Use yapf for code formatting

Black is restrictive as a matter of design/feature and the [driving
philosophy](psf/black#2156 (comment))
seems to favor very limited changes to Black over addressing
extraordinarily bad formatting.

yapf is not perfect as it requires a commenting pattern to readable
formats for long logical and mathematical expressions. It's far more
flexible and succeeds at generating maintainable code in cases that
Black cannot.

.github/workflows/check-source-code.yml

@@ -20,7 +20,7 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        pip install pylint black
+        pip install pylint yapf
         pip install -r requirements.txt
         pip install -r guild/tests/requirements.txt
 
@@ -32,4 +32,4 @@ jobs:
     - name: Check source code format
       if: ${{ success() || failure() }}
       run: |
-        python -m black --diff --check setup.py guild examples
+        python -m yapf --diff -r setup.py tools.py guild examples

Makefile

@@ -101,8 +101,8 @@ README.html: README.md
 	markdown_py README.md > README.html
 
 format-code:
-	@echo Formatting guild code
-	@black setup.py tools.py guild examples
+	@echo Formating Guild source code
+	@yapf -ri setup.py tools.py guild examples
 
 coverage-check:
 	if [ -z "$(TESTS)" ]; then \

examples/bias/gen_schema.py

@@ -4,7 +4,6 @@
 --help` for command help.
 """
 
-
 import argparse
 import os
 
@@ -81,7 +80,6 @@ class TFDVSchemaDisplay(object):
 
     Patch is unapplied when the context manager exits.
     """
-
     def __init__(self):
         self._display_save = None
         self.display_obj = []

examples/detectron2/demo.py

@@ -97,8 +97,7 @@ def get_parser():
                 "{}: {} in {:.2f}s".format(
                     path,
                     "detected {} instances".format(len(predictions["instances"]))
-                    if "instances" in predictions
-                    else "finished",
+                    if "instances" in predictions else "finished",
                     time.time() - start_time,
                 )
             )

examples/detectron2/predictor.py

@@ -139,7 +139,6 @@ class AsyncPredictor:
     Because rendering the visualization takes considerably amount of time,
     this helps improve throughput when rendering videos.
     """
-
     class _StopToken:
         pass
 

examples/dvc/eval_models.py

@@ -36,7 +36,6 @@ def plot_contours(ax, clf, xx, yy, **params):
 
 X, y = load_data()
 
-
 models = (
     joblib.load("model-1.joblib"),
     joblib.load("model-2.joblib"),

examples/dvc/train_models.py

@@ -29,10 +29,8 @@
     svm.SVC(kernel="poly", degree=3, gamma="auto", C=C),
 )
 
-
 X, y = load_data()
 
-
 for i, m in enumerate(models):
     print("Saving model-%i.joblib" % (i + 1))
     m.fit(X, y)

examples/erlang-plugin/erlang_guild/plugins/escript.py

@@ -22,7 +22,9 @@ def __init__(self, script_path):
             operations={
                 script_name: {
                     "exec": exec_spec,
-                    "sourcecode": {"dest": "."},
+                    "sourcecode": {
+                        "dest": "."
+                    },
                     "flags": flagdefs,
                 },
             },

examples/pytorch-lightning/mnist.py

@@ -69,7 +69,10 @@ def add_model_specific_args(parent_parser):
 args = parser.parse_args()
 
 transform = transforms.Compose(
-    [transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))]
+    [
+        transforms.ToTensor(),
+        transforms.Normalize((0.1307,), (0.3081,)),
+    ]
 )
 
 mnist_train = MNIST(os.getcwd(), train=True, download=True, transform=transform)

examples/tensorboard/audio.py

@@ -12,6 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
+
 """Sample data exhibiting audio summaries, via a waveform generator."""
 
 import inspect
@@ -25,7 +26,6 @@
 
 tf.compat.v1.disable_eager_execution()
 
-
 FLAGS = flags.FLAGS
 
 flags.DEFINE_string(

examples/tensorboard/custom_scalar.py

@@ -33,8 +33,8 @@
 x_train, y_train = x[:train_size], y[:train_size]
 x_test, y_test = x[train_size:], y[train_size:]
 
-logdir = os.getenv("LOGDIR") or "logs/scalars/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+logdir = (
+    os.getenv("LOGDIR") or ("logs/scalars/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 file_writer = tf.summary.create_file_writer(logdir + "/metrics")
 file_writer.set_as_default()

examples/tensorboard/graphs.py

@@ -33,8 +33,9 @@
 train_images = train_images / 255.0
 
 # Define the Keras TensorBoard callback.
-logdir = os.getenv("TRAIN_LOGDIR") or "logs/fit/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+logdir = (
+    os.getenv("TRAIN_LOGDIR")
+    or ("logs/fit/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 tensorboard_callback = keras.callbacks.TensorBoard(log_dir=logdir)
 

examples/tensorboard/hparams.py

@@ -33,7 +33,6 @@
 logdir = os.getenv("LOGDIR") or "logs/image/" + datetime.now().strftime("%Y%m%d-%H%M%S")
 tensorboard_callback = keras.callbacks.TensorBoard(log_dir=logdir)
 
-
 model.fit(
     x_train,
     y_train,

examples/tensorboard/images.py

@@ -45,8 +45,9 @@
 img = np.reshape(train_images[0], (-1, 28, 28, 1))
 
 # Sets up a timestamped log directory.
-logdir = os.getenv("TRAIN_DATA_LOGDIR") or "logs/train_data/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+logdir = (
+    os.getenv("TRAIN_DATA_LOGDIR")
+    or ("logs/train_data/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 # Creates a file writer for the log directory.
 file_writer = tf.summary.create_file_writer(logdir)
@@ -60,8 +61,9 @@
     images = np.reshape(train_images[0:25], (-1, 28, 28, 1))
     tf.summary.image("25 training data examples", images, max_outputs=25, step=0)
 
-logdir = os.getenv("PLOTS_LOGDIR") or "logs/plots/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+logdir = (
+    os.getenv("PLOTS_LOGDIR")
+    or ("logs/plots/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 file_writer = tf.summary.create_file_writer(logdir)
 
@@ -104,7 +106,6 @@ def image_grid():
 with file_writer.as_default():
     tf.summary.image("Training data", plot_to_image(figure), step=0)
 
-
 model = keras.models.Sequential(
     [
         keras.layers.Flatten(input_shape=(28, 28)),
@@ -149,8 +150,9 @@ def plot_confusion_matrix(cm, class_names):
     return figure
 
 
-logdir = os.getenv("IMAGE_LOGDIR") or "logs/image/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+logdir = (
+    os.getenv("IMAGE_LOGDIR")
+    or ("logs/image/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 # Define the basic TensorBoard callback.
 tensorboard_callback = keras.callbacks.TensorBoard(log_dir=logdir)

examples/tensorboard/loss_scalar.py

@@ -33,8 +33,8 @@
 x_train, y_train = x[:train_size], y[:train_size]
 x_test, y_test = x[train_size:], y[train_size:]
 
-logdir = os.getenv("LOGDIR") or "logs/scalars/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+logdir = (
+    os.getenv("LOGDIR") or ("logs/scalars/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 tensorboard_callback = keras.callbacks.TensorBoard(log_dir=logdir)
 

examples/tensorboard/pr_curve.py

@@ -12,6 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
+
 """Create sample PR curve summary data.
 
 We have 3 classes: R, G, and B. We generate colors within RGB space from 3
@@ -152,22 +153,16 @@ def start_runs(logdir, steps, run_name, thresholds, mask_every_other_prediction=
     # tail of the normal distribution.
     examples = tf.concat([true_reds, true_greens, true_blues], axis=0)
     probabilities_colors_are_red = (
-        1
-        - red_predictor.cdf(
-            tf.norm(tensor=examples - tf.constant([255.0, 0, 0]), axis=1)
-        )
+        1 - red_predictor
+        .cdf(tf.norm(tensor=examples - tf.constant([255.0, 0, 0]), axis=1))
     ) * 2
     probabilities_colors_are_green = (
-        1
-        - green_predictor.cdf(
-            tf.norm(tensor=examples - tf.constant([0, 255.0, 0]), axis=1)
-        )
+        1 - green_predictor
+        .cdf(tf.norm(tensor=examples - tf.constant([0, 255.0, 0]), axis=1))
     ) * 2
     probabilities_colors_are_blue = (
-        1
-        - blue_predictor.cdf(
-            tf.norm(tensor=examples - tf.constant([0, 0, 255.0]), axis=1)
-        )
+        1 - blue_predictor
+        .cdf(tf.norm(tensor=examples - tf.constant([0, 0, 255.0]), axis=1))
     ) * 2
 
     predictions = (

examples/tensorboard/projector.py

@@ -51,8 +51,9 @@
 )
 
 # Set up a logs directory, so Tensorboard knows where to look for files
-log_dir = os.getenv("LOGDIR") or "logs/imdb-example/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+log_dir = (
+    os.getenv("LOGDIR")
+    or ("logs/imdb-example/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 if not os.path.exists(log_dir):
     os.makedirs(log_dir)
@@ -65,7 +66,6 @@
     for unknown in range(1, encoder.vocab_size - len(encoder.subwords)):
         f.write("unknown #{}\n".format(unknown))
 
-
 # Save the weights we want to analyse as a variable. Note that the first
 # value represents any unknown word, which is not in the metadata, so
 # we will remove that value.

examples/tensorboard/projector2.py

@@ -15,8 +15,9 @@
 vectors = np.array([[0, 0, 1], [0, 1, 0], [1, 0, 0], [1, 1, 1]])
 metadata = ['001', '010', '100', '111']  # labels
 
-log_dir = os.getenv("LOGDIR") or "logs/projector/" + datetime.now().strftime(
-    "%Y%m%d-%H%M%S"
+log_dir = (
+    os.getenv("LOGDIR")
+    or ("logs/projector/" + datetime.now().strftime("%Y%m%d-%H%M%S"))
 )
 with SummaryWriter(log_dir) as writer:
     writer.add_embedding(vectors, metadata)

guild/_lex.py

@@ -384,8 +384,7 @@ def token(self):
                 if not self.lexoptimize:
                     if newtok.type not in self.lextokens_all:
                         raise LexError(
-                            "%s:%d: Rule '%s' returned an unknown token type '%s'"
-                            % (
+                            "%s:%d: Rule '%s' returned an unknown token type '%s'" % (
                                 func.__code__.co_filename,
                                 func.__code__.co_firstlineno,
                                 func.__name__,
@@ -417,10 +416,11 @@ def token(self):
                     self.lexpos = lexpos
                     newtok = self.lexerrorf(tok)
                     if lexpos == self.lexpos:
-                        # Error method didn't change text position at all. This is an error.
+                        # Error method didn't change text position at
+                        # all. This is an error.
                         raise LexError(
-                            "Scanning error. Illegal character '%s'"
-                            % (lexdata[lexpos]),
+                            "Scanning error. Illegal character '%s'" %
+                            (lexdata[lexpos]),
                             lexdata[lexpos:],
                         )
                     lexpos = self.lexpos

guild/_test.py

@@ -130,8 +130,7 @@ def _run_test(name, fail_fast, force):
         _log_test_not_found(name)
         return False
     if (
-        not force
-        and os.getenv("FORCE_TEST") != "1"
+        not force and os.getenv("FORCE_TEST") != "1"
         and front_matter_skip_test(filename)
     ):
         _log_skipped_test(name)
@@ -362,7 +361,6 @@ class Checker(doctest.OutputChecker):
     environment variable to `1`.
 
     """
-
     def check_output(self, want, got, optionflags):
         if optionflags & STRICT:
             optionflags -= optionflags & doctest.ELLIPSIS
@@ -538,7 +536,7 @@ def _parse_example(self, m, name, lineno):
         source_lines = m.group("source").split("\n")
         _check_prompt_blank(source_lines, indent, name, lineno)
         self._check_prefix(source_lines[1:], " " * indent + ">", name, lineno)
-        source = "\n".join([sl[indent + 2 :] for sl in source_lines])
+        source = "\n".join([sl[indent + 2:] for sl in source_lines])
 
         want = m.group("want")
         want_lines = want.split("\n")
@@ -766,7 +764,8 @@ def _filter_ignored(paths, ignore):
     if isinstance(ignore, str):
         ignore = [ignore]
     return [
-        p for p in paths if not any((fnmatch.fnmatch(p, pattern) for pattern in ignore))
+        p for p in paths
+        if not any((fnmatch.fnmatch(p, pattern) for pattern in ignore))
     ]
 
 
@@ -813,8 +812,7 @@ def cat_json(*parts):
 def dir(path=".", ignore=None):
     return sorted(
         [
-            name
-            for name in os.listdir(path)
+            name for name in os.listdir(path)
             if ignore is None or not any((fnmatch.fnmatch(name, p) for p in ignore))
         ]
     )
@@ -887,7 +885,6 @@ class Project:
     facility should be refactored to use the pattern described in
     `guild/tests/test-template.md` using `use_project()`.
     """
-
     def __init__(self, cwd, guild_home=None, env=None):
         from guild import index as indexlib  # expensive
 
@@ -990,8 +987,7 @@ def _filter_output(out, ignore):
             ignore = [ignore]
         return "\n".join(
             [
-                line
-                for line in out.split("\n")
+                line for line in out.split("\n")
                 if all(s and s not in line for s in ignore)
             ]
         )
@@ -1148,7 +1144,7 @@ def _strip_error_module(last_line):
 
 
 def _strip_class_module(class_name):
-    return class_name[class_name.rfind(".") + 1 :]
+    return class_name[class_name.rfind(".") + 1:]
 
 
 def _normlf(s):
@@ -1359,8 +1355,7 @@ def _strip_lines(out, patterns):
     if isinstance(patterns, str):
         patterns = [patterns]
     stripped_lines = [
-        line
-        for line in out.split("\n")
+        line for line in out.split("\n")
         if not any((re.search(p, line) for p in patterns))
     ]
     return "\n".join(stripped_lines)
@@ -1376,7 +1371,7 @@ def _cut_line(line, to_cut):
     cut_line = []
     cols = line.split()
     for i in to_cut:
-        cut_line.extend(cols[i : i + 1])
+        cut_line.extend(cols[i:i + 1])
     return cut_line