Unicorn MARC upload procedures

Uploading MARC records into Unicorn is a three-step process involving:

1. data preparation within the external MARC editor (LaserQuest, OCLC, etc.)
2. file transfer from the client machine (your PC) to the Unicorn server (chess) using IMPORT_EXPORT, and finally,
3. the actual record load itself using SIRSI's REPORT utility.

Once this process is completed the records should be immediately accessible within Unicorn by either barcode, call number or control number. If you have answered "Yes" to the question "Index now?" you can also find the records using a keyword (NOT a browse) search of the shadow catalogue. After ADUTEXT (the bib indexing programme) is run (currently each day at 8:00PM), the records should be retrievable by the public as well.

Preparing records for IMPORT_EXPORT

Bibliographic and authority records may be imported into Unicorn from any source capable of producing USMARC communications format output. For those of you accustomed to editing in LaserQuest, Bibliofile, or OCLC, either online or through the Microenhancer, treat the MARC record as you always have, except that we will no longer key our local call number into 090 except on serial records (until MARC holdings is implemented). If item information is desired when loading a new record, all holdings information formerly entered into the 090 and 035 (barcode only) may be input into the 949, a field designed for generating associated call and copy records.

Field 949

This field may be used in external MARC editors, such as LaserQuest, Passport or Connexion, to generate holdings within Unicorn. The 949 is repeatable when you need to create multiple volumes/locations. Multiple copies of a call record (volume) may be generated from a single 949 by adding a second (third, etc.) |c and |i to the record (see the example below). The format is as follows (asterisks mark subfields of particular interest to cataloguers):

|a* Call number- preserve New call number spacing
|v* Volume/Part
|c* Copy- implicit if only one copy being sent; must be explicit for multiple copies
|i* Barcode
|t* Item type
|l* Home location
|k Current location- if assigning the work to a temporary location other than |l
|m* Library
|p Price
|u Acq_date
|d Last activity date
|e Date last charged
|f Date inventoried
|n Total charges
|g Times inventoried
|q Inhouse charges
|r Circulate flag
|s Permanence flag
|x Item Cat1
|z Item Cat2
|c Next copy
|i Next barcode

Multiple copies of a single volume owned by a single branch:

     949  DK 254 K52 T7
          1935|c1|i31162006329172|tIN-LIBRARY|lRARE|mQEII|c2
          |i31162006329180|c3|i31162006329198
     949  LE 3 M4214g 1990|c1|tIN-LIBRARY|lBOOKSTACKS|mCNS|c2|c3
    

An analytic owned by a single branch (note lack of |z before "no.467"; Unicorn will supply this on loading):

     949  QA 3 L28 NO.467|i31162006329206|tBOOK|lBOOKSTACKS|mQEII
    

Government document style numbering:

     949  Fs23- 235 \1993|i31162006329149|tBOOK|lGOVDOC|mQEII
    

Microfiche with an automatic barcode (leave |i out to have system generate this for you):

     949  Micro Fiche 4930|tMICROFORM|lMICROFORM|mQEII
    

A title held in multiple branches under different call numbers:

     949  WQ 100 A4R 1975|i31162002018175|p55.00|n21
     |tBOOK|lBOOKSTACKS|mHEALTH
     949  RG 526 A4 1975|i31162006329230
     |tBOOK|lBOOKSTACKS|mGRENFELL
     949  RG 526 A4 1975|i31162006329248|tBOOK|lBOOKSTACKS|mQEII
    

A multi-volume set held in a single branch:

     949  QL 737 U5 L55 1994|vV.1|i31162006329206
     |tBOOK|lBOOKSTACKS|mQEII 
     949  QL 737 U5 L55 1994|vV.2|i31162006329214
     |tBOOK|lBOOKSTACKS|mQEII
     949  QL 737 U5 L55 1994|vV.3|i31162006329222
     |tBOOK|lBOOKSTACKS|mQEII
    

A reference annual where back issues are not physically barcoded:

     949  QH 1 B75|v1993|tIN-LIBRARY|lREFERENCE|mQEII
     949  QH 1 B75|v1994|tIN-LIBRARY|lREFERENCE|mQEII
     949  QH 1 B75|v1995|tIN-LIBRARY|lREFERENCE|mQEII
     949  QH 1 B75|v1996|i31162006329214|tIN-LIBRARY
     |lREFERENCE|mQEII
    

The same example except latest edition is kept in Reference and superseded ones go to the stacks. Note here that the chronology is placed in |v. Unicorn does not automatically recognize dates as belonging to the separate volume field and hence these must be separately subfielded if you wish them to be treated as volume information:

