Help with the regular expression

[Dysonco]

Hi All,

New member of the forum, so hello everyone! I thought I'd kick off with a conundrum that I'm having trouble solving.

I'm a great fan of Calibre, its a great bit of software, I'm just having a little trouble configuring it to correctly identify the information from the filenames in my book collection.

All my books have their filenames in this format:

AuthorLastname, AuthorFirstnames - BookSeries SeriesNumber - BookTitle.FileExtension

I've tweaked the regular expression (mostly by trial and error as I'm most definately not a programmer) to this:

(?P<author>[^_]+) - (?P<series>[^_]+) - (?P<title>[^_]+)

Now this works fine on the filename example as above, but unfortunately fails when the BookSeries and SeriesNumber parts are missing (when its a single book and not a part of a series).

So for example:

Pratchett, Terry - Discworld 01 - The Colour Of Magic.pdf

Would work okay, but:

Pratchett, Terry - Strata.pdf

Wouldn't. I almost need a way to set the expression to realise that if theres only two groups to recognise that it is Author and Title and ignore the series bit.

Any suggestions from the gurus?

Many thanks,

Mike

[Starson17]

Originally Posted by Dysonco

AuthorLastname, AuthorFirstnames - BookSeries SeriesNumber - BookTitle.FileExtension

I've tweaked the regular expression (mostly by trial and error as I'm most definately not a programmer) to this:

(?P<author>[^_]+) - (?P<series>[^_]+) - (?P<title>[^_]+)

Now this works fine on the filename example as above, but unfortunately fails when the BookSeries and SeriesNumber parts are missing (when its a single book and not a part of a series).

So for example:

Pratchett, Terry - Discworld 01 - The Colour Of Magic.pdf

Would work okay, but:

Pratchett, Terry - Strata.pdf

Wouldn't.

This is what I'm using now:

^((?P<author>([^\-_0-9]+)(?=\s*-\s*)(?!\s*-\s*[0-9.]+)|\b))(\s*-\s*)?((?P<series>[^0-9\-]+) (- )?(?P<series_index>[0-9.]+)\s*-\s*)?(?P<title>.+)

It will parse your two examples and many others. Yours doesn't get the series_index. There are several regex threads here if you want more options.

[Dysonco]

Hi Starson17,

Wow that works great!

Many thanks,

Mike

[Ozzy]

@ Starson17 -- That's a great expression, exactly what I was looking for. Thnx.

Originally Posted by Starson17

This is what I'm using now:

^((?P<author>([^\-_0-9]+)(?=\s*-\s*)(?!\s*-\s*[0-9.]+)|\b))(\s*-\s*)?((?P<series>[^0-9\-]+) (- )?(?P<series_index>[0-9.]+)\s*-\s*)?(?P<title>.+)

It will parse your two examples and many others. Yours doesn't get the series_index. There are several regex threads here if you want more options.

[kapaka]

Originally Posted by Starson17

This is what I'm using now:

^((?P<author>([^\-_0-9]+)(?=\s*-\s*)(?!\s*-\s*[0-9.]+)|\b))(\s*-\s*)?((?P<series>[^0-9\-]+) (- )?(?P<series_index>[0-9.]+)\s*-\s*)?(?P<title>.+)

It will parse your two examples and many others. Yours doesn't get the series_index. There are several regex threads here if you want more options.

Thank You So Much!!!
I have been trying to figure out regular expression for just this, was about to quit and edit manually

[qlfwyyd]

This is awesome! Thanks so much.

[qlfwyyd]

Actually - any chance this could be adapted so that it would still work if the series index is not included? At present, if there is no index, the series name is added to the title.

[Starson17]

Originally Posted by qlfwyyd

Actually - any chance this could be adapted so that it would still work if the series index is not included? At present, if there is no index, the series name is added to the title.

Yes, but that breaks other things. The numbers of the series_index are used to help identify the series name.

[qlfwyyd]

Ah well, it's pretty damn good anyway. I'll just have to add in the series numbers by hand first. Thanks again!

[Starson17]

Originally Posted by qlfwyyd

Ah well, it's pretty damn good anyway. I'll just have to add in the series numbers by hand first. Thanks again!

Or you can change the regex when you have books to add that don't have the series number, but have some other way to identify the series. If the series name is always in brackets, or always is after the first space-hyphen-space, then you can find it that way. I haven't found any single regex that will do all the different file naming schemes I've seen.