Copy date from Title to Pubdate field

[rlh3]

I have a significant number of items in my various Calibre databases that are periodicals. The way I catalog them is to have the name of the publication and the year-Month of publication in the title field (E.g. National Geographic [1994-04].)

What I am trying to figure out, is there any way extract the [1994-04] and use that to make the entry in the published field as "Apr 1994."

I spend a fair amount of time getting the title field correct, and would really like to be able to automate this part.

RLH3

[BetterRed]

Can't help with automate at add time, that probably needs the Actions Chain plugin.

However, you should be able to make use of the Bulk Metadata Edit->Search and replace tool's Regular expression mode to deal with existing books. Maybe a Saved MDE S&R spec can be used in an Action Chains script.

BR

[rlh3]

My workflow would not allow it to be done at the point of adding the books to the database. I add them, then there is a fairly extensive editing process to clean everything up (and checking the information against published indices of the magazines in question.)
I guess what I was hoping for was someone who has a clue with RegEx would be able to come up with something that I could use during the editing process as a saved S&R.

rlh3

[BetterRed]

FWIW in my Periodicals library, I put the publication name in Title - e.g. "The Atlantic", Granta" etc , and the issue date in a custom column — #issue_date/Issue date formatted as YYYY MMMM. If I wanted a composite containing both title and #issue_date I'd create one with a template- {title} {#issue_date).

Yes I have duplicate titles, but Calibre can handle that okay. Can't remember why I created a custom columns for the issue date rather than use pubdate.

IMO it is better to store data in the lowest form that makes sense… it's why Lego works.

BR

[rlh3]

I understand where you are coming from, but I have 30-40000 issues already formatted and in my libraries (I collect scans of comics and pulps in addition to magazines) and I really do not want to have to go back and change all of those.

I probably could have designed this whole thing better in the first place, but what I've got works for me.

As I mentioned up-thread, I am hoping that someone who understands RegEx well will take pity on me and help me out with this.

rlh3

[theducks]

I believe Calibre will accept any valid date format. You set the Display format it Tweeks.

National Geographic [1994-04]


Test this in the MDE bulk mode S&R REGEX

search {title}
.+?\s[(\d\d\d\d-\d\d)]

Replace {pubdate}
\1

That should populate Published. If that does not work try Replace \1-01
(pads with day 1)

Cleanup pass if wanted
search {title}
(.+?)\s[\d\d\d\d-\d\d]

Replace {title}
NOTE I moved the ( ) pair

[rlh3]

ducks:

Both options give me an "Unknown string format: National Geographic [1889-04]"

The date format of the published (pubdate) field is already set to the correct format (mmm yyyy)

rlh3

calibre, version 7.19.0
ERROR: Failed: Unknown string format: National Geographic [1889-04]

Traceback (most recent call last):
  File "calibre\gui2\dialogs\metadata_bulk.py", line 198, in do_it
  File "calibre\gui2\dialogs\metadata_bulk.py", line 535, in do_all
  File "calibre\db\cache.py", line 85, in call_func_with_lock
  File "calibre\db\cache.py", line 1636, in set_field
  File "calibre\db\write.py", line 568, in set_books
  File "calibre\db\write.py", line 568, in <dictcomp>
  File "calibre\db\write.py", line 81, in adapt_date
  File "calibre\utils\date.py", line 138, in parse_only_date
  File "calibre\utils\date.py", line 114, in parse_date
  File "dateutil\parser\_parser.py", line 1368, in parse
  File "dateutil\parser\_parser.py", line 643, in parse
dateutil.parser._parser.ParserError: Unknown string format: National Geographic [1889-04]

[theducks]

Sorry, that was my best guess (I don't normally do dates that way).
Maybe one of the developers will step in.

[rlh3]

Ducks:

Just for fun this morning I tried running what you gave me but sent the result to the comments field and it pasted the entire title field (E.g. National Geographic [1994-04]) rather than just the date. Is there a way to make the search more selective?

rlh3

[theducks]

There are REGEX tutorials here on MR. That is where I learned what I do.
Calibre (and Sigil) use the PCRE dialect.

There is also a link to a REGEX cheatsheet, just remember The PCRE does not support all those.
\d is a digit
.+? is a not greedy wild card for 1 or more characters
\s is any kind of space
S the wild card should stop at the first space

try escaping the dash \- and the square brackets \[ \]

[A-Z], [0-9] is a range AFAIK it should have been treated as a literal in this case

[rlh3]

Escaping the brackets and the dash solved the problem.
Thank you very much for your help on this, you have just saved me many, many hours of work.

rlh3