feat: support appstream data keywords and categories for FTS#204
Open
feat: support appstream data keywords and categories for FTS#204
Conversation
m1rm
changed the title
m1rm
force-pushed
the
feat/integrate-appstream-data
branch
from
01c46f3 to
9157312
Compare
…a (summary, description etc.) was too messy and thus messed up search rankings
beautify
…est into main exists and pushes are added to branch with open PR piggyback: improve ci setup; account for duplicate runs tryout to fix duplicate ci runs the other way
flush already hands ownership off to the caller (p.cur is nilled), so append-clone of keywords/categories was pure overhead per <component>. Strict=false hid real malformed-input errors; the upstream feed is well-formed XML.
Two parallel maps keyed by pkgname plus a third union map to iterate
was wasted memory and an extra pass. One map of {kw, cat} slices covers
it.
The http.Client parameter on Update was always passed nil from main; the fallback client timeouts (15m / 2m) were pure dead code since the ctx deadline (10m in runCommand) always wins. Match the convention of the other update commands: no client param, no hand-rolled timeouts. Also unexport latestRelease — it's not called from outside the package.
*gzip.Reader already satisfies io.Reader.
Replace the manual stack + 6 skip flags with the decoder's own Skip(): when a <keywords>/<keyword>/<categories>/<category> tag has a non-en/de xml:lang, skip its entire subtree in one call. The remaining state is five booleans tracking the enter/leave of accepted elements. No behavior change; all existing tests pass.
pierres
force-pushed
the
feat/integrate-appstream-data
branch
from
9157312 to
3c5646d
Compare
Upstream archlinux-appstream-data publishes core/extra/multilib only, so core-testing/extra-testing/multilib-testing packages never get keywords or categories columns populated. Document the asymmetry so search-ranking surprises don't lead down a wrong debugging path.
Extract the DB-facing portion of Update into applyTerms so tests can drive it directly against in-memory SQLite without mocking the HTTP fetches. Covers: - Keywords + categories land on matching package rows; non-mentioned packages stay empty. - FTS matches on both the new keyword and category columns after the rebuild — catches column-order drift between the schema and the query. - A second run clears stale data from rows no longer in the accumulator. - Duplicates and stopwords are stripped by dedupeWords. - All-stopword input does not update the row at all.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
TL;DR
Integrates Arch Linux AppStream metadata into the Go app: keywords and categories columns on package, filled from upstream Components-x86_64.xml.gz on sources.archlinux.org, exposed to FTS5 search with tuned BM25 weights. Adds a update-appstream CLI command and just update-appstream / inclusion in just update-data.
Motivation
Improve package search (including German-oriented terms) using AppStream and data.
Keep the implementation streaming and aligned with existing update jobs (pacmandb-style callback parsing).
Data source & versioning
https://archlinux.org/packages/extra/any/archlinux-appstream-data/json/ (pkgver).{APPSTREAM_SOURCES_BASE}{pkgver}/{core|extra|multilib}/Components-x86_64.xml.gzBehaviour
What is indexed
Database & search
Operations
Testing
Follow-ups (optional)