     949  QH 1 B75|v1993|i31162006329206|tBOOK|lBOOKSTACKS|mQEII
     949  QH 1 B75|v1994|i31162006329214|tBOOK|lBOOKSTACKS|mQEII
     949  QH 1 B75|v1995|i31162006329222|tBOOK|lBOOKSTACKS|mQEII
     949  QH 1 B75|v1996|i31162006329230|tIN-LIBRARY
     |lREFERENCE|mQEII
    

Saving your edited records in USMARC

Before Unicorn can even recognize the records you are attempting to import they must be in an acceptable format, namely either MARC21 communications format, or the flat ASCII file format described in the Unicorn Cataloguing module manual. Technical service operations at the Memorial libraries generally produce external catalogue records from BookWhere, LaserQuest or OCLC. Of these, the BookWhere and OCLC records saved using the xpo command may be loaded directly without further preparation. On most machines these are saved to disk with the filename cdtable.mrc.

To import LQ and Bibliofile records, you must first pass the finished file through their respective utility programmes where they will be converted to USMARC. The LaserQuest utilities are outside of the main programme and are entered through the file utility.exe in the c:\quest or c:\questlan directory of your machine. Choose the SAVE TO MARC option and give the output file a name that you can remember.

Bibliofile utilities are entered through the main menu. Select MARC Record Export, Export method CFILE, Communication format MARC, Path A:\. The default filename for these will be com.cp.

Importing MARC files into Unicorn

Note that the load programmes listed below make use of control numbers from your saved records to update or overlay existing Unicorn records. If no 001, 010, 020, 022, or 035 is present in the incoming record, an overlay cannot be accomplished and a new control number will be automatically assigned based on the load argument you requested. If the fields requested in your load argument (ex.: lig) do not exist, an automatic control number is assigned based on the record's system-assigned "catkey". These are recognizable by their "a" prefix. To effect an overlay, the presentation of the control number in the incoming record must exactly match the control number in Unicorn, including punctuation, spaces, and prefixes. Please be careful when uploading, as it is possible to lose records in a database of over one million records quite easily!

Besides the 001, 010, 020, 022, and 035 the other MARC field of consequence in the load programmes is the 949, described earlier in this document. This field enables the creation of item records based on the call number, barcode, item type, home location, library, volume and copy number present. The addition of 949 allows the cataloguer to complete all cataloguing activities except authority work outside of Unicorn, uploading the edited records to the Library's OPAC. Since not all loaders enable this field, descriptions of each loader are given below.

Use the following to guide you through the import of a file of MARC records from drive A: Default responses (e.g. hitting the ENTER key will automatically select this option) are shown using square brackets. For verb responses, typing in the first few characters of the word is sufficient to invoke the selection. Where several possible responses are indicated, an explanation of the differences between selections follows this step-by-step guide. To retrace your steps at any point in the IMPORT_EXPORT or REPORT utilities type or [exit]. The model used below is that of the report named loadbiblio. Some of the other loaders will also ask you to select the default library and one of two other questions. Since these are self-explanatory, they are not included here.

IMPORT_EXPORT

  - Select what you want to work with:  3) marcimport

  - Select the type of files to use:  2) bibliographic data

  - Select operation:  import

  - Select the type of file:

     1) Bibliofile- use for all non-OCLC USMARC records
     6) Other- use for OCLC records to truncate "ocm" to "o"; this passes
	OCLC records through an oclcprep report prior to load

  - What are you importing from?  2) pcdiskette

  - Enter the pcdiskette drive name:  [A:]

  - Enter the pcdiskette file name:              

     filename.ext   This is the name by which you saved the file
     on your diskette.  For Bibliofile records this will always
     be com.cp unless you rename the file after creation.  For
     OCLC records it is the Passport for Windows default filename,
     which usually ends with the ".dat" file extension (orders.dat,
     robin.dat, etc.). Most other utilities will allow you to name
     the file when saved.

  - Enter the imported file name:           

     filename.ext   This is the name you will use to save the
     records to the Unicorn server; it should be recognizable as
     to contents and source; e.g. scores.cp, nfldmaps.suz, ooa113, etc.

  - Is pcdiskette ready?  2) yes

At the end of this process, [esc] or exit your way back to the
GUI interface

REPORTS

For OCLC records

For LaserQuest, Bibliofile, or BookWhere? records see instructions below. To create item records for original uploads with 949s see below.

  - From the dropdown menu, select "Utilities", then "Report"

  - Select what you want to use: (1) Running reports

  - Select operation: Create

  - Select a group for scheduling: (10) Marcimport

  - Select a report from the MARCIMPORT group: (1) Bibload

  - Select operation: [REVIEW]

  - Select the phase to review and/or modify the answers: (3) Selection

  - Load error records for review?: [NO]

  - Library name: QEII (or whatever library you are loading for)

  - Load mode: Create (if new; update/create if some are new, some are
	meant to overlay previously existing Unicorn records such as
	skelrecs)

  - If you chose an UPDATE option, answer YES to the Question: "Should
	update happen only if DATE CATALOGED is never?"  This prevents
	you from overlaying an already catalogued title. If, on the other
	hand, you want to overlay a previously catalogued title (perhaps
	to bring in an updated CONSER record for a dead serial), answer
	NO.

