I want to see *all* the codes in code view

[sjchase]

I am hoping this is a simple setting that is just hiding from me somewhere. I have been editing my book and adding the full html character codes for things like ellipses, right and left quotes, em and en dashes, etc. They appear to display correctly in book view, but in code view they vanish into their "real" vs. code form. For example, & hellip; (ignore the space) shows up as ... in both code view and book view. I would like to be able to confirm that I have successfully substituted all the special characters, but code view is "helping" me just a little too much

Thanks in advance,
Sabrina

[GrannyGrump]

I hope someone will jump in here if I am giving bad advice ---

*In most cases* it will help to go to Edit >> Preferences >> Clean Source. Then under "Automatically Clean..." UNCHECK both Open and Save. Select "Pretty Print" rather than "HTML Tidy" so that if/when it does run, it uses the Pretty Print mode --- HTML Tidy can sometimes be destructive of your code and layout, and can actually lose data.

That said, I have found with version .074, it will automatically "Pretty Print" or "Tidy" your (x)html pages when you run certain procedures. One that happens to me all the time is when I "split at cursor" or "split at markers". All my preferred indentations and line spacing on the code page are changed to Sigil's preferred layout, and my careful usage of named/numbered entities (like your hellip) are changed to the actual characters. (I would prefer the entities for certain Search and Replace procedures.)

I think v .072 did not do this, and keep meaning to load it up and test it, but haven't checked that yet.

[Doitsu]

Originally Posted by sjchase

I have been editing my book and adding the full html character codes for things like ellipses, right and left quotes, em and en dashes, etc. They appear to display correctly in book view, but in code view they vanish into their "real" vs. code form.

AFAIK, this is a special "feature" of Sigil 0.7.4. Sigil will automatically change named entities to their Unicode equivalents, whenever text is changed in Book View mode, even if the Open and Save Clean Source options are disabled.
The only way to avoid this behavior is to change text only in Code View mode or to downgrade to an earlier Sigil version that doesn't exhibit this behavior.

If you only use a limited number of named entities you could also use the Saved Searches feature to define a group that contains Unicode characters and their named entity equivalents and then use this group to replace Unicode characters with named entities.

However, since named entities are not required in XHTML, this would only make sense, if you prefer named entities for proofreading purposes.

[GrannyGrump]

@Doitsu said

The only way to avoid this behavior is to change text only in Code View mode

As I noted, sometimes being in Code View is no protection. In version .074, Splitting files will always cause Pretty Print to kick in -- I wish there was a way to disable it completely. I'm using named entities for special font characters (fleurons, logos) in the "private use area", so in code view they convert to "missing character" glyphs, which is not at all helpful when searching or replacing or proofing. I'll bet there are other Sigil users who grit their teeth over this behavior.

[eschwartz]

The calibre editor has an option to control "Replace HTML entities as they are typed".

I don't think that will happen when splitting filesyes it does but I bet Kovid would be more responsive to a fix, since calibre is being developed.

[Toxaris]

Originally Posted by sjchase

I am hoping this is a simple setting that is just hiding from me somewhere. I have been editing my book and adding the full html character codes for things like ellipses, right and left quotes, em and en dashes, etc. They appear to display correctly in book view, but in code view they vanish into their "real" vs. code form. For example, & hellip; (ignore the space) shows up as ... in both code view and book view. I would like to be able to confirm that I have successfully substituted all the special characters, but code view is "helping" me just a little too much

As … is exactly the same as … for example, can you explain what the problem is? Why do you want to substitute the special characters to named entities? If it is about the conversion of ... to …, that can be done with some simple search and replace actions.

[DiapDealer]

Originally Posted by Toxaris

As … is exactly the same as … for example, can you explain what the problem is? Why do you want to substitute the special characters to named entities? If it is about the conversion of ... to …, that can be done with some simple search and replace actions.

When it comes to typographic quotes (single and double) and apostrophes I can see the desire to distinguish, with entities, that which can be ambiguous using characters (’ for closing single quote VS ’ for apostrophes). Think about formatting documents that will have both British and American editions (single quotes vs double-quotes) and how much easier the transition would be--or just about any situation where you need to algorithmically distinguish between a single closing quote and an apostrophe.

But other than that ... I agree, I find it just as easy to search for (and replace with) characters as I do entities. *shrugs*

Either way, users really should be able to choose not to have their entities be eaten.

That's why I stick with v0.7.3. Yes, it barks about missing Doctypes, but I find that noisy (and easily fixed) issue much less destructive and annoying than 0.7.4's penchant for eating markup.

[GrannyGrump]

Toxaris said:

Why do you want to substitute the special characters to named entities?

In my case, it is a mixture of what DiapDealer replied

When it comes to typographic quotes (single and double) and apostrophes I can see the desire to distinguish, with entities,

and the wish to be able to easily find and replace my *really special* characters, such as those fleurons I mentioned above from the "private usage area." When you are using entities like
& # xf021; and & # xf029; and & # 61477;
it helps to be able to see the actual codes instead of just a generic missing-glyph character, when I want to replace one or more of those with a different decorative character.

Not all "special character" entities convert automatically to the actual character if they depend on a special or embedded font that is not availabe to the code editor.

PS-- And I *still* wish we could search in book view, like the old versions. It was the fastest simplest way to find stray straight-quote marks that fell in the text outside of tags. Now you have to mess around with regex to hunt them down.

[KevinH]

Hi,

I see you were complaining about the exact same thing I was troubled by.

FYI: I have made a patch to Sigil to enable you to set a Preference for exactly which Entities you want to preserve in Code View.

Please see this post for the patch against Sigil trunk:

https://www.mobileread.com/forums/sho...5&postcount=46

And this thread for one other bug fix you may be interested in and some discussion.

https://www.mobileread.com/forums/sho...d.php?t=242391

So if you know how to build from source, please test this patch and let me know if it does what you need. If you can not, I can compile a Sigil.dmg for Mac with this patch and post it for you and probably do a Linux build as well, but I can not do a Windows build so you are out of luck there.

Hope this helps,

KevinH

[DiapDealer]

In case some Windows people want to experiment (and aren't following the Maverick Crashing thread). I've posted (Google Drive link) the Windows build of Sigil patched with KevinH's "Preserve Entities" feature (and with Qt5.3.1). The named entities you wish to be preserved (when editing in Book View) must be entered in the new Edit > Preferences > Preserve Entities section of Sigil's settings.

I'm sure there's probably some sort of daily download limit for files on Google Drive, but I'm thinking we probably won't have to worry about it (but I thought I'd mention it just in case).

Anyway ... make of it what you will.
https://drive.google.com/file/d/0B5C...it?usp=sharing

[sjchase]

Thank you very much DiapDealer!

[DiapDealer]

Originally Posted by sjchase

Thank you very much DiapDealer!

KevinH did all the heavy-lifting. I just compiled it for Windows. But you're welcome anyway.

[mvo]

Thank you very much KevinH and DiapDealer.