.Q.fs[x;y]
.Q.fsn[x;y;z]

Where

x is a unary value
y is a filepath
z is an integer

loops over file y, grabs z-sized lumps of complete "\n" delimited records, applies x to each record, and returns the size of the file as given by hcount. This enables you to implement a streaming algorithm to convert a large CSV file into an on-disk database without holding the data in memory all at once.

.Q.fsn is almost identical to .Q.fs but takes an extra argument z, the size in bytes that chunks will be read in. This is particularly useful for balancing load speed and RAM usage.

.Q.fs is a projection of .Q.fsn with the chunk size set to 131000 bytes.

For example, assume that the file potamus.csv contains the following:

Take, a,   hippo, to,   lunch, today,        -1, 1941-12-07
A,    man, a,     plan, a,     hippopotamus, 42, 1952-02-23

If you call .Q.fs on this file with the function 0N!, you get the following list of rows:

q).Q.fs[0N!]`:potamus.csv
("Take, a,   hippo, to,   lunch, today,        -1, 1941-12-07";"A,    man, a,..
120

.Q.fs can also be used to read the contents of the file into a list of columns.

q).Q.fs[{0N!("SSSSSSID";",")0:x}]`:potamus.csv
(`Take`A;`a`man;`hippo`a;`to`plan;`lunch`a;`today`hippopotamus;-1 42i;1941.12..
120