Content Comparison

Weekly record files are downloaded from OCLC Worldshare Collection Manager

1. Log in here: https://usu.share.worldcat.org/wms

2. Metadata --> My Files --> Downloads

3. Be extremely careful when downloading:

   1. Download ONLY the files of with OReillyOnlineForCheryl at the end:

      1. metacoll.UUS.deletes.W20230922.T104438.OReillyOnlineForCheryl.1.mrc

      2. metacoll.UUS.new.W20230922.T104438.OReillyOnlineForCheryl.1.mrc

   2. These may include 4 file types: deletes, new, merges, updates)

NOTE: If you accidentally click Download for a non-OReilly file, OCLC WS will remove it from the My Files list for everybody in THE USU library, and you will need to email the file to Tyler in Acquisitions.

• Files appended with MerrillCazierApproval, Price, or with FIRM/APPROVAL in the middle, should ONLY be downloaded by the designated Acquisitions staff

4. Download deletes and new files

   1. And go ahead and download the OReilly merges and updates files, but don’t load (they don’t work well) just add their names to tracking, and then delete them from your computer 

   2. We process & load just the deletes and new files

6. We will use two different load tables for delete and new files

7. Order of loadingSafari Online OReilly is as usual, important, but is different than most:

• Always load Delete file first (see processing workflow below)

• Load New file second(see separate processing workflow below)

• Load each week's Delete & New before loading other dated files from O’Reilly; do not load weekly files out-of-order

8. In the tracker, copy/paste each file name, in order, to help keep the processing/loading order clear: https://usu.app.box.com/file/1318206846659

Image Added

• It will be easier to keep loading order if the files are downloaded and processed each week (unlike my backlog shown above) – but files can be done late, so long as you keep to the date order

9. OCLC WorldShare will keep about 2-3 weeks of files, so an alternate will need do the O’Reilly downloads (at the least) for prolonged leave times (or you can download them remotely for working on later)

Make sure Delete file is loaded first; then New

Processing Delete File UTF8

1. MarcEditing for 001

• Find: =001 o <--with the usual 2 blank spaces

• Replace: =001 safario

• Save

• Since this file is going to Overlay, do a quick Find all

  • =001 (for QC)

add prefix safari

• Compile into MARC with same name + DONE

or without, if you prefer

2. load with CAUTION Special use for overlay and delete (.overlaytodelete) load table

• Select Use Review Files

• Test (rejected records will have been already deleted)

• then Load (the records that were not rejected will overlay the Sierra records)

Image Added

If there more than 1 or 2 rejected records on no match: copy/paste into Notebook or excel and search the titles in Sierra.

• some may not match because the 001 did not get the prefix of safari

  • if so, search in Sierra Cataloging and manual delete them

    • 001 should have the same OCLC number (may have on, ocn, ocm, as a prefix)

    • no order should be attached

    • item location should be cosaf

    • and have 830 Safari Books Online

3. Make a review file in Create Lists: Copy-->[find load file], then export selected fields to make sure there are no purchased titles in the delete record, before Deleting:

   1. Export Records from Create Lists Load review file: 

   2. Image Added

      item location (79)

   3. order record number (81)

   4. bib MARC (! 830)

   5. bib MARC (! 910)

4. Open in Excell and check values:

• item location: should be cosaf

• order record number: (should be blank-no orders)

• bib record 830 fields: should have bothSafari Books Online and O'Reilly for Higher Education (may also have regular series titles in there as well)

• 910 in bib record should be empty

• If no signs of purchase/order is found, go on to Deleting

4a. Alternate method of checking for purchased titles:

• Create a second list (type: bibliographic; search method: review [your list of items to be deleted])

• Construct your queries using the JSON interface [see below]

• Copy and paste into the window the following the following block of text
  

{"queries": [ { "target": { "record": { "type": "item" }, "id": 79 }, "expr": [ { op": "not_equal", "operands": [ "cosaf" ] } ] }, "or", { "target": { record": { "type": "bib" }, "field": { "marcTag": "910" } }, "expr": [ { op": "not_equal", "operands": [ "" ] } ] }, "or", { "target": { "record": { type": "bib" }, "field": { "marcTag": "830" } }, "expr": [ { "op": "equals", operands": [ "o\u0027reilly for higher education" ] }, "and", { "op": "equals", operands": [ "safari books online" ] } ] } ]}

• Run the search. If any items populate the second list, examine them to learn if they are in fact purchases.

