Specifying build sequence

[Craiglea]

I am trying to build a series of books into an iSilo document from a series of html files on my PC. The end book will consist of series of individual books, some of which have chapters. When I move to the next page in the iSilo document I want to go through a sequence like BOOK1, BOOK2, CHAP1, CHAP2 etc, BOOK3, CHAP1 etc, BOOK4 etc.

The top-level html consists of a list of the individual books but not chapters in each. Links to books with chapters are to another html file with the individual chapters listed.

When I build the document I get a sequence that runs BOOK1, BOOK2, BOOK3, BOOK4, CHAP1, CHAP2, CHAP1 etc which clearly does not work in the same way as reading a real book. IT does matter whether I pull in the top-level document as the only source or include all the individual books in addition. Bookmarks are created with an external html file and this is working fine.

I know that I could get the sequence I want by creating a top-level html that defines:

BOOK1
BOOK2
CHAP1
CHAP2
BOOK3
CHAP1
CHAP2
BOOK4

etc.

But I would rather not.

Any other options before I have to do this?

Cheers

[iSilo]

You can add to the source list all the source files in the order that you want them to appear in the document. So instead of just adding the single top-level file, you will need to add all the files. Your source list should then look something like:

Source Page
------- -----
TOP.HTML 1
BOOK1.HTML 2
BOOK2.HTML 3
CHAP1.HTML 4
CHAP2.HTML 5
BOOK3.HTML 6
CHAP1.HTML 7
CHAP2.HTML 8
BOOK4.HTML 9
etc.

[Craiglea]

That seems to work well. The downside seems to be that the final PDB file has ended up about 25% bigger. It seems that the first part of the book consists of the pages in the order specified in the list of source files. After this I get the same content again - presumably pulled in through the links specified in the source files.

It seems to be that maybe I need to reduce the link depth with the source files explicitly defined in this manner(?)

Cheers

[Voltage Spike]

It seems to be that maybe I need to reduce the link depth with the source files explicitly defined in this manner(?)
It sounds like a bug. iSiloX is supposed to keep track of what it has already converted and not convert the same material twice.

Could this be an issue of case? For example, the .ixl specifies "BOOK1.HTML" while the link inside the document specifies "book1.html"?

[Craiglea]

I thought it shouldn't convert twice. I could try it under Windows but the IXL files are a bit too Mac-specific wrt paths in order to convert easily.

Cheers

[iSilo]

Make sure that all file names that should be the same agree in case, as the behavior should be that case does matter in URLs. That would explain why you would be getting files pulled in more than once. So if you have "Book1.html" in the source list. Make sure that the URLs of the links in the files are "Book1.html" exactly as is and not something where one or more characters differ in case (e.g., "book1.html" would cause the file to be pulled in more than once" since the 'b' is now lowercased whereas it was uppercase 'B' in the source list).

[Craiglea]

I don't think it is anything to do with mixing case in the hyperlinks. I've been pretty consistent in keeping track of the case used in hyperlinks although there might be the odd slip-up. If it were a problem with hyperlinks I would have expected that only the odd link with the incorrect case would have been included - it looked to me as though all the 'chapters' (i.e. third level of the hierarchy) were included after the specified list of pages.

I moved the source files over to my Windows machine and changed the path definitions (Easier than I first thought. The structured nature of the *.ixl file meant that I could just do a search and replace on path definitions using a text editor) and rebuilt the iSilo file.

This time the additional files were not pulled-in to the build and the file size went down to the smaller size that I had started with.

Cheers