allow to name the downloaded files according to formReference (#178)

* allow to name the downloaded files according to formReference; fixed compatibility issue with audio/wav which should be audio/x-wav

* added test

---------

Co-authored-by: Robert Forkel <xrotwang@googlemail.com>

CHANGELOG.md

@@ -2,6 +2,12 @@
 
 The `pycldf` package adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.html).
 
+
+## Unreleased
+
+- Added option to `downloadmedia` subcommand to customize file naming.
+
+
 ## [1.38.1] - 2024-05-06
 
 - Fixed bug whereby foreign keys for the ParameterNetwork component were not added correctly because

src/pycldf/commands/downloadmedia.py

@@ -11,6 +11,11 @@
 
 def register(parser):
     add_dataset(parser)
+    parser.add_argument(
+        '--use-form-id',
+        help='Use the value in formReference as name of downloaded file',
+        action='store_true',
+        default=False)
     parser.add_argument(
         'output',
         help='Existing local directory to download the files to',
@@ -27,6 +32,6 @@ def run(args):
     for s in args.filters:
         col, _, substring = s.partition('=')
         filters.append((col, substring))
-    for item in MediaTable(get_dataset(args)):
+    for item in MediaTable(get_dataset(args), args.use_form_id):
         if all(substring in item[col] for col, substring in filters):
             item.save(args.output)

src/pycldf/media.py

@@ -57,7 +57,7 @@ class File:
     """
     def __init__(self, media: 'MediaTable', row: dict):
         self.row = row
-        self.id = row[media.id_col.name]
+        self.id = row[media.filename_col.name]
         self._mimetype = row[media.mimetype_col.name]
         self.url = None
         self.scheme = None
@@ -184,7 +184,7 @@ class MediaTable(pycldf.ComponentWithValidation):
     """
     Container class for a `Dataset`'s media items.
     """
-    def __init__(self, ds: pycldf.Dataset):
+    def __init__(self, ds: pycldf.Dataset, use_form_id: bool = False):
         super().__init__(ds)
         self.url_col = ds.get(('MediaTable', 'http://cldf.clld.org/v1.0/terms.rdf#downloadUrl'))
         self.path_in_zip_col = ds.get(
@@ -196,6 +196,8 @@ def __init__(self, ds: pycldf.Dataset):
                     self.url_col = col
                     break
         self.id_col = ds[self.component, 'http://cldf.clld.org/v1.0/terms.rdf#id']
+        self.filename_col = ds[self.component, 'http://cldf.clld.org/v1.0/terms.rdf#formReference']\
+            if use_form_id else self.id_col
         self.mimetype_col = ds[self.component, 'http://cldf.clld.org/v1.0/terms.rdf#mediaType']
 
     @functools.cached_property
@@ -271,6 +273,11 @@ def __init__(self, s):
         mtype, _, param = self.string.partition(';')
         param = param.strip()
         self.type, _, self.subtype = mtype.partition('/')
+
+        # for compatibility reasons
+        if self.type == 'audio' and self.subtype.lower() in {'wav'}:
+            self.subtype = 'x-wav'
+
         if param.startswith('charset='):
             self.encoding = param.replace('charset=', '').strip()
         else:

tests/test_media.py

@@ -101,6 +101,10 @@ def test_read_data_url(url, data):
         urllib.parse.urlparse(url), Mimetype('text/plain;charset=US-ASCII')) == data
 
 
+def test_fixed_mimetypes():
+    assert Mimetype('audio/wav').extension == '.wav'
+
+
 def test_read_file_url(tmp_path):
     p = tmp_path / 'test.txt'
     p.write_text('äöü', encoding='utf16')