Image Added

5. use FUNCTION-->Delete Records to batch delete the file

• Left dropdown: Review

• Review File: CreateLists Load for Deletes file

• Start, to bring in Records

  • Scroll down to check number of records (minus no match) is correct

  • Delete the listed Bibliographic record AND all attached records

• Delete Records

  • Are you very sure?

    • Yes

• Finish recording in: https://usu.app.box.com/file/1318206846659

• Statistics: record number of records deleted for CAP stats (Batch: Deletes)

6. Clean up Data Exchange; Create Lists review file will disappear on its own

7. finish Recording

8. Take down statistics (number deleted)

Unicode: UTF8

1. edit 001

• Find/Replace

• Find: =001 o (2 spaces between 1 & o)

• Replace: =001 safario

• Replace All

QC: number modified should match number of records

add prefix safari

2. While learning, it’s a good idea to do another find to check that the replace didn’t deform the 001 field

• Find All =001

• scroll down

I like to Save between edits

• Edit prepend to our new prepend

• Find/Replace

old prepend: http://dist.lib.usu.edu/login?url=

Replacement: https://library2.usu.edu/eval_dist/url856.php?url=

there’s no other place in the bib that these combinations could be found, so it’s safe to just do a straightforward find/replace

• copy/paste old into Find:

=856 40$uhttps://dist.lib.usu.edu/login?url=

• copy/paste new into Replace:

=856 40$uhttps://library2.usu.edu/eval_dist/url856.php?url=

QC: same number modified as number of records

QC: do a Find All on =856 to check all went well, if desired

4. edit subfield data - 856|z

• Field: 856 Subfield: z

• make sure Delete Subfield is checked

• Remove Text

• It will ask if you are sure

• Yes

Doing both step 4 & 5 as one step tends to mess up the 856 structure in Edit Subfield Data, so I do them separately

delete supplied |z

5. Edit Subfield Data

• Field: 856 Subfield: z

• Replace with: [updated text]

• New subfield only

• Replace Text

• Save

Available online; click to view. Select “Institution not listed” and enter your USU email address.

Do a Find All for 856, if desired

6. delete fields:

• Add/Delete Field

037

365

655

887

902

936

949 - important! (We need to replace any with our version)

994

7. Find all and delete if needed

• Find: =###

• If found, then

• Add/Delete Field tool

  • Delete Field

773

506

533

538

583

830 Edits

8. add 830 field

• Add/Delete Field tool

  • Add Field

  • Copy/paste text

\0$aSafari Books Online.

9. add another 830

• Add/Delete Field tool

\0$aO'Reilly for Higher Education.

10. add 949

• Add/Delete Field tool

• Save before next step

\\$lcosaf$t20$zn

11. check 245

• Find/Replace

• Find: =245

• Find All

look for non-Roman letter title displaying as <>. with no useful text and delete from file

This find works even better:

• Find: <>.

• Then note the number of records with this problem

• Scroll down to bottom to make sure these are all in the 245 field

• If there are any, the records will need to be deleted from the file, as below:

Deleting individual records

11a. Selecting/deleting MARC Records -->Close MarcEditor:

• Tools → Select MARC Records → Delete Selected Records

• Source MARC File:

  • [OReilly .mrk file]

• Display: 245$a

• Click Import File to bring in file records

• Search: Field F#:245$a <>.

  • Click search icon

    • check results

  • Scroll to check for unmarked

• Delete Selected

  • Select new file name or keep old one--I add a “b”

13. Go back to MarcEditor:

• Find: =245 to check results

• Note down the new number of records for Quality Checking

Or use Reports → Field Count:

13a. Run MARCValidator at this point to look for any malformed records or fields

• Save first, then

• Reports → MARCValidator

14. Compile File into MARC

I like to put DONE or FINAL at the end of the new mrc file