In order to use JoliTypo in other contexts, it could be useful to release a .phar version, which can be configured with a .jolitypo configuration file.

It could be used with all static site generators, after build phase to fix the generated HTML files.

It looks Box could help use to generate the .phar

Hi !

By using the french floating number notation (eg: 1,2), the NoSpaceBeforeComma fixer add an unwanted space after comma.

[...] « seule » 1,7 million de personnes [...]

is converted to :

[...] « seule » 1, 7 million de personnes [...]

This has been noticed using fr_FR locale and the following options :

'Ellipsis', 'Dimension', 'Numeric', 'Dash',  'SmartQuotes', 'FrenchNoBreakSpace', 'NoSpaceBeforeComma',  'CurlyQuote', 'Hyphen', 'Trademark'

But it seems not to be a conflict between fixers nor a locale trouble (tested both in my own project and using the demo page.)

When reformatting in set language, have the ability to harmonize spacing around punctuation:
word ! Word? Word! Word ?
To
Word! Word? Word! Word?
Or
Word ! Word ? Word ! Word ?

When converting the following example using JoliTypo, nested quotes (quotes inside quotes) are not parsed and the closing quote is mistaken for an apostrophe.

Input

"This 'magic' piece of code fixes dump quotes and apostrophes."

Output

“This 'magic’ piece of code fixes dump quotes and apostrophes.” 
------^

Expected

“This ‘magic’ piece of code fixes dump quotes and apostrophes.”
------^ 

Allow live testing / configuration of the lib for demonstration and utility purpose.

Example:
"The man was 5'6" and 120 lbs." returns “The man was 5'6” and 120 lbs."
Should return “The man was 5'6" and 120 lbs.”

We must have something more specific than protectedTag

(cf. reddit comment)

Hello, recently the commit 0729612 fixes a namespace problem in Twig extension PHP file. Since you didn't released it in a stable yet, each time we require the Symfony bundle using the flex recipe in Symfony 3.4 or 4, it just fails and breaks the Symfony app. When is the next release planned ?

See https://en.wikipedia.org/wiki/Diacritic

in src/JoliTypo/Fixer/Numeric.php, line 19

This name is a reserved word since PHP 7 and should not be used as class, interface or trait name.

use JoliTypo\StateBag;

/**
 * Add nbsp between numeric and units.
 */
class Numeric implements FixerInterface
{
    public function fix($content, StateBag $stateBag = null)
    {
        // Support a wide range of currencies
        $content = preg_replace('@([\dº])('.Fixer::ALL_SPACES.')+([º°%Ω฿₵¢₡$₫֏€ƒ₲₴₭£₤₺₦₨₱៛₹$₪৳₸₮₩¥\w]{1})@', '$1'.Fixer::NO_BREAK_SPACE.'$3', $content);

Posted from SensioLabsInsight

in src/JoliTypo/Fixer/FrenchNoBreakSpace.php#L27-L28

Maybe there is a reason
have a "+" instead a "*" ? because we can't force a nbsp before a ":" if there is no space

$content = preg_replace('@['.Fixer::ALL_SPACES.']+(:)@mu', Fixer::NO_BREAK_SPACE.'$1', $content);
$content = preg_replace('@['.Fixer::ALL_SPACES.']+([;!\?])@mu', Fixer::NO_BREAK_THIN_SPACE.'$1', $content);

$content = preg_replace('@['.Fixer::ALL_SPACES.']*(:)@mu', Fixer::NO_BREAK_SPACE.'$1', $content);
$content = preg_replace('@['.Fixer::ALL_SPACES.']*([;!\?])@mu', Fixer::NO_BREAK_THIN_SPACE.'$1', $content);

See

• https://github.com/actions/upload-release-asset#example-workflow---upload-a-release-asset
• https://github.com/Cecilapp/Cecil/blob/master/.github/workflows/release.yml#L62

We need to fix the Configuration.php

with something like this

doctrine/DoctrineCacheBundle#148

To use JoliTypo on big chunks of text we could use this protip: https://twitter.com/fabpot/status/1099930119500693505?s=19

Documentation to update.

• There is some new stuffs to checks: https://github.com/heiglandreas/Org_Heigl_Hyphenator/commits/master
• Waiting for a release.

