zzstoatzz.io
search for standard sites
pub-search.waow.tech
search
zig
blog
atproto
feat: search documents by publication subdomain (base_path)
- add base_path to publications_fts for FTS search
- add DocsByPubBasePath queries to find docs via publication match
- deduplicate results when doc matches both content and base_path
- fix use-after-free by duping URIs for hash map
searching "gyst" now returns all docs on gyst.leaflet.pub, not just
those with "gyst" in their content.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
+5
-1
backend/src/db/schema.zig
+5
-1
backend/src/db/schema.zig
···
44
44
\\CREATE VIRTUAL TABLE IF NOT EXISTS publications_fts USING fts5(
45
45
\\ uri UNINDEXED,
46
46
\\ name,
47
-
\\ description
47
+
\\ description,
48
+
\\ base_path
48
49
\\)
49
50
, &.{});
50
51
···
162
163
// URL path field for documents (e.g., "/001" for zat.dev)
163
164
// used to build full URL: publication.url + document.path
164
165
client.exec("ALTER TABLE documents ADD COLUMN path TEXT", &.{}) catch {};
166
+
167
+
// note: publications_fts was rebuilt with base_path column via scripts/rebuild-pub-fts
168
+
// new publications will include base_path via insertPublication in indexer.zig
165
169
}
+3
-3
backend/src/indexer.zig
+3
-3
backend/src/indexer.zig
···
81
81
&.{ uri, did, rkey, name, description orelse "", base_path orelse "" },
82
82
);
83
83
84
-
// update FTS index
84
+
// update FTS index (includes base_path for subdomain search)
85
85
c.exec("DELETE FROM publications_fts WHERE uri = ?", &.{uri}) catch {};
86
86
c.exec(
87
-
"INSERT INTO publications_fts (uri, name, description) VALUES (?, ?, ?)",
88
-
&.{ uri, name, description orelse "" },
87
+
"INSERT INTO publications_fts (uri, name, description, base_path) VALUES (?, ?, ?, ?)",
88
+
&.{ uri, name, description orelse "", base_path orelse "" },
89
89
) catch {};
90
90
}
91
91
+62
-2
backend/src/search.zig
+62
-2
backend/src/search.zig
···
159
159
\\ORDER BY d.created_at DESC LIMIT 40
160
160
);
161
161
162
+
// Find documents by their publication's base_path (subdomain search)
163
+
// e.g., searching "gyst" finds all docs on gyst.leaflet.pub
164
+
const DocsByPubBasePath = zql.Query(
165
+
\\SELECT d.uri, d.did, d.title, '' as snippet,
166
+
\\ d.created_at, d.rkey,
167
+
\\ p.base_path,
168
+
\\ 1 as has_publication,
169
+
\\ d.platform, COALESCE(d.path, '') as path
170
+
\\FROM documents d
171
+
\\JOIN publications p ON d.publication_uri = p.uri
172
+
\\JOIN publications_fts pf ON p.uri = pf.uri
173
+
\\WHERE publications_fts MATCH :query
174
+
\\ORDER BY d.created_at DESC LIMIT 40
175
+
);
176
+
177
+
const DocsByPubBasePathAndPlatform = zql.Query(
178
+
\\SELECT d.uri, d.did, d.title, '' as snippet,
179
+
\\ d.created_at, d.rkey,
180
+
\\ p.base_path,
181
+
\\ 1 as has_publication,
182
+
\\ d.platform, COALESCE(d.path, '') as path
183
+
\\FROM documents d
184
+
\\JOIN publications p ON d.publication_uri = p.uri
185
+
\\JOIN publications_fts pf ON p.uri = pf.uri
186
+
\\WHERE publications_fts MATCH :query AND d.platform = :platform
187
+
\\ORDER BY d.created_at DESC LIMIT 40
188
+
);
189
+
162
190
/// Publication search result (internal)
163
191
const Pub = struct {
164
192
uri: []const u8,
···
219
247
const has_tag = tag_filter != null;
220
248
const has_platform = platform_filter != null;
221
249
222
-
// search documents - handle all filter combinations
250
+
// track seen URIs for deduplication (content match + base_path match)
251
+
var seen_uris = std.StringHashMap(void).init(alloc);
252
+
defer seen_uris.deinit();
253
+
254
+
// search documents by content (title, content) - handle all filter combinations
223
255
var doc_result = if (has_query and has_tag and has_platform)
224
256
