Merge branch 'main' into json-coders

HISTORY.md

@@ -1,6 +1,10 @@
 
 # Release history
 
+### Aider v0.50.1
+
+- Bugfix for provider API exceptions.
+
 ### Aider v0.50.0
 
 - Infinite output for DeepSeek Coder, Mistral models in addition to Anthropic's models.

aider/__init__.py

@@ -1 +1 @@
-__version__ = "0.50.1-dev"
+__version__ = "0.50.2-dev"

aider/coders/base_coder.py

@@ -1244,6 +1244,7 @@ class Coder:
 
         self.io.log_llm_history("TO LLM", format_messages(messages))
 
+        completion = None
         try:
             hash_object, completion = send_completion(
                 model.name,
@@ -1263,6 +1264,8 @@ class Coder:
         except KeyboardInterrupt as kbi:
             self.keyboard_interrupt()
             raise kbi
+        except Exception as e:
+            self.io.tool_error(f"Error during API call: {str(e)}")
         finally:
             self.io.log_llm_history(
                 "LLM RESPONSE",

aider/coders/editblock_prompts.py

@@ -125,8 +125,8 @@ Every *SEARCH/REPLACE block* must use this format:
 7. The end of the replace block: >>>>>>> REPLACE
 8. The closing fence: {fence[1]}
 
-Every *SEARCH* section must *EXACTLY MATCH* the existing source code, character for character, including all comments, docstrings, etc.
+Every *SEARCH* section must *EXACTLY MATCH* the existing file content, character for character, including all comments, docstrings, etc.
-
+If the file contains code or other data wrapped/escaped in json/xml/quotes or other containers, you need to propose edits to the literal contents of the file, including the container markup.
 
 *SEARCH/REPLACE* blocks will replace *all* matching occurrences.
 Include enough lines to make the SEARCH blocks uniquely match the lines to change.

aider/commands.py

@@ -724,7 +724,7 @@ class Commands:
             add = result.returncode != 0
         else:
             response = self.io.prompt_ask(
-                "Add the output to the chat?\n(y/n/instructions)", default=""
+                "Add the output to the chat?\n(Y/n/instructions)", default=""
             ).strip()
 
             if response.lower() in ["yes", "y"]:

aider/main.py

@@ -327,6 +327,17 @@ def main(argv=None, input=None, output=None, force_git_root=None, return_coder=F
     parser = get_parser(default_config_files, git_root)
     args, unknown = parser.parse_known_args(argv)
 
+    if args.verbose:
+        print("Config files search order, if no --config:")
+        for file in default_config_files:
+            exists = "(exists)" if Path(file).exists() else ""
+            print(f"  - {file} {exists}")
+
+    default_config_files.reverse()
+
+    parser = get_parser(default_config_files, git_root)
+    args, unknown = parser.parse_known_args(argv)
+
     # Load the .env file specified in the arguments
     loaded_dotenvs = load_dotenv_files(git_root, args.env_file)
 

aider/website/HISTORY.md

@@ -16,6 +16,10 @@ cog.out(text)
 
 # Release history
 
+### Aider v0.50.1
+
+- Bugfix for provider API exceptions.
+
 ### Aider v0.50.0
 
 - Infinite output for DeepSeek Coder, Mistral models in addition to Anthropic's models.

