## Chapter 6 — Twelve characters are not the digest ### Lesson map Lesson 06. Map, example, EX052. ```mermaid flowchart TD H["sha256 of the bytes"] --> X["hexdigest, 64 characters"] X --> S["slice the first 12"] S --> Q{"Is the slice the key?"} Q -->|No| K["The lookup used the full digest"] ``` The page shows a prefix. The dict key is the whole fingerprint. ### Logical drawing ```text computed hexdigest, 64 hex characters printed the first 12 only lookup the full digest, before the slice three prefixes on the page 0edf30e2fa30 3e5d3b83407c ffcab6e8630a ``` ### Learning outcomes After this chapter, you should be able to: 1. Separate `hexdigest` from the slice `h[:12]`. 2. Say which value is the dict key. 3. Refuse a note that treats a prefix as the stored fingerprint. ### 6.1 The slice is a reading aid. The key is the full digest. `hashlib.sha256(data).hexdigest()` returns 64 hexadecimal characters. That length is the whole SHA-256 result written in hex: 256 bits, four bits to a character. The print uses `h[:12]`, so the page keeps the first 12 characters and drops the other 52. Twelve hex characters are 48 bits. They are enough to tell these three rows apart on a classroom page. They are not the value the script stored for the lookup. The lookup happens before the slice is a problem for a careful reader, because the code computes `h` once and then uses that same `h` twice. `known_bad.get(h, 'no match')` searches for the 64-character string. The f-string then prints `h[:12]`. A classmate who copies only the visible prefix has not copied the key. Two different files can share a 12-character opening in principle. This run's three openings differ, and that fact does not promote the prefix into an identity. Kornblum (2006) describes a way to notice files that are almost the same. This script does not do that. It asks whether the full digest is exactly the one key. A one-byte change in the byte string would miss. The chapter does not print the hidden 52 characters, because the lab did not print them. | Piece | Role in this file | |---|---| | `hexdigest()` | 64-character fingerprint | | `h[:12]` | The characters on the page | | `known_bad.get(h, ...)` | Lookup of the full fingerprint | | The other 52 characters | Computed, not printed | ### 6.2 Worked example | Offered reading | What the file does | Decision | |---|---|---| | `3e5d3b83407c` is on the installer row | First 12 of that row's digest | Keep | | The key is those 12 characters | The key is the unprinted full digest | Refuse | | "Paste the prefix into a new list as the key." | The list was built from `hexdigest()` | Refuse | | The three prefixes differ | True of this run, and still a prefix | Keep | > **Synthetic case.** The lab page shows prefixes only. ### Errors and limits Do not invent the hidden characters in order to look complete. ### Apply - **EX052 — Prefix.** Each row shows 12 hex characters. Is that slice the dict key, and does this page print the other 52 characters? *Rubric:* 3 = no; `hexdigest` is 64 characters; the lookup uses the full digest; the page prints `h[:12]` only; do not fill in the rest;; 2 = the key is the full digest and the rest is unprinted;; 1 = the prefix only;; 0 = none. ### Continue studying — lesson 06 (chapter 06 of 20) [← Chapter 05](#ch-05) · [Course chapter index](../SUBJECTS.md) · [Chapter 07 →](#ch-07) Continue to the next authored lesson. ## Worked answers **EX052.** No. `hexdigest` is 64 characters. The lookup uses that full digest. The page prints the first 12 only. Do not fill in the rest.