Replace notation-within-number-span with plain notation

[nichtich]

E.g. analysis of 700.90440747471 includes

T1--0901-0905:074 Museen, Sammlungen, Ausstellungen; Sammeln von Objekten

which should become T1--074 instead. I marked it as bug because notation T1--0901-0905:074 does not exist in the schedules but T1--074 does. The full notation with : could be of use for later analysis but not needed for now and for converting to PICA.

The change can directly be done in https://github.com/gbv/coli-ana/blob/dev/lib/parseInputStream.js.

[stefandesu]

notation T1--0901-0905:074 does not exist in the schedules

But it does, doesn't it? https://deweyde.pansoft.de/webdewey/index_11.html?recordId=int%3aT1--0901-0905%3b1%3b074

[ulsw]

stefan is right! Don't change!

[nichtich]

Well, it's an internal notation. Should we reduce to T1--074 instead only in the list of basic semantic elements as encoded in PICA?

[nichtich]

I've included the change in the conversion to PICA, so 700.90440747471 results in:

045H/10 $eDDC23ger$a700.90440747471$c7$f09044$f074$g7471$Acoli-ana

which consists of DDC notations

• 2
• T1--09044
• T1--074
• T2--7471

This will allow to find all publications about T1--074 in one search.

We can close this issue but should work on bringing tabular display and PICA result more aligned (see also #42).

[stefandesu]

Now I'm confused though. What's the difference between T1--074 and T1--0901-0905:074? Why can the same number be built with both of those?

[ulsw]

it depends on how the ddc numbers are built, in 709.04007479494 "T1--074" is applied, but in 759.030904107477595 "T1--0901-T1--0905:074" is applied

[stefandesu]

it depends on how the ddc numbers are built

Maybe I didn't make my confusion clear. Let's keep to 700.90440747471 as an example. In my previous comment, there's a screenshot from WebDewey where it is built with "T1--0901-T1--0905:074". However, you can also build the exact same number with "T1--074":

@nichtich also said "it's an internal notation." which is only more confusion. It seems like we're all talking past each other.

[nichtich]

However, you can also build the exact same number with "T1--074"

Either it is possible to build the same number in different ways or T1--0901-0905:074 can be mapped as semantically equivalent (although not structurally) to T1--074.

[ulsw]

Either it is possible to build the same number in different ways

There is only one way to synthesize DDC numbers, e.g. cf.

"https://www.oclc.org/content/dam/oclc/dewey/versions/print/intro.pdf "Classifying with the DDC 5.1 . . . TABLE OF LAST RESORT 5.9 When several numbers have been found for the work in hand, and each seems as good as the next, the following table of last resort (in order of preference) may be used as a guideline in the absence of any other rule:"

[nichtich]

I'm closing this issue with following solution:

• analysis uses full URIs from tables such as T1--0901-0905:074
• In PICA these are reduced to basic form such as T1-074 because these are more used for retrieval

[UmaB7]

Jakob - As discussed agree with you :-)