html to epub and chap detection

[Trigun]

hello
i use the tool ebook-convert for convert a html file in a epub
in my html file i have

<mbp:pagebreak /><mbp:section>

at the end of each chapter but after the convert i lose the chap division.... how i can prevent it?

[..] </div> 
</div><mbp:pagebreak /><mbp:section><header></header> 
<div class="entry-content"> 
 <div id="chapterContent" class="innerContent fr-view"> 
  <p dir="ltr">Chapter [..]

i use :
ebook-convert "something.html" "something.mobi"

PS the format of the chapter isn't always the same but the pagebreak and section is always at the end of the chapter
sometimes the chapter is in a <h2> tag and is detected, but i don't have a standard

[kovidgoyal]

<mbpagebreak> is not valid html. Use the search and replace options to remove it and use the structure detection options to detect the chapter headings

[Trigun]

i know where the chapter start (i merge different urls in a single html and after convert it to mobi)
i'm the one who add that tag for separate the chapters but i don't know the structure of the text inside of the urls
sometimes the url is formatted with h2 title
sometimes is different and when that happens i don't have the chapter end (i don't care about the title of the chap but atleast i want a division from the previous chapter)
how i can do that?

i can add a h2 title too instead of the mbp pagebreak
but in that case if the text have already a h2 title i have double chapter

[jackie_w]

If you know the mbp pagebreak tags are all already in the right places, you could try search/replacing all the mbp pagebreak tags with something like <h6></h6> (if your text doesn't contain any other h6 tags). Then use calibre to split on h6 tags. Finally remove the empty h6 pairs. Crude but functional.