.
We need to be able to use this tool everywhere, you can help by providing:
- Wordpress plugin (to replace or complete
wptexturize
) - Dotclear plugin ...
Also, there is a Todo list 😙
This piece of code is under MIT License. See the LICENSE file.
There is already quite a bunch of tools like this one (including good ones). Sadly, some are only for one language, some are running regexp on the whole HTML code (which is bad), some are not tested, some are bundled inside a CMS or a Library, some are not using proper auto-loading, some do not have an open bug tracker... Have a look by yourself:
- https://michelf.ca/projets/php-smartypants/
- https://michelf.ca/projets/php-smartypants/typographer/
- https://www.bioinformatics.org/phplabware/internal_utilities/htmLawed/
- https://github.com/Cerdic/textwheel/blob/master/typographie/fr.php
- https://github.com/spip/SPIP/blob/master/ecrire/typographie/fr.php
- https://github.com/dg/texy/blob/master/Texy/modules/TexyTypographyModule.php
- https://github.com/scoates/lexentity
- https://github.com/nofont/Typesetter.js
- https://github.com/judbd/php-typography (fork of php-typography, you can test it here: https://www.roxane-company.com/typonerd/)
- https://mdash.ru/
- https://blot.im/typeset/ (Server side Javascript pre-processor)
Thanks to theses online resources for helping a developer understand typography:
- [FR] https://typographisme.net/post/Les-espaces-typographiques-et-le-web
- https://daringfireball.net/projects/smartypants/
- [FR] https://www.uzine.net/article1802.html
- [FR] https://dascritch.net/post/2011/05/09/Les-espacements-unicodes
- https://www.punctuationmatters.com/ is a must-read
- https://practicaltypography.com/
- [FR] "Abrégé du code typographique à l'usage de la presse", ISBN: 9782351130667
- https://en.wikipedia.org/wiki/Non-English_usage_of_quotation_marks