Page Page
WebCheck, a product of epixtech, Inc., is a Windows-based URL checker. Used in conjunction with Dynix cataloging software, it allows catalogers to maintin 856 tags in the MARC record.
WebCheck validation is divided into several &SAVEDLISTS& files to separate out bibs "owned" by consortia agency (AXE, KTC, IRC, USD, FSCC, PPL) as well as by narrower KFP holdings code (049 in the MARC record) within the AXE primary agency. This allows catalogers in documents to work specifically on documents; periodicals to work on periodicals; cataloging to work on ebooks or annual reports, etc. This is also necessary organization of a high-growth collection: as of January 2002 we have over 15,000 bibliographic records containing 856 tags.
As part of the cataloger validation, an &SAVEDLISTS& file is used to identify particular sets of records for an agency or department. These &SAVEDLISTS& file names must be communicated to the cataloger and typed exactly as they are stored on the Dynix server (uppercase).
A cataloger can select an &SAVEDLISTS& file to review or validate. However, when an &SAVEDLISTS& of items has actually completed validation and is ready for updating, a separate step must be undertaken in order to review only those URLs that are bad. This information is found in the BIB field BADURL.
If a cataloger wishes to work on just the invalid URLs from &SAVEDLISTS&, pull down "BIB" in the first combo-box, then pull down the "SELECT=" option in the second combo-box. Finally, enter in the URL box (default says "SELECT BIB complete select statement" - delete this) and type the following statement:
GET.LIST URL.XX; SELECT BIB SAVING UNIQUE BADURL
, where URL.XX is the name of the specific &SAVEDLISTS& file that a cataloger may be working with.
This will display the date validated, the redirected ULR, or 404/405 message or other error messages, giving the cataloger a better idea of what is wrong with the URL and what the server response was when the validation occurred.
This list of files should be used by the catalogers for WebCheck validation. The &SAVEDLISTS& files break down into groups of 500 records or less per material type and agency.
| &SAVEDLISTS& Name | Description |
| URL.CLOUD | Community Resource Agencies |
| URL.FSCC | Fort Scott Community College Library, Fort Scott, KS |
| URL.IRC | Instructional Resource Center, School of Education, PSU |
| URL.KFPI | PSU Axe Library State Documents |
| URL.KFPJ | PSU Axe Library Federal Documents |
| URL.KFPJ.AB | Federal A and B |
| URL.KFPJ.C.1 | Federal C1 |
| URL.KFPJ.C.2 | Federal C2 |
| URL.KFPJ.C.3 | Federal C3 |
| URL.KFPJ.D | Federal D |
| URL.KFPJ.E.1 | Federal E1 |
| URL.KFPJ.E.2 | Federal E2 |
| URL.KFPJ.F | Federal F |
| URL.KFPJ.G.1 | Federal G1 |
| URL.KFPJ.G.2 | Federal G2 |
| URL.KFPJ.G.3 | Federal G3 |
| URL.KFPJ.H | Federal H |
| URL.KFPJ.I | Federal I |
| URL.KFPJ.J | Federal J |
| URL.KFPJ.KLM | Federal K, L and M |
| URL.KFPJ.NOPR | Federal N, O, P and R |
| URL.KFPJ.S | Federal S |
| URL.KFPJ.TVX | Federal T, U, V, W, X |
| URL.KFPJ.Y.1 | Federal Y1 |
| URL.KFPJ.Y.2 | Federal Y2 |
| URL.KFPJ.Y.3 | Federal Y3 |
| URL.KFPJ.Z | Federal Z |
| URL.KTC | Kansas Technology Center Library, PSU |
| URL.NETLIBRARY | eBooks from NetLibrary |
| URL.NEWS | Community Resource Newspaper Files |
| URL.NON.DOCS | Any KFP Holdings Code Minus External Agencies, KFPJ, KFPI, and KFPX) |
| URL.PPL | Pittsburg Public Library, Pittsburg, KS |
| URL.USD | Unified School District #250, Pittsburg, KS |
The following directions pertain mainly to the person creating the select lists on the Dynix server, not the catalogers. The SUZY.URL.PARA should be run at least monthly to include new additions to the database (particularly from docs loads). If a site has added new records to the database, but if SUZY.URL.PARA has not been run, the new record's URL will not be in the &SAVEDLISTS& file automatically until SUZY.URL.PARA has been run.
