Well this fault occurs relatively frequently on EpubCheck validation whenever you use the file splitter in Sigil. Here is the error message on EpubCheck from the IDPF validator:
ERROR(RSC-005): Error while parsing file 'value of attribute "id" is invalid; must be an XML name without colons'.
Of course, there are no colons in the id line to speak of so that error message from the Validator is complete hogwash and no help at all.
I kept getting this annoying error on Sigil file split then one day as I was browsing the rules and regs for epubs on the IDPF site I read something interesting which said this, more or less: if you use a 32 char hex id in an 8-4-4-4-12 configuration to denote xml structure ids in the epub then the first hex character in the uid must be an alphabet character. To illustrate this more broadly(with emphasis):
This id will fail IDPF EpubCheck because it start with a numeric digit:
7d0d5c28-5743-40c1-bafa-048c5bba8e6f
But this id will pass because it starts with an alphabet character:
ed0d5c28-5743-40c1-bafa-048c5bba8e6f
So if you get this EpubCheck error on file split, just check the file split idref in the opf spine and manifest and, if necessary, change the first character from a numeric digit to an alphabet character in the range of a to f(because its hex). Do this for both ids in the spine and manifest and the problem will be resolved.
And it would also be quite nice if this problem was fixed in Sigil since it has been with us for such a long time. Can someone please fix this problem?
By the way, the book id in the metadata section is a different can of beans because it isn't part of the epub structure -- so it doesn't matter if this uid starts with a numeric digit or an alpha character.
ERROR(RSC-005): Error while parsing file 'value of attribute "id" is invalid; must be an XML name without colons'.
Of course, there are no colons in the id line to speak of so that error message from the Validator is complete hogwash and no help at all.
I kept getting this annoying error on Sigil file split then one day as I was browsing the rules and regs for epubs on the IDPF site I read something interesting which said this, more or less: if you use a 32 char hex id in an 8-4-4-4-12 configuration to denote xml structure ids in the epub then the first hex character in the uid must be an alphabet character. To illustrate this more broadly(with emphasis):
This id will fail IDPF EpubCheck because it start with a numeric digit:
7d0d5c28-5743-40c1-bafa-048c5bba8e6f
But this id will pass because it starts with an alphabet character:
ed0d5c28-5743-40c1-bafa-048c5bba8e6f
So if you get this EpubCheck error on file split, just check the file split idref in the opf spine and manifest and, if necessary, change the first character from a numeric digit to an alphabet character in the range of a to f(because its hex). Do this for both ids in the spine and manifest and the problem will be resolved.
And it would also be quite nice if this problem was fixed in Sigil since it has been with us for such a long time. Can someone please fix this problem?
By the way, the book id in the metadata section is a different can of beans because it isn't part of the epub structure -- so it doesn't matter if this uid starts with a numeric digit or an alpha character.