Page 1 of 1

Verification of the work page auto-tagger

Posted: Sat Jun 16, 2007 5:44 am
by imslp
After some wrestling I've finished writing the work page auto-tagging system... though I'd like some confirmation whether it works right. The logic is as follows:

First assume all publications are public domain in all three places (Canada, US and EU).

If death date is less than 50 years ago, and it was first published less than 50 years ago (to account for posthumous publication), it is not public domain in Canada.

If published after 1963, the publication is not public domain in the US. (I'm not sure how to work around the 1963/1923 issue; maybe a new template is in order?) The public domain status in the US is independent of composer death date.

If death date is less than 70 years ago, it is not public domain in the EU. The public domain status in the EU is independent of date of first publication.

The code for determining the tags are as follows:

Code: Select all

array( '000' => '',
			'001' => '{{WorkNonPD-EU}}',
			'010' => '{{WorkNonPD-US}}',
			'011' => '{{WorkNonPD-USandEU}}',
			'100' => '{{Work1923}}',
			'101' => '{{Work1923}}',
			'110' => '{{WorknotPD}}',
			'111' => '{{WorknotPD}}' );
Where 1 is not-pd and 0 is pd. The order of the numbers is Canada, US then EU.

Posted: Wed Jun 27, 2007 8:51 am
by Peter
I noticed an error. when no date is provided for a work by a composer that is only in canada in public domain (example here) the auto tagger assumes it is published before 1923 and says that the work is in public domain in U.S.

Posted: Wed Jun 27, 2007 8:53 pm
by imslp
Well, this is actually the intended effect (see here). Basically, when no year of publication (or year of composition) is given, the auto-tagger assumes the year zero. Testing is also done on the composer's year of birth, but that is relatively lax (1909 and 1949 respectively). It is like this because there are many pages with no year of publication/composition, and I'm not sure how else it can be done; I think manual correction of the exceptions is easier than trying to untag all the tags... plus, the work could really be PD in the US if it was published before 1923.

If you have a better idea about this, please tell me :)