mirror of
https://github.com/Kaelio/ktx.git
synced 2026-06-22 08:38:08 +02:00
feat(cli): show embedding coverage in ktx status, drop duplicate disk counts
Inline `(N embedded)` next to the Wiki scope counts and Semantic-layer source counts, computed with `SUM(embedding_json IS NOT NULL)` over `knowledge_pages` and `local_sl_sources`. Rename the "Knowledge" label to "Wiki" (canonical per `docs/terminology.md`) and rename the matching `localStats.knowledgePages` field to `localStats.wikiPages`. Drop `wiki=N md` and `semantic-layer=N yaml` from the Disk row — those duplicated the per-surface rows above. Disk now reports only actual byte usage (db, cache, raw-sources). The unused `wikiGlobalMarkdownCount` / `semanticLayerYamlCount` fields, the `isMarkdownEntry` / `isYamlEntry` helpers, and the `filter` arg on `summarizeDir` are removed.
This commit is contained in:
Andrey Avtomonov
parent
abdca58884
commit
0915b21a1a
2 changed files with 64 additions and 70 deletions
|
|
@ -268,9 +268,9 @@ describe('buildLocalStatsStatus', () => {
|
|||
embedding_json TEXT
|
||||
);
|
||||
INSERT INTO knowledge_pages VALUES
|
||||
('a.md', 'a', 'GLOBAL', NULL, '', '', '[]', '', NULL),
|
||||
('a.md', 'a', 'GLOBAL', NULL, '', '', '[]', '', '[0.1,0.2]'),
|
||||
('b.md', 'b', 'GLOBAL', NULL, '', '', '[]', '', NULL),
|
||||
('c.md', 'c', 'PROJECT', NULL, '', '', '[]', '', NULL);
|
||||
('c.md', 'c', 'PROJECT', NULL, '', '', '[]', '', '[0.3,0.4]');
|
||||
|
||||
CREATE TABLE local_sl_sources (
|
||||
connection_id TEXT NOT NULL,
|
||||
|
|
@ -282,7 +282,7 @@ describe('buildLocalStatsStatus', () => {
|
|||
PRIMARY KEY (connection_id, source_name)
|
||||
);
|
||||
INSERT INTO local_sl_sources VALUES
|
||||
('analytics', 'orders', '', NULL, NULL, '2026-05-10T10:00:00Z'),
|
||||
('analytics', 'orders', '', '[0.1,0.2]', NULL, '2026-05-10T10:00:00Z'),
|
||||
('analytics', 'users', '', NULL, NULL, '2026-05-10T10:00:00Z');
|
||||
|
||||
CREATE TABLE local_sl_dictionary_values (
|
||||
|
|
@ -308,18 +308,21 @@ describe('buildLocalStatsStatus', () => {
|
|||
{ connectionId: 'analytics', adapter: 'live-database', lastCompletedAt: '2026-05-10T10:00:00Z' },
|
||||
{ connectionId: 'docs', adapter: 'notion', lastCompletedAt: '2026-05-01T10:00:00Z' },
|
||||
]);
|
||||
expect(stats.knowledgePages).toEqual([
|
||||
{ scope: 'GLOBAL', count: 2 },
|
||||
{ scope: 'PROJECT', count: 1 },
|
||||
expect(stats.wikiPages).toEqual([
|
||||
{ scope: 'GLOBAL', count: 2, embeddedCount: 1 },
|
||||
{ scope: 'PROJECT', count: 1, embeddedCount: 1 },
|
||||
]);
|
||||
expect(stats.semanticLayer).toEqual([
|
||||
{ connectionId: 'analytics', sourceCount: 2, dictionaryValueCount: 2 },
|
||||
{
|
||||
connectionId: 'analytics',
|
||||
sourceCount: 2,
|
||||
embeddedSourceCount: 1,
|
||||
dictionaryValueCount: 2,
|
||||
},
|
||||
]);
|
||||
expect(stats.projectDir.dbSqliteBytes).toBeGreaterThan(0);
|
||||
expect(stats.projectDir.ktxCacheBytes).toBe(2048);
|
||||
expect(stats.projectDir.rawSources).toEqual({ fileCount: 2, bytes: 612 });
|
||||
expect(stats.projectDir.wikiGlobalMarkdownCount).toBe(2);
|