c.query(DocsByFtsAndTagAndPlatform.positional, DocsByFtsAndTagAndPlatform.bind(.{
225
257
.query = fts_query,
···
244
276
if (doc_result) |*res| {
245
277
defer res.deinit();
246
278
for (res.rows) |row| {
247
-
try jw.write(Doc.fromRow(row).toJson());
279
+
const doc = Doc.fromRow(row);
280
+
// dupe URI for hash map (outlives result)
281
+
const uri_dupe = try alloc.dupe(u8, doc.uri);
282
+
try seen_uris.put(uri_dupe, {});
283
+
try jw.write(doc.toJson());
284
+
}
285
+
}
286
+
287
+
// also search documents by publication base_path (subdomain search)
288
+
// e.g., "gyst" finds all docs on gyst.leaflet.pub even if content doesn't contain "gyst"
289
+
// skip if tag filter is set (tag filter is content-specific)
290
+
if (has_query and !has_tag) {
291
+
var basepath_result = if (has_platform)
292
+
c.query(DocsByPubBasePathAndPlatform.positional, DocsByPubBasePathAndPlatform.bind(.{
293
+
.query = fts_query,
294
+
.platform = platform_filter.?,
295
+
})) catch null
296
+
else
297
+
c.query(DocsByPubBasePath.positional, DocsByPubBasePath.bind(.{ .query = fts_query })) catch null;
298
+
299
+
if (basepath_result) |*res| {
300
+
defer res.deinit();
301
+
for (res.rows) |row| {
302
+
const doc = Doc.fromRow(row);
303
+
// deduplicate: skip if already found by content search
304
+
if (!seen_uris.contains(doc.uri)) {
305
+
try jw.write(doc.toJson());
306
+
}
307
+
}
248
308
}
249
309
}
250
310
+86
scripts/rebuild-pub-fts
+86
scripts/rebuild-pub-fts
···
1
+
#!/usr/bin/env -S uv run --script --quiet
2
+
# /// script
3
+
# requires-python = ">=3.12"
4
+
# dependencies = ["httpx", "pydantic-settings"]
5
+
# ///
6
+
"""Rebuild publications_fts with base_path column for subdomain search."""
7
+
import os
8
+
import httpx
9
+
from pydantic_settings import BaseSettings, SettingsConfigDict
10
+
11
+
12
+
class Settings(BaseSettings):
13
+
model_config = SettingsConfigDict(
14
+
env_file=os.environ.get("ENV_FILE", ".env"), extra="ignore"
15
+
)
16
+
turso_url: str
17
+
turso_token: str
18
+
19
+
@property
20
+
def turso_host(self) -> str:
21
+
url = self.turso_url
22
+
if url.startswith("libsql://"):
23
+
url = url[len("libsql://") :]
24
+
return url
25
+
26
+
27
+
settings = Settings() # type: ignore
28
+
29
+
print("Rebuilding publications_fts with base_path column...")
30
+
31
+
response = httpx.post(
32
+
f"https://{settings.turso_host}/v2/pipeline",
33
+
headers={
34
+
"Authorization": f"Bearer {settings.turso_token}",
35
+
"Content-Type": "application/json",
36
+
},
37
+
json={
38
+
"requests": [
39
+
{"type": "execute", "stmt": {"sql": "DROP TABLE IF EXISTS publications_fts"}},
40
+
{
41
+
"type": "execute",
42
+
"stmt": {
43
+
"sql": """
44
+
CREATE VIRTUAL TABLE publications_fts USING fts5(
45
+
uri UNINDEXED,
46
+
name,
47
+
description,
48
+
base_path
49
+
)
50
+
"""
51
+
},
52
+
},
53
+
{
54
+
"type": "execute",
55
+
"stmt": {
56
+
"sql": """
57
+
INSERT INTO publications_fts (uri, name, description, base_path)
58
+
SELECT uri, name, COALESCE(description, ''), COALESCE(base_path, '')
59
+
FROM publications
60
+
"""
61
+
},
62
+
},
63
+
{"type": "execute", "stmt": {"sql": "SELECT COUNT(*) FROM publications_fts"}},
64
+
{"type": "close"},
65
+
]
66
+
},
67
+
timeout=60,
68
+
)
69
+
response.raise_for_status()
70
+
data = response.json()
71
+
72
+
for i, result in enumerate(data["results"][:-1]): # skip close
73
+
if result["type"] == "error":
74
+
print(f"Step {i} error: {result['error']}")
75
+
elif result["type"] == "ok":
76
+
if i == 3: # count query
77
+
rows = result["response"]["result"].get("rows", [])
78
+
if rows:
79
+
count = (
80
+
rows[0][0].get("value", rows[0][0])
81
+
if isinstance(rows[0][0], dict)
82
+
else rows[0][0]
83
+
)
84
+
print(f"Rebuilt with {count} publications")
85
+
86
+
print("Done!")
+111
-1
site/index.html
+111
-1
site/index.html
···
111
111
.result-title {
112
112
color: #fff;
113
113
margin-bottom: 0.5rem;
114
+
/* prevent long titles from breaking layout */
115
+
display: -webkit-box;
116
+
-webkit-line-clamp: 2;
117
+
-webkit-box-orient: vertical;
118
+
overflow: hidden;
119
+
word-break: break-word;
114
120
}
115
121
116
122
.result-title a { color: inherit; }
···
383
389
.active-filter .clear:hover {
384
390
color: #c44;
385
391
}
392
+
393
+
/* mobile improvements */
394
+
@media (max-width: 600px) {
395
+
body {
396
+
padding: 0.75rem;
397
+
font-size: 13px;
398
+
}
399
+
400
+
.container {
401
+
max-width: 100%;
402
+
}
403
+
404
+
/* ensure minimum 44px touch targets */
405
+
.tag, .platform-option, .suggestion {
406
+
min-height: 44px;
407
+
display: inline-flex;
408
+
align-items: center;
409
+
padding: 0.5rem 0.75rem;
410
+
}
411
+
412
+
button {
413
+
min-height: 44px;
414
+
padding: 0.5rem 0.75rem;
415
+
}
416
+
417
+
/* stack search box on very small screens */
418
+
.search-box {
419
+
flex-direction: column;
420
+
gap: 0.5rem;
421
+
}
422
+
423
+
.search-box input[type="text"] {
424
+
width: 100%;
425
+
}
426
+
427
+
.search-box button {
428
+
width: 100%;
429
+
}
430
+
431
+
/* result card mobile tweaks */
432
+
.result {
433
+
padding: 0.75rem 0;
434
+
}
435
+
436
+
.result:hover {
437
+
margin: 0 -0.75rem;
438
+
padding: 0.75rem;
439
+
}
440
+
441
+
.result-title {
442
+
font-size: 14px;
443
+
line-height: 1.4;
444
+
}
445
+
446
+
.result-snippet {
447
+
font-size: 12px;
448
+
line-height: 1.5;
449
+
}
450
+
451
+
/* badges inline on mobile */
452
+
.entity-type, .platform-badge {
453
+
font-size: 9px;
454
+
padding: 2px 5px;
455
+
margin-right: 6px;
456
+
vertical-align: middle;
457
+
}
458
+
459
+
/* tags wrap better on mobile */
460
+
.tags-list, .platform-filter-list {
461
+
gap: 0.5rem;
462
+
}
463
+
464
+
/* suggestions responsive */
465
+
.suggestions {
466
+
line-height: 2;
467
+
}
468
+
469
+
/* related items more compact */
470
+
.related-item {
471
+
max-width: 150px;
472
+
font-size: 11px;
473
+
padding: 0.5rem;
474
+
}
475
+
}
476
+
477
+
/* ensure touch targets on tablets too */
478
+
@media (hover: none) and (pointer: coarse) {
479
+
.tag, .platform-option, .suggestion, .related-item {
480
+
min-height: 44px;
481
+
display: inline-flex;
482
+
align-items: center;
483
+
}
484
+
}
386
485
</style>
387
486
</head>
388
487
<body>
···
460
559
if (results.length === 0) {
461
560
resultsDiv.innerHTML = `
462
561
<div class="empty-state">
463
-
<p>no results${query ? ` for "${escapeHtml(query)}"` : ''}${tag ? ` in #${escapeHtml(tag)}` : ''}${platform ? ` on ${escapeHtml(platform)}` : ''}</p>
562
+
<p>no results${query ? ` for ${formatQueryForDisplay(query)}` : ''}${tag ? ` in #${escapeHtml(tag)}` : ''}${platform ? ` on ${escapeHtml(platform)}` : ''}</p>
464
563
<p>try different keywords</p>
465
564
</div>
466
565
`;
···
524
623
return str.replace(/[&<>"']/g, c => ({
525
624
'&': '&', '<': '<', '>': '>', '"': '"', "'": '''
526
625
})[c]);
626
+
}
627
+
628
+
// display query without adding redundant quotes
629
+
function formatQueryForDisplay(query) {
630
+
if (!query) return '';
631
+
const escaped = escapeHtml(query);
632
+
// if query is already fully quoted, don't add more quotes
633
+
if (query.startsWith('"') && query.endsWith('"')) {
634
+
return escaped;
635
+
}
636
+
return `"${escaped}"`;
527
637
}
528
638
529
639
// platform-specific URL patterns