Running calibredb from a cron script

[rolgiati]

Complete newbee in the use of calibredb.

Simple question: I want to run a calibredb list -f ...... once a month from a cron job to generate an up-to-date csv list of all my books.

Do I need to run calibre-server before running calibredb ?
and is there somewhere a howto to guide the hesitant newbees ?

[BetterRed]

You don't have to run calibre_server to use calibredb, see ==>> calibredb command

calibredb can be used to manipulate either a calibre database specified by a path or a calibre Content server running either on the local machine or over the internet

BR

[rolgiati]

Thanks; that's just what I needed.

[rolgiati]

I am looking for an answer to either of the following two questions:

When running calibredb catalog is there a way torestrict the fields listed in the output ?

OR

When running calibredb list -f "authors, title, series" is there a way to avoid the fields in the output being padded with spaces to a fixed length, and the lines prefixed with a reference number ?

[theducks]

id is the reference number in Calibre the matches a specific line item

[PeterT]

See the manual https://manual.calibre-ebook.com/gen...calibredb.html

[rolgiati]

Originally Posted by PeterT

See the manual https://manual.calibre-ebook.com/gen...calibredb.html

Thank you for our kind help, but I have already read that page carefully before asking my questions, and I did not see there anything that answers them.

Maybe I'll have to stick with calibredb list, followed by a sed job to remove the unwanted padding spaces....

[BetterRed]

Originally Posted by rolgiati

Thank you for our kind help, but I have already read that page carefully before asking my questions, and I did not see there anything that answers them.

@rogliati - you couldn't have read it very carefully. Had you, then you would have seen that the list sub-command has a --separator option, and that the --machine-readable option will give you a json file.

I don't know anything about the catalog sub command. But if it's like the one in the GUI then I think it's aimed at producing glossy presentation catalogues with cover images, fancy typesetting etc.

BR

[rolgiati]

Originally Posted by BetterRed

@rogliati - you couldn't have read it very carefully. Had you, then you would have seen that the list sub-command has a --separator option, and that the --machine-readable option will give you a json file.

I have read it very carefully; but found that the --for-machine (not --machine-readable) option outputs a file with on line for each field of the book; useless for my need, which is to get a file I can multiply grep to get a list of all the books that have the different words anywhere in title, authors or series.
Mind that the --separator is ignored when --for-machine is used.

Originally Posted by BetterRed

I don't know anything about the catalog sub command. But if it's like the one in the GUI then I think it's aimed at producing glossy presentation catalogues with cover images, fancy typesetting etc.

The catalog option is completely different from that of the GUI, it outputs a csv file with all the fields of a book on one line.
Still I might go that way, with an awk job to remove all the unwanted fluff.

So although our help did not, thank you for taking the time.

PS I am rolgiati...

[kovidgoyal]

calibredb catalog test.csv --fields=title,authors

For more information:

calibredb catalog test.csv --help

[rolgiati]

Originally Posted by kovidgoyal

calibredb catalog test.csv --fields=title,authors
For more information:
calibredb catalog test.csv --help

Thank you kind sir, that did it perfectly.
Just adding now a mixture of sed, grep, sort and awk jobs to modify the formatting, and I'll have what I wanted / needed.

Something like

sed 's/,\"\"//p'| sed 's/\",\"/ \# /p' | sed 's/\",\"/ \/ /gp'| sed 's/\"//gp' | sort -u  | grep -v authors | awk -F# '{print $2}' > Catalog$LIB

seems to work well.

[rolgiati]

Now that my script is running as wished to produce the catalog files I wanted, a new problem has arisen:

For the calibredb-based script to run, calibre must not be running.

So just in case it is running when the cron-job is launched, I have tried to stop it from the script, with pkill calibre, pkill /opt/calibre/bin/calibre, etc run either as user or as root. only pkill /opt/calibre/bin/calibre-parallel stopped something, nothing else works.

Any idea ?

[BetterRed]

try calibre -s

And try RingTFM for heaven's sake

Moderator Notice
I merged your three threads so that anyone coming along later can read the saga of cron and calibre in one fell swoop.

BR

[rolgiati]

Originally Posted by BetterRed

try calibre -s

Perfect, ta