aider/website/_data/edit_leaderboard.yml

@@ -577,6 +577,7 @@
   pass_rate_2: 77.4
   percent_cases_well_formed: 99.2
   error_outputs: 23
+  released: 2024-06-20
   num_malformed_responses: 4
   num_with_malformed_responses: 1
   user_asks: 2
@@ -603,6 +604,7 @@
   num_malformed_responses: 0
   num_with_malformed_responses: 0
   user_asks: 0
+  released: 2024-03-13
   lazy_comments: 0
   syntax_errors: 0
   indentation_errors: 0
@@ -644,6 +646,7 @@
   commit_hash: d31eef3-dirty
   pass_rate_1: 40.6
   pass_rate_2: 55.6
+  released: 2024-07-18
   percent_cases_well_formed: 100.0
   error_outputs: 1
   num_malformed_responses: 0
@@ -668,6 +671,7 @@
   pass_rate_1: 60.9
   pass_rate_2: 69.9
   percent_cases_well_formed: 97.7
+  released: 2024-06-28
   error_outputs: 58
   num_malformed_responses: 13
   num_with_malformed_responses: 3
@@ -690,6 +694,7 @@
   commit_hash: f7ce78b-dirty
   pass_rate_1: 46.6
   pass_rate_2: 63.9
+  released: 2024-07-23
   percent_cases_well_formed: 92.5
   error_outputs: 84
   num_malformed_responses: 19
@@ -716,6 +721,7 @@
   percent_cases_well_formed: 100.0
   error_outputs: 0
   num_malformed_responses: 0
+  released: 2024-07-23
   num_with_malformed_responses: 0
   user_asks: 0
   lazy_comments: 0
@@ -738,6 +744,7 @@
   pass_rate_2: 72.9
   percent_cases_well_formed: 97.7
   error_outputs: 13
+  released: 2024-07-24
   num_malformed_responses: 3
   num_with_malformed_responses: 3
   user_asks: 1
@@ -763,6 +770,7 @@
   error_outputs: 3
   num_malformed_responses: 0
   num_with_malformed_responses: 0
+  released: 2024-07-24
   user_asks: 3
   lazy_comments: 0
   syntax_errors: 1
@@ -785,6 +793,7 @@
   percent_cases_well_formed: 100.0
   error_outputs: 27
   num_malformed_responses: 0
+  released: 2024-07-23
   num_with_malformed_responses: 0
   user_asks: 23
   lazy_comments: 8
@@ -810,6 +819,7 @@
   num_malformed_responses: 0
   num_with_malformed_responses: 0
   user_asks: 0
+  released: 2024-07-23
   lazy_comments: 0
   syntax_errors: 0
   indentation_errors: 0
@@ -838,9 +848,34 @@
   indentation_errors: 2
   exhausted_context_windows: 0
   test_timeouts: 5
+  released: 2024-08-06
   command: aider --model openai/gpt-4o-2024-08-06
   date: 2024-08-06
   versions: 0.48.1-dev
   seconds_per_case: 6.5
   total_cost: 0.0000
+
+- dirname: 2024-08-14-13-07-12--chatgpt-4o-latest-diff
+  test_cases: 133
+  model: chatgpt-4o-latest
+  edit_format: diff
+  commit_hash: b1c3769
+  pass_rate_1: 53.4
+  pass_rate_2: 69.2
+  percent_cases_well_formed: 97.7
+  error_outputs: 27
+  num_malformed_responses: 5
+  num_with_malformed_responses: 3
+  user_asks: 7
+  lazy_comments: 0
+  syntax_errors: 0
+  indentation_errors: 0
+  exhausted_context_windows: 0
+  test_timeouts: 0
+  command: aider --model openai/chatgpt-4o-latest
+  date: 2024-08-14
+  released: 2024-08-08
+  versions: 0.50.2-dev
+  seconds_per_case: 26.3
+  total_cost: 3.6113
   

aider/website/docs/config.md

@@ -28,7 +28,7 @@ Using a `.aider.conf.yml` file:
 dark-mode: true
 ```
 
-By setting an environgment variable:
+By setting an environment variable:
 
 ```
 export AIDER_DARK_MODE=true

aider/website/docs/leaderboards/index.md

@@ -321,6 +321,6 @@ mod_dates = [get_last_modified_date(file) for file in files]
 latest_mod_date = max(mod_dates)
 cog.out(f"{latest_mod_date.strftime('%B %d, %Y.')}")
 ]]]-->
-August 10, 2024.
+August 14, 2024.
 <!--[[[end]]]-->
 </p>

aider/website/docs/usage/images-urls.md

@@ -23,7 +23,7 @@ You can add images to the chat just like you would
 add any other file:
 
 - Use `/add <image-filename>` from within the chat
-- Use `/add-clipboard-image` to paste an image from your clipboard into the chat.
+- Use `/clipboard` to paste an image from your clipboard into the chat.
 - Launch aider with image filenames on the command line: `aider <image-filename>` along with any other command line arguments you need.
 
 ## Web pages

benchmark/benchmark.py

@@ -28,8 +28,6 @@ from aider.coders import Coder
 from aider.dump import dump  # noqa: F401
 from aider.io import InputOutput
 
-load_dotenv()
-
 BENCHMARK_DNAME = Path(os.environ.get("AIDER_BENCHMARK_DIR", "tmp.benchmarks"))
 
 EXERCISES_DIR_DEFAULT = "exercism-python"
@@ -39,6 +37,8 @@ app = typer.Typer(add_completion=False, pretty_exceptions_enable=False)
 
 NUM_TESTS = (89, 133)
 
+load_dotenv(override=True)
+
 
 def show_stats(dirnames, graphs):
     raw_rows = []
@@ -378,7 +378,7 @@ def summarize_results(dirname):
         pass_rate = 100 * passed_tests[i] / res.completed_tests
         percents[i] = pass_rate
         # console.print(f"{pass_rate:.1f}% correct after try {i+1}")
-        setattr(res, f"pass_rate_{i+1}", f"{pass_rate:.1f}")
+        setattr(res, f"pass_rate_{i + 1}", f"{pass_rate:.1f}")
 
     print(f"- dirname: {dirname.name}")
     style = None if res.completed_tests in NUM_TESTS else "red"
@@ -393,10 +393,10 @@ def summarize_results(dirname):
         console.print(f"  {key}: {val}", style=style)
 
     for i in range(tries):
-        print(f"  pass_rate_{i+1}: {percents[i]:.1f}")
+        print(f"  pass_rate_{i + 1}: {percents[i]:.1f}")
 
     pct_well_formed = 1.0 - res.num_with_malformed_responses / res.completed_tests
-    print(f"  percent_cases_well_formed: {pct_well_formed*100:.1f}")
+    print(f"  percent_cases_well_formed: {pct_well_formed * 100:.1f}")
 
     show("error_outputs")
     show("num_malformed_responses")
@@ -564,7 +564,6 @@ def run_test_real(
         fnames=fnames,
         use_git=False,
         stream=False,
-        pretty=False,
         verbose=verbose,
     )
     coder.max_apply_update_errors = max_apply_update_errors
@@ -591,7 +590,7 @@ def run_test_real(
 
             coder.apply_updates()
         else:
-            response = coder.run(with_message=instructions)
+            response = coder.run(with_message=instructions, preproc=False)
         dur += time.time() - start
 
         if not no_aider:

benchmark/over_time.py

@@ -3,6 +3,8 @@ import yaml
 from imgcat import imgcat
 from matplotlib import rc
 
+from aider.dump import dump  # noqa: 401
+
 
 def plot_over_time(yaml_file):
     with open(yaml_file, "r") as file:
@@ -12,49 +14,97 @@ def plot_over_time(yaml_file):
     pass_rates = []
     models = []
 
+    print("Debug: Raw data from YAML file:")
+    print(data)
+
     for entry in data:
         if "released" in entry and "pass_rate_2" in entry:
             dates.append(entry["released"])
             pass_rates.append(entry["pass_rate_2"])
             models.append(entry["model"].split("(")[0].strip())
 
+    print("Debug: Processed data:")
+    print("Dates:", dates)
+    print("Pass rates:", pass_rates)
+    print("Models:", models)
+
+    if not dates or not pass_rates:
+        print(
+            "Error: No data to plot. Check if the YAML file is empty or if the data is in the"
+            " expected format."
+        )
+        return
+
     plt.rcParams["hatch.linewidth"] = 0.5
     plt.rcParams["hatch.color"] = "#444444"
 
     rc("font", **{"family": "sans-serif", "sans-serif": ["Helvetica"], "size": 10})
     plt.rcParams["text.color"] = "#444444"
 
-    fig, ax = plt.subplots(figsize=(10, 5))
+    fig, ax = plt.subplots(figsize=(12, 6))  # Increase figure size for better visibility
+
+    print("Debug: Figure created. Plotting data...")
     ax.grid(axis="y", zorder=0, lw=0.2)
     for spine in ax.spines.values():
         spine.set_edgecolor("#DDDDDD")
         spine.set_linewidth(0.5)
 
     colors = [
-        "red" if "gpt-4" in model else "green" if "gpt-3.5" in model else "blue" for model in models
+        (
+            "purple"
+            if "-4o" in model and "gpt-4o-mini" not in model
+            else "red" if "gpt-4" in model else "green" if "gpt-3.5" in model else "lightblue"
+        )
+        for model in models
     ]
+
+    # Separate data points by color
+    purple_points = [(d, r) for d, r, c in zip(dates, pass_rates, colors) if c == "purple"]
+    red_points = [(d, r) for d, r, c in zip(dates, pass_rates, colors) if c == "red"]
+    green_points = [(d, r) for d, r, c in zip(dates, pass_rates, colors) if c == "green"]
+
+    # Plot lines for purple, red, and green points
+    if purple_points:
+        purple_dates, purple_rates = zip(*sorted(purple_points))
+        ax.plot(purple_dates, purple_rates, c="purple", alpha=0.5, linewidth=1)
+    if red_points:
+        red_dates, red_rates = zip(*sorted(red_points))
+        ax.plot(red_dates, red_rates, c="red", alpha=0.5, linewidth=1)
+    if green_points:
+        green_dates, green_rates = zip(*sorted(green_points))
+        ax.plot(green_dates, green_rates, c="green", alpha=0.5, linewidth=1)
+
+    # Plot all points
     ax.scatter(dates, pass_rates, c=colors, alpha=0.5, s=120)
 
     for i, model in enumerate(models):
         ax.annotate(
             model,
             (dates[i], pass_rates[i]),
-            fontsize=12,
+            fontsize=8,
             alpha=0.75,
             xytext=(5, 5),
             textcoords="offset points",
         )
 
     ax.set_xlabel("Model release date", fontsize=18, color="#555")
-    ax.set_ylabel("Aider code editing benchmark,\npercent completed correctly", fontsize=18, color="#555")
+    ax.set_ylabel(
+        "Aider code editing benchmark,\npercent completed correctly", fontsize=18, color="#555"
+    )
     ax.set_title("LLM code editing skill by model release date", fontsize=20)
-    ax.set_ylim(0, 30)
+    ax.set_ylim(0, 100)  # Adjust y-axis limit to accommodate higher values
-    plt.xticks(fontsize=14)
+    plt.xticks(fontsize=14, rotation=45, ha="right")  # Rotate x-axis labels for better readability
     plt.tight_layout(pad=3.0)
+
+    print("Debug: Saving figures...")
     plt.savefig("tmp_over_time.png")
     plt.savefig("tmp_over_time.svg")
+
+    print("Debug: Displaying figure with imgcat...")
     imgcat(fig)
 
+    print("Debug: Figure generation complete.")
+
 
 # Example usage
-plot_over_time("_data/edit_leaderboard.yml")
+plot_over_time("aider/website/_data/edit_leaderboard.yml")

tests/basic/test_main.py

@@ -226,9 +226,10 @@ class TestMain(TestCase):
 
     def test_main_exit_calls_version_check(self):
         with GitTemporaryDirectory():
-            with patch("aider.main.check_version") as mock_check_version, patch(
+            with (
-                "aider.main.InputOutput"
+                patch("aider.main.check_version") as mock_check_version,
-            ) as mock_input_output:
+                patch("aider.main.InputOutput") as mock_input_output,
+            ):
                 main(["--exit"], input=DummyInput(), output=DummyOutput())
                 mock_check_version.assert_called_once()
                 mock_input_output.assert_called_once()
@@ -373,6 +374,67 @@ class TestMain(TestCase):
             self.assertRegex(relevant_output, r"AIDER_DARK_MODE:\s+on")
             self.assertRegex(relevant_output, r"dark_mode:\s+True")
 
+    def test_yaml_config_file_loading(self):
+        with GitTemporaryDirectory() as git_dir:
+            git_dir = Path(git_dir)
+
+            # Create fake home directory
+            fake_home = git_dir / "fake_home"
+            fake_home.mkdir()
+            os.environ["HOME"] = str(fake_home)
+
+            # Create subdirectory as current working directory
+            cwd = git_dir / "subdir"
+            cwd.mkdir()
+            os.chdir(cwd)
+
+            # Create .aider.conf.yml files in different locations
+            home_config = fake_home / ".aider.conf.yml"
+            git_config = git_dir / ".aider.conf.yml"
+            cwd_config = cwd / ".aider.conf.yml"
+            named_config = git_dir / "named.aider.conf.yml"
+
+            cwd_config.write_text("model: gpt-4-32k\nmap-tokens: 4096\n")
+            git_config.write_text("model: gpt-4\nmap-tokens: 2048\n")
+            home_config.write_text("model: gpt-3.5-turbo\nmap-tokens: 1024\n")
+            named_config.write_text("model: gpt-4-1106-preview\nmap-tokens: 8192\n")
+
+            with (
+                patch("pathlib.Path.home", return_value=fake_home),
+                patch("aider.coders.Coder.create") as MockCoder,
+            ):
+                # Test loading from specified config file
+                main(
+                    ["--yes", "--exit", "--config", str(named_config)],
+                    input=DummyInput(),
+                    output=DummyOutput(),
+                )
+                _, kwargs = MockCoder.call_args
+                self.assertEqual(kwargs["main_model"].name, "gpt-4-1106-preview")
+                self.assertEqual(kwargs["map_tokens"], 8192)
+
+                # Test loading from current working directory
+                main(["--yes", "--exit"], input=DummyInput(), output=DummyOutput())
+                _, kwargs = MockCoder.call_args
+                print("kwargs:", kwargs)  # Add this line for debugging
+                self.assertIn("main_model", kwargs, "main_model key not found in kwargs")
+                self.assertEqual(kwargs["main_model"].name, "gpt-4-32k")
+                self.assertEqual(kwargs["map_tokens"], 4096)
+
+                # Test loading from git root
+                cwd_config.unlink()
+                main(["--yes", "--exit"], input=DummyInput(), output=DummyOutput())
+                _, kwargs = MockCoder.call_args
+                self.assertEqual(kwargs["main_model"].name, "gpt-4")
+                self.assertEqual(kwargs["map_tokens"], 2048)
+
+                # Test loading from home directory
+                git_config.unlink()
+                main(["--yes", "--exit"], input=DummyInput(), output=DummyOutput())
+                _, kwargs = MockCoder.call_args
+                self.assertEqual(kwargs["main_model"].name, "gpt-3.5-turbo")
+                self.assertEqual(kwargs["map_tokens"], 1024)
+
     def test_map_tokens_option(self):
         with GitTemporaryDirectory():
             with patch("aider.coders.base_coder.RepoMap") as MockRepoMap: