Tuesday, July 30, 2013

Checking Out the FamilySearch "Export Search Results" Feature - Fail!

FamilySearch described the "Export Search Results" feature in the post Search Gets Major Feature Enhancements by Robert Kehrer on 22 July 2013.  The description said:

"A new tool visible above the search results for logged in users allows the export of a page of results to a spreadsheet file. This file can be opened in most common spreadsheet applications. The exported file contains the search parameters used for the search and a date and time stamp for reference. The search results are presented in rows, and each column represents a unique piece of indexed data for the results."


"The Export Results button will export the current page of results and is only available when you are logged into FamilySearch.org."

According to several reports I read, the "Export Search Results" to spreadsheet was activated on 29 July 2013.  I wanted to try it out, because it has the potential to capture information for a selected collection in one file which can be very useful for researchers performing a one name study or a one place study.

1)  I went to the FamilySearch "Search" page, signed in, and entered the surname "Seaver" in the search field, clicked the "Exact" match box, and then clicked the "Search" button, and saw the Search Results in the "Records" tab:

There are 46,485 matches for my search in all of the searchable databases, so I need to collect them one database at a time.

2)  To see the results for each database, I clicked on the "Collections" tab on the screen above.  That provided the list of 249 collections that had an entry for "Seaver:"

My goal is to be able to collect the information one collection at a time.

3)  I started with the "Vermont, Vital Records, 1760-1954."  Here is the first screen of results for that collection:

I picked the largest number of records to display on the page - 75 (the other options are 25 and 50).
In the upper right corner of the screen above, there is an "Export Results 1-75" button.  That is where you can create a file that can be read by a Spreadsheet program.

4)  I clicked on the "Export Results" button, and an Excel spreadsheet file (extension .xls) was downloaded to my Download file folder.  I clicked on the file name and saw:

The spreadsheet file has indexed record titles in row 6, and then provides 75 lines of indexed information.

I expanded some of the columns so that I could see the names and other information better:

Wait!!  The spreadsheet starts out with names Warren W. Seaver, Helen Seaver, George M. Seaver ... etc.  

This list is not from the Vermont, Vital Records, 1760-1954 collection (if it did, it would start out with Abbie E. Seaver, Jane A. Seaver, Bertha Lillian Seaver ... etc.).

The spreadsheet reflects the first 75 matches from the "Records" tab results (which are from the 1940 U.S. Census), not from the Vermont collection.  That would be great for search results for a specific name, but doesn't work for a specific surname.

5)  I hope that FamilySearch fixes this problem as soon as possible.  

Having only 75 search results in each Excel file has benefits and drawbacks - I can probably run through one list of 75 in an hour or so.  But having a list of X (in the Vermont case, 1,764 entries, would be more efficient from a file naming standpoint and easier to store.  With the current FamilySearch file naming convention (with a day and time unique identifier), I will have to rename every file of 75 to identify the collection, the search parameter, and a sequential number).  

The spreadsheet file does not seem to include the name of the collection being downloaded.  However, it provides a link to each record summary in the collection.  

There is a "score" in Column 1 of the spreadsheet which ranks the matches by an algorithm.  I'll have to play with the search filters a bit to see how that works!

I will edit this blog post if and when FamilySearch fixes this perceived problem.

The URL for this post is:  http://www.geneamusings.com/2013/07/checking-out-familysearch-export-search.html

Copyright (c) 2013, Randall J. Seaver


Catherine Pendleton said...

Hope they fix this soon. I just tried this out and the same thing happened to me. I was exporting my search results for the Revolutionary War Pension and Bounty Land Warrant Applications collection---I got a list of names from the 1940 census instead.

Anonymous said...


Even before you got to the problem you're really talking about I saw another possible issue. When the file went to your "Downloads" file was that automatic, or the default but you had the choice to redirect it?

I really don't like programs that tell ME where I want to put something (whether downloading to the Downloads folder, or automatically installing on the C drive when I have a P drive set up for programs). If they're just going automatically to Downloads, then maybe that's something else they can fix.

Dave L

Les Hardy said...

Be sure to change the filename before saving.

The default filename contains colons.
A colon is an illegal character in filenames, and should never be used.

This does not bother Linux users, but anyone trying to manage these files in Windows will have problems.

Also, I am currently adding a merge option to GenScriber to allow multiple search results to be joined automatically.

I would be interested to know of any other feature suggestions that would be useful when downloading/opening these files.

Les Hardy

Anonymous said...

Dave L (Anonymous)
That is not an issue. It's not something familysearch.org can fix.
Downloads are saved where you setup your browser to save to.
The option to go to a specific folder or to have the browser ask where to save is your decision.

Not quite sure why you setup a P drive for programs, but there is a purpose to storing programs in the correct folders, which I won't go into here.

If an application insists it wants to install to your default drive, there is usually a good reason.

Les Hardy