Minor bug fixes

README.md

@@ -101,24 +101,10 @@ Download platform-specific binaries:
 # Extract and install
 unzip oai_vx.x.x_mac_arm64.zip  # or `oai_vx.x.x-linux-x86_64.zip`
 chmod +x oai
-mkdir -p ~/.local/bin
+mkdir -p ~/.local/bin # Remember to add this to your path. Or just move to folder already in your $PATH
 mv oai ~/.local/bin/
 ```
 
-### Option 3: Build Your Own Binary
-
-```bash
-# Install build dependencies
-pip install -r requirements.txt
-pip install nuitka ordered-set zstandard
-
-# Run build script
-chmod +x build.sh
-./build.sh
-
-# Binary will be in dist/oai
-cp dist/oai ~/.local/bin/
-```
 
 ### Alternative: Shell Alias
 
@@ -508,6 +494,7 @@ Full license: https://opensource.org/licenses/MIT
 
 **Rune Olsen**
 
+- Homepage: https://ai.fubar.pm/
 - Blog: https://blog.rune.pm
 - Project: https://iurl.no/oai
 
@@ -527,3 +514,7 @@ Contributions welcome! Please:
 ---
 
 **Star ⭐ this project if you find it useful!**
+
+---
+
+Did you really read all the way down here? WOW! You deserve a 🍾 🥂!

oai.py

@@ -4650,25 +4650,55 @@ def chat():
             file_attachments = []
             content_blocks = []
             
-            # Smart file detection with extension whitelist
-            # Only matches files with known extensions or clear path prefixes
-            # 
-            # Valid matches:
-            #   @/path/to/file
-            #   @~/file.txt
-            #   @./script.py
-            #   @report.pdf
-            #
-            # Excludes:
-            #   @domain.com (no whitelisted extension)
-            #   @mail.server.eu (TLD not in whitelist)
-            #   user@domain.com (email pattern)
-            #   Diagnostic-Code: @server.eu (TLD not in whitelist)
+            # Smart file detection: Simple pattern + extension validation
+            # This avoids extremely long regex that can cause binary signing issues
             
-            file_pattern = r'(?:^|\s)@((?:[~/][\S]+)|(?:\.[\S]+)|(?:[\w][\w-]*\.(?:py|txt|md|log|json|csv|pdf|png|jpg|jpeg|gif|bmp|webp|svg|ico|zip|tar|gz|bz2|7z|rar|xz|js|ts|jsx|tsx|vue|html|css|scss|sass|less|xml|yaml|yml|toml|ini|conf|cfg|env|properties|sh|bash|zsh|fish|bat|cmd|ps1|c|cpp|cc|cxx|h|hpp|hxx|java|class|jar|war|rb|go|rs|swift|kt|kts|php|sql|db|sqlite|sqlite3|lock|gitignore|dockerignore|editorconfig|eslintrc|prettierrc|babelrc|nvmrc|npmrc|pyc|pyo|pyd|so|dll|dylib|exe|app|dmg|pkg|deb|rpm|apk|ipa|wasm|proto|graphql|graphqls|grpc|avro|parquet|orc|feather|arrow|hdf5|h5|mat|r|rdata|rds|pkl|pickle|joblib|npy|npz|safetensors|onnx|pt|pth|ckpt|pb|tflite|mlmodel|coreml|rknn)))(?=\s|$)'
+            # Common file extensions we support
+            ALLOWED_FILE_EXTENSIONS = {
+                # Code
+                '.py', '.js', '.ts', '.jsx', '.tsx', '.vue', '.java', '.c', '.cpp', '.cc', '.cxx',
+                '.h', '.hpp', '.hxx', '.rb', '.go', '.rs', '.swift', '.kt', '.kts', '.php',
+                '.sh', '.bash', '.zsh', '.fish', '.bat', '.cmd', '.ps1',
+                # Data
+                '.json', '.csv', '.yaml', '.yml', '.toml', '.xml', '.sql', '.db', '.sqlite', '.sqlite3',
+                # Documents
+                '.txt', '.md', '.log', '.conf', '.cfg', '.ini', '.env', '.properties',
+                # Images
+                '.png', '.jpg', '.jpeg', '.gif', '.bmp', '.webp', '.svg', '.ico',
+                # Archives
+                '.zip', '.tar', '.gz', '.bz2', '.7z', '.rar', '.xz',
+                # Config files
+                '.lock', '.gitignore', '.dockerignore', '.editorconfig', '.eslintrc', 
+                '.prettierrc', '.babelrc', '.nvmrc', '.npmrc',
+                # Binary/Compiled
+                '.pyc', '.pyo', '.pyd', '.so', '.dll', '.dylib', '.exe', '.app', 
+                '.dmg', '.pkg', '.deb', '.rpm', '.apk', '.ipa',
+                # ML/AI
+                '.pkl', '.pickle', '.joblib', '.npy', '.npz', '.safetensors', '.onnx',
+                '.pt', '.pth', '.ckpt', '.pb', '.tflite', '.mlmodel', '.coreml', '.rknn',
+                # Data formats
+                '.wasm', '.proto', '.graphql', '.graphqls', '.grpc', '.avro', '.parquet',
+                '.orc', '.feather', '.arrow', '.hdf5', '.h5', '.mat', '.r', '.rdata', '.rds',
+                # Other
+                '.pdf', '.class', '.jar', '.war'
+            }
+            
+            # Simple pattern: @filepath where filepath starts with ~, /, . or has an extension
+            # Much shorter than listing all extensions in the regex
+            file_pattern = r'(?:^|\s)@([~/\.][\S]+|[\w][\w-]*\.[\w]+)(?=\s|$)'
             
             for match in re.finditer(file_pattern, user_input, re.IGNORECASE):
                 file_path = match.group(1)
+                
+                # Get the extension
+                file_ext = os.path.splitext(file_path)[1].lower()
+                
+                # Skip if it doesn't start with a path char and has no allowed extension
+                if not file_path.startswith(('/', '~', '.', '\\')):
+                    if not file_ext or file_ext not in ALLOWED_FILE_EXTENSIONS:
+                        # This looks like a domain name, not a file
+                        continue
+                
                 expanded_path = os.path.expanduser(os.path.abspath(file_path))
                 
                 if not os.path.exists(expanded_path) or os.path.isdir(expanded_path):
@@ -4681,7 +4711,6 @@ def chat():
                     continue
                 
                 mime_type, _ = mimetypes.guess_type(expanded_path)
-                file_ext = os.path.splitext(expanded_path)[1].lower()
                 
                 try:
                     with open(expanded_path, 'rb') as f:
@@ -4729,7 +4758,7 @@ def chat():
                     app_logger.error(f"File read error for {expanded_path}: {e}")
                     continue
             
-            # Remove file attachments from text (use same pattern as file detection)
+            # Remove file attachments from text (use same simple pattern)
             text_part = re.sub(file_pattern, lambda m: m.group(0)[0] if m.group(0)[0].isspace() else '', text_part).strip()
 
             # Build message content