xindex 1.0.5 → 1.0.6

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/.tmp/watch.log ADDED
@@ -0,0 +1,9 @@
1
+ [xindex-1776976173089-3ab4dd18cfbfc8] started, indexing and watching: .
2
+ indexing: notes.md
3
+ indexing: util.ts
4
+ initial index complete, acquiring watcher lock...
5
+ done. indexed: {"indexedAmount":2}
6
+ watching for changes: .
7
+ index: wombat.md
8
+ remove: functional
9
+ remove batch failed: Error saving index: Error: ENOENT: no such file or directory, open '.xindex/semantic/index.json'
package/README.md CHANGED
@@ -6,6 +6,8 @@ Local semantic code search for your codebase — plus an MCP server so Claude Co
6
6
 
7
7
  ## Install
8
8
 
9
+ Requires **Node ≥22**.
10
+
9
11
  ```bash
10
12
  npm i -g xindex
11
13
  ```
@@ -47,6 +49,7 @@ Open the project in Claude Code — it picks up the xindex MCP server and can ca
47
49
  - **Watch mode** — keeps the index warm while you code
48
50
  - **Gitignore-aware** — respects `.gitignore` + custom ignore rules
49
51
  - **Zero config** — works with defaults; `.xindex.json` is optional
52
+ - **Tolerant** — tolerates unreadable files, oversize files, empty files, binary files, and symlinks; each is skipped with a log line so the run always finishes
50
53
 
51
54
  ## Claude Code skills (`@xi`)
52
55
 
@@ -171,7 +174,8 @@ Project-root file. All fields optional; unknown keys ignored; missing/empty →
171
174
  - **`ignoreKeywords`** — `string[]`, default `[]`. Tokens stripped before embedding — add project slang/boilerplate polluting results. Entries ≤1 char warn.
172
175
  - **`ignoreFiles`** — `string[]`, default `[]`. Extra globs excluded during walk/watch, on top of `.gitignore` — add vendored/generated folders.
173
176
  - **`maxLines`** — `number`, default `30`. Lines per chunk — tune if chunks feel over/under-sized.
174
- - **`maxFileBytes`** — `number`, default `5000000`. Skip files over this (5 MB) — lower for faster indexing on huge generated files.
177
+ - **`maxFileBytes`** — `number`, default `1000000`. Skip files over this (1 MB) — raise to index larger generated files.
178
+ - **`followSymlinks`** — `boolean`, default `false`. When `false`, symbolic links encountered during walk/watch are skipped with a log line. Set `true` to follow them (cycles are broken via `realpath` dedup).
175
179
 
176
180
  ```json
177
181
  {
package/apps/indexApp.ts CHANGED
@@ -1,4 +1,3 @@
1
- import {readFile} from "fs/promises";
2
1
  import {from} from "../packages/streamx/src/from.js";
3
2
  import {batch} from "../packages/streamx/src/batch.js";
4
3
  import {map} from "../packages/streamx/src/map.js";
@@ -8,23 +7,24 @@ import {IWalkFiles} from "../componets/walkFiles.js";
8
7
  import {IIndexContent} from "../features/indexContent.js";
9
8
  import {ILogger} from "../componets/logger.js";
10
9
  import {INDEXING_BATCH_SIZE} from "../componets/config/INDEXING_BATCH_SIZE";
10
+ import {SafeIndexBatch} from "../componets/io/safeIndexBatch.js";
11
11
 
12
12
  export type IIndexApp = (inputs: string[]) => Promise<void>;
13
13
 
14
- export function IndexApp({walkFiles, indexContent, log}: {
14
+ export function IndexApp({walkFiles, indexContent, log, maxFileBytes}: {
15
15
  walkFiles: IWalkFiles,
16
16
  indexContent: IIndexContent,
17
17
  log: ILogger,
18
+ maxFileBytes: number,
18
19
  }): IIndexApp {
20
+ const safeIndexBatch = SafeIndexBatch({indexContent, log, maxFileBytes});
19
21
  return async function indexApp(inputs) {
20
22
  await run(
21
23
  from(walkFiles(inputs))
22
24
  .pipe(tap(id => log(`indexing: ${id}`)))
23
25
  .pipe(batch(INDEXING_BATCH_SIZE))
24
26
  .pipe(map<string[], string[]>(async (ids) => {
25
- const texts = await Promise.all(ids.map(id => readFile(id, "utf8")));
26
- const items = ids.map((id, i) => ({id, content: `${texts[i]}. ${id}`}));
27
- await indexContent(items);
27
+ await safeIndexBatch(ids);
28
28
  return ids;
29
29
  }))
30
30
  );
package/apps/run.index.ts CHANGED
@@ -8,8 +8,8 @@ const appId = AppId();
8
8
  const cwd = process.cwd();
9
9
  const log = BufferedLoggerToStdOut();
10
10
  const {indexContent, getIndexStats, config} = await BuildComponents({log});
11
- const walkFiles = WalkFiles({cwd, log, ignoreFiles: config.ignoreFiles});
12
- const indexApp = IndexApp({walkFiles, indexContent, log});
11
+ const walkFiles = WalkFiles({cwd, log, ignoreFiles: config.ignoreFiles, followSymlinks: config.followSymlinks});
12
+ const indexApp = IndexApp({walkFiles, indexContent, log, maxFileBytes: config.maxFileBytes});
13
13
 
14
14
  const inputs = process.argv.slice(2);
15
15
  if (!inputs.length) inputs.push(".");
package/apps/run.mcp.ts CHANGED
@@ -17,8 +17,8 @@ const watchDisabled = args.includes("--watch-disabled");
17
17
  const cwd = process.cwd();
18
18
  const log = BufferedLoggerToStdErr();
19
19
  const {indexContent, removeContent, getIndexStats, searchContentIndex, resetIndex, config} = await BuildComponents({log});
20
- const walkFiles = WalkFiles({cwd, log, ignoreFiles: config.ignoreFiles});
21
- const indexApp = IndexApp({walkFiles, indexContent, log});
20
+ const walkFiles = WalkFiles({cwd, log, ignoreFiles: config.ignoreFiles, followSymlinks: config.followSymlinks});
21
+ const indexApp = IndexApp({walkFiles, indexContent, log, maxFileBytes: config.maxFileBytes});
22
22
  const search = SearchApp({searchContentIndex});
23
23
 
24
24
  const appId = AppId();
@@ -31,7 +31,7 @@ const watcherLock = WatcherLock({
31
31
  const watch = watchDisabled ? undefined : {
32
32
  watchFiles: WatchFiles({cwd, log, ignoreFiles: config.ignoreFiles}),
33
33
  watchDir: watchDirArg ? watchDirArg.split("=")[1] : ".",
34
- handleFileEvents: HandleFileEvents({indexContent, removeContent, log}),
34
+ handleFileEvents: HandleFileEvents({indexContent, removeContent, log, maxFileBytes: config.maxFileBytes}),
35
35
  watcherLock,
36
36
  };
37
37
 
package/apps/run.watch.ts CHANGED
@@ -11,9 +11,9 @@ import {join} from "path";
11
11
  const cwd = process.cwd();
12
12
  const log = BufferedLoggerToStdOut();
13
13
  const {indexContent, removeContent, getIndexStats, config} = await BuildComponents({log});
14
- const walkFiles = WalkFiles({cwd, log, ignoreFiles: config.ignoreFiles});
14
+ const walkFiles = WalkFiles({cwd, log, ignoreFiles: config.ignoreFiles, followSymlinks: config.followSymlinks});
15
15
  const watchFiles = WatchFiles({cwd, log, ignoreFiles: config.ignoreFiles});
16
