[GUI-Plugin] Library Codes-SRU

[davidjoseph1]

New Version: Author-Title search language tweaked for greater accuracy.

[davidjoseph1]

Some questions about future development of this version:

A) How accurate and authoritative do users want the call number return to today?
There is a mechanical piece To this: right now, the SRU look up does not discriminate between multiple record returns when using Title author – the identifier Search will either return one record or none – and because I have not Implemented a rules based parsing of multiple applicable records the title author search will return the first record’s Library of Congress classification number and Dewey decimal Number, even if it is incorrect. Obviously my first development effort beyond merely getting it to work should be to par multiple records And return the best one.

OK, the next part of the mechanical piece is what if there is no applicable record at the library of Congress? There are many books which are published in many additions which have different ISBN’s and may have cataloging in publication information with actual library of Congress control numbers, but that particular edition may not exist Within the catalog of the Library of Congress to be returned as a record. But there may be another applicable record, which would serve. What should I do in this case? Is this a manual thing for the user to figure out on their own? Or should I create a truncated title/author search that may bring up an applicable record with a good classification, Even if the Edition is not exactly the same.
See this is the kind of problem that cataloging librarians encounter every day. I encountered this while testing to particular books that I have in electronic form. Let me compare and contrast – there is a 2009 reprint of the 1970 Harper and row introduction to population genetics theory by Kimura and crow by the Blackburn press, which is a facsimile Of the 1970 Harper and row edition, but with newly issued 10 and 13 digit ISBN’s and a Library of Congress control number issued in 2005, but neither the library of Congress Control number, nor the ISBN numbers are responsive to records at the library of Congress. The Reprint was simply never acquired and catalogued. What OCLC world cat does is bundle together all of the ISBN’s and title author information in related records so that librarians can distinguish set that along together with common classifications. Without the OCLC data Aggregation, a cataloger would have to research the provenance And history of a work and then select classifications based on Name and subject authority records, and the classifications of research libraries applying the same criteria To the work. In the crow – kimura Example given above, the librarian might simply take the classification of the 1970 edition which exists under a separate ISBN at the library of Congress And append year of publication at the end to distinguish the two editions.

Another example:
There is a 2023 reprint electronic edition of a 1949 Pseudonymous memoir entitled “the autobiography of a Sydney barmaid” by ‘Caddie’ . There was a 1975 reprint, which was the source of the electronic 2023 reprint with an introduction by a named person and a change to the original title, and the 2023 reprint has a slightly different title again. The 1949 and 1955 printing are in the library of Congress with Classifications. The 1975 is not, and the 2023 reprint was done without cataloging in publication and is collected in only a few Libraries With electronic classification records in the hands of the OCLC and gate-fenced data, Available only For staff use enforced by password in those institutions.


Is this a user situation they can handle themselves or should I work towards trying to figure this piece out?

B) Metadata for use in calibre beyond LCC and DDC - The Library of Congress classifies subject heading, which is available. That is the next piece that I want to make active often they provide table, contents, and summary information as well, sometimes urls To publisher provided data cards, describing author and title. What Dalton’s plug-in did before the old API was sunset Was that it returned author details, which were based on name authority records, aggregated by the OCLC, but which Originate in national name, authority, databases created by the library of Congress and other central depository, libraries, and other countries, Which are collected in several different ways, including The virtual international authority file, the International standard name identifier, and the national name and subject authority databases of the various countries. OCLC has its own version of that as well. Should there be a look up that encompasses VIAF and ISNI and the authority services at the library of Congress? Those name authority services exist to correlate variance of the Names of the same author. Of the same title. They correlate Subjects across various languages.

How much metadata do you want?

C) When I looked through the Development history of DaltonST’s Library codes plug-in - Yes, I used the way back machine to find the original version of the classify look up API script page which DaltonST adapted Originally in 2015, I came to the conclusion that though he was an enthusiastic and prolific developer of library organization, and Medida analysis tools, he unfortunately, didn’t have a library science background, or the interest in researching the Knowledge appropriate to the field, it seems like he found classify adventitiously because it was a very convenient and easy way to grab Acceptable Metadata Quickly.

I’m not a librarian, I’ve never done any formal library science studies. Nor am I a programmer, coding in python is very new to me. But I am an avid user of research library systems, and I have an interest in ontologies of knowledge. Does library codes need to embody that latter interest? Should it be “good Enough“? Is it OK if it doesn’t come up with the right answer all the time and you need to have a head on your shoulders to figure out why? Or does that need to be in the code?

If it does need to be in the code, I’m gonna need help coding it.

[AmyHNJ]

Thank you! Someone pointed me to this in the thread I created a while ago looking for replacements for LC and the manual website.

[lonegunman]

Thanks for your work
I change the separators from > to . in the db to allow hierarchical classification, may it is useful fore someone

[Reader4106]

I've been using "Library Codes" for a long time and therefore miss the fact that #lcc, #ddc, #lcead and lc_genre can no longer be recorded.

I was very pleased that davidjoseph1 made the effort to not only restore the old functionality, but also improve it.
Thanks for that.