WebCheck (client end) is simply maintained by pulling down the software from the epixtech site. Permissions are identified within the Webcheck directions, i.e., the OS login must be identified to be able to execute, edit, and display WebCheck; in addition the Dynix login must have the ability to edit bib records in UBR if in fact they are to go ahead and correct the URL on the bib record.
Axe uniquely parses out the URLs and distributes the maintenance among several departments and sites. Clearly it is impossible to validate 15,000 or more URLs in a single validation run.
In ACC.CAT1 a paragraph exists called SUZY.URL.PARA, which does several things: searches the bib file for occurrences of 856 tags in the MARC record (BIB field ELECTLOC); sorts the bib file by JW049 (holdings code) to move bibs into responsible areas (such as by each agency; and within PSU, defined by docs, non-docs (annual reports and others), ebooks, periodicals, com res, KTC, IRC, etc.
Government Documents are further broken down by broad SuDoc Number (A-Z), since many of the URLs reside in doc records. SuDoc ranges C, G, and Y are further broken down by groups of 500, and should be examined frequently to determine if further breakdown is required. Optimally, no more than 500 items should be in a given &SAVEDLISTS& file in order to efficiently validate through WebCheck. Lastly, SUZY.URL.PARA sorts the URLs within an &SAVEDLISTS& file by the URL proper, making it easier for the cataloger to find similar problem web domain names all in the same validation session.
To list all filenames in &SAVEDLISTS& beginning like "URL...", type
LIST &SAVEDLISTS& WITH @ID LIKE "URL..." BY @ID
When validating the records, especially large quantities of records which may reside on the same domain, directory, or location, all &SAVEDLISTS& should use SSELECT (sort-selected) by ELECTLOC (856 tag) to keep like-named documents and like-named servers together, the theory being that if one domain name has changed route or path, generally all on that server, path, or directory will have changed.
URL.KFPJ holds the largest number of records on the Axe Library Dynix Server because it contains all the federal documents in the online catalog with 856 tags.
An I-Descriptor in BIB is set up to truncate the bib call number by the first letter of the document's call number. The I-Descriptor is called SJ.CALL.TRIM, and looks like this:
If the call number needs to be further segmented down, simply change the value in field 2 from CALL[1,2] to CALL[1,3] or CALL[1,4], etc. as necessary.
The results of the trimmed, or truncated call number display will result thusly:
GET.LIST URL.KFPJ
LIST BIB BY SJ.CALL.TRIM.1 BREAK.ON SJ.CALL.TRIM.1 TOTAL COUNTER DET-SUPP BIB....... CALL
*** 9230
9230 records listed.
This breakdown clearly defines which primary call number areas of the SuDoc records have minimal 856 tags and which have volumnious records. Areas such as C, E, G, and Y need to then be sampled off at 500 records per &SAVEDLISTS& filename for optimal effectiveness.
For &SAVEDLISTS& with more than 500 records, a naming scheme should be devised so that when 500 records are sampled, then saved to a &SAVEDLISTS& (such as KFPJ.C.1), these 500 in C.1 need to be removed from the original "C" list using LIST.DIFF. Thus:
Send comments to: suzyq@pittstate.edu
Susan M. Johns-Smith
Axe Library
Pittsburg State University
1605 South Joplin Street
Pittsburg, KS 66762
Phone: 620-235-4115
This page last updated Monday, 28-Oct-2002 15:58:26 CST