Merge pull request #15 from PoroCYon/master

Enhance CLI interface

unzip-http

@@ -1,20 +1,35 @@
 #!/usr/bin/env python3
 
-'''
-Usage:
-    unzip-http <url.zip> <filenames..>
+"""
+usage: unzip-http [-h] [-l] [-f] [-o] url [files ...]
 
-Extract <filenames> from a remote .zip at <url> to stdout.
-If no filenames given, displays .zip contents (filenames and sizes).
-Each filename can be a wildcard glob; all matching files are concatenated and sent to stdout in zipfile order.
+Extract individual files from .zip files over http without downloading the
+entire archive. HTTP server must send `Accept-Ranges: bytes` and
+`Content-Length` in headers.
 
-HTTP server must send `Accept-Ranges: bytes` and `Content-Length` in headers.
-'''
+positional arguments:
+  url                   URL of the remote zip file
+  files                 Files to extract. If no filenames given, displays .zip
+                        contents (filenames and sizes). Each filename can be a
+                        wildcard glob.
+
+options:
+  -h, --help            show this help message and exit
+  -l, --list            List files in the remote zip file
+  -f, --full-filepaths  Recreate folder structure from zip file when extracting
+                        (instead of extracting the files to the current
+                        directory)
+  -o, --stdout          Write files to stdout (if multiple files: concatenate
+                        them to stdout, in zipfile order)
+"""
 
 import sys
 import io
+import math
 import time
 import fnmatch
+import argparse
+import pathlib
 
 import unzip_http
 
@@ -44,19 +59,65 @@ class StreamProgress:
         return r
 
 
-def main(url, *globs):
-    rzf = unzip_http.RemoteZipFile(url)
+def list_files(rzf):
+    def safelog(x):
+        return 1 if x == 0 else math.ceil(math.log10(x))
+
+    digits_compr = max(safelog(f.compress_size) for f in rzf.infolist())
+    digits_plain = max(safelog(f.file_size    ) for f in rzf.infolist())
+    fmtstr = f'%{digits_compr}d -> %{digits_plain}d\t%s'
     for f in rzf.infolist():
-        if not globs:
-            print(f'{f.compress_size} -> {f.file_size}  {f.filename}')
-        elif any(fnmatch.fnmatch(f.filename, g) for g in globs):
-            fp = StreamProgress(rzf.open(f), name=f.filename, total=f.compress_size)
-            while r := fp.read(2**18):
-                sys.stdout.buffer.write(r)
-
-
-args = sys.argv[1:]
-if not args:
-    print(__doc__, file=sys.stderr)
-else:
-    main(*args)
+        print(fmtstr % (f.compress_size, f.file_size, f.filename), file=sys.stderr)
+
+
+def extract_one(outfile, rzf, f, ofname):
+    print(f'Extracting {f.filename} to {ofname}...', file=sys.stderr)
+
+    fp = StreamProgress(rzf.open(f), name=f.filename, total=f.compress_size)
+    while r := fp.read(2**18):
+        outfile.write(r)
+
+
+def download_file(f, rzf, args):
+    if not any(fnmatch.fnmatch(f.filename, g) for g in args.files):
+        return
+
+    if args.stdout:
+        extract_one(sys.stdout.buffer, rzf, f, "stdout")
+    else:
+        path = pathlib.Path(f.filename)
+        if args.full_filepaths:
+            path.parent.mkdir(parents=True, exist_ok=True)
+        else:
+            path = path.name
+
+        with open(str(path), 'wb') as of:
+            extract_one(of, rzf, f, str(path))
+
+
+def main(args):
+    rzf = unzip_http.RemoteZipFile(args.url[0])
+    if args.list or len(args.files) == 0:
+        list_files(rzf)
+    else:
+        for f in rzf.infolist():
+            download_file(f, rzf, args)
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(prog='unzip-http', \
+        description="Extract individual files from .zip files over http without downloading the entire archive. HTTP server must send `Accept-Ranges: bytes` and `Content-Length` in headers.")
+
+    parser.add_argument('-l', '--list', action='store_true', default=False,
+                        help="List files in the remote zip file")
+    parser.add_argument('-f', '--full-filepaths', action='store_true', default=False,
+                        help="Recreate folder structure from zip file when extracting (instead of extracting the files to the current directory)")
+    parser.add_argument('-o', '--stdout', action='store_true', default=False,
+                        help="Write files to stdout (if multiple files: concatenate them to stdout, in zipfile order)")
+
+    parser.add_argument("url", nargs=1, help="URL of the remote zip file")
+    parser.add_argument("files", nargs='*', help="Files to extract. If no filenames given, displays .zip contents (filenames and sizes). Each filename can be a wildcard glob.")
+
+    args = parser.parse_args()
+    main(args)
+