feat: complete remaining features — training, FTS, bot groups, repair, re-tag
continuous-integration/drone/push Build is passing

Manual override training (GroupingRule):
- Learn patterns from manual group creation (common filename prefix or creator)
- Apply learned rules as first auto-grouping pass (highest confidence after albums)
- GroupingRule model stores pattern, channel, signal type, confidence

Hash verification after upload:
- Re-hash upload files on disk before indexing to catch disk corruption
- Creates HASH_MISMATCH notification on discrepancy

Grouping conflict detection:
- After all grouping passes, check if grouped packages match rules from different groups
- Creates GROUPING_CONFLICT notification for manual review

Per-channel grouping flags:
- Add autoGroupEnabled boolean to TelegramChannel (default true)
- Auto-grouping passes (all except album) gated behind this flag
- Album grouping always runs as it reflects Telegram's native behavior

Full-text search (tsvector):
- Add searchVector tsvector column with GIN index and auto-update trigger
- Backfill 1870 existing packages
- FTS with ts_rank for ranked results, ILIKE fallback for short/failed queries
- Applied to both web app and bot search

Bot group awareness:
- /group <query> — view group info or search groups by name
- /sendgroup <id> — send all packages in a group to linked Telegram account

Bulk repair:
- repairPackageAction clears dest info and resets watermark for re-processing
- Repair button in notification bell for MISSING_PART and HASH_MISMATCH alerts
- /api/notifications/repair endpoint

Retroactive category re-tagging:
- When channel category changes, auto-update tags on all existing packages
- Removes old category tag, adds new one

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-03-30 14:34:14 +02:00
co-authored by Claude Opus 4.6
parent 7f9a03d4ee
commit f4aa9d9a2f
12 changed files with 788 additions and 25 deletions
+121
View File
@@ -79,6 +79,69 @@ export async function processAlbumGroups(
}
}
/**
* Apply learned GroupingRules from manual overrides.
* For each rule, find ungrouped packages whose fileName contains the pattern.
*/
export async function processRuleBasedGroups(
sourceChannelId: string,
indexedPackages: IndexedPackageRef[]
): Promise<void> {
const rules = await db.groupingRule.findMany({
where: { sourceChannelId },
orderBy: { confidence: "desc" },
});
if (rules.length === 0) return;
const ungrouped = await db.package.findMany({
where: {
id: { in: indexedPackages.map((p) => p.packageId) },
packageGroupId: null,
},
select: { id: true, fileName: true, creator: true },
});
if (ungrouped.length < 2) return;
for (const rule of rules) {
const matches = ungrouped.filter((pkg) => {
const lower = rule.pattern.toLowerCase();
return pkg.fileName.toLowerCase().includes(lower) ||
(pkg.creator && pkg.creator.toLowerCase().includes(lower));
});
if (matches.length < 2) continue;
// Check if any are already grouped (by a previous rule in this loop)
const stillUngrouped = await db.package.findMany({
where: {
id: { in: matches.map((m) => m.id) },
packageGroupId: null,
},
select: { id: true },
});
if (stillUngrouped.length < 2) continue;
try {
const groupId = await createAutoGroup({
sourceChannelId,
name: rule.pattern,
packageIds: stillUngrouped.map((m) => m.id),
groupingSource: "MANUAL",
});
log.info(
{ groupId, ruleId: rule.id, pattern: rule.pattern, memberCount: stillUngrouped.length },
"Applied learned grouping rule"
);
} catch (err) {
log.warn({ err, ruleId: rule.id }, "Failed to apply grouping rule");
}
}
}
/**
* After album grouping, cluster remaining ungrouped packages from the same channel
* that were posted within a configurable time window.
@@ -525,6 +588,64 @@ function extractRootFolder(paths: string[]): string | null {
return maxSegment;
}
/**
* Detect packages that could have been grouped differently.
* Checks if any grouped package's filename matches a GroupingRule
* that would place it in a different group.
*/
export async function detectGroupingConflicts(
sourceChannelId: string,
indexedPackages: IndexedPackageRef[]
): Promise<void> {
const rules = await db.groupingRule.findMany({
where: { sourceChannelId },
});
if (rules.length === 0) return;
const grouped = await db.package.findMany({
where: {
id: { in: indexedPackages.map((p) => p.packageId) },
packageGroupId: { not: null },
},
select: {
id: true,
fileName: true,
packageGroupId: true,
packageGroup: { select: { name: true, groupingSource: true } },
},
});
for (const pkg of grouped) {
for (const rule of rules) {
if (pkg.fileName.toLowerCase().includes(rule.pattern.toLowerCase())) {
// Check if the rule's source group is different from current group
if (rule.createdByGroupId && rule.createdByGroupId !== pkg.packageGroupId) {
try {
await db.systemNotification.create({
data: {
type: "GROUPING_CONFLICT",
severity: "INFO",
title: `Potential grouping conflict: ${pkg.fileName}`,
message: `Grouped by ${pkg.packageGroup?.groupingSource ?? "unknown"} into "${pkg.packageGroup?.name}", but also matches rule "${rule.pattern}" from a different manual group`,
context: {
packageId: pkg.id,
fileName: pkg.fileName,
currentGroupId: pkg.packageGroupId,
matchedRuleId: rule.id,
matchedPattern: rule.pattern,
},
},
});
} catch {
// Best-effort
}
break; // One notification per package
}
}
}
}
}
/**
* Find the longest common prefix among a list of filenames,
* trimming trailing separators and partial words.