- Rename the Go module from codeberg.org/squidd/parsimony to
git.kmalloc.org/squid/parsimony in go.mod and every import.
- Fix .gitignore: the bare `parsimony` pattern also ignored the cmd/parsimony
folder, so the main package was never committed and a clean checkout could
not be built. Anchor it to the root (`/parsimony`) and add cmd/parsimony.
- the API + admin UI server and the public web server get ReadTimeout 30s,
WriteTimeout 60s and IdleTimeout 120s next to ReadHeaderTimeout
- the archive import extends its own read and write deadlines to 30
minutes with http.ResponseController, so a big mbox is not cut
- tests: a stalled body and an idle connection are cut on both servers,
an import longer than the default timeouts succeeds, fields are pinned
- README: document the limits and timeouts
- walk at most 10 nested multipart levels and 100 parts, then use what
was found: the post is still delivered, the archive stores less
- read a text part only while it can still become the body, and stop at
the first non-blank plain part, so memory stays near MaxBody
- a Date more than 24 hours after the time of parsing becomes that time;
a Date before 1990 is treated as missing
- tests: 20000-level nesting finishes fast, exact limits, allocation
bound, pathological HTML, date clamps
- Add ImportMbox: batches of 200, idempotent by Message-ID, threads rebuilt at the end
- Date fallbacks: Date header, then the From line, then the import time
- Report failed and oversized messages without stopping, keep progress on a read error
- Import regardless of the list's archive policy
- Add web.listen config (default 127.0.0.1:8081, empty disables)
- Start the public web in app, close it with the rest
- Add end-to-end test: archive stays private, becomes public, hides addresses, one-click unsubscribe works
- Cover mutation survivors: attachments, first text part, multibyte cap, member id 0, page clamp, API limit cap, closing connections
- Update OpenAPI, README and example config
- Decode subject, sender name, date, Message-ID and In-Reply-To
- Pick the plain text body from nested multipart, fall back to stripped HTML
- Decode transfer encodings and charsets, repair invalid UTF-8, cap at 1 MiB
- Synthesize a stable Message-ID when missing
- Never expose a sender address domain; add email obfuscation for public pages