Replacing code without replacing text?

[ElMiko]

So I've now run into this problem twice. Hopefully, the reg-ex wizards among you will have a head-slappingly simple solution to my problem.

First Case
Every new paragraph is formatted as:

<p class="calibre2"><span class="none">A Bunch of Text</span></p>

I want to change the "p class" to "calibre5" for Chapter headings only, which are currently formatted like this (1., 2., 3., ..., 21., 22., etc):

<p class="calibre2"><span class="none">1.</span></p>

Finding all the instances with reg-ex is easy enough by searching for:

<p class="calibre2"><span class="none">[0-9]+\.</span></p>

But that's as far as I can get. Replacing the above with:

<p class="calibre5"><span class="none">[0-9]+\.</span></p>

changes the number to "[0-9]+\."!

Second Case
In another book, there are random paragraph breaks mid-sentence.
My solution to that would be to find all the new paragraphs that begin with a lower-case letter, and replace all the code between it and the previous paragraph. Again, finding the instances via reg-ex proved pretty intuitive:

</p>[\s]+<p class="calibre2">[a-z]

But again, i just can't figure out how to alter the code (in this case replacing it with a blank space) without altering the first letter of the first word in that paragraph.

Looking over what I've written, I am humbled by how utterly inarticulate this all sounds, but if someone understands my gibberish, I'd be most grateful for any light you could shine on my problem.

[theducks]

Originally Posted by ElMiko

So I've now run into this problem twice. Hopefully, the reg-ex wizards among you will have a head-slappingly simple solution to my problem.

First Case
Every new paragraph is formatted as:

<p class="calibre2"><span class="none">A Bunch of Text</span></p>

I want to change the "p class" to "calibre5" for Chapter headings only, which are currently formatted like this (1., 2., 3., ..., 21., 22., etc):

<p class="calibre2"><span class="none">1.</span></p>

Finding all the instances with reg-ex is easy enough by search for:

<p class="calibre2"><span class="none">([0-9]+\.)</span></p>

But that's as far as I can get. Replacing the above with:

Add the Red

<p class="calibre5"><span class="none">\1</span></p>

Now make the replace look like the above:
The red \1 is the found value inside the ( ) above

changes the number to "[0-9]+\."!

Second Case
In another book, there are random paragraph breaks mid-sentence.
My solution to that would be to find all the new paragraphs that begin with a lower-case letter, and replace all the code between it and the previous paragraph. Again, finding the instances via reg-ex proved pretty intuitive:

</p>[\s]+<p class="calibre2">([a-z])

But again, i just can't figure out how to alter the code (in this case replacing it with a blank space) without altering the first letter of the first word in that paragraph.
[/QUOTE]
the replace is: (space)\1

Looking over what I've written, I am humbled by how utterly inarticulate this all sounds, but if someone understands my gibberish, I'd be most grateful for any light you could shine on my problem.

[ElMiko]

@theducks - Life. Saver. If I could impose on you for just one clarification which might help me understand the nature of my mistakes (and your solutions) better, i'd be even more grateful:

What was the significance of the parentheses? That is, how are they acting upon the expression in a way that would be distinct without them? Is it entirely so that the \1 knows what to refer to?

[theducks]

Originally Posted by ElMiko

@theducks - Life. Saver. If I could impose on you for just one clarification which might help me understand the nature of my mistakes (and your solutions) better, i'd be even more grateful:

What was the significance of the parentheses? That is, how are they acting upon the expression in a way that would be distinct without them? Is it entirely so that the \1 knows what to refer to?

forward reference
\1 is the first
\2 is the second and so on
example:
FN LN:
(.+) (.+)
want LN, FN
\2, \1

[sellew]

I'm not very much experienced in regex but maybe I can throw some light. The parentheses are used to create a group, so that when replacing you can refer to it. That's the meaning of '\1': keep the original set of characters (the group) as it is.
You can use more than one group in the expression you want to match, for instance:
<p class="calibre2"><span class="none">([0-9]+\.)</span>blablabla([0-9])</p>
Then, when replacing, you can refer to each group using \1 and \2.
Of course you can use as many groups as required depending on the complexity of the search.

[theducks]

Here is the Link to Manichean's tutorial that finally got REGEX through my thick skull

http://www.mobileread.com/forums/sho...REGEX+Tutorial

[ElMiko]

@sellew - That's exactly the explanation I was looking for. Thank you for putting it so clearly! Manichean's tutorial is indeed incredibly helpful, but sometimes (oftentimes... most times...) I get lost anyway.

@ theducks - Thanks for the additional example. And sorry for being so obtuse sometimes (oftentimes... most times...)!