fix issues after merge!

Author: TheoMoins

load_corpus.py

@@ -159,12 +159,6 @@
             else:
                 print(".......unknown feature list format. Ignoring.......")
 
-    elif args.config:
-        # Load from new-style JSON config file
-        config = Config.from_json(args.config)
-        # Override paths if provided via CLI
-        if args.s:
-            config.corpus.paths = args.s
 
     else:
         if not args.s:

superstyl/config.py

@@ -244,7 +244,7 @@ def validate(self) -> None:
             raise ValueError("No paths specified for corpus loading.")
 
         # Validate paths type
-        if not isinstance(self.corpus.paths, list):
+        if not isinstance(self.corpus.paths, (list, str)):
             raise TypeError("Paths in config must be either a list or a glob pattern string.")
 
         for feat_config in self.features:

superstyl/load.py

@@ -6,13 +6,12 @@
 import pandas
 from typing import Optional, List, Tuple, Union
 
-from superstyl.config import Config, FeatureConfig, NormalizationConfig
+from superstyl.config import Config, FeatureConfig
 
 
 def _load_single_feature(
     myTexts: List[dict],
     feat_config: FeatureConfig,
-    norm_config: NormalizationConfig,
     use_provided_feat_list: bool = False,
 ) -> Tuple[pandas.DataFrame, List]:
     """
@@ -183,7 +182,7 @@ def load_corpus(
     if len(config.features) == 1:
         feat_config = config.features[0]
         feats_df, feat_list = _load_single_feature(
-            myTexts, feat_config, config.normalization, use_provided_feat_list
+            myTexts, feat_config, use_provided_feat_list
         )
         corpus = pandas.concat([metadata, feats_df], axis=1)
         return corpus, feat_list
@@ -199,7 +198,7 @@ def load_corpus(
         print(f".......processing {prefix}.......")
 
         feats_df, feat_list = _load_single_feature(
-            myTexts, feat_config, config.normalization, use_provided_feat_list
+            myTexts, feat_config, use_provided_feat_list
         )
 
         # Prefix columns to avoid collisions

superstyl/preproc/pipe.py

@@ -297,7 +297,7 @@ def extract_tokens(path: str, config: Config=Config()) -> List[str]:
             return nltk.tokenize.wordpunct_tokenize(text)
 
         elif config.corpus.format == "tei":
-            return LOADERS['tei'].extract_units(path, config.corpus.units, feats)
+            return LOADERS['tei'].extract_units(path, config.sampling.units, feats)
 
         elif config.sampling.units == "verses" and config.corpus.format == "txm":
             return LOADERS['txm'].extract_units(path, config.sampling.units, feats)