- const handleFileEvents = HandleFileEvents({indexContent, removeContent, log});
16
+ const handleFileEvents = HandleFileEvents({indexContent, removeContent, log, maxFileBytes: config.maxFileBytes});
17
17
 
18
18
  const appId = AppId();
19
19
  const watcherLock = WatcherLock({
@@ -22,7 +22,7 @@ const watcherLock = WatcherLock({
22
22
  log,
23
23
  });
24
24
 
25
- const app = WatchApp({walkFiles, watchFiles, handleFileEvents, indexContent, log, watcherLock});
25
+ const app = WatchApp({walkFiles, watchFiles, handleFileEvents, indexContent, log, watcherLock, maxFileBytes: config.maxFileBytes});
26
26
 
27
27
  const inputs = process.argv.slice(2);
28
28
  if (!inputs.length) inputs.push(".");
package/apps/watchApp.ts CHANGED
@@ -3,7 +3,6 @@ import {batch} from "../packages/streamx/src/batch.js";
3
3
  import {map} from "../packages/streamx/src/map.js";
4
4
  import {tap} from "../packages/streamx/src/tap.js";
5
5
  import {run} from "../packages/streamx/src/index.js";
6
- import {readFile} from "fs/promises";
7
6
  import {IWalkFiles} from "../componets/walkFiles.js";
8
7
  import {IWatchFiles} from "../componets/watchFiles.js";
9
8
  import {IHandleFileEvents} from "../componets/index/handleFileEvent.js";
@@ -12,20 +11,23 @@ import {IWatcherLock} from "../componets/index/watcherLock.js";
12
11
  import {WatchFileEventsApp} from "./watchFileEventsApp.js";
13
12
  import {IIndexContent} from "../features/indexContent.js";
14
13
  import {INDEXING_BATCH_SIZE} from "../componets/config/INDEXING_BATCH_SIZE";
14
+ import {SafeIndexBatch} from "../componets/io/safeIndexBatch.js";
15
15
 
16
16
  export type IWatchApp = {
17
17
  run: (inputs: string[]) => Promise<void>;
18
18
  stop: () => void;
19
19
  };
20
20
 
21
- export function WatchApp({walkFiles, watchFiles, handleFileEvents, indexContent, log, watcherLock}: {
21
+ export function WatchApp({walkFiles, watchFiles, handleFileEvents, indexContent, log, watcherLock, maxFileBytes}: {
22
22
  walkFiles: IWalkFiles,
23
23
  watchFiles: IWatchFiles,
24
24
  handleFileEvents: IHandleFileEvents,
25
25
  indexContent: IIndexContent,
26
26
  log: ILogger,
27
27
  watcherLock: IWatcherLock,
28
+ maxFileBytes: number,
28
29
  }): IWatchApp {
30
+ const safeIndexBatch = SafeIndexBatch({indexContent, log, maxFileBytes});
29
31
  return {
30
32
  async run(inputs) {
31
33
  // Phase 1: initial index — walk all files
@@ -34,9 +36,7 @@ export function WatchApp({walkFiles, watchFiles, handleFileEvents, indexContent,
34
36
  .pipe(tap(id => log(`indexing: ${id}`)))
35
37
  .pipe(batch(INDEXING_BATCH_SIZE))
36
38
  .pipe(map<string[], string[]>(async (ids) => {
37
- const texts = await Promise.all(ids.map(id => readFile(id, "utf8")));
38
- const items = ids.map((id, i) => ({id, content: `${texts[i]}. ${id}`}));
39
- await indexContent(items);
39
+ await safeIndexBatch(ids);
40
40
  return ids;
41
41
  }))
42
42
  );
@@ -33,7 +33,8 @@ export async function BuildComponents({log}: { log: ILogger }) {
33
33
  extractKeywords,
34
34
  cleanUpKeywords,
35
35
  locateInFile,
36
- scoreThreshold: SCORE_THRESHOLD
36
+ scoreThreshold: SCORE_THRESHOLD,
37
+ log,
37
38
  });
38
39
 
39
40
  return {
@@ -1 +1 @@
1
