Matching script¶

Note

To best explain this process, we are going to go through the matching workflow for subjects but there are no differences for persons, families, or corporate entities.

Preparation¶

Place a copy of your template CSV (subject-template.csv) and your script export CSV (allSubjects.csv) in the folder helping-scripts in your repository.

Running the script¶

Determine which column in each spreadsheet contain the name of entity. In our example, the name of the subject is in a column called term for subject-template.csv and in a column called title for allSubjects.csv. We want these two column to match, so we are going to quickly rename title with term in allSubjects.csv and save.

After that, we will run the script matchingNames.py. Open your terminal and cd into the helping-scripts folder and type the following.

python matchingNames.py -f subject-template.csv -f2 allSubjects.csv -c term

In the command, -f will always be your template spreadsheet and -f2 will always be your export spreadsheet, while -c will be the name of your shared column header. Type the entire name of the spreadsheets, including any datetime stamp.

Reviewing results¶

The script should produce a new CSV called subject-template-matches.csv which is a copy of your subject-template.csv with rows from allSubjects.csv inserted if the names matched. Reviewing this, any row without additional information did not match an existing entity, but rows with additional information from allSubjects.csv did match based on names. You can also look at the column _merge to determine which entities already exist - left_only means a match was not found while both means a match was found.

Example match CSV

Ad this point, we also recommend briefly visually reviewing both original spreadsheet to see if any matches with slight spelling differences were missed (like Cat and Cats) and manually correcting them if so.

If any subjects matched (aka already exist), delete their rows from subject-template.csv and add their names and URIs to alreadyExistingEntities.csv. This should leave you with a subject-template.csv that only contains new subjects.