The goal of this tool is to make it possible to search some common sources of good/consistent MARC data, and, after retrieving the data, attempt to map the MARC record into RDA and display the result using RIMMF.
The whole thing started as a 'tool' that we could use to develop and practise the MARC-to-RDA mapping supported by RIMMF2. It continues to be used for this purpose.
It also seems to be handy from a cataloging point of view, i.e., in our use of RIMMF to teach 'RDA thinking'. It gives you an extra option when you are building a 'Work' record, and the cursor is on 'Creator'. You can still select 'Create new Person', open a Person template, and fill out the template to create a Person record; now you can also press <F3>, enter a name, fetch the record, press Import, and … an RDA record is created for the Person and your Creator is linked to the new record.
The default setup in RIMMF is to search [NAF] Name as a simple keyword. To change the search target, and/or search options, goto the 'Search options' panel on the left.
The top 9 options search LC's Z39 interface using the SRU protocol; the next group of options search the ID.LOC.GOV versions of the NAF and SAF, and VIAF, using an autoSuggest request. The final option browses to Wikipedia using the current search term; such a valuable resource!
The reason LC NAF is split into /Name and /Title is because we are yet unable to simultaneously retrieve Name and Title records using this interface.
The reason for using two different protocols to search NAF and SAF is that one (SRU) provides much more granularity and comprehensive results, while the other (ID) are much faster. The former allows the selection of indexes to search against, application of modifiers to the search terms (eg. left-anchor, truncation, phrase vs. keyword), control over the number of records output, and retrieval via variant versions of a heading.
On the other hand, an ID.LOC.GOV search:
The Search Options are applicable only to the SRU search sources (the top 8 items).
Indexed search: if checked, the request will be made against a specific index, rather than as a google-type keyword. Here is a list of indexes1) we are using:
[BIB] Any - [BIB] Name bath.name [BIB] Title bath.title [BIB] Lccn bath.lccn [BIB] Isbn bath.isbn [NAF] Name bath.name [NAF] Title bath.uniformTitle [NAF] Lccn bath.lccn [SAF] Subject dc.subject
If 'Indexed search' is not ticked, then no index is specified, and the results are interesting, but fuzzy.
Truncation: if ticked, an '*' is appended to each search request.
As phrase: if ticked, the search request is wrapped in quotes (to keep the words 'together').
Left-anchored: if ticked, a '^' is prefixed to the search request.
Enter the words, etc., that you want to search here, then press <Enter>, or press the 'Search' button on the left. Even after pressing <Enter>, the cursor will remain in the box until you move it, so its easy to change a search.
After a search completes, if any results are retrieved, a hit list will be created and displayed in the main body of the form.
The usual action at this point is to select an item in the hitlist and click 'View MARC'; the corresponding record will be fetched 2) and displayed as Tagged Text in your default text editor. From there you may determine if the retrieved data is what you are looking for, or not.
As a shortcut, double-clicking on an item in the hitlist will do the same thing as clicking View Marc.
After looking at a MARC record and deciding that it matches your search, you may want to add the record to your RIMMF datastore. To do this, select the entry in the hitlist (this may already have been done in the previous step) and press 'Import'.
This will launch the following sequence of processing inside the program:
The above happens pretty quick, except perhaps when importing a Manifestation, when you might be able to see the program 'thinking a bit' about the task. The reason for this is simply the large size of the default template for this entity.
If a search seems hung-up you may use this button to abort it.
The default time-out for the search tool is 30 seconds.
Another way to abort a search is to click on the clock widget in the top right corner.
After a search, a blue caption will display between the search terms and the hitlist, and two little widgets will be shown at the top right side of the screen:
On the left ('16' in the snapshot) is the hit counter for the current search. On the right ('1 s' in the current snap) is the time taken for the search; the second widget will indicate a search is in progress by spinning a clockwise direction. You can click on the second widget to stop a search.
Local search
Every time you use the tool to perform a search, RIMMF performs the same search locally against your own Entity Index (EI). If there are any hits against the EI, it will be announced by the blue caption at the top of the hitlist; and if there are, simply click on the caption to switch over to the EI and view the matching items. The purpose of this is to prevent one from adding duplicates to the system.
Maximum hits
By default, the tool retrieves 15 items for an SRU request. If there are more hits than 15 (as evidenced by the hit counter), simply click the Search button a second time without doing or changing anything: RIMMF will return to the server with a request for all of the matching records (up to a maximum of 100).
Bottom splitter
The splitter bar at the bottom of the form opens to reveal the URL that we sent to the server; someday this might be editable so that the user might roll their own search; at present its use is so that we can figure out what the search tool is doing.
Problem searching names with first initial
When searching a name using the default options, the LC Z39 interface does not like it if you supply only one letter for the forename; for example:
Freud, A
This search will always time out3). Strangely, to avoid this problem, simply remove the 'A' from the search term. Other ways to avoid this problem: