Tabular files.
WHAT CSV IS
Plain text, one record per line, values separated by commas.
The most common interchange format for tabular data.
WHY NOT TO PARSE IT YOURSELF
Values containing commas, quotes or line breaks break naive splitting.
Use the library built for it.
WHAT IT HANDLES
Quoting Embedded separators Headers
READING AS RECORDS
The library can return each row as a dictionary keyed by the header.
Far clearer than working with positions.
WHAT TO CHECK
The encoding, which is a frequent source of mangled characters Whether a header row exists Whether the separator is genuinely a comma
THAT LAST POINT
Some systems export with semicolons or tabs.
WHAT TO BE CAREFUL WITH
Values that look like numbers but should stay as text, such as reference codes with leading zeros Dates, which arrive as text and need converting
FOR SPREADSHEET FILES
Libraries exist for reading them directly.
WHEN TO PREFER CSV
Whenever possible. Simpler and more reliable.
WHAT TO VALIDATE
That the row count matches expectations.