  - Set date catalogued: (2) no change for orders or either (1) or (3)
	for finished original cataloguing records such as recon,
	theses, or HonsDiss being uploaded for the first time.

  - Select the file to load: pick the number of your file

  - Load file format: [MARC]

  - Preprocessing method: [NONE] since we don't preprocess records here

  - Default record format: 4 (MARC) if loading books, otherwise choose the
	format appropriate to the materials being loaded.  In most cases
	the record leader will tell Unicorn what format the record is
	anyway.

**- Title control key: 12 (OCLC# in 001)[END]

  - Default holding code: [DEFAULT] unless you are wanting to generate
	item records at the next step.  If so, the holding codes used
	here are the same as those used in receiving materials when you
	tell Unicorn to "load" the record upon receipt. "MAIN" will
	generate an ITEM TYPE of CAT-STACKS and a LOCATION of IN-PROCESS.
	Note that selecting a holding code will apply that code's 
	default TYPE/LOCATION to all titles in the batch.  If this is
	problematic, you will probably want to select (1) at the next
	step.

  - Copy processing: 1 for orders, 2 for post-receipt (if you want item
	records generated with the ITEM TYPE/LOCATION of the holding
	code selected at the previous step. If you wish to load recon
	or original catalogue records from diskette, go to the section
	below on copy processing for original records.

  - Strip unwanted entries: [NO]

  - Remove MeSH entries: [NO]

  - Call number load rules: this should read "AUTO,999,All
	Subfields,First". If it says "LC,50,All Subfields,First" 
	Follow these steps:

	Call number load rules: (2) Modify
	Modify a rule for call number: (1) LC,50,All Subfields,First
	Modification choice: (1) Classification
		Classification scheme: (5) AUTO
	Modification choice: (2) Entry ID
		Entry ID: 999
	Modification choice: [END]
	Call number load rules: [END]
	Call number load rules: [END]

  - Update call number: NO

  - Update bib record: [YES]

  - Update price: [YES]

  - Update pub year: [YES]

  - Set ITEM TYPE from GMD: [NO]

  - Immediate indexing: YES or [NO] depending on whether you want to
        be able to search them by author/title in the shadow catalogue
        or just find them by control# in the regular catalogue

  - Review: [EXIT]

  - Schedule: [SCHEDULE]

  - Give it a name you can remember

  - Select an output device: [NONE] unless you want to print the report
        for this job

  - Enter login name: [ENTER] unless you want to email yourself the 
        report for this job

For LQ/Bibliofile/BookWhere? records

  - From the dropdown menu, select "Utilities", then "Report"

  - Select what you want to use: (1) Running reports

  - Select operation: Create

  - Select a group for scheduling: (10) Marcimport

  - Select a report from the MARCIMPORT group: (1) Bibload

  - Select operation: [REVIEW]

  - Select the phase to review and/or modify the answers: (3) Selection

  - Load error records for review?: [NO]

  - Library name: QEII (or whatever library you are loading for)

  - Load mode: Create (if new; update/create if some are new, some are
        meant to overlay previously existing Unicorn records such as skelrecs)

  - If you chose an UPDATE option, answer YES to the Question: "Should
        update happen only if DATE CATALOGED is never?"  This prevents
	you from overlaying an already catalogued title. If, on the other
        hand, you want to overlay a previously catalogued title (perhaps
	to bring in an updated CONSER record for a dead serial), answer
        NO.

  - Set date catalogued: (2) no change for orders or either (1) or (3)
	for finished original cataloguing records such as recon,
        theses, or HonsDiss being uploaded for the first time.

  - Select the file to load: pick the number of your file

  - Load file format: [MARC]

  - Preprocessing method: [NONE] since we don't preprocess records here

  - Default record format: 4 (MARC) if loading books, otherwise choose the
        format appropriate to the materials being loaded.  In most cases
	the record leader will tell Unicorn what format the record is
        anyway.

**- Title control key: for monographs (8) (LCCN in 010- do not use LCCN in
	001 or Melvyl records will be rejected and you will end up with OCLC
	and U of Alberta numbers having an "L" prefix), (4) (ISBN).  For
	serials use (6) ISSN. [END]  Any other control number (001, 035,
	etc.) will generate a "a + catkey" system-generated CONTROL
	NUMBER.

