Merge 62ed9d4621 into 3caab85931

aider/commands.py

@@ -443,6 +443,11 @@ class Commands:
 
         self.coder.choose_fence()
 
+        # Show progress indicator
+        total_files = len(self.coder.abs_fnames) + len(self.coder.abs_read_only_fnames)
+        if total_files > 20:
+            self.io.tool_output(f"Calculating tokens for {total_files} files...")
+
         # system messages
         main_sys = self.coder.fmt_system_prompt(self.coder.gpt_prompts.main_system)
         main_sys += "\n" + self.coder.fmt_system_prompt(self.coder.gpt_prompts.system_reminder)
@@ -474,8 +479,20 @@ class Commands:
         fence = "`" * 3
 
         file_res = []
-        # files
-        for fname in self.coder.abs_fnames:
+        # Process files with progress indication
+        total_editable_files = len(self.coder.abs_fnames)
+        total_readonly_files = len(self.coder.abs_read_only_fnames)
+        
+        # Display progress for editable files
+        if total_editable_files > 0:
+            if total_editable_files > 20:
+                self.io.tool_output(f"Calculating tokens for {total_editable_files} editable files...")
+            
+            # Calculate tokens for editable files
+            for i, fname in enumerate(self.coder.abs_fnames):
+                if i > 0 and i % 20 == 0 and total_editable_files > 20:
+                    self.io.tool_output(f"Processed {i}/{total_editable_files} editable files...")
+                
                 relative_fname = self.coder.get_rel_fname(fname)
                 content = self.io.read_text(fname)
                 if is_image_file(relative_fname):
@@ -486,8 +503,16 @@ class Commands:
                     tokens = self.coder.main_model.token_count(content)
                 file_res.append((tokens, f"{relative_fname}", "/drop to remove"))
         
-        # read-only files
-        for fname in self.coder.abs_read_only_fnames:
+        # Display progress for read-only files
+        if total_readonly_files > 0:
+            if total_readonly_files > 20:
+                self.io.tool_output(f"Calculating tokens for {total_readonly_files} read-only files...")
+            
+            # Calculate tokens for read-only files
+            for i, fname in enumerate(self.coder.abs_read_only_fnames):
+                if i > 0 and i % 20 == 0 and total_readonly_files > 20:
+                    self.io.tool_output(f"Processed {i}/{total_readonly_files} read-only files...")
+                
                 relative_fname = self.coder.get_rel_fname(fname)
                 content = self.io.read_text(fname)
                 if content is not None and not is_image_file(relative_fname):
@@ -496,6 +521,9 @@ class Commands:
                     tokens = self.coder.main_model.token_count(content)
                     file_res.append((tokens, f"{relative_fname} (read-only)", "/drop to remove"))
 
+        if total_files > 20:
+            self.io.tool_output("Token calculation complete. Generating report...")
+
         file_res.sort()
         res.extend(file_res)
 
@@ -510,7 +538,7 @@ class Commands:
         def fmt(v):
             return format(int(v), ",").rjust(width)
 
-        col_width = max(len(row[1]) for row in res)
+        col_width = max(len(row[1]) for row in res) if res else 0
 
         cost_pad = " " * cost_width
         total = 0

aider/io.py

@@ -129,7 +129,19 @@ class AutoCompleter(Completer):
             return
         self.tokenized = True
         
-        for fname in self.all_fnames:
+        # Performance optimization for large file sets
+        if len(self.all_fnames) > 100:
+            # Skip tokenization for very large numbers of files to avoid input lag
+            self.tokenized = True
+            return
+            
+        # Limit number of files to process to avoid excessive tokenization time
+        process_fnames = self.all_fnames
+        if len(process_fnames) > 50:
+            # Only process a subset of files to maintain responsiveness
+            process_fnames = process_fnames[:50]
+
+        for fname in process_fnames:
             try:
                 with open(fname, "r", encoding=self.encoding) as f:
                     content = f.read()
@@ -1120,6 +1132,21 @@ class InputOutput:
                 self.chat_history_file = None  # Disable further attempts to write
 
     def format_files_for_input(self, rel_fnames, rel_read_only_fnames):
+        # Optimization for large number of files
+        total_files = len(rel_fnames) + len(rel_read_only_fnames or [])
+        
+        # For very large numbers of files, use a summary display
+        if total_files > 50:
+            read_only_count = len(rel_read_only_fnames or [])
+            editable_count = len([f for f in rel_fnames if f not in (rel_read_only_fnames or [])])
+            
+            summary = f"{editable_count} editable file(s)"
+            if read_only_count > 0:
+                summary += f", {read_only_count} read-only file(s)"
+            summary += " (use /ls to list all files)\n"
+            return summary
+            
+        # Original implementation for reasonable number of files
         if not self.pretty:
             read_only_files = []
             for full_path in sorted(rel_read_only_fnames or []):

aider/utils.py

@@ -126,6 +126,17 @@ def format_messages(messages, title=None):
                 else:
                     output.append(f"{role} {item}")
         elif isinstance(content, str):  # Handle string content
+            # For large content, especially with many files, use a truncated display approach
+            if len(content) > 5000:
+                # Count the number of code blocks (approximation)
+                fence_count = content.count("```") // 2
+                if fence_count > 5:
+                    # Show truncated content with file count for large files to improve performance
+                    first_line = content.split("\n", 1)[0]
+                    output.append(f"{role} {first_line} [content with ~{fence_count} files truncated]")
+                else:
+                    output.append(format_content(role, content))
+            else:
                 output.append(format_content(role, content))
         function_call = msg.get("function_call")
         if function_call: