Put one word pair per line in a plain text file, separated by a tab, and Paira turns it into a study set. You do not have to tell it which separator you used. It tries five of them, keeps the one that reads the most lines, and shows you the count of lines it could not parse before anything is saved.
The format, in one line
A line is a term, a separator, and a definition. That is the whole requirement. Written with a spaced hyphen as the separator, a working file looks like this:
- ephemeral - geçici
- tenacious - inatçı
- ubiquitous - her yerde bulunan
The same content works with a tab, a semicolon, a comma or a colon in place of that hyphen. A tab is the best choice and the examples here only use a hyphen because a tab is invisible on a web page. You do not need a header row, and you do not need quotes around anything.
Which separators does it accept?
Five, tried in this order: tab, spaced hyphen, semicolon, comma, colon. Order matters when two of them tie, and tab wins ties because it is what Quizlet exports by default. Paira counts how many lines each separator splits into at least two non-empty fields and keeps the winner.
That count is the whole decision. If your definitions contain commas but your columns are separated by tabs, the tab still wins, because splitting on commas would produce a mess of short fragments while splitting on tabs produces clean pairs on every line.
Note that a plain space is not on the list. Separating with a single space would make every multi-word definition ambiguous, so the spaced hyphen exists to do that job unambiguously instead.
How do I add an example sentence?
Add a third column using the same separator. Paira reads it as an example sentence attached to that word.
- ephemeral - geçici - Fame is ephemeral.
- tenacious - inatçı - She is tenacious about deadlines.
There is a rule behind this that is worth knowing, because it explains a surprise. A third column is only treated as an example when more than half of your parsed lines actually have three fields. If ten lines have two columns and one line has three, that eleventh line's extra text is joined back onto the definition instead of becoming an example. The majority decides what the file means, not the exception.
What happens to lines it cannot read?
They are skipped, and you are told how many. A line is skipped when splitting it on the winning separator does not produce at least two non-empty fields. In practice that means a line with no separator at all, a line whose first field is empty, or a line where everything after the first field is blank.
Empty lines are not skipped lines. Blank lines and surrounding whitespace are removed before parsing starts, so a file with double-spaced entries does not report half its lines as failures.
If the skipped count is higher than you expected, the usual cause is a mixed file: some lines use a tab, others use a comma. Only one separator wins, so the minority format fails. Open the file, make the separators consistent, and import again.
What if my definition contains the separator?
Extra fields are joined back together rather than thrown away. On a two-column file, everything after the first separator becomes the definition, separator characters included. On a three-column file, everything after the second separator becomes the example. So a definition like "to give up, to abandon" survives a comma-separated file as long as the file is genuinely comma-separated and the first comma is the column break.
The safest habit is to use tabs. A tab almost never appears inside a word or a definition, which removes the ambiguity entirely.
Where the file can come from
Anywhere that produces plain text. A spreadsheet column pair saved as tab-separated text. A word list you kept in a notes app. An export from another flashcard tool. A list a teacher handed out. Paira does not care what made the file, only that each line holds a pair.
Your sets work in any language pair, in both directions, and nothing about the import assumes English on one side.