Note: I use "set" and "batch" interchangeably.

1. Create a folder for the record batch.
2. Rename .mrc file Vendor_mm-dd-yy
3. Open .mrk file
4. Dedupe based on OCLC number
   
   1. Tools→Record Deduplication→Control number→OCLC number
5. Run a Field Count report
   1. Reports→Field Count→Tab delimited→Generate (name it after the MARC file by adding _fieldcount.txt)
   2. Create a spreadsheet and save it as .xlsx
   3. Use the spreadsheet to track work and take notes
6. Skim the report and highlight the following fields
   
   • 006
     
   • 007
   • 040
   • 245 $h
   • 260
   • 264
   • 300
   • 336
   • 337
   • 338
   • 533
   • 538
   • 583
   • 655
   • 773
7. Add the following fields to the spreadsheet
   • 852
   • 876
   • 956
   • XPT
8. Verify or edit as follows (File→Select Records for Edit; Tools→)
   
   1. 006 (Verify if present, add if missing. Can copy and paste from below or Edit→Insert/Edit 006→Electronic Resource)
      

      • 006 m\\\\\o\\d\\\\\\\\

        OR

        006 m\\\\\o\\u\\\\\\\\

   2. 007 (If the 007 is present, verify the first two subfield; disregard any additional subfields. Paste from below if missing)
      • cr\-n-
   3. 040 (Add new $d MUB)
   4. 245 $h (not necessary in RDA records; if missing, there MUST be a 336/337/338 combo)
      • electronic resource in square brackets
   5. 260/264 (one of these must be present; there may be more than one 264)
   6. 300$a (Verify if present, additional information is fine; add if missing)
      • \\$a1 online resource
   7. 336/337/338 (Required if there is no 245 $h; verify if present, add if missing. These fields are allowed even if there is a 245 $h; do not delete these fields if present.)
      • =336  \\$atext$btxt$2rdacontent
      • =337  \\$acomputer$bc$2rdamedia (this field is optional, but recommended)
      • =338  \\$aonline resource$bcr$2rdacarrier
   8. 533/538/583 (Used only for records for DLF Registry of Digital Masters, HathiTrust Digital Library and other digital preservation projects; used with $5)
      • Delete if it's not used as above
   9. 655
      • Delete all 655s with the text "electronic books"
      • Add \7$aElectronic books.$2lcsh
   10. 773
       • Delete
   11. 852

       • (general ebook) \\$bBC-BC$cWWW$x[vendorMM-DD-YYBC]$zClick on Find It button

       • (reference ebook) \\$bBC-BC$cRWWW$x[vendorMM-DD-YYBC]$zClick on Find It button

   12. 876
       • (purchased) \\$j04$p $xMT:EBOOK$zClick on Find It button
       • (PDA) \\$j04$p $xMT:EBOOK$xIPS:PA$zClick on Find It button
   13. XPT (most records will not need this field added)
       • \\$aY$bMUB
   14. 856/956
       • If there is only one 856 on each record, swap $u to 956 40 $u, then add $7BC by editing the subfield data to all. (When using field swap, omit $ from subfield.)
       • If there are multiple 856s, find a common element to the ones we’ll use, and use that to further define the swap fields function. (i.e. umbc or dx.doi) If this creates more than one 956 in each record, skim to see if you can delete any by using words like “proxy” in your search.
9. Convert to .mrc
10. Combine multiple .mrc files using Tools→MarcJoin
    1. Create and place Destination file: Vendor_combined_mm-dd-yy
    2. Select all files to join
    3. Click Process
11. Save a copy and rename for DSS
    1. bc_vendor_additions_MM-DD-YY.mrc (All lowercase)
12. Move the file to BC folder on Dropbox.
13. Submit an Rx
    1. Include the name of file that is in Dropbox
    2. Note that the records are OCLC sourced
    3. Say how many records are in the file (the maximum number for one load is ≈ 15,000 or 30MB)
    4. Ask DSS to load the records in TEST
       1. example: Please load the file bc_asp_03-24-15.mrc into Test. It contains approx. 1700 OCLC sourced records.
14. Once you receive notice that the records have been loaded in TEST, examine the loader report linked in the Rx message
    
    1. Spot check bibliographic records
    2. Spot check Items and HOLS, both in rejected and overlaid records
    3. Spot check Find It links (Note: Rejected records will NOT have a 956 field, so Find It will not work.)
    4. Note any problems and see if they are problems with the records as you sent them. If so, make any corrections to the set and resubmit it on the same Rx. If you can't figure out the problem, ask questions in response to the Rx.
    5. If you are satisfied with the records in TEST, respond to the Rx and request the set be loaded in LIVE.
       1. example: These look good. Can you load them into Live? And can you add the 956s for the rejected records?
15. Once you receive notice that the records have been loaded in LIVE, examine the loader report linked in the Rx message as above.

Submit an Rx for your load request and say that you are adding the file to the BC Dropbox folder (also good to say how many records are in the file; the max number of records is about 15,000, or a file of 30MB). Ask for them to be loaded in Test first so that you can see if everything went as expected. If it all looks good, then you can give me the go ahead to load in Live.

Any records that are rejected won’t have the 956 carry over. For small numbers, go in and manually change them? Work with ITD for larger sets of problems?

Figure out how to track on stats

Deletions

(deletion wording? I've put the file bc_ebldda_deletions 10-4-14.txt in Dropbox. Please delete any UMBC items with collection WWW + IPS:PA and any linked HOLs. Please also delete any 956 fields containing "UMBC.eblib." If there are no other items/HOLs on the record, please delete the bib record.)

Notes:

Convert .dat or .bin to .mrc by changing the extension.

Make sure the entire file is loaded; if it's not loaded, there will be a colored bar across the top of the editor with an option to load the entire file.

Use the Extract Selected records tool -- set the display field to 856, and search for |end| -- which is a delimiter that is placed between fields.  If it's present, there are two fields in the record.