Build a CLI command that prints each distinct line of a file once while preserving first-seen order. The graded follow-ups push you toward handling a file too large to fit in memory and preserving global output order when partitioning or spilling data to disk.

Requirements - Implement a command-line tool that reads a file and outputs each distinct line once (a global de-duplication utility, not merely adjacent-line collapsing). Preserve the order in which unique lines first appear. It must run end-to-end in your own IDE / on your own machine. - A working in-memory solution typically combines a set/hashmap to track seen lines with a queue or simple pass to preserve output order — but getting there is mostly about decomposing the steps and discussing each with the interviewer, not about a hard algorithmic trick. - Follow-up 1 — file too large for memory: the input file no longer fits in RAM. Discuss an external approach: stream the file, hash lines and partition/spill to disk (e.g. bucket by hash into multiple temp files), then de-duplicate per bucket. - Follow-up 2 — preserve global order: explain how the external-memory design still emits unique lines in their original first-seen order even though records are partitioned. Continue reasoning about how your chosen language handles file-system access and memory loading of large inputs. Notes - There is no provided online coding environment . You write, build, and run locally; the round leans on real file-system I/O, so make sure you can run code on your machine before the interview. - At the end you package and upload your code (a Google Form has been used for submission). Budget time accordingly. - Pace yourself for the two follow-ups — several candidates ran out of time on the second one. When preparing, focus on your chosen language's file-system and large-input/memory-loading APIs; that is the direction the follow-ups push. - Talk through each step with the interviewer and converge on one solution together — the round rewards clear incremental problem-solving over a clever one-liner.

Implement `uniq` — Unique Lines in a File — Vanta Interview Question

Requirements

Implement a command-line tool that reads a file and outputs each distinct line once (a global de-duplication utility, not merely adjacent-line collapsing). Preserve the order in which unique lines first appear. It must run end-to-end in your own IDE / on your own machine.
A working in-memory solution typically combines a set/hashmap to track seen lines with a queue or simple pass to preserve output order — but getting there is mostly about decomposing the steps and discussing each with the interviewer, not about a hard algorithmic trick.
Follow-up 1 — file too large for memory: the input file no longer fits in RAM. Discuss an external approach: stream the file, hash lines and partition/spill to disk (e.g. bucket by hash into multiple temp files), then de-duplicate per bucket.
Follow-up 2 — preserve global order: explain how the external-memory design still emits unique lines in their original first-seen order even though records are partitioned. Continue reasoning about how your chosen language handles file-system access and memory loading of large inputs.

Notes

There is no provided online coding environment. You write, build, and run locally; the round leans on real file-system I/O, so make sure you can run code on your machine before the interview.
At the end you package and upload your code (a Google Form has been used for submission). Budget time accordingly.
Pace yourself for the two follow-ups — several candidates ran out of time on the second one. When preparing, focus on your chosen language's file-system and large-input/memory-loading APIs; that is the direction the follow-ups push.
Talk through each step with the interviewer and converge on one solution together — the round rewards clear incremental problem-solving over a clever one-liner.

Requirements

Notes

Comments