ATM, the phar is super heavy! (34.7Mb)

• Make it much lighter (drop everything not needed)
• use castor to build it (update the CI too)

Hi,

the quotes for the de-CH locale are not set correctly. The quotes for Germany and Austria are applied wich is not correct for Switzerland. In Switzerland they use French-style angle quotation marks, see: https://en.wikipedia.org/wiki/Quotation_mark#German

In Switzerland, however, the French-style angle quotation mark sets are also used for German printed text: «A ‹B›?»
Andreas fragte mich: «Hast du den Artikel ‹EU-Erweiterung› gelesen?»
Andreas asked me: ‘Have you read the “EU Expansion” article?’

When running the composer test, I get following deprecating message:

Remaining indirect deprecation notices (1)

  1x: Since symfony/framework-bundle 5.1: Not setting the "framework.router.utf8" configuration option is deprecated, it will default to "true" in version 6.0.
    1x in FunctionalTest::testRenderTwigViaFilter from JoliTypo\Tests\Bridge

Legacy deprecation notices (1)

I want to add a toggle theme option to the website to let the user toggle between light and dark mode.

Hi !

I'm using this library to save user provided content to a database (utf8mb4 encoded field). The content is added by the user to a textarea that is then converted from markdown text to html using the markdown-it library.
I've been struggling to find out why the images included in my html didn't show up in my browser when their urls seemed to be right. Here's what I found out :

• using JoliTypo, the content in the database, converted to ISO 8859-1 looks like this :

<p><a href="http://pubpeer.dev/stor­age/image-1492678775687.jpg" target="_self"><img src="http://pubpeer.dev/stor­age/image-1492678775687.jpg" alt="file"></a></p>

• when I don't use JoliTypo :

<p><a href="http://pubpeer.dev/stor­age/image-1492678775687.jpg" target="_self"><img src="http://pubpeer.dev/stor­age/image-1492678775687.jpg" alt="file"></a></p>

So it seems that JoliTypo is adding UTF... characters in my content. Am I doing something wrong or is there a bug somewhere ?

Thanks,

Xavier

à and &#8239 are displayed as a messed up char and /
libxml fail? a regexp capturing too much and breaking the à?
Do not know yet.

https://travis-ci.org/jolicode/JoliTypo/jobs/13911015

Composer with version 1.0.5 of joliTypo gives the following notice, seems concerning mainly Tests Classes.

Deprecation Notice: Class JoliTypo\Fixer\FrenchQuotes located in ./vendor/jolicode/jolitypo/src/JoliTypo/Fixer/FrenchQuotes.php does not comply with psr-4 autoloading standard. It will not autoload anymore in Composer v2.0. in phar:///var/www/PROJECT/composer.phar/src/Composer/Autoload/ClassMapGenerator.php:185

• separate the fix method into fixHTML and fixString, keep BC;
• update documentation

I get a strange result when I have the character œ in a string the string seems to be returned in a different encoding.
Currently I reproduce the issue on a server but not in local. I think the issue come from the php configuration but I can't find what's wrong yet.

I reproduce the issue on the demo too https://jolitypo.jolicode.com/, using the french locale, applying all fixes and pasting the string <p>des œuvres d'art.</p>

Hi, I've been trying to use JoliTypo for personnal use on http://borisschapira.com/ but it provokes encoding errors for accented characters. Here is an example of what I give to JoliTypo fixer (with encoding determined via mb_detect_encoding) and what JoliTypo responds :

Mentions Légales (UTF-8)
Mentions Légales (ASCII) 

