Left Menu Right Menu

This procedure was put together by Bob Thomas to update the community on the issues.

If you can create a set of records in Alma, you can use the combine sets functionality to make sure those records are not included in a separate set of records. In our current efforts to clean up bib records in Alma that don’t have inventory attached, we need to be able to create sets of bib records and make sure those sets do not include records that are part of a boundwith set of records. These instructions show one method of creating a set of boudwith records so they can be excluded from other sets.

The boundwith record structure in Alma uses a ‘host’ bibliographic record (the bib record with the holding and item record attached) which contains 774 fields which point to the bib records for each title that are contained in the boundwith item. Alma uses the MMS IDs that are stored in the $w subfield of the 774 fields to create the virtual links in Primo that link the title bibs to the inventory (i.e., the host item). Here are some fields taken from a host bibliographic records in Alma:

001  9983196820001453

245  00 |a Host bibliographic record for boundwith item barcode i1857042

774  1_ |t Methods of estimating recharge to the Floridan aquifer in northeast Florida / by G.G. Phelps ; prepared in cooperation with city of Jacksonville, Florida |w 9984636650001453

774  1_ |t Water-Resources investigations / U.S. Geological Survey. |w 9982433750001453

The record above points to two bib records, each describing one of titles included in this boundwith. The first step in creating a set of all a library’s boundwith records is to create a set of bib records that contain MMS IDs in 774 fields. The proposed search strategy will included all bib records in the set that include a “99” in any value in the “Other System Number” index. This includes 774 $w with MMS IDs, but is also includes any other “Other System Number”, such as 035 fields, that also contain “99”. 

Note: An alternate search strategy, to just create a set of all bib records that contain “Host bibliographic record” is not used as Alma doesn’t require that host bibliographic records contain that string. The title in the host bibliographic record can be changed, and Alma allows directly linking between a title bib record in a boundwith, and other title bib records in the boundwith, without using a separate host bibliographic record.

Since our original boundwith set will contain records we don’t need, the bib records in the set are exported and MarcEdit is used to extract those records with MMS IDs in 774 $w fields from the other, undesired bib records. Next, MarcEdit is used to extract the MMS IDs from the 774 $w and 001 fields, so that both the host bibliographic records and the linked title records will be included in the final set of boundwiths.

Although we now have a file that contains the MMS IDs of all the records associated with the boundwiths, the file is not in the format needed to use in Alma to create a set. Therefore, MS Word and Excel are used to reformat the file (into a file with a single column of all the MMS IDs) and to get rid of any duplicate MMS IDs.

Finally, a set of all the records associated with boundwiths will be created in Alma based on the file of MMS IDs created above.

Step 1 – Create a set of bib records that might be boundwith records, either the host bibliographic records or the records that the host bib records link to.

Host bibliographic records don’t have to use the default 245 field “Host bibliographic record for boundwith item barcode . . .” Alma allows a bib record for a title in a boundwith volume to be directly linked to other title bib records in that volume. Additionally, the records that a host bib record links to have no indication that they are part of a boundwith relationship. The key is that the host bibliographic records contain the MMS ID for each linked record in individual 774 $w subfields. So the first step is to find all records in Alma that contain a 774 $w subfield. Not easy to do, since Alma does not allow searching for specific fields or subfields. So this is one workaround.

Create an All Titles set where All titles (Other System Number contains keywords "99*"). And save the resulting set with a file name (I used “Boundwiths Step 1”). Alma online help shows that the 774 $w subfield is included in the All Titles Other System Number index. Although the search ‘Other System Number starts with "99"’ should work, it does not currently work in my environment (no records returned; Salesforce case submitted).

Step 2 – Export the set of bib records created in step 1.

Go to ‘Run a job’ and use the ‘Export Bibliographic Records’ standard job. Run it against the set you created in step 1, using Output format = MARC21 Binary. 

When the job has completed, go to the job history tab (in Monitor Jobs), and on the line for your export job, select Actions -> Report. The report will have a hyperlink with a file name something like “BIBLIOGRAPHIC_8987926080001453_1.mrc”. Click on the file name link and on the next screen select Actions -> Download and save the file of bib records on your computer.

Step 3 – Extract the bib records that contain 774 $w subfields from the file downloaded in step 2.

Open MarcEdit and select Tools (along the top row) -> Select MARC Records -> Extract Selected Records. Change ‘Display Field’ to ‘774$w’, click on the folder for the ‘Source MARC File’ and select the file downloaded in step 2, and click the ‘Import File’ button.

Next, click on the “Display Field (774$w)” column heading to sort by that column by the values in the 774 $w subfield. Check the ‘Retain Checked Items’ box, click the top line with a value in 774$w (not the ones that start with (CKB)), hold down shift key and click the last line with a value in 774$w. If clicking on the line doesn’t check the appropriate boxes, click on one of the boxes (while the lines are still highlighted), and all the lines should be checked. Check to make sure all lines with a value in 774$w (other than CKB) are now checked.

Next, you need to create a file of the records with a value of interest in the 774$w subfield. To do this, click the ‘Export Selected’ button and tell MarcEdit where to put the file. 

Finally, you need the convert the MarcEdit ‘.mrk’ file to a ‘.mrc’ MARC file. Go to MARC Tools (the big icon – not the top row), choose MarcMaker, select the file you just created as the input file (it will have a ‘.mrk’ extension), select the output file and allow MarcEdit to fill in the output file name, and click on the ‘Execute’ button.

Next, you need the convert the MarcEdit ‘.mrk’ file to a ‘.mrc’ MARC file.

Step 4 – Extract the MMS IDs from the 001 fields and 774 $w subfields in the file of records created in step 3.

In MarcEdit, select Tools (on the top row) -> Export -> Export Tab Delimited Records. For the top file name box, click on the folder and select the ‘.mrc’ file you created in step 3. Then click on the folder next to the second file name box and provide a file name for the ‘.txt’ file that will be created as part of this process. Leave the default values for the ‘Select Field Delimiter’ and the ‘In field delimiter’ fields and click on the ‘Next’ button. 

Enter “001” (without quotes) in the ‘Field’ box and click the ‘Add Field’ link. Next, enter “774” in the ‘Field’ box and “w” in the Subfield box, then click the ‘Add Field’ link. Finally, click the ‘Export’ button.

Step 5 – Reformat the file of MMS IDs created in step 4.

Open the file of MMS IDs created in step 4 in MS Word. Replace “ (quote) with nothing (remove all quotes) and click the ‘Replace All’ button. Replace ^t (tab) with ^p (carriage return) and click the ‘Replace All’ button. Replace ; (semicolon) with ^p (carriage return) and click the ‘Replace All’ button.

Finally, replace the top lines (001 and 774$w) with a single line that contains only “MMS ID” (without the quotes). Save the file and close MS Word.

Open the ‘.txt’ file from the previous step in in Excel. When opening the ‘.txt’ file in Excel, in step 3 of the ‘Text Import Wizard’, change the ‘Column data format’ selection from ‘General’ to ‘Text’ (this prevents Excel from displaying pure number values like MMS IDs in scientific notation). Next, select the entire spreadsheet and use Data -> Sort to sort the MMS IDs. Next, use Data -> Remove Duplicates to retain only unique MMS IDs. Finally, click save the file.

Step 6 – Split the file of MMS IDs.

If your file of MMS IDs contained less than 5,000 values, you can skip this step. If your file contained more than 5,000 MMS IDs, you have to split the file into multiple files, as when creating sets based on MMS IDs (which is the next step), Alma can only accept files of 5,000 or fewer MMS IDs.

Open the reformatted file of MMS IDs from step 4 in Excel. Save as different name (e.g., BoundwithsFile1).
File #1 – remove all lines 5,001 and greater, and save the file.

Open the original reformatted file of MMS IDs from step 4 in Excel. Save as different name (e.g., BoundwithsFile2).
File #2 –remove lines 2-5000 and then 5,001 and greater, and save the file.

Open the original reformatted file of MMS IDs from step 4 in Excel. Save as different name (e.g., BoundwithsFile3).
File #3 –remove lines 2-9,999 and then 5,001 and greater, and save the file.

Continue this process until you have all your MMS IDs in files of 5,000 or less.

Step 7 – Create sets of bib records in Alma based on the files of MMS IDs.

In Alma, go to Manage Sets and click on the ‘Add Set button’ and click the ‘Itemized’ option. Name the set and set content type = All Titles. Browse for your file of MMS IDs (or your first file) and click the 
‘Save’ button.

If you have more than one file of MMS IDs, repeat the above process for each file, giving each file a different file name.

If you had multiple files of MMS IDs, you then must combine all sets of 5,000 into a single set. To so do, go to Manage sets, and for your first file, select Actions -> Combine sets. Modify the resulting merged file name, if desired (I used “Boundwiths File 1-2”), change the ‘Operation’ to “Or”, click on the blue arrow at the right end of the ‘With’ box, select your second file of MMS IDs, click the ‘Select’ button, then the ‘Submit’ button. Repeat this process, adding an additional set, until all your sets have been combined into one larger set (I named my final set “Boundwiths”).

Step 8 – Add a bib record marker for your boundwiths.

This step is entirely optional. The set created in step 7 should be a set containing all your boundwith records – both the host bibliographic records and the linked/related records. However, the set is an itemized set, so any boundwith cluster of records created in the future will not be added to this set. To make future maintenance (possibly) easier, I added a marker to each bib record in my set of boundwith records. If you want to do so, here’s how.

Select a local 9XX field to use for a marker (we had previously defined 965 as our field to use for any load or maintenance markers).

In the Metadata Editor, create a normalization rule to add a boundwith marker. My normalization rule was:

rule "Add Local Field"


   (not exists "965.a.boundwiths")


   addField "965.a.boundwiths."

   addSubField "965.9.LOCAL" if (not exists "965.9.LOCAL")


If you haven’t already, create a job to run the normalization rule. Alma -> Resource Configuration -> General -> Processes, click the ‘Add Process’ button, set Business Entity = Bibliographic title, and Type = Marc 21 Bib normalization, click the ‘Next’ button, add name and description, click the ‘Next’ button, check MarcDroolNormalization and click the ‘Add to Selection’ button, click the ‘Next’ button, set the Drools File Key = [your normalization rule], click the ‘Save’ button.

Run your normalization rule. Alma -> Run a Job, check the job that runs your normalization rule, click the ‘Next’ button, select your file of boundwiths, click Next, Next, Next, and finally Submit.

Checking for broken boundwith links.

A simple way to check for bound with links is to do the following:

For each file created in “Step 7” above, go to the “Members” of the set and click on the “Tools” button. This currently only works for files of 5,000 or fewer records, so if you have more than 5,000 boundwith records, you must use the separate files created in “Step 6” above (which were used to create the Alma sets in “Step 7”). It takes Alma quite a few minutes to create the spreadsheet, so don’t do this step when you’re in a hurry.

In the resulting Excel spreadsheet, a blank cell in column “P” indicates a problem with a link (not necessarily a broken link).

Software: Alma

Current phase: NA Written by: Bob Thomas

Approved by: NA
Last updated: 12/30/2015
Staff Contact: Cassie Schmitt Nature of last update: NA

Document History: First draft written on 2/19/2015.