However, I was unable to get the new plug-in to work, probably due to my lack of English language skills. I was able to install it, but it doesn't work yet. I didn't find any help, explanation, etc. either.
What do I have to pay attention to, what do I have to set where to get back to the information. Since I almost exclusively read German-language books, there is only a limited selection of information available anyway.

(Translated with Google Translator)

[davidjoseph1]

Originally Posted by Reader4106

I've been using "Library Codes" for a long time and therefore miss the fact that #lcc, #ddc, #lcead and lc_genre can no longer be recorded.

I was very pleased that davidjoseph1 made the effort to not only restore the old functionality, but also improve it.
Thanks for that.

However, I was unable to get the new plug-in to work, probably due to my lack of English language skills. I was able to install it, but it doesn't work yet. I didn't find any help, explanation, etc. either.
What do I have to pay attention to, what do I have to set where to get back to the information. Since I almost exclusively read German-language books, there is only a limited selection of information available anyway.

(Translated with Google Translator)

Guten Morgen, Reader4106!

Meine Änderungen an „Codes“ haben die Art und Weise, wie das Plugin mit Calibre interagiert, (noch) nicht verändert. Die Änderungen betreffen ausschließlich die Datenbank und die Erstellung der URL-Abfrage, die mit den Menüoptionen verknüpft ist, die Dalton-ST ursprünglich codiert hat.
Es funktioniert also weitgehend genauso wie vorher, sodass jedes Problem beim Abrufen von Daten mit der Abfrage zu tun hat, die verschiedene Ursachen haben kann:

1) Die ISBN ist in der Library of Congress nicht katalogisiert. Suchen nach ISBN schlagen vollständig fehl und liefern keinerlei Ergebnisse. Ich habe dafür keine Lösung mit der aktuellen bibliografischen Datenbank der Library of Congress. Die fehlgeschlagene ISBN wäre bei WorldCat kreuzindiziert worden, aber diese API funktioniert nicht mehr. Es gibt alternative APIs: VIAF – „Virtual International Authority File“, ISNI – „International Standard Name Identifier“ und die Name and Subject Authority-Server der Library of Congress. Es *gibt* bibliografische Datensätze, die über SRU bei der Deutschen Nationalbibliothek durchsucht werden können - "http://sru.k10plus.de/gvk?", aber ich habe weder die Konfigurierbarkeit für "Bibliothekscodes" implementiert, um eine richtige Such-URL zu erstellen, noch die PICA-Analysefunktion. Ebenso hat die Französische Nationalbibliothek eine API, aber ich habe noch nicht herausgefunden, wie sie funktioniert.

2) Die Boolesche Suche nach Autor und Titel kann mehrere verschiedene Datensätze als Ergebnisse zurückgeben, die derzeit noch nicht analysiert und voneinander unterschieden werden können, um ein einziges Ergebnis zurückzugeben. Daran arbeite ich derzeit und lerne dabei die Python-Programmierung. Es gibt viele verschiedene Skripte, die richtig codiert werden müssen, um mehrere Ergebnisse zu analysieren und das am besten geeignete auszuwählen.

Die Vereinigung der bibliografischen Datensätze zwischen dem Katalog der Library of Congress und dem Satz von Büchern, die Sie katalogisieren, ist zahlenmäßig viel kleiner als der frühere Vergleich zwischen dem alten OCLC Discovery-Katalog und demselben Satz.

[translation into English for everyone else -Good morning Reader4106:

My changes to "Library Codes" have not changed the way the plugin interacts with Calibre (yet). The changes are purely in the database and the creation of the URL query associated with the menu options that Dalton-ST originally coded.

So it works pretty much the same as before, so any problem retrieving data is related to the query, which can have a number of causes:

1) The ISBN is not cataloged in the Library of Congress. Searches for ISBN fail completely and return no results. I have no solution for this with the current Library of Congress bibliographic database. The failed ISBN would have been cross-indexed in WorldCat, but that API no longer works. There are alternative APIs: VIAF - "Virtual International Authority File", ISNI - "International Standard Name Identifier" and the Library of Congress Name and Subject Authority servers. There *are* bibliographic records searchable via SRU at the German National Library - "http://sru.k10plus.de/gvk?", but I haven't implemented the configurability for "library codes" to create a proper search URL, nor the PICA parsing feature. Likewise, the French National Library has an API, but I haven't figured out how it works yet.

2) Boolean search by author and title can return multiple different records as results, which currently cannot be parsed and distinguished from each other to return a single result. I'm currently working on this, learning Python programming in the process. There are many different scripts that need to be coded properly to parse multiple results and select the most appropriate one.

The union of the set of bibliographic records between the Library of Congress catalogue and the set of books that you are cataloguing is much smaller in number than the former comparison between the old OCLC Discovery catalogue and the same set. ]

[Reader4106]

Thanks for your answer.
I can very well imagine that it is because all the books you are looking for (ISBN, title) are written in German. And also very often by authors who have not, or not yet, been translated into English.
I'm also very inept at a lot of programming/analysis/logic stuff.
So I can't help you solve the problem.
So I will wait until there is a definitive and easy solution to update my missing LCC and DDC data.
Good luck with your programming.