French text in Code View

[AUDRY Roger]

As you know, French comprises a lot of accented characters.When a html file is viewed in Code View, words appear either normal or with accented letters replaced by an ASCII code : for instance "président" is viewed as
"président".
The consequence of this is that the spellchecker will red underline the word
although there is no spelling mistake; and there are so many occurence of this
that the spellcheck becomes useless.
What disturb me is that it seems that there is no rule to determine such
different behaviours; and it happens that the same file appears normal in one session and different in the next.
Can somebody can explain this to me, and may be, give a me way of being
sure that the "normal" behavious will be selected.

Thanks for your help

[roger64]

Bonjour

This is what I see in code view with Sigil 0.6. So, I presume you are doing something wrong?

See attachment below.

[user_none]

Why not just write président instead of président? A valid epub must be utf-8 or utf-16 encoded which encompasses all characters in the French alphabet. As long as the reader has a font supporting French characters there should be no issue with using the characters themselves instad of the entities. That said if they don't have a font with French characters installed the entities shouldn't display properly either.

[HarryT]

The most common reasons for using entities rather than Unicode characters are:

• If your keyboard does not support the character you need to type, eg many keyboards do not have em-dash or the copyright symbol.
• If your editor does not support unicode (not an issue for Sigil, of course!).
• When you want to make it explicit in the source what is happening, eg the   code is clearer than the corresponding whitespace character.
• When you need to escape HTML special characters like >, &, or ".

[user_none]

Originally Posted by HarryT

The most common reasons for using entities rather than Unicode characters are: ...

All very good and valid reasons but they don't really apply in this situation... I would hope the character é would be relativity easy for a French Speaker (presumably using a French keyboard) to type.

[HarryT]

Originally Posted by user_none

All very good and valid reasons but they don't really apply in this situation... I would hope the character é would be relativity easy for a French Speaker (presumably using a French keyboard) to type.

I agree, but it can be quicker to type the entity than to copy and paste the accented letter from the Windows "character map" accessory if one is using an English-language keyboard and you need to use an accented letter. Not that that would apply to the OP, of course, but as a general principle.

[meme]

One approach is to just convert all your entities to characters, do the spell check, and then convert them back to entities if needed. Of course, you don't have to do this by hand with 0.6.0 - you can create a Saved Searches group to do it for you. The examples included have a group with several of the blank space to show how it can be done.

It takes a bit to enter them, but do it once and its done - and if you export the group to a .ini file you can share it for others to use. A list like this could be scripted - maybe something I'll look into - the issue is deciding what lists to use (seems like 'everything' is needed, grouped appropriately).

There's also a request to add all special characters to the special characters dialog so an external character map wouldn't be necessary, but that may or may not get done.

[theducks]

replacing entities should be fairly simple using the new (Sigil 6) 'Insert Special Characters' in conjunction with Search and replace.

If this is frequent, add those to the ones already in the Saved search list