Reestablishing smallcaps

[roger64]

Hi

I wish to change hundreds of names for smallcap use They are all within their own span.
I would like to go from

<span class="smcpIncise">SMITH</span>

to

<span class="smcpIncise">Smith</span>

I can change them one by one, using a right click with Sigil.
I would like to use a batch command to do it...

[Doitsu]

Originally Posted by roger64

I would like to go from

<span class="smcpIncise">SMITH</span>

to

<span class="smcpIncise">Smith</span>

That's relatively easy with Sigil.

In Regex mode, search for:

<span class="smcpIncise">([[:upper:]])([[:upper:]]{2,})</span>

and replace with:

<span class="smcpIncise">\1\L\2\E</span>

For more examples see this very old post.

[Turtle91]

If you have any spans with spaces, or any character other than A-Z, then you would also have to check for that. You can also do it with two passes:

pass 1
search:
<span class="smcpIncise">(.*?)</span>

<span class="smcpIncise">\L\1\E</span>

pass 2
search:
<span class="smcpIncise">(.*?)</span>

replace:
<span class="smcpIncise">\u\1</span>


This two-pass technique can be saved as a "Saved Search" and run together with a single click.


After that you would need to individually check each span for words you wish to be capitalized like proper names, "I", "I've", "I'm", etc. Although you can create individual S/R for a bunch of those.

It would be really nice if we could have some kind of "Sentence case" or "Title Case" functionality in the S/R, but as I understand it, Sigil's flavor of Regex doesn't support that. If you absolutely need to have it, then you could try Calibre Editor which has the ability to do Regex Functions.

edit: You can get a "Sentence case" like effect by modifying Doitsu's search slightly. Adding a space or punctuation capture into the second group works well:

search:
<span class="smcpIncise">([[:upper:]])([[:upper:][:space:][:punct:]]{2,})</span>

replace:
<span class="smcpIncise">\1\L\2\E</span>

Cheers,

[roger64]

Hi

Thank you both for your detailed replies. I will proceed carefully.

Edit: I used the first regex with Sigil (it does not work with the Calibre editor). It proceeded about 95% of all entries. The remaining ones had one or more accented letters of some sort (â, é, ë, ê, ô, ç, etc.). But it was easy to correct. No need to adapt the regex.

[JSWolf]

How would you handle the regex if you had something like...

<span class="italic>Mark <span class="smcpIncise">SMITH</span> Jr.</span>

[Doitsu]

Originally Posted by roger64

Edit: I used the first regex with Sigil (it does not work with the Calibre editor). It proceeded about 95% of all entries. The remaining ones had one or more accented letters of some sort (â, é, ë, ê, ô, ç, etc.). But it was easy to correct. No need to adapt the regex.

The regular expressions suggested by Turtle91 and me only work with non-accented characters. To add support for Unicode characters simply prefix the search expression with (*UCP). E.g.

Find:(*UCP)<span class="smcpIncise">([[:upper:]])([[:upper:]]{2,})</span>
Replace:<span class="smcpIncise">\1\L\2\E</span>

will find:

<span class="smcpIncise">ÉPILOGUE</span>

and replace it with:

<span class="smcpIncise">Épilogue</span>

[roger64]

@Doitsu

Thank you very much for this very useful tip and congratulations for your extensive knowledge!

[Turtle91]

Originally Posted by roger64

@Doitsu

Thank you very much for this very useful tip and congratulations for your extensive knowledge!

Ditto! Thanks!!

[Jellby]

I guess you are lucky you don't have names like CALDERÓN DE LA BARCA -> Calderón de la Barca