- export const INDEXING_BATCH_SIZE = 5;
1
+ export const INDEXING_BATCH_SIZE = 16;
@@ -1 +1 @@
1
- export const WATCH_FLUSH_MS = 500;
1
+ export const WATCH_FLUSH_MS = 300;
@@ -3,13 +3,15 @@ import {IXindexConfig} from "./xindexConfig.js";
3
3
  import {ILogger} from "../logger.js";
4
4
 
5
5
  const DEFAULT_MAX_LINES = 30;
6
- const DEFAULT_MAX_FILE_BYTES = 5_000_000;
6
+ const DEFAULT_MAX_FILE_BYTES = 1_000_000;
7
+ const DEFAULT_FOLLOW_SYMLINKS = false;
7
8
 
8
9
  const DEFAULTS: IXindexConfig = {
9
10
  ignoreKeywords: [],
10
11
  ignoreFiles: [],
11
12
  maxLines: DEFAULT_MAX_LINES,
12
13
  maxFileBytes: DEFAULT_MAX_FILE_BYTES,
14
+ followSymlinks: DEFAULT_FOLLOW_SYMLINKS,
13
15
  };
14
16
 
15
17
  export type ILoadConfig = () => Promise<IXindexConfig>;
@@ -41,11 +43,18 @@ export function LoadConfig({configPath, log}: { configPath: string, log: ILogger
41
43
  ignoreFiles: toStrings(parsed.ignoreFiles),
42
44
  maxLines: toNum(parsed.maxLines, DEFAULT_MAX_LINES),
43
45
  maxFileBytes: toNum(parsed.maxFileBytes, DEFAULT_MAX_FILE_BYTES),
46
+ followSymlinks: typeof parsed.followSymlinks === "boolean" ? parsed.followSymlinks : DEFAULT_FOLLOW_SYMLINKS,
44
47
  };
45
48
 
46
49
  for (const kw of config.ignoreKeywords) {
47
50
  if (kw.length <= 1) log(`warning: ignoreKeywords entry "${kw}" is <=1 char`);
48
51
  }
52
+ if (config.maxFileBytes < 1024) {
53
+ log(`warning: maxFileBytes (${config.maxFileBytes}) < 1024 — likely a typo`);
54
+ }
55
+ if (parsed.followSymlinks !== undefined && typeof parsed.followSymlinks !== "boolean") {
56
+ log(`warning: followSymlinks must be boolean — using default (${DEFAULT_FOLLOW_SYMLINKS})`);
57
+ }
49
58
 
50
59
  return config;
51
60
  };
@@ -3,4 +3,5 @@ export type IXindexConfig = {
3
3
  ignoreFiles: string[];
4
4
  maxLines: number;
5
5
  maxFileBytes: number;
6
+ followSymlinks: boolean;
6
7
  };
@@ -9,6 +9,7 @@ import {ResetIndex, IResetIndex} from "../../features/resetIndex.js";
9
9
  import {VectraIndex} from "./vectraIndex.js";
10
10
  import {IndexApi} from "./indexApi.js";
11
11
  import {ILocateInFile} from "../locate/locateInFile.js";
12
+ import {ILogger} from "../logger.js";
12
13
 
13
14
  export type IContentIndexDriver = Readonly<{
14
15
  getIndexStats: IGetIndexStats,
@@ -19,20 +20,21 @@ export type IContentIndexDriver = Readonly<{
19
20
  flush: () => Promise<void>,
20
21
  }>;
21
22
 
22
- export async function ContentIndexDriver({path, embed, extractKeywords, cleanUpKeywords, locateInFile, scoreThreshold}: {
23
+ export async function ContentIndexDriver({path, embed, extractKeywords, cleanUpKeywords, locateInFile, scoreThreshold, log}: {
23
24
  path: string,
24
25
  embed: IEmbed,
25
26
  extractKeywords: IExtractKeywords,
26
27
  cleanUpKeywords: ICleanUpKeywords,
27
28
  locateInFile: ILocateInFile,
28
29
  scoreThreshold: number,
30
+ log: ILogger,
29
31
  }): Promise<IContentIndexDriver> {
30
32
  const index = await VectraIndex(path + "/semantic");
31
33
  const indexApi = IndexApi({index});
32
34
 
33
35
  return {
34
36
  getIndexStats: GetIndexStats({index}),
35
- indexContent: IndexContent({extractKeywords, cleanUpKeywords, embed, indexApi}),
37