Here is my (pretty simple) code :

``` php`
function typofr($text)
{
static $fixer;
if (!isset($fixer)) {
$fixer = new Fixer(array(
'Trademark'));
$fixer->setLocale('fr_FR');
}
$fixed = $fixer->fix($text);
return $text."<script>console && console.log('-------');console && console.log('".$text." (".mb_detect_encoding($text).")'); console && console.log('".$fixed." (".mb_detect_encoding($fixed).")')</script>";
}

And you can temporarily see the result here, in the console : http://borisschapira.com/

I sometimes use JoliTypo to fix a text which has already been fixed before. If my text contains Personnalisation, after multiple saves, as many ­ will be added, which ends up giving me this kind of content :

Person­­­­­­­­na­­­­­­­­li­­­­­­­­sa­­­­­­­­tion

A lazy solution would be to replace (with regexp) ­­ with ­.

Todo :

• Check if this is happening with other fixers
• Check if this is du to JoliTypo or org_heigl/hyphenator

It could be useful to make available a standalone binary of JoliTypo, to make it usable outside of the PHP ecosystem (for all JS/Go static site generators for example).

May be Box may help us to craft the PHAR version.

If, for some reason, I want to get the ugly keyboard typed text back, then I would like to use something like

$fixer->unfix($fixedText);

Using this method, I could even "unfix" some parts of the fixed text and keep the rest. (Like removing the hyphenation while keeping the Smart Quotes)

Bravo pour ce projet ! Ce serait bien d'ajouter un insécable automatiquement entre un chiffre et son unité, en suivant l'exemple de php-typography.

The FrenchQuote, EnglishQuote and GermanQuotes fixers could be refactored to be configurable,
that would also allow more languages to be supported.

Other fixers can also be implemented, following what php-typography do.

• diacritics
• marks

See https://github.com/Melindrea/wp-typography/blob/1bc281cd82673fd293689d1a6871afd2718c86d1/php-typography/php-typography.php#L826

I have an issue with some fixers , \n are removed .
Example with this text :

## Verrückt und Harakiri 2
Am Tag nach dem

the fixer fiw the text like this

## Verrückt und Harakiri 2 Am Tag

Here the preg_match output of the regex of the fixer :

I found the problem is that \s match \n .
So to fix this we should replace \s by his equivalent minus \n .

I foud this in the python doculentation of re lib :

When the UNICODE flag is not specified, it matches any whitespace character, this is equivalent to the set [ \t\n\r\f\v]. The LOCALE flag has no extra effect on matching of the space. If UNICODE is set, this will match the characters [ \t\n\r\f\v] plus whatever is classified as space in the Unicode character properties database.

So replace \s by \t\r\f\v can be a solution

• Add php-cs-fixer
• Add a .php_cs rules file
• Maybe rename some badly named variables

While curly quotes are common in German, especially in script use, angular quote marks (reversed french guillemets or »Möwchen«) are more common in printed german text, books and newspapers. Please note that they are used in the reversed direction compared to French or Swiss German:

This is an »example«.
And this is an »example with another ›single quote‹ inside«.

Please provide an option for these quote marks.

Looking at this commit: 9cece66

We should add the corresponding conflict section in composer to avoid time wasting.

Hi,

I'm applying JoliTypo on Wordpress contents but sometimes, these contents contains shortcodes like [[ poney ]]. Is there a way to tell Jolitypo not to replace content inside these shortcodes, like a Regex Exclusion Pattern for example ?

Hi,

i would be great to be able to control the options of the Org\Heigl\Hyphenator\Hyphenator instance, like wordMin, LeftMin and so on. I created my own fixer to do this but I think it would be nice to have an interface in JoliTypo to control this.

Replaces the usual dot . with an interpunct · in gender-neutral words.

musicien.ne.s would become musicien·ne·s

See this Wikipedia article

Regular expressions in FrenchNoBreakSpace for : ; ! and ? characters are based on the fact that there should be already at least one space before the character. This fixer should also work if the space has been forgotten by the writer.

in src/JoliTypo/Fixer/BaseOpenClosePair.php, line 5
> The class JoliTypo\Fixer is declared but never used. You should remove the use statement.

<?php

namespace JoliTypo\Fixer;

use JoliTypo\Fixer;
use JoliTypo\StateBag;
use JoliTypo\StateNode;

abstract class BaseOpenClosePair
{

Posted from SensioLabsInsight

To allow JoliTypo to fix complete HTML documents, with doctype & body tag, the exportDOMDocument must be rewritten.

The loadDOMDocument hack needs to be considered.

PHP Warning:  DOMDocument::loadHTML(): Tag section invalid in Entity

DOMDocument doesn't know any HTML5 tag, even if the Doctype is provided.
The end result is still OK (the unknown tags are not removed) but the warning is annoying.

the plugin can correct the straight quotes to add the correct french quotes and nbsp, but if a french quote is added directly, with or without space, the quote is not corrected.

example : http://fortuitement.com/test-typo/

nbsp are missing in the first « test » and «test»