While implementing csv-streamify I found out that there was a strange character (or wrongly encoded file) in the CSV that returned on an dirty column name.
example.csv
weird_char,ok_col,some_emoj馃
1,1,1
The object returning included the exact same string as in the keys.
{
'weird_char': 1,
ok_col: 1,
'some_emoj馃': 1
}
My request is to add the ability to map or clean this column names to get a cleaner object, like
{
weird_char: 1,
ok_col: 1,
some_emoj: 1
}
One solution would be to add a regex replace in https://github.com/klaemo/csv-stream/blob/master/csv-streamify.js#L56, example
if (state.lineNo === 0) {
state._columns = state._line.map(col => col.replace(/[^a-zA-Z0-9_]/g,''))
state.lineNo += 1
reset()
return
}
A nicer alternative would be to allow to pass a function as value for columns
const csv = require('csv-streamify')
const parser = csv({
columns: (cols) => cols.map(col => col.replace(/[^a-zA-Z0-9_]/g,'')),
objectMode: true,
})
While implementing
csv-streamifyI found out that there was a strange character (or wrongly encoded file) in the CSV that returned on an dirty column name.example.csv
The object returning included the exact same string as in the keys.
My request is to add the ability to map or clean this column names to get a cleaner object, like
One solution would be to add a regex replace in https://github.com/klaemo/csv-stream/blob/master/csv-streamify.js#L56, example
A nicer alternative would be to allow to pass a function as value for
columns