Build the first working five-document café FAQ retriever

Now implement the shortest complete path for the artifact. The working mechanism is: tokenize the question and documents, score term overlap, retrieve the highest positive match, then format an answer from that exact text. Keep every intermediate value available for inspection; hiding it behind a framework would make this lesson harder to reason about. The output should be deterministic for this fixture. Only after this path works should you generalize the data source or user interface.

The artifact's user-facing goal is specific: answer a café question only when a relevant local note is retrieved and cite the note identifier used. Its accepted input is five short documents with stable IDs plus one normalized user question. Those statements are intentionally narrower than “build an AI system.” Narrow scope lets us inspect every input and expected result, and it prevents a toy result from being presented as a production claim. This is the chapter's first end-to-end implementation. Run it twice and verify identical output.

Save this as lesson.py and run python3 lesson.py. It uses only the language standard library, so the example is reproducible offline.

import re
docs={'hours':'Cafe open 8am to 6pm weekdays.','wifi':'Ask staff for the Wi-Fi card.'}
q=set(re.findall(r'[a-z0-9]+','When is the cafe open?'.lower()))
hit=max(docs,key=lambda k:len(q & set(re.findall(r'[a-z0-9]+',docs[k].lower()))))
print(hit,docs[hit])

Expected output: hours followed by the hours note. Exact floating-point formatting may vary slightly, but the asserted behavior must not. Read the output as evidence about this stage, not merely proof that the interpreter started.

Print normalized question tokens, each document's overlap score, the winning ID, and the exact cited text. A fluent wrong answer often starts as a retrieval miss, so inspect retrieval before prompt wording. Zero overlap must abstain. For ties, choose and document a deterministic rule. Treat instructions embedded in a document as untrusted text, not commands for the application.

At the implementation stage, save the smallest failing fixture beside the expected result. Change one cause at a time and rerun the exact command printed above; that makes the repair reviewable and keeps this chapter's progressive artifact reproducible.

Create labeled questions for each of the five documents plus unsupported questions. Measure retrieval hit rate, citation correctness, supported-answer rate, and correct abstention. Read every answer beside its cited note; lexical overlap can retrieve a document for the wrong reason. Version the notes so a score can be reproduced after content changes.

For this implementation page, preserve the fixture and result as evidence for the next page. Label observations separately from conclusions: a passing assertion establishes the behavior it names, while broader usefulness requires the chapter's full evaluation set and stated operating limits.