perf: combined feed index for multi-user global feed

Instead of the browser resolving 20+ user shards recursively (~27 MB),
generate a pre-sorted feed.json at merge time with 50 activities per
page. The global feed loads one ~30 KB file on first paint; "Load more"
fetches subsequent pages (feed-2.json, feed-3.json, etc.).

Per-user profile pages still use year-sharded loadIndexPaged as before.
This commit is contained in:
Davide Scaini
2026-04-20 15:31:35 +02:00
parent e8a5fbbaba
commit db7047f210
4 changed files with 160 additions and 8 deletions
+5
View File
@@ -133,6 +133,11 @@ def _write_root_manifest(data: Path) -> None:
root.write_text(json.dumps(manifest, indent=2))
console.print(f"Root manifest updated: [cyan]{len(users)}[/cyan] user shard(s)")
if len(users) > 1:
from bincio.render.merge import write_combined_feed
n = write_combined_feed(data)
console.print(f"Combined feed: [cyan]{n}[/cyan] activities across all users")
def _link_data(site: Path, data: Path) -> None:
"""Symlink site/public/data → data root (each user has their own _merged/)."""
+82
View File
@@ -320,3 +320,85 @@ def _write_year_shards(merged_dir: Path, activities: list[dict], index_meta: dic
"activities": [],
}
(merged_dir / "index.json").write_text(json.dumps(root_doc, indent=2, ensure_ascii=False))
FEED_PAGE_SIZE = 50
# Extra fields stripped from the combined feed — preview_coords is the biggest
# contributor (~24% of shard size) but the feed cards need it for thumbnails,
# so we keep it. mmp is never displayed in feed cards.
_COMBINED_FEED_STRIP = _FEED_STRIP | {"mmp"}
def write_combined_feed(data_dir: Path) -> int:
"""Build data_dir/feed.json — the N most recent activities across all users.
The global feed page loads this single file instead of resolving 20+ user
shards recursively. Returns the number of activities written.
"""
user_dirs = sorted(
p for p in data_dir.iterdir()
if p.is_dir() and (p / "activities").exists()
)
all_activities: list[dict] = []
for user_dir in user_dirs:
handle = user_dir.name
merged = user_dir / "_merged"
index_path = merged / "index.json" if merged.exists() else user_dir / "index.json"
if not index_path.exists():
continue
index = json.loads(index_path.read_text(encoding="utf-8"))
shards = index.get("shards", [])
activities = index.get("activities", [])
if shards:
year_shards = sorted(
[s for s in shards if re.match(r"index-\d{4}\.json$", s.get("url", ""))],
key=lambda s: s["url"],
reverse=True,
)
base = index_path.parent
for shard in year_shards[:2]:
shard_path = base / shard["url"]
if shard_path.exists():
shard_data = json.loads(shard_path.read_text(encoding="utf-8"))
for a in shard_data.get("activities", []):
a_tagged = {**a, "handle": handle}
detail_url = a_tagged.get("detail_url", "")
if detail_url and not detail_url.startswith("http") and not detail_url.startswith("/"):
merged_rel = f"{handle}/_merged/" if merged.exists() else f"{handle}/"
a_tagged["detail_url"] = merged_rel + detail_url
track_url = a_tagged.get("track_url", "")
if track_url and not track_url.startswith("http") and not track_url.startswith("/"):
merged_rel = f"{handle}/_merged/" if merged.exists() else f"{handle}/"
a_tagged["track_url"] = merged_rel + track_url
all_activities.append(a_tagged)
else:
for a in activities:
all_activities.append({**a, "handle": handle})
all_activities.sort(key=lambda a: a.get("started_at", ""), reverse=True)
# Remove stale feed pages
for f in data_dir.glob("feed*.json"):
f.unlink()
if not all_activities:
return 0
pages = [all_activities[i:i + FEED_PAGE_SIZE] for i in range(0, len(all_activities), FEED_PAGE_SIZE)]
for page_num, page in enumerate(pages):
slim = [{k: v for k, v in a.items() if k not in _COMBINED_FEED_STRIP} for a in page]
fname = "feed.json" if page_num == 0 else f"feed-{page_num + 1}.json"
doc = {
"bas_version": "1.0",
"page": page_num + 1,
"total_pages": len(pages),
"total_activities": len(all_activities),
"activities": slim,
}
(data_dir / fname).write_text(json.dumps(doc, indent=2, ensure_ascii=False))
return len(all_activities)
+30 -8
View File
@@ -2,7 +2,7 @@
import { onMount } from 'svelte';
import type { ActivitySummary, BASIndex, Sport } from '../lib/types';
import { formatDistance, formatDuration, formatElevation, formatDate, isUnlisted, sportIcon, sportColor, sportLabel } from '../lib/format';
import { loadIndexPaged, loadShardActivities } from '../lib/dataloader';
import { loadIndexPaged, loadShardActivities, loadCombinedFeed, loadCombinedFeedPage } from '../lib/dataloader';
/** Render preview_coords as an SVG polyline path string. */
function trackPath(coords: [number, number][] | null, w: number, h: number): string {
@@ -45,6 +45,9 @@
let error = '';
let mounted = false;
let pendingShards: string[] = [];
/** Remaining combined-feed pages (multi-user global feed). */
let feedNextPage = 0;
let feedTotalPages = 0;
/** Logged-in handle — resolved async via bincio:me event. */
let me: string = '';
@@ -61,20 +64,26 @@
$: filtered = sport === 'all' ? withPrivacy : withPrivacy.filter(a => a.sport === sport);
$: visible = filtered.slice(0, shown);
$: canShowMore = shown < filtered.length;
$: hasMore = canShowMore || pendingShards.length > 0;
$: hasMore = canShowMore || pendingShards.length > 0 || feedNextPage > 0;
async function loadMore() {
if (canShowMore) {
shown += PAGE_SIZE;
return;
}
if (!pendingShards.length) return;
loadingMore = true;
const url = pendingShards[0];
pendingShards = pendingShards.slice(1);
try {
const fresh = await loadShardActivities(url);
// Merge avoiding duplicates (IDB activities may already be present)
let fresh: ActivitySummary[] = [];
if (feedNextPage > 0) {
fresh = await loadCombinedFeedPage(base, feedNextPage);
feedNextPage = feedNextPage < feedTotalPages ? feedNextPage + 1 : 0;
} else if (pendingShards.length) {
const url = pendingShards[0];
pendingShards = pendingShards.slice(1);
fresh = await loadShardActivities(url);
} else {
return;
}
const existing = new Map(all.map(a => [a.id, a]));
for (const a of fresh) if (!existing.has(a.id)) existing.set(a.id, a);
all = [...existing.values()].sort((a, b) =>
@@ -82,7 +91,7 @@
);
shown += PAGE_SIZE;
} catch {
// shard load failed — don't block the user
// load failed — don't block the user
} finally {
loadingMore = false;
}
@@ -109,6 +118,17 @@
}
try {
const isGlobalFeed = !profileIndexUrl && !filterHandle;
if (isGlobalFeed) {
const combined = await loadCombinedFeed(base);
if (combined) {
all = combined.activities;
feedTotalPages = combined.remainingPages + 1;
feedNextPage = combined.remainingPages > 0 ? 2 : 0;
loading = false;
return;
}
}
const indexUrl = profileIndexUrl
? `${base}data/${profileIndexUrl}`
: `${base}data/index.json`;
@@ -263,6 +283,8 @@
Loading…
{:else if canShowMore}
Load more ({filtered.length - shown} remaining)
{:else if feedNextPage > 0}
Load more activities
{:else}
Load older activities ({pendingShards.length} more {pendingShards.length === 1 ? 'year' : 'years'})
{/if}
+43
View File
@@ -253,6 +253,49 @@ export async function loadShardActivities(shardUrl: string): Promise<ActivitySum
}
}
interface FeedPage {
page: number;
total_pages: number;
total_activities: number;
activities: ActivitySummary[];
}
/**
* Load the combined feed (multi-user global feed). Returns the first page of
* activities pre-sorted across all users, plus remaining page count.
*
* Falls back to the full shard-resolution path if feed.json doesn't exist
* (single-user installs, older data).
*/
export async function loadCombinedFeed(
baseUrl: string,
): Promise<{ activities: ActivitySummary[]; remainingPages: number } | null> {
try {
const feed = await fetchJSON<FeedPage>(`${baseUrl}data/feed.json`);
return {
activities: feed.activities ?? [],
remainingPages: (feed.total_pages ?? 1) - 1,
};
} catch {
return null;
}
}
/**
* Load a subsequent page of the combined feed (feed-2.json, feed-3.json, etc.).
*/
export async function loadCombinedFeedPage(
baseUrl: string,
page: number,
): Promise<ActivitySummary[]> {
try {
const feed = await fetchJSON<FeedPage>(`${baseUrl}data/feed-${page}.json`);
return feed.activities ?? [];
} catch {
return [];
}
}
/**
* Load a single activity detail, checking IndexedDB first so locally-converted
* activities are available offline.