  - Default holding code: [DEFAULT] unless you are wanting to generate
	item records at the next step.  If so, the holding codes used
        here are the same as those used in receiving materials when you
	tell Unicorn to "load" the record upon receipt. "MAIN" will
	generate an ITEM TYPE of CAT-STACKS and a LOCATION of IN-PROCESS.
	Note that selecting a holding code will apply that code's
	default TYPE/LOCATION to all titles in the batch.  If this is
	problematic, you will probably want to select (1) at the next
	step.

  - Copy processing: 1 for orders, 2 for post-receipt (if you want item
        records generated with the ITEM TYPE/LOCATION of the holding
	code selected at the previous step. If you wish to load recon
        or original catalogue records from diskette, go to the section
	below on copy processing for original records.

  - Strip unwanted entries: [NO]

  - Remove MeSH entries: [NO]

  - Call number load rules: this should read "AUTO,999,All
	Subfields,First". If it says "LC,50,All Subfields,First" use
	(2) Modify to change it.

	Call number load rules: (2) Modify
	Modify a rule for call number: (1) LC,50,All Subfields,First
	Modification choice: (1) Classification
	Classification scheme: (5) AUTO
	Modification choice: (2) Entry ID
	Entry ID: 999
	Modification choice: [END]
	Call number load rules: [END]
	Call number load rules: [END]

  - Update call number: NO

  - Update bib record: [YES]

  - Update price: [YES]

  - Update pub year: [YES]

  - Set ITEM TYPE from GMD: [NO]

  - Immediate indexing: YES or [NO] depending on whether you want to
	be able to search them by author/title in the shadow catalogue
        or just find them by control# in the regular catalogue

  - Review: [EXIT]

  - Schedule: [SCHEDULE]

  - Give it a name you can remember

  - Select an output device: [NONE] unless you want to print the report
       for this job

  - Enter login name: [ENTER] unless you want to email yourself the
       report for this job

Copy processing for original records with 949

For uploading recon, original, or downloaded and finished utility (OCLC, LQ, Bibliofile or BookWhere?) records created in MARC format with call number and item information in 949, follow the instructions given above for the source of the records, with the following exceptions:

**- Title control key: for Recon records, the defaults here should be
	the same as for LQ/Bibliofile/BookWhere? records, namely
	(8) LCCN in 010 and (4) ISBN in 020 for	monographs, or (6) 
	ISSN for serials.  

	For records exported from LQ, Bibliofile or BookWhere? and
	edited locally with 949, treat the same as Recon books above.

	For records exported from OCLC which have been catalogued
	with 949s, choose (12) OCLC number in 001.

	For records where you are trying to effect an overlay of an
	existing MUN record on the basis of an old RSN or other control
	number you must use (10) Local control number in 001 to effect an
	overlay.  The numbers much match exactly, including inputting any
	alphabetic characters, spaces or punctuation present in the
	control number of the record to be overwritten.  

	If no control number occurs in the field you have specified here,
	your new records will be assigned a system-generated "a + catkey" 
	number.

  - Copy processing: (3) Create and Update copies using SIRSI holdings
	statement

  - Call number load rules: this should read "LC,949,Subfield a,First"
    unless you are loading records for another class scheme, such
	as ALPHANUM for CMC, GovDocs or Micro Fiche, AUTOTHES for
	MUN theses, or AUTOHON for Honours Dissertations.  Simply 
	substitute the other class scheme in place of 949 above.
	Choose Entry ID: (230) 949

  - Update call number: [YES]

How to find your records in Unicorn immediately after loading
(if you have not turned on Immediate Indexing)

To find your unindexed records in Unicorn immediately after loading and before ADUTEXT is run, you need to be able to predict what the control number or call number will be so that searching using DISPLAY1 ITEM is possible or you can use a barcode if you used 949 to carry this information over. Here is a guide to help you.

CONTROL NUMBER

The control number assigned to incoming records can be predicted based on the load arguments you gave under "Title control key" above. Base your search on the answers to the Title control key question below:

• OCLC number in 001: These will have the OCLC number as their control number with the "ocm" replaced with a prefix of "o".
• LCCN in 010: These will have the LC card number from the 010 appear with a prefix of "l" (lower case "L"). If the 010 has an alpha prefix (serial records and some earlier music numbers have such prefixes) these will appear after the "l" and before the number. UKM numbers in 010 may result in the prefix "lgb".
• ISBN in 020: Control numbers for these will be the ISBN from the first occuring 020 in the incoming record with a prefix of "i".
• ISSN in 022: Control numbers for these will be the ISSN from the |a of the 022 with an "s" prefix.

CALL NUMBER

• LC call number: This will be used exactly as it appears in the 050 only if you selected "LC,50,All Subfields,First" under Call number load rules.
• Local call number from 949: Search exactly as presented in loaded 949(s)

BARCODE: may be searched directly using either DISPLAY1 ITEM or EDIT ITEM screens.