discord: resolve the city people actually typed #62

Merged

emi merged 1 commit from fix/discord-city-resolution into dev

2026-08-02 01:18:50 +00:00

admin_emi commented

2026-07-24 21:48:34 +00:00

Owner

Reported live from #general:

@Thermograph Vilnius, Lithuania
I don't track a city called 'Vilnius, Lithuania' yet. Try a major city name.

Vilnius has been in cities.json the whole time, with country: "Lithuania" sitting right there in the record. _resolve_city matched the query against name alone, exact-or-prefix — so the comma form could never match anything, and the country field was never consulted at all.

The same exchange showed the two neighbouring failures:

Vilinus — one transposition — fell through to the same refusal, which reads as we've never heard of Vilnius rather than you typed it wrong.
Tell me about weather in lithuania was quoted back verbatim as a city we don't track, which is what made it look broken rather than merely unmatched.

What it does now

In descending order of confidence: exact name honouring a City, Country/Region qualifier → the same ignoring an unrecognised qualifier (better to grade Paris and name it in the reply than refuse over "Paris, Wherever") → prefix → close-enough name for typos → a city or country named somewhere inside a sentence, so "weather in lithuania" lands on Vilnius. Accents fold both ways, since people type Zurich and Sao Paulo far more often than Zürich and São Paulo.

Free-text matching is whole-word and ignores names under four characters, so ordinary chatter doesn't become a weather report — how are you today still resolves to nothing. The unmatched reply also stops reading a whole sentence back as a place name.

This fixes /grade identically — it shares the resolver and had the same bug.

Verified

Against the exact strings from the report:

typed	before	after
`Tell me about weather in lithuania`	refused	Vilnius, Lithuania
`Vilinus, Lithuania`	refused	Vilnius, Lithuania
`Vilnius, Lithuania`	refused	Vilnius, Lithuania
`san fran`	San Francisco	San Francisco
`Zurich` / `Sao Paulo`	refused	Zürich / São Paulo
`how are you today`	refused	refused

Full backend suite: 416 passed, 8 skipped. Seven new regression tests, each named after a real failure.

Not covered

A typo inside a sentence (whats it like in Toyko) still misses — fuzzy matching applies to the query as a whole, not per word, and loosening that invites false positives. The conversational agent handles that case; this is the always-on deterministic path.

https://claude.ai/code/session_015Z1ebLbhUxeZ9ozpNrVTCP

Reported live from #general: > **@Thermograph** Vilnius, Lithuania > *I don't track a city called 'Vilnius, Lithuania' yet. Try a major city name.* Vilnius has been in `cities.json` the whole time, with `country: "Lithuania"` sitting right there in the record. `_resolve_city` matched the query against `name` alone, exact-or-prefix — so the comma form could never match anything, and the country field was never consulted at all. The same exchange showed the two neighbouring failures: - **`Vilinus`** — one transposition — fell through to the same refusal, which reads as *we've never heard of Vilnius* rather than *you typed it wrong*. - **`Tell me about weather in lithuania`** was quoted back verbatim as a city we don't track, which is what made it look broken rather than merely unmatched. ## What it does now In descending order of confidence: exact name honouring a `City, Country/Region` qualifier → the same ignoring an unrecognised qualifier (better to grade Paris and name it in the reply than refuse over "Paris, Wherever") → prefix → close-enough name for typos → a city or country named somewhere inside a sentence, so "weather in lithuania" lands on Vilnius. Accents fold both ways, since people type Zurich and Sao Paulo far more often than Zürich and São Paulo. Free-text matching is whole-word and ignores names under four characters, so ordinary chatter doesn't become a weather report — `how are you today` still resolves to nothing. The unmatched reply also stops reading a whole sentence back as a place name. **This fixes `/grade` identically** — it shares the resolver and had the same bug. ## Verified Against the exact strings from the report: | typed | before | after | |---|---|---| | `Tell me about weather in lithuania` | refused | Vilnius, Lithuania | | `Vilinus, Lithuania` | refused | Vilnius, Lithuania | | `Vilnius, Lithuania` | refused | Vilnius, Lithuania | | `san fran` | San Francisco | San Francisco | | `Zurich` / `Sao Paulo` | refused | Zürich / São Paulo | | `how are you today` | refused | refused | Full backend suite: 416 passed, 8 skipped. Seven new regression tests, each named after a real failure. ## Not covered A typo *inside* a sentence (`whats it like in Toyko`) still misses — fuzzy matching applies to the query as a whole, not per word, and loosening that invites false positives. The conversational agent handles that case; this is the always-on deterministic path. https://claude.ai/code/session_015Z1ebLbhUxeZ9ozpNrVTCP

admin_emi added 1 commit 2026-07-24 21:48:35 +00:00

discord: resolve the city people actually typed

PR build (required check) / changes (pull_request) Successful in 6s

Details

secrets-guard / encrypted (pull_request) Successful in 5s

Details

PR build (required check) / build-frontend (pull_request) Has been skipped

Details

shell-lint / shellcheck (pull_request) Successful in 7s

Details

PR build (required check) / validate-observability (pull_request) Successful in 38s

Details

PR build (required check) / build-backend (pull_request) Successful in 1m19s

Details

PR build (required check) / gate (pull_request) Successful in 2s

Details

1b862aa049

Reported from #general: "@Thermograph Vilnius, Lithuania" came back "I don't
track a city called 'Vilnius, Lithuania' yet." Vilnius has been in cities.json
the whole time, with country "Lithuania" sitting right there in the record. The
resolver matched the query against `name` alone, exact-or-prefix, so the comma
form could never match anything and the country field was never consulted.

The same message also showed the two neighbouring failures: "Vilinus" (one
transposition) fell through to the same refusal, which reads as "we've never
heard of Vilnius" rather than "you typed it wrong"; and "Tell me about weather
in lithuania" was quoted back verbatim as a city we don't track, which is what
made it look broken rather than merely unmatched.

So the resolver now goes, in descending order of confidence: exact name
honouring a "City, Country/Region" qualifier; the same ignoring an unrecognised
qualifier (better to grade Paris and name it in the reply than refuse over
"Paris, Wherever"); prefix; close-enough name for typos; and finally a city or
country named somewhere inside a sentence, so "weather in lithuania" lands on
Vilnius. Accents fold both ways, since people type Zurich and Sao Paulo far more
often than Zürich and São Paulo.

Free-text matching is whole-word and ignores names under four characters, so
ordinary chatter doesn't turn into a weather report -- "how are you today"
still resolves to nothing. And the unmatched reply no longer reads a whole
sentence back as a place name.

This fixes /grade identically; it shared the resolver and the same bug.

Claude-Session: https://claude.ai/code/session_015Z1ebLbhUxeZ9ozpNrVTCP