FEAT: added adapter for .csv.gz files

Author: gdementen

larray_editor/arrayadapter.py

@@ -1001,8 +1001,9 @@ def cell_activated(self, row_idx, column_idx):
 @adapter_for('pathlib.Path')
 def get_path_suffix_adapter(fpath):
     logger.debug(f"get_path_suffix_adapter('{fpath}')")
-    if fpath.suffix.lower() in PATH_SUFFIX_ADAPTERS:
-        path_adapter_cls, required_module = PATH_SUFFIX_ADAPTERS[fpath.suffix]
+    suffix = fpath.suffix.lower()
+    if suffix in PATH_SUFFIX_ADAPTERS:
+        path_adapter_cls, required_module = PATH_SUFFIX_ADAPTERS[suffix]
         if required_module is not None:
             if required_module not in sys.modules:
                 import importlib
@@ -1013,7 +1014,15 @@ def get_path_suffix_adapter(fpath):
                                 f"which is required to handle {fpath.suffix} "
                                 f"files")
                     return None
-        return path_adapter_cls
+        # 2 options:
+        # - either there is a single adapter for that suffix
+        if (isinstance(path_adapter_cls, type) and
+                issubclass(path_adapter_cls, AbstractAdapter)):
+            return path_adapter_cls
+        # - different adapters handle that suffix and/or not all instances can
+        #   be handled
+        else:
+            return path_adapter_cls(fpath)
     elif fpath.is_dir():
         return DirectoryPathAdapter
     else:
@@ -2771,8 +2780,12 @@ def _detect_encoding(self, chunk):
         try:
             import charset_normalizer
             chartset_match = charset_normalizer.from_bytes(chunk).best()
-            self._encoding = chartset_match.encoding
-            logger.debug(f"encoding detected as {self._encoding}")
+            if chartset_match is None:
+                self._encoding = None
+                logger.debug("could not detect encoding from chunk")
+            else:
+                self._encoding = chartset_match.encoding
+                logger.debug(f"encoding detected as {self._encoding}")
         except ImportError:
             logger.debug("could not import 'charset_normalizer' => cannot detect encoding")
 
@@ -3198,6 +3211,31 @@ def cell_activated(self, row_idx, column_idx):
             # return self.data.open(info.filename)
 
 
+class CSVGZPathAdapater(CsvFileAdapter):
+    @classmethod
+    def open(cls, fpath):
+        import gzip
+        # not specifying an encoding is not an option because in that case
+        # we would get bytes and not str, which makes csv reader unhappy
+        return gzip.open(fpath, mode='rt', encoding='utf-8')
+
+    @property
+    def _binary_file(self):
+        import gzip
+        return gzip.open(self.data.name, mode='rb')
+
+
+@path_adapter_for('.gz', 'gzip')
+def dispatch_gzip_path_adapter(gz_path):
+    # strip .gz extension and dispatch to appropriate adapter
+    fpath = gz_path.with_name(gz_path.stem)
+    suffix = fpath.suffix.lower()
+    if suffix == '.csv':
+        return CSVGZPathAdapater
+    else:
+        return None
+
+
 @path_adapter_for('.zip', 'zipfile')
 class ZipPathAdapter(ZipFileAdapter):
     @classmethod