+ indexContent: IndexContent({extractKeywords, cleanUpKeywords, embed, indexApi, log}),
36
38
  removeContent: RemoveContent({indexApi}),
37
39
  searchContentIndex: SearchIndex({extractKeywords, cleanUpKeywords, embed, index, locateInFile, scoreThreshold}),
38
40
  resetIndex: ResetIndex({indexApi}),
@@ -1,22 +1,22 @@
1
- import {readFile} from "fs/promises";
2
1
  import {IIndexContent} from "../../features/indexContent.js";
3
2
  import {IRemoveContent} from "../../features/removeContent.js";
4
3
  import {ILogger} from "../logger.js";
5
4
  import {FileEventType, IFileEvent} from "../watchFiles.js";
5
+ import {SafeIndexBatch} from "../io/safeIndexBatch.js";
6
6
 
7
7
  export type IHandleFileEvent = (event: IFileEvent) => Promise<void>;
8
8
  export type IHandleFileEvents = (events: IFileEvent[]) => Promise<void>;
9
9
 
10
- export function HandleFileEvent({indexContent, removeContent, log}: {
10
+ export function HandleFileEvent({indexContent, removeContent, log, maxFileBytes}: {
11
11
  indexContent: IIndexContent,
12
12
  removeContent: IRemoveContent,
13
13
  log: ILogger,
14
+ maxFileBytes: number,
14
15
  }): IHandleFileEvent {
16
+ const safeIndexBatch = SafeIndexBatch({indexContent, log, maxFileBytes});
15
17
  return async function handleFileEvent(event) {
16
18
  if (event.type === FileEventType.index) {
17
- const text = await readFile(event.path, "utf8");
18
- await indexContent([{id: event.path, content: `${text}. ${event.path}`}]);
19
- log(`index: ${event.path}`);
19
+ await safeIndexBatch([event.path]);
20
20
  } else {
21
21
  try { await removeContent([event.path]); } catch (e) { log(`remove failed: ${event.path} — ${(e as any)?.message ?? e}`); }
22
22
  log(`remove: ${event.path}`);
@@ -24,42 +24,22 @@ export function HandleFileEvent({indexContent, removeContent, log}: {
24
24
  };
25
25
  }
26
26
 
27
- export function HandleFileEvents({indexContent, removeContent, log}: {
27
+ export function HandleFileEvents({indexContent, removeContent, log, maxFileBytes}: {
28
28
  indexContent: IIndexContent,
29
29
  removeContent: IRemoveContent,
30
30
  log: ILogger,
31
+ maxFileBytes: number,
31
32
  }): IHandleFileEvents {
33
+ const safeIndexBatch = SafeIndexBatch({indexContent, log, maxFileBytes});
32
34
  return async function handleFileEvents(events) {
33
- const indexEvents = events.filter((event) => event.type === FileEventType.index);
34
- const removeEvents = events.filter((event) => event.type === FileEventType.remove);
35
+ const indexPaths = events.filter(e => e.type === FileEventType.index).map(e => e.path);
36
+ const removePaths = events.filter(e => e.type === FileEventType.remove).map(e => e.path);
35
37
 
36
- if (indexEvents.length > 0) {
37
- const indexItems = (await Promise.all(indexEvents.map(async (event) => {
38
- log(`index: ${event.path}`);
39
- try {
40
- const text = await readFile(event.path, "utf8");
41
- return {id: event.path, content: `${text}. ${event.path}`};
42
- } catch (e) {
43
- log(`index failed: ${event.path} — ${(e as any)?.message ?? e}`);
44
- return undefined;
45
- }
46
- }))).filter((item): item is { id: string, content: string } => !!item);
47
-
48
- if (indexItems.length > 0) {
49
- try {
50
- await indexContent(indexItems);
51
- } catch (e) {
52
- log(`index batch failed: ${(e as any)?.message ?? e}`);
53
- }
54
- }
55
- }
56
-
57
- if (removeEvents.length > 0) {
58
- const removePaths = removeEvents.map((event) => {
59
- log(`remove: ${event.path}`);
60
- return event.path;
61
- });
38
+ for (const p of indexPaths) log(`index: ${p}`);
39
+ if (indexPaths.length > 0) await safeIndexBatch(indexPaths);
62
40
 
41
+ if (removePaths.length > 0) {
42
+ for (const p of removePaths) log(`remove: ${p}`);
63
43
  try {
64
44
  await removeContent(removePaths);
65
45
  } catch (e) {
@@ -0,0 +1,30 @@
1
+ import {safeReadFile, SafeReadResultType} from "./safeReadFile.js";
2
+ import {IIndexContent} from "../../features/indexContent.js";
3
+ import {ILogger} from "../logger.js";
4
+
5
+ export type ISafeIndexBatch = (ids: string[]) => Promise<void>;
6
+
7
+ export function SafeIndexBatch({indexContent, log, maxFileBytes}: {
8
+ indexContent: IIndexContent,
9
+ log: ILogger,
10
+ maxFileBytes: number,
11
+ }): ISafeIndexBatch {
12
+ return async function safeIndexBatch(ids) {
13
+ const results = await Promise.all(ids.map(id => safeReadFile(id, maxFileBytes)));
14
+ const ok: {id: string, content: string}[] = [];
15
+ for (const r of results) {
16
+ if (r.type === SafeReadResultType.ok) {
17
+ ok.push({id: r.id, content: `${r.content}. ${r.id}`});
18
+ } else {
19
+ log(`skip ${r.id}: ${r.reason}`);
20
+ }
21
+ }
22
+ if (ok.length > 0) {
23
+ try {
24
+ await indexContent(ok);
25
+ } catch (e: any) {
26
+ log(`index batch failed: ${e?.message ?? e}`);
27
+ }
28
+ }
29
+ };
30
+ }
@@ -0,0 +1,36 @@
1
+ import {stat, readFile} from "fs/promises";
2
+ import {IType} from "../IType.js";
3
+
4
+ export enum SafeReadResultType {
5
+ ok = "ok",
6
+ skip = "skip",
7
+ error = "error",
8
+ }
9
+
10
+ export enum SafeReadSkipReason {
11
+ tooLarge = "tooLarge",
12
+ empty = "empty",
13
+ binary = "binary",
14
+ }
15
+
16
+ export type ISafeReadResult =
17
+ | IType<{ type: SafeReadResultType.ok, id: string, content: string }>
18
+ | IType<{ type: SafeReadResultType.skip, id: string, reason: SafeReadSkipReason }>
19
+ | IType<{ type: SafeReadResultType.error, id: string, reason: string }>;
20
+
21
+ export async function safeReadFile(id: string, maxBytes: number): Promise<ISafeReadResult> {
22
+ try {
23
+ const s = await stat(id);
24
+ if (s.size > maxBytes) return {type: SafeReadResultType.skip, id, reason: SafeReadSkipReason.tooLarge};
25
+ const buf = await readFile(id);
26
+ const probe = buf.subarray(0, Math.min(10, buf.length));
27
+ for (let i = 0; i < probe.length; i++) {
28
+ if (probe[i] === 0) return {type: SafeReadResultType.skip, id, reason: SafeReadSkipReason.binary};
29
+ }
30
+ const content = buf.toString("utf8");
31
+ if (!content.trim()) return {type: SafeReadResultType.skip, id, reason: SafeReadSkipReason.empty};
32
+ return {type: SafeReadResultType.ok, id, content};
33
+ } catch (e: any) {
34
+ return {type: SafeReadResultType.error, id, reason: e?.code ?? e?.message ?? String(e)};
35
+ }
36
+ }
@@ -11,9 +11,9 @@ import {filter} from "../../packages/streamx/src/filter.js";
11
11
  import {map} from "../../packages/streamx/src/map.js";
12
12
  import {flatMap} from "../../packages/streamx/src/flatMap.js";
13
13
  import {tap} from "../../packages/streamx/src/tap.js";
14
- import {scaleSync} from "../../packages/streamx/src/scaleSync.js";
15
14
  import {run} from "../../packages/streamx/src/index.js";
16
15
  import {DEFAULT_LOCATE_BATCH_SIZE} from "../config/DEFAULT_LOCATE_BATCH_SIZE";
16
+ import {scaleSync} from "../../packages/streamx/src/scaleSync";
17
17
 
18
18
  export type ILocateInFile = (
19
19
  query: string,
@@ -61,12 +61,17 @@ export function LocateInFile({
61
61
  return false;
62
62
  }
63
63
  }))
64
- .pipe(scaleSync(SCALE_FILE_READS, async (id: string) => {
65
- const text = await readFile(id, "utf8");
66
- return {id, text};
64
+ .pipe(scaleSync(SCALE_FILE_READS, async (id: string): Promise<{ id: string, text: string } | null> => {
65
+ try {
66
+ const text = await readFile(id, "utf8");
67
+ return {id, text};
68
+ } catch {
69
+ return null;
70
+ }
67
71
  }))
68
- .pipe(flatMap(({id, text}: { id: string, text: string }): IWindow[] => {
69
- return windowsOf({text, id, windowLines});
72
+ .pipe(filter(async (r: { id: string, text: string } | null) => r !== null))
73
+ .pipe(flatMap((r: { id: string, text: string } | null): IWindow[] => {
74
+ return windowsOf({text: r!.text, id: r!.id, windowLines});
70
75
  }));
71
76
 
72
77
  const withVectors = from<IWindow>(windows)
@@ -1,11 +1,16 @@
1
- import {readdir, stat, readFile} from "fs/promises";
1
+ import {readdir, stat, readFile, realpath} from "fs/promises";
2
2
  import {join, relative} from "path";
3
3
  import ignore from "ignore";
4
4
  import {ILogger} from "./logger.js";
5
5
 
6
6
  export type IWalkFiles = (inputs: string[]) => AsyncIterable<string>;
7
7
 
8
- export function WalkFiles({cwd, log, ignoreFiles = []}: {cwd: string, log: ILogger, ignoreFiles?: string[]}): IWalkFiles {
8
+ export function WalkFiles({cwd, log, ignoreFiles = [], followSymlinks = false}: {
9
+ cwd: string,
10
+ log: ILogger,
11
+ ignoreFiles?: string[],
12
+ followSymlinks?: boolean,
13
+ }): IWalkFiles {
9
14
 
10
15
  async function tryReadGitignore(dir: string): Promise<string> {
11
16
  try {
@@ -15,6 +20,8 @@ export function WalkFiles({cwd, log, ignoreFiles = []}: {cwd: string, log: ILogg
15
20
  }
16
21
  }
17
22
 
23
+ const visited = new Set<string>();
24
+
18
25
  async function* walk(dir: string, parentRules: string[]): AsyncIterable<string> {
19
26
  const localGitignore = await tryReadGitignore(dir);
20
27
  const rules = localGitignore ? [...parentRules, localGitignore] : parentRules;
@@ -27,11 +34,35 @@ export function WalkFiles({cwd, log, ignoreFiles = []}: {cwd: string, log: ILogg
27
34
  const abs = join(dir, entry.name);
28
35
  const rel = relative(cwd, abs);
29
36
 
37
+ if (entry.isSymbolicLink()) {
38
+ if (!followSymlinks) {
39
+ log(`skip symlink ${rel}`);
40
+ continue;
41
+ }
42
+ try {
43
+ const real = await realpath(abs);
44
+ if (visited.has(real)) continue;
45
+ visited.add(real);
46
+ const s = await stat(real);
47
+ if (s.isDirectory()) {
48
+ if (entry.name.startsWith(".")) continue;
49
+ if (ig.ignores(rel + "/")) continue;
50
+ yield* walk(abs, rules);
51
+ } else if (s.isFile()) {
52
+ if (ig.ignores(rel)) continue;
53
+ yield rel;
54
+ }
55
+ } catch (e: any) {
56
+ log(`skip symlink ${rel}: ${e?.code ?? e?.message ?? e}`);
57
+ }
58
+ continue;
59
+ }
60
+
30
61
  if (entry.isDirectory()) {
31
62
  if (entry.name.startsWith(".")) continue;
32
63
  if (ig.ignores(rel + "/")) continue;
33
64
  yield* walk(abs, rules);
34
- } else {
65
+ } else if (entry.isFile()) {
35
66
  if (ig.ignores(rel)) continue;
36
67
  yield rel;
37
68
  }
@@ -2,22 +2,31 @@ import {IndexCommandType, IIndexApi} from "../componets/index/indexApi.js";
2
2
  import {IExtractKeywords} from "../componets/keywords/extractKeywords.js";
3
3
  import {ICleanUpKeywords} from "../componets/keywords/cleanUpKeywords.js";
4
4
  import {IEmbed} from "../componets/llm/embed.js";
5
+ import {ILogger} from "../componets/logger.js";
5
6
 
6
7
  export type IIndexContent = (items: Array<{id: string, content: string}>) => Promise<void>;
7
8
 
8
- export function IndexContent({extractKeywords, cleanUpKeywords, embed, indexApi}: {
9
+ export function IndexContent({extractKeywords, cleanUpKeywords, embed, indexApi, log}: {
9
10
  extractKeywords: IExtractKeywords,
10
11
  cleanUpKeywords: ICleanUpKeywords,
11
12
  embed: IEmbed,
12
13
  indexApi: IIndexApi,
14
+ log: ILogger,
13
15
  }): IIndexContent {
14
16
  return async function indexContent(items) {
15
- const indexItems = await Promise.all(items.map(async (item) => {
16
- const keywords = cleanUpKeywords(extractKeywords(item.content)).join(", ");
17
- const vector = await embed(keywords);
18
- return {id: item.id, vector, keywords};
19
- }));
17
+ const indexItems = (await Promise.all(items.map(async (item) => {
18
+ try {
19
+ const keywords = cleanUpKeywords(extractKeywords(item.content)).join(", ");
20
+ const vector = await embed(keywords);
21
+ return {id: item.id, vector, keywords};
22
+ } catch (e: any) {
23
+ log(`skip ${item.id}: keywords/embed failed — ${e?.code ?? e?.message ?? e}`);
24
+ return null;
25
+ }
26
+ }))).filter((x): x is NonNullable<typeof x> => !!x);
20
27
 
21
- await indexApi({type: IndexCommandType.index, items: indexItems});
28
+ if (indexItems.length > 0) {
29
+ await indexApi({type: IndexCommandType.index, items: indexItems});
30
+ }
22
31
  }
23
32
  }
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "xindex",
3
- "version": "1.0.5",
3
+ "version": "1.0.6",
4
4
  "description": "Local semantic code search — index codebase, search by meaning or keywords",
5
5
  "type": "module",
6
6
  "main": "xindex.ts",
@@ -15,11 +15,16 @@
15
15
  "index": "tsx apps/run.index.ts",
16
16
  "search": "tsx apps/run.search.ts",
17
17
  "reset": "tsx apps/run.reset.ts",
18
+ "test": "npm run test.functional && npm run test.compilation",
18
19
  "mcp": "tsx apps/run.mcp.ts",
19
20
  "watch": "tsx apps/run.watch.ts",
21
+ "test.functional": "bash test/functional.sh",
20
22
  "test.compilation": "npx -y tsc --ignoreConfig --noEmit --target ES2022 --module ESNext --moduleResolution bundler --esModuleInterop --skipLibCheck --strict false $(git ls-files '*.ts')"
21
23
  },
22
24
  "private": false,
25
+ "engines": {
26
+ "node": ">=22"
27
+ },
23
28
  "keywords": [
24
29
  "semantic-search",
25
30
  "code-search",