||||
expect(stats.projectDir.semanticLayerYamlCount).toBe(2);
|
||||
});
|
||||
|
||||
it('tolerates a SQLite DB missing some tables', async () => {
|
||||
|
|
@ -344,7 +347,7 @@ describe('buildLocalStatsStatus', () => {
|
|||
const stats = await buildLocalStatsStatus(projectIn(tempDir));
|
||||
expect(stats.unavailable).toBeUndefined();
|
||||
expect(stats.ingest.totalCompletedRuns).toBe(1);
|
||||
expect(stats.knowledgePages).toEqual([]);
|
||||
expect(stats.wikiPages).toEqual([]);
|
||||
expect(stats.semanticLayer).toEqual([]);
|
||||
});
|
||||
});
|
||||
|
|
@ -360,31 +363,36 @@ describe('renderProjectStatus Local data', () => {
|
|||
{ connectionId: 'analytics', adapter: 'live-database', lastCompletedAt: new Date(Date.now() - 60 * 60 * 1000).toISOString() },
|
||||
],
|
||||
},
|
||||
knowledgePages: [
|
||||
{ scope: 'GLOBAL', count: 2 },
|
||||
{ scope: 'PROJECT', count: 1 },
|
||||
wikiPages: [
|
||||
{ scope: 'GLOBAL', count: 2, embeddedCount: 2 },
|
||||
{ scope: 'PROJECT', count: 1, embeddedCount: 0 },
|
||||
],
|
||||
semanticLayer: [
|
||||
{ connectionId: 'analytics', sourceCount: 12, dictionaryValueCount: 200 },
|
||||
{
|
||||
connectionId: 'analytics',
|
||||
sourceCount: 12,
|
||||
embeddedSourceCount: 10,
|
||||
dictionaryValueCount: 200,
|
||||
},
|
||||
],
|
||||
projectDir: {
|
||||
dbSqliteBytes: 4096,
|
||||
ktxCacheBytes: 1_048_576,
|
||||
rawSources: { fileCount: 5, bytes: 200 },
|
||||
wikiGlobalMarkdownCount: 7,
|
||||
semanticLayerYamlCount: 3,
|
||||
},
|
||||
};
|
||||
const rendered = renderProjectStatus(status, { useColor: false });
|
||||
expect(rendered).toContain('Local data');
|
||||
expect(rendered).toContain('Wiki');
|
||||
expect(rendered).not.toContain('Knowledge');
|
||||
expect(rendered).toContain('3 completed runs');
|
||||
expect(rendered).toContain('GLOBAL=2');
|
||||
expect(rendered).toContain('PROJECT=1');
|
||||
expect(rendered).toContain('12 sources · 200 dictionary values');
|
||||
expect(rendered).toContain('GLOBAL=2 (2 embedded)');
|
||||
expect(rendered).toContain('PROJECT=1 (0 embedded)');
|
||||
expect(rendered).toContain('12 sources (10 embedded) · 200 dictionary values');
|
||||
expect(rendered).toContain('db=4.00 KiB');
|
||||
expect(rendered).toContain('cache=1.00 MiB');
|
||||
expect(rendered).toContain('wiki=7 md');
|
||||
expect(rendered).toContain('semantic-layer=3 yaml');
|
||||
expect(rendered).not.toMatch(/wiki=\d+ md/);
|
||||
expect(rendered).not.toMatch(/semantic-layer=\d+ yaml/);
|
||||
});
|
||||
|
||||
it('renders unavailable note when DB is missing', async () => {
|
||||
|
|
@ -392,14 +400,12 @@ describe('renderProjectStatus Local data', () => {
|
|||
const status = await buildProjectStatus(project, { claudeCodeAuthProbe: stubClaudeCodeAuthProbe });
|
||||
status.localStats = {
|
||||
ingest: { totalCompletedRuns: 0, perConnection: [] },
|
||||
knowledgePages: [],
|
||||
wikiPages: [],
|
||||
semanticLayer: [],
|
||||
projectDir: {
|
||||
dbSqliteBytes: null,
|
||||
ktxCacheBytes: 0,
|
||||
rawSources: { fileCount: 0, bytes: 0 },
|
||||
wikiGlobalMarkdownCount: 0,
|
||||
semanticLayerYamlCount: 0,
|
||||
},
|
||||
unavailable: 'no .ktx/db.sqlite yet',
|
||||
};
|
||||
|
|
|
|||
|
|
@ -1,4 +1,3 @@
|
|||
import type { Dirent } from 'node:fs';
|
||||
import { stat as statAsync, readdir as readdirAsync } from 'node:fs/promises';
|
||||
import { basename, join } from 'node:path';
|
||||
import { runClaudeCodeAuthProbe } from './context/llm/claude-code-runtime.js';
|
||||
|
|
@ -105,20 +104,20 @@ interface LocalStatsIngestPerConnection {
|
|||
interface LocalStatsSemanticLayerEntry {
|
||||
connectionId: string;
|
||||
sourceCount: number;
|
||||
embeddedSourceCount: number;
|
||||
dictionaryValueCount: number;
|
||||
}
|
||||
|
||||
interface LocalStatsKnowledgeEntry {
|
||||
interface LocalStatsWikiEntry {
|
||||
scope: string;
|
||||
count: number;
|
||||
embeddedCount: number;
|
||||
}
|
||||
|
||||
interface LocalStatsProjectDir {
|
||||
dbSqliteBytes: number | null;
|
||||
ktxCacheBytes: number;
|
||||
rawSources: { fileCount: number; bytes: number };
|
||||
wikiGlobalMarkdownCount: number;
|
||||
semanticLayerYamlCount: number;
|
||||
}
|
||||
|
||||
/** @internal */
|
||||
|
|
@ -127,7 +126,7 @@ export interface LocalStatsStatus {
|
|||
totalCompletedRuns: number;
|
||||
perConnection: LocalStatsIngestPerConnection[];
|
||||
};
|
||||
knowledgePages: LocalStatsKnowledgeEntry[];
|
||||
wikiPages: LocalStatsWikiEntry[];
|
||||
semanticLayer: LocalStatsSemanticLayerEntry[];
|
||||
projectDir: LocalStatsProjectDir;
|
||||
unavailable?: string;
|
||||
|
|
@ -774,16 +773,12 @@ interface DirSummary {
|
|||
bytes: number;
|
||||
}
|
||||
|
||||
async function summarizeDir(
|
||||
dir: string,
|
||||
filter?: (entry: Dirent, fullPath: string) => boolean,
|
||||
maxDepth = 10,
|
||||
): Promise<DirSummary> {
|
||||
async function summarizeDir(dir: string, maxDepth = 10): Promise<DirSummary> {
|
||||
let fileCount = 0;
|
||||
let bytes = 0;
|
||||
const walk = async (current: string, depth: number): Promise<void> => {
|
||||
if (depth > maxDepth) return;
|
||||
let entries: Dirent[];
|
||||
let entries;
|
||||
try {
|
||||
entries = await readdirAsync(current, { withFileTypes: true });
|
||||
} catch {
|
||||
|
|
@ -796,7 +791,6 @@ async function summarizeDir(
|
|||
continue;
|
||||
}
|
||||
if (!entry.isFile()) continue;
|
||||
if (filter && !filter(entry, full)) continue;
|
||||
try {
|
||||
const s = await statAsync(full);
|
||||
fileCount += 1;
|
||||
|
|
@ -810,14 +804,6 @@ async function summarizeDir(
|
|||
return { fileCount, bytes };
|
||||
}
|
||||
|
||||
function isMarkdownEntry(entry: Dirent): boolean {
|
||||
return entry.isFile() && /\.mdx?$/i.test(entry.name);
|
||||
}
|
||||
|
||||
function isYamlEntry(entry: Dirent): boolean {
|
||||
return entry.isFile() && /\.ya?ml$/i.test(entry.name);
|
||||
}
|
||||
|
||||
async function fileSizeOrNull(filePath: string): Promise<number | null> {
|
||||
try {
|
||||
const s = await statAsync(filePath);
|
||||
|
|
@ -844,18 +830,12 @@ export async function buildLocalStatsStatus(project: KtxLocalProject): Promise<L
|
|||
dbSqliteBytes,
|
||||
ktxCacheBytes: (await summarizeDir(join(project.projectDir, '.ktx', 'cache'))).bytes,
|
||||
rawSources: await summarizeDir(join(project.projectDir, 'raw-sources')),
|
||||
wikiGlobalMarkdownCount: (
|
||||
await summarizeDir(join(project.projectDir, 'wiki', 'global'), isMarkdownEntry)
|
||||
).fileCount,
|
||||
semanticLayerYamlCount: (
|
||||
await summarizeDir(join(project.projectDir, 'semantic-layer'), isYamlEntry)
|
||||
).fileCount,
|
||||
};
|
||||
|
||||
if (dbSqliteBytes === null) {
|
||||
return {
|
||||
ingest: { totalCompletedRuns: 0, perConnection: [] },
|
||||
knowledgePages: [],
|
||||
wikiPages: [],
|
||||
semanticLayer: [],
|
||||
projectDir: projectDirSummary,
|
||||
unavailable: 'no .ktx/db.sqlite yet',
|
||||
|
|
@ -905,28 +885,34 @@ export async function buildLocalStatsStatus(project: KtxLocalProject): Promise<L
|
|||
left.connectionId.localeCompare(right.connectionId),
|
||||
);
|
||||
|
||||
const knowledgeRows = tryQuery(
|
||||
const wikiRows = tryQuery(
|
||||
() =>
|
||||
db
|
||||
.prepare(
|
||||
`SELECT scope, COUNT(*) AS n FROM knowledge_pages GROUP BY scope ORDER BY scope`,
|
||||
`SELECT scope, COUNT(*) AS n, SUM(CASE WHEN embedding_json IS NOT NULL THEN 1 ELSE 0 END) AS embedded
|
||||
FROM knowledge_pages
|
||||
GROUP BY scope
|
||||
ORDER BY scope`,
|
||||
)
|
||||
.all() as Array<{ scope: string; n: number }>,
|
||||
[] as Array<{ scope: string; n: number }>,
|
||||
.all() as Array<{ scope: string; n: number; embedded: number | null }>,
|
||||
[] as Array<{ scope: string; n: number; embedded: number | null }>,
|
||||
);
|
||||
const knowledgePages: LocalStatsKnowledgeEntry[] = knowledgeRows.map((row) => ({
|
||||
const wikiPages: LocalStatsWikiEntry[] = wikiRows.map((row) => ({
|
||||
scope: row.scope,
|
||||
count: row.n,
|
||||
embeddedCount: row.embedded ?? 0,
|
||||
}));
|
||||
|
||||
const sourceRows = tryQuery(
|
||||
() =>
|
||||
db
|
||||
.prepare(
|
||||
`SELECT connection_id, COUNT(*) AS n FROM local_sl_sources GROUP BY connection_id`,
|
||||
`SELECT connection_id, COUNT(*) AS n, SUM(CASE WHEN embedding_json IS NOT NULL THEN 1 ELSE 0 END) AS embedded
|
||||
FROM local_sl_sources
|
||||
GROUP BY connection_id`,
|
||||
)
|
||||
.all() as Array<{ connection_id: string; n: number }>,
|
||||
[] as Array<{ connection_id: string; n: number }>,
|
||||
.all() as Array<{ connection_id: string; n: number; embedded: number | null }>,
|
||||
[] as Array<{ connection_id: string; n: number; embedded: number | null }>,
|
||||
);
|
||||
const dictionaryRows = tryQuery(
|
||||
() =>
|
||||
|
|
@ -942,6 +928,7 @@ export async function buildLocalStatsStatus(project: KtxLocalProject): Promise<L
|
|||
slMap.set(row.connection_id, {
|
||||
connectionId: row.connection_id,
|
||||
sourceCount: row.n,
|
||||
embeddedSourceCount: row.embedded ?? 0,
|
||||
dictionaryValueCount: 0,
|
||||
});
|
||||
}
|
||||
|
|
@ -949,6 +936,7 @@ export async function buildLocalStatsStatus(project: KtxLocalProject): Promise<L
|
|||
const existing = slMap.get(row.connection_id) ?? {
|
||||
connectionId: row.connection_id,
|
||||
sourceCount: 0,
|
||||
embeddedSourceCount: 0,
|
||||
dictionaryValueCount: 0,
|
||||
};
|
||||
existing.dictionaryValueCount = row.n;
|
||||
|
|
@ -960,14 +948,14 @@ export async function buildLocalStatsStatus(project: KtxLocalProject): Promise<L
|
|||
|
||||
return {
|
||||
ingest: { totalCompletedRuns, perConnection },
|
||||
knowledgePages,
|
||||
wikiPages,
|
||||
semanticLayer,
|
||||
projectDir: projectDirSummary,
|
||||
};
|
||||
} catch (error) {
|
||||
return {
|
||||
ingest: { totalCompletedRuns: 0, perConnection: [] },
|
||||
knowledgePages: [],
|
||||
wikiPages: [],
|
||||
semanticLayer: [],
|
||||
projectDir: projectDirSummary,
|
||||
unavailable: failureDetail(error),
|
||||
|
|
@ -1116,7 +1104,7 @@ function renderLocalStats(
|
|||
|
||||
const localLabelWidth = Math.max(
|
||||
'Ingest'.length,
|
||||
'Knowledge'.length,
|
||||
'Wiki'.length,
|
||||
'Semantic layer'.length,
|
||||
'Disk'.length,
|
||||
);
|
||||
|
|
@ -1138,13 +1126,13 @@ function renderLocalStats(
|
|||
}
|
||||
}
|
||||
|
||||
if (stats.knowledgePages.length === 0) {
|
||||
lines.push(` ${lLabel('Knowledge')} ${dim('no pages yet')}`);
|
||||
if (stats.wikiPages.length === 0) {
|
||||
lines.push(` ${lLabel('Wiki')} ${dim('no pages yet')}`);
|
||||
} else {
|
||||
const knowledgeText = stats.knowledgePages
|
||||
.map((entry) => `${entry.scope}=${entry.count}`)
|
||||
const wikiText = stats.wikiPages
|
||||
.map((entry) => `${entry.scope}=${entry.count} ${dim(`(${entry.embeddedCount} embedded)`)}`)
|
||||
.join(` ${dim('·')} `);
|
||||
lines.push(` ${lLabel('Knowledge')} ${knowledgeText}`);
|
||||
lines.push(` ${lLabel('Wiki')} ${wikiText}`);
|
||||
}
|
||||
|
||||
if (stats.semanticLayer.length === 0) {
|
||||
|
|
@ -1154,8 +1142,10 @@ function renderLocalStats(
|
|||
let firstLine = true;
|
||||
for (const entry of stats.semanticLayer) {
|
||||
const prefix = firstLine ? lLabel('Semantic layer') : ' '.repeat(localLabelWidth);
|
||||
const sourcesText = `${entry.sourceCount} source${entry.sourceCount === 1 ? '' : 's'} (${entry.embeddedSourceCount} embedded)`;
|
||||
const dictText = `${entry.dictionaryValueCount} dictionary value${entry.dictionaryValueCount === 1 ? '' : 's'}`;
|
||||
lines.push(
|
||||
` ${prefix} ${entry.connectionId.padEnd(nameWidth)} ${dim(`${entry.sourceCount} source${entry.sourceCount === 1 ? '' : 's'} · ${entry.dictionaryValueCount} dictionary value${entry.dictionaryValueCount === 1 ? '' : 's'}`)}`,
|
||||
` ${prefix} ${entry.connectionId.padEnd(nameWidth)} ${dim(`${sourcesText} · ${dictText}`)}`,
|
||||
);
|
||||
firstLine = false;
|
||||
}
|
||||
|
|
@ -1168,8 +1158,6 @@ function renderLocalStats(
|
|||
diskBits.push(
|
||||
`raw-sources=${disk.rawSources.fileCount} file${disk.rawSources.fileCount === 1 ? '' : 's'} (${formatBytes(disk.rawSources.bytes)})`,
|
||||
);
|
||||
diskBits.push(`wiki=${disk.wikiGlobalMarkdownCount} md`);
|
||||
diskBits.push(`semantic-layer=${disk.semanticLayerYamlCount} yaml`);
|
||||
lines.push(` ${lLabel('Disk')} ${dim(diskBits.join(` ${dim('·')} `))}`);
|
||||
lines.push('');
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue