Add parse/rename CLI options

renamer.py

@@ -2,9 +2,11 @@
 
 import argparse
 import collections
+import enum
 import logging
 import os
 import os.path
+import pprint
 import re
 import sys
 
@@ -42,12 +44,48 @@ PATTERNS = (
     ("unknown", r".*")
 )
 
+
+# noinspection PyInterpreter
+class EnumAction(argparse.Action):
+    """
+    Argparse action for handling Enums
+    """
+    def __init__(self, **kwargs):
+        # Pop off the type value
+        enum_type = kwargs.pop("type", None)
+
+        # Ensure an Enum subclass is provided
+        if enum_type is None:
+            raise ValueError("type must be assigned an Enum when using EnumAction")
+        if not issubclass(enum_type, enum.Enum):
+            raise TypeError("type must be an Enum when using EnumAction")
+
+        # Generate choices from the Enum
+        kwargs.setdefault("choices", tuple(e.value for e in enum_type))
+
+        super(EnumAction, self).__init__(**kwargs)
+
+        self._enum = enum_type
+
+    def __call__(self, parser, namespace, values, option_string=None):
+        # Convert value back into an Enum
+        value = self._enum(values)
+        setattr(namespace, self.dest, value)
+
+
+class CliAction(enum.Enum):
+    parse = "parse"
+    rename = "rename"
+
+
 _lg = logging.getLogger("spqr.movie-renamer")
 
 
 def main():
     parser = argparse.ArgumentParser(description="Rename media files.")
-    parser.add_argument("target", type=str,
+    parser.add_argument("action", type=CliAction, action=EnumAction, metavar="ACTION",
+                        help="what to do with media file/directory (%(choices)s)")
+    parser.add_argument("target", type=str, metavar="TARGET",
                         help="path to the media file/directory")
     parser.add_argument("-v", "--verbose", action="store_true", default=False,
                         help="verbose output")
@@ -56,54 +94,61 @@ def main():
     loglevel = logging.DEBUG if args.verbose else logging.INFO
     logging.basicConfig(level=loglevel)
 
-    if os.path.isdir(args.target):
-        process_dir(args.target)
-    else:
-        process_file(args.target)
+    process_path(args.action, args.target)
 
     return 0
 
 
-def process_dir(dir_path):
-    for fname in os.listdir(dir_path):
-        fpath = os.path.join(dir_path, fname)
-        process_file(fpath)
-
-
-def process_file(fpath):
+def process_path(action: CliAction, path):
     # process only files
-    if not os.path.isfile(fpath):
-        _lg.debug("Not a file: %s", fpath)
-        return
+    if os.path.isdir(path):
+        for child_path in sorted(os.listdir(path)):
+            process_path(action, os.path.join(path, child_path))
 
     # split filepath to dir path, title, and extension
-    dir_path, fname = os.path.split(fpath)
+    dir_path, fname = os.path.split(path)
     title, ext = os.path.splitext(fname)
     ext = ext[1:]
     if ext not in PROCESSED_FILETYPES:
-        _lg.debug("Extension is not supported: %s", fpath)
+        _lg.debug("Extension is not supported: %s", path)
         return
 
     parsed_title = parse_title(title)
+    if action == CliAction.parse:
+        print_parsed_title(title, parsed_title)
+        return
 
-    # create file name from parsed chunks
+    if action == CliAction.rename:
+        pretty_title = generate_pretty_name(parsed_title)
+        pretty_title += ".%s" % ext
+        if pretty_title != fname:
+            _lg.warning("%s -> %s", fname, pretty_title)
+        return
+
+
+def print_parsed_title(title, parsed):
+    print(title)
+    pprint.pprint(parsed, indent=4)
+
+
+def generate_pretty_name(parsed_title):
+    """ Create file name from parsed chunks. """
     chunk_order = [k for k, _ in PATTERNS]
     chunk_order = ["name"] + chunk_order
-    episode_idx = chunk_order.index("episode") + 1
-    chunk_order = chunk_order[:episode_idx] + ["episode_name"] + chunk_order[episode_idx:]
+    ep_idx = chunk_order.index("episode") + 1
+    chunk_order = chunk_order[:ep_idx] + ["episode_name"] + chunk_order[ep_idx:]
+
     result = []
     for chunk_type in chunk_order:
         if not parsed_title.get(chunk_type, []):
             continue
         result.append(".".join(parsed_title[chunk_type]))
-    result.append(ext)
     result = ".".join(result)
-
-    if result != fname:
-        _lg.warning("%s -> %s", fname, result)
+    return result
 
 
 def _get_parsed_title_dict(chunk_list, chunk_map):
+    """ Get {chunk_type: [chunk_value_1, ...,  chunk_value_n]} dictionary. """
     p_title = collections.defaultdict(list)
     for idx, chunk in enumerate(chunk_list):
         chunk_type = chunk_map[idx]
@@ -112,7 +157,7 @@ def _get_parsed_title_dict(chunk_list, chunk_map):
 
 
 def _guess_combined(chunk_values, chunk_map):
-    """ Try to combine unknown chunks in pairs and parse them """
+    """ Try to combine unknown chunks in pairs and parse them. """
     is_changed = False
     p_title = _get_parsed_title_dict(chunk_values, chunk_map)
     if len(p_title["unknown"]) < 2:
@@ -165,14 +210,14 @@ def parse_title(title):
     # remove non-word chunks (like single hyphens), but leave ampersands (&)
     chunk_values = list(filter(lambda ch: re.search(r"(\w|&)+", ch), chunk_values))
 
+    chunk_map = []  # list of chunk_types
     # parse each chunk
-    chunk_map = []
     for ch_value in chunk_values:
         chunk_map.append(guess_part(ch_value))
 
     _, chunk_values, chunk_map = _guess_combined(chunk_values, chunk_map)
 
-    # # try to parse unknown chunks, replacing all hyphens in them with dots
+    # try to parse unknown chunks, replacing all hyphens in them with dots
     p_title = _get_parsed_title_dict(chunk_values, chunk_map)
     is_changed = False
     if p_title.get("unknown"):
@@ -230,11 +275,12 @@ def parse_title(title):
     return dict(p_title)
 
 
-def guess_part(fname_part):
-    for pat_type, pattern in PATTERNS:
+def guess_part(chunk_value):
+    """ Return chunk type for given chunk value. """
+    for chunk_type, pattern in PATTERNS:
         full_match_pat = r"^" + pattern + r"$"
-        if re.match(full_match_pat, fname_part, flags=re.I):
-            return pat_type
+        if re.match(full_match_pat, chunk_value, flags=re.I):
+            return chunk_type
     raise RuntimeError("unhandled pattern type")