Problem:

• 100000 line CSV file.
• header row includes unnecessary type definitions in it (i.e. "header1:String,header2:Decimal,header3:INT32")
• CSV is too large to open in Numbers/Google Sheets for simple find/replace

Solution:

• get only the header row

head -n 1 orig.csv > headers.csv

• open headers.csv, use find and replace to delte all the ":String" cruft

• copy the revised row into a text editor, put quotation marks around all the strings, copy again

• get everything but the header row

tail -n +2 orig.csv > headless.csv

• insert the new header row into the csv file

( echo "headers","as","comma","delimited","strings" ; cat headless.csv ) > final.csv && rm headless.csv

• convert csv file into a json file

csvtojson final.csv > converted.json

Tools:

• bash (MacOS command line/terminal)

• a spreadsheet app (MacOS Numbers.app)

• csvtojson npm package