Refactor file exclusion logic by moving default ignore patterns to a dedicated constant file and applying them consistently across all relevant utilities

Author: robertpiosik

packages/vscode/package.json

@@ -540,52 +540,6 @@
         "codeWebChat.ignorePatterns": {
           "type": "array",
           "description": "List of glob patterns to ignore files and folders in the context.",
-          "default": [
-            "**/*.svg",
-            "**/package-lock.json",
-            "**/pnpm-lock.yaml",
-            "**/*.png",
-            "**/*.jpg",
-            "**/*.jpeg",
-            "**/*.gif",
-            "**/*.bmp",
-            "**/*.ico",
-            "**/*.webp",
-            "**/*.tiff",
-            "**/*.psd",
-            "**/*.avif",
-            "**/*.mp4",
-            "**/*.webm",
-            "**/*.avi",
-            "**/*.mov",
-            "**/*.wmv",
-            "**/*.flv",
-            "**/*.mkv",
-            "**/*.mp3",
-            "**/*.wav",
-            "**/*.ogg",
-            "**/*.m4a",
-            "**/*.aac",
-            "**/*.flac",
-            "**/*.zip",
-            "**/*.rar",
-            "**/*.7z",
-            "**/*.tar",
-            "**/*.gz",
-            "**/*.bz2",
-            "**/*.ttf",
-            "**/*.otf",
-            "**/*.woff",
-            "**/*.woff2",
-            "**/*.eot",
-            "**/*.pdf",
-            "**/*.exe",
-            "**/*.dll",
-            "**/*.so",
-            "**/*.dylib",
-            "**/*.bin",
-            "**/*.vsidx"
-          ],
           "items": {
             "type": "string",
             "description": "A glob pattern."

packages/vscode/src/commands/checkpoints-command/utils/clean-directory.ts

@@ -12,12 +12,8 @@ export const clean_directory = async (params: {
     const entry_uri = vscode.Uri.joinPath(params.dir_uri, name)
     const relative_path = path.relative(params.root_path, entry_uri.fsPath)
 
-    if (params.workspace_provider.is_excluded(relative_path)) {
-      continue
-    }
-
     if (
-      type != vscode.FileType.Directory &&
+      params.workspace_provider.is_excluded(relative_path) ||
       params.workspace_provider.is_ignored_by_patterns(entry_uri.fsPath)
     ) {
       continue

packages/vscode/src/commands/checkpoints-command/utils/copy-optimised-recursively.ts

@@ -13,7 +13,10 @@ const directory_contains_ignored = async (
     const entry_uri = vscode.Uri.joinPath(dir_uri, name)
     const relative_path = path.relative(root_path, entry_uri.fsPath)
 
-    if (workspace_provider.is_excluded(relative_path)) {
+    if (
+      workspace_provider.is_excluded(relative_path) ||
+      workspace_provider.is_ignored_by_patterns(entry_uri.fsPath)
+    ) {
       return true
     }
 
@@ -35,10 +38,6 @@ const directory_contains_ignored = async (
       ) {
         return true
       }
-    } else if (entry_stat.type == vscode.FileType.File) {
-      if (workspace_provider.is_ignored_by_patterns(entry_uri.fsPath)) {
-        return true
-      }
     }
   }
   return false
@@ -51,7 +50,10 @@ export const copy_optimised_recursively = async (
   workspace_provider: WorkspaceProvider
 ) => {
   const relative_path = path.relative(root_path, source_uri.fsPath)
-  if (relative_path && workspace_provider.is_excluded(relative_path)) {
+  if (
+    (relative_path && workspace_provider.is_excluded(relative_path)) ||
+    workspace_provider.is_ignored_by_patterns(source_uri.fsPath)
+  ) {
     return
   }
 
@@ -89,9 +91,6 @@ export const copy_optimised_recursively = async (
       )
     }
   } else if (source_stat.type == vscode.FileType.File) {
-    if (workspace_provider.is_ignored_by_patterns(source_uri.fsPath)) {
-      return
-    }
     await fs.copyFile(source_uri.fsPath, dest_uri.fsPath)
   }
   // Other file types are ignored. fs.stat resolves symlinks.

packages/vscode/src/commands/checkpoints-command/utils/sync-directory.ts

@@ -30,19 +30,8 @@ export const sync_directory = async (params: {
       const dest_uri = vscode.Uri.joinPath(params.dest_dir, name)
       const relative_path = path.relative(params.root_path, dest_uri.fsPath)
 
-      if (params.workspace_provider.is_excluded(relative_path)) {
-        continue
-      }
-
-      let dest_stat
-      try {
-        dest_stat = await vscode.workspace.fs.stat(dest_uri)
-      } catch (e) {
-        continue
-      }
-
       if (
-        dest_stat.type !== vscode.FileType.Directory &&
+        params.workspace_provider.is_excluded(relative_path) ||
         params.workspace_provider.is_ignored_by_patterns(dest_uri.fsPath)
       ) {
         continue

packages/vscode/src/constants/ignore-patterns.ts

@@ -0,0 +1,46 @@
+export const IGNORE_PATTERNS = [
+  '**/package-lock.json',
+  '**/pnpm-lock.yaml',
+  '**/node_modules',
+  '**/*.png',
+  '**/*.jpg',
+  '**/*.jpeg',
+  '**/*.gif',
+  '**/*.bmp',
+  '**/*.ico',
+  '**/*.webp',
+  '**/*.tiff',
+  '**/*.psd',
+  '**/*.avif',
+  '**/*.mp4',
+  '**/*.webm',
+  '**/*.avi',
+  '**/*.mov',
+  '**/*.wmv',
+  '**/*.flv',
+  '**/*.mkv',
+  '**/*.mp3',
+  '**/*.wav',
+  '**/*.ogg',
+  '**/*.m4a',
+  '**/*.aac',
+  '**/*.flac',
+  '**/*.zip',
+  '**/*.rar',
+  '**/*.7z',
+  '**/*.tar',
+  '**/*.gz',
+  '**/*.bz2',
+  '**/*.ttf',
+  '**/*.otf',
+  '**/*.woff',
+  '**/*.woff2',
+  '**/*.eot',
+  '**/*.pdf',
+  '**/*.exe',
+  '**/*.dll',
+  '**/*.so',
+  '**/*.dylib',
+  '**/*.bin',
+  '**/*.vsidx'
+]

packages/vscode/src/context/providers/workspace-provider.ts

@@ -3,6 +3,7 @@ import * as fs from 'fs'
 import * as path from 'path'
 import ignore, { Ignore } from 'ignore'
 import { CONTEXT_CHECKED_PATHS_STATE_KEY } from '../../constants/state-keys'
+import { IGNORE_PATTERNS } from '../../constants/ignore-patterns'
 import { natural_sort } from '../../utils/natural-sort'
 import { Logger } from '@shared/utils/logger'
 import { display_token_count } from '@/utils/display-token-count'
@@ -1242,9 +1243,6 @@ export class WorkspaceProvider
       }
     }
 
-    // Add default exclusions (e.g., node_modules at the root)
-    this.combined_gitignore.add(['node_modules/'])
-
     // After updating gitignore rules, clear token caches since exclusions may have changed
     this.file_token_counts.clear()
     this.directory_token_counts.clear()
@@ -1274,6 +1272,8 @@ export class WorkspaceProvider
       this.user_ignore.add(patterns)
     }
 
+    this.user_ignore.add(IGNORE_PATTERNS)
+
     // Clear token caches since exclusions have changed
     this.file_token_counts.clear()
     this.directory_token_counts.clear()

packages/vscode/src/views/panel/backend/message-handlers/handle-commit-changes/utils/file-utils.ts

@@ -3,6 +3,7 @@ import * as path from 'path'
 import * as fs from 'fs'
 import { Logger } from '@shared/utils/logger'
 import { GitRepository } from '@/utils/git-repository-utils'
+import { IGNORE_PATTERNS } from '@/constants/ignore-patterns'
 import ignore from 'ignore'
 
 export interface FileData {
@@ -20,8 +21,8 @@ export const collect_affected_files_with_metadata = async (params: {
   const staged_files = params.repository.state.indexChanges || []
   const files_data: FileData[] = []
   const config = vscode.workspace.getConfiguration('codeWebChat')
-  const ignorePatterns = config.get<string[]>('ignorePatterns') ?? []
-  const ig = ignore().add(ignorePatterns)
+  const user_ignore_patterns = config.get<string[]>('ignorePatterns') ?? []
+  const ig = ignore().add(user_ignore_patterns).add(IGNORE_PATTERNS)
 
   for (const change of staged_files) {
     const file_path = change.uri.fsPath