cloud-files 5.8.2__py3-none-any.whl → 5.9.0__py3-none-any.whl

{cloud_files-5.8.2.dist-info → cloud_files-5.9.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cloud-files
-Version: 5.8.2
+Version: 5.9.0
 Summary: Fast access to cloud storage and local FS.
 Home-page: https://github.com/seung-lab/cloud-files/
 Author: William Silversmith

{cloud_files-5.8.2.dist-info → cloud_files-5.9.0.dist-info}/RECORD

@@ -1,10 +1,10 @@
 cloudfiles/__init__.py,sha256=pLB4CcV2l3Jgv_ni1520Np1pfzFj8Cpr87vNxFT3rNI,493
-cloudfiles/cloudfiles.py,sha256=tPG1PBLEjABPu-KLe93yf6xW_zbafPsQ6z5NuofyUoU,56743
+cloudfiles/cloudfiles.py,sha256=eUFf_PKaLtOIkDmGjDRggPMkMY46BHrXOvNSoAnsDWU,57930
 cloudfiles/compression.py,sha256=WXJHnoNLJ_NWyoY9ygZmFA2qMou35_9xS5dzF7-2H-M,6262
 cloudfiles/connectionpools.py,sha256=aL8RiSjRepECfgAFmJcz80aJFKbou7hsbuEgugDKwB8,4814
 cloudfiles/exceptions.py,sha256=N0oGQNG-St6RvnT8e5p_yC_E61q2kgAe2scwAL0F49c,843
 cloudfiles/gcs.py,sha256=unqu5KxGKaPq6N4QeHSpCDdtnK1BzPOAerTZ8FLt2_4,3820
-cloudfiles/interfaces.py,sha256=M62UdugtWcF-J4iQMClHNDEQYu7xxCSc1aT7WW2C1lU,44942
+cloudfiles/interfaces.py,sha256=Eurpmwv6sbn44AfPGp1Pahb2drhqN9lo5J7CRDTyzWU,47190
 cloudfiles/lib.py,sha256=HHjCvjmOjA0nZWSvHGoSeYpxqd6FAG8xk8LM212LAUA,5382
 cloudfiles/monitoring.py,sha256=N5Xq0PYZK1OxoYtwBFsnnfaq7dElTgY8Rn2Ez_I3aoo,20897
 cloudfiles/paths.py,sha256=FLdShqkOg1XlkHurU9eiKzLadx2JFYG1EmleCpOFsYQ,12229
@@ -16,12 +16,12 @@ cloudfiles/threaded_queue.py,sha256=Nl4vfXhQ6nDLF8PZpSSBpww0M2zWtcd4DLs3W3BArBw,
 cloudfiles/typing.py,sha256=f3ZYkNfN9poxhGu5j-P0KCxjCCqSn9HAg5KiIPkjnCg,416
 cloudfiles_cli/LICENSE,sha256=Jna4xYE8CCQmaxjr5Fs-wmUBnIQJ1DGcNn9MMjbkprk,1538
 cloudfiles_cli/__init__.py,sha256=Wftt3R3F21QsHtWqx49ODuqT9zcSr0em7wk48kcH0WM,29
-cloudfiles_cli/cloudfiles_cli.py,sha256=JlP9ocqxZbMANAZhZCQSvvjwe6syovQ1asUzSeAlNYk,38459
-cloud_files-5.8.2.dist-info/AUTHORS,sha256=BFVmobgAhaVFI5fqbuqAY5XmBQxe09ZZAsAOTy87hKQ,318
-cloud_files-5.8.2.dist-info/LICENSE,sha256=Jna4xYE8CCQmaxjr5Fs-wmUBnIQJ1DGcNn9MMjbkprk,1538
-cloud_files-5.8.2.dist-info/METADATA,sha256=iMhQdNleZM5bNnHKDZ97QNjVuA-7GIWIMHa_wZtePLU,30530
-cloud_files-5.8.2.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
-cloud_files-5.8.2.dist-info/entry_points.txt,sha256=xlirb1FVhn1mbcv4IoyMEGumDqKOA4VMVd3drsRQxIg,51
-cloud_files-5.8.2.dist-info/pbr.json,sha256=geeaELiKgs-Cl3LBIxRry_acNuF2kxgqEmZPNhjainY,46
-cloud_files-5.8.2.dist-info/top_level.txt,sha256=xPyrST3okJbsmdCF5IC2gYAVxg_aD5AYVTnNo8UuoZU,26
-cloud_files-5.8.2.dist-info/RECORD,,
+cloudfiles_cli/cloudfiles_cli.py,sha256=GTQj0UZB34Cfy4q-hIbXqRUnbLYCTQ6OeXjAb930i5Q,38602
+cloud_files-5.9.0.dist-info/AUTHORS,sha256=BFVmobgAhaVFI5fqbuqAY5XmBQxe09ZZAsAOTy87hKQ,318
+cloud_files-5.9.0.dist-info/LICENSE,sha256=Jna4xYE8CCQmaxjr5Fs-wmUBnIQJ1DGcNn9MMjbkprk,1538
+cloud_files-5.9.0.dist-info/METADATA,sha256=4qhGrbkuqEdwCuq-Nqedo7nBNn_QkA5qHFLxfskqid4,30530
+cloud_files-5.9.0.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+cloud_files-5.9.0.dist-info/entry_points.txt,sha256=xlirb1FVhn1mbcv4IoyMEGumDqKOA4VMVd3drsRQxIg,51
+cloud_files-5.9.0.dist-info/pbr.json,sha256=9M5V77fSgk_LF2IUco2G8NcksQ_1cmz7cGYU3OSqRzY,46
+cloud_files-5.9.0.dist-info/top_level.txt,sha256=xPyrST3okJbsmdCF5IC2gYAVxg_aD5AYVTnNo8UuoZU,26
+cloud_files-5.9.0.dist-info/RECORD,,

cloud_files-5.9.0.dist-info/pbr.json

@@ -0,0 +1 @@
+{"git_version": "623052c", "is_release": true}

cloudfiles/cloudfiles.py

@@ -18,6 +18,7 @@ import platform
 import posixpath
 import re
 import shutil
+import threading
 import types
 import time
 
@@ -1007,6 +1008,34 @@ class CloudFiles:
       return results
     return first(results.values())
 
+  def subtree_size(self, prefix:GetPathType = "") -> int:
+    """High performance size calculation for directory trees."""
+    prefix, return_multiple = toiter(prefix, is_iter=True)
+    total_bytes = 0
+
+    total = totalfn(prefix, None)
+
+    lock = threading.Lock()
+
+    def size_thunk(prefix):
+      nonlocal total_bytes
+      nonlocal lock
+
+      with self._get_connection() as conn:
+        subtree_bytes = conn.subtree_size(prefix)
+        with lock:
+          total_bytes += subtree_bytes
+    
+    schedule_jobs(
+      fns=( partial(size_thunk, path) for path in prefix ),
+      concurrency=self.num_threads,
+      progress=self.progress,
+      green=self.green,
+      total=total,
+    )
+
+    return total_bytes
+
   @parallelize(desc="Delete")
   def delete(
     self, paths:GetPathType, total:Optional[int] = None, 
@@ -1666,6 +1695,12 @@ class CloudFiles:
       return os.path.join(*paths)
     return posixpath.join(*paths)
 
+  @property
+  def sep(self) -> str:
+    if self._path.protocol == "file":
+      return os.sep
+    return posixpath.sep
+
   def dirname(self, path:str) -> str:
     if self._path.protocol == "file":
       return os.path.dirname(path)
@@ -1706,11 +1741,17 @@ class CloudFiles:
 
 class CloudFile:
   def __init__(
-    self, path:str, cache_meta:bool = False, 
+    self,
+    path:str,
+    cache_meta:bool = False, 
     secrets:SecretsType = None,
     composite_upload_threshold:int = int(1e8),
     locking:bool = True,
     lock_dir:Optional[str] = None,
+    endpoint:Optional[str] = None, 
+    no_sign_request:bool = False,
+    request_payer:Optional[str] = None,
+    use_https:bool = False,
   ):
     path = paths.normalize(path)
     self.cf = CloudFiles(
@@ -1719,6 +1760,10 @@ class CloudFile:
       composite_upload_threshold=composite_upload_threshold,
       locking=locking,
       lock_dir=lock_dir,
+      use_https=use_https,
+      endpoint=endpoint,
+      request_payer=request_payer,
+      no_sign_request=no_sign_request,
     )
     self.filename = paths.basename(path)
     
@@ -1726,6 +1771,10 @@ class CloudFile:
     self._size:Optional[int] = None
     self._head = None
 
+  @property
+  def sep(self) -> str:
+    return self.cf.sep
+
   @property
   def protocol(self):
     return self.cf.protocol

cloudfiles/interfaces.py

@@ -48,6 +48,7 @@ MEM_POOL = None
 
 S3_ACLS = {
   "tigerdata": "private",
+  "nokura": "public-read",
 }
 
 S3ConnectionPoolParams = namedtuple('S3ConnectionPoolParams', 'service bucket_name request_payer')
@@ -303,6 +304,19 @@ class FileInterface(StorageInterface):
 
     return self.io_with_lock(do_size, path, exclusive=False)
 
+  def subtree_size(self, prefix:str = "") -> int:
+    total_bytes = 0
+
+    subdir = self.get_path_to_file("")
+    if prefix:
+      subdir = os.path.join(subdir, os.path.dirname(prefix))
+
+    for root, dirs, files in os.walk(subdir):
+      files = ( os.path.join(root, f) for f in files )
+      total_bytes += sum(( os.path.getsize(f) for f in files ))
+
+    return total_bytes
+
   def exists(self, file_path):
     path = self.get_path_to_file(file_path)
     def do_exists():
@@ -580,8 +594,7 @@ class MemoryInterface(StorageInterface):
 
     Returns: iterator
     """
-    layer_path = self.get_path_to_file("")        
-    path = os.path.join(layer_path, prefix) + '*'
+    layer_path = self.get_path_to_file("")
 
     remove = layer_path
     if len(remove) and remove[-1] != '/':
@@ -615,6 +628,21 @@ class MemoryInterface(StorageInterface):
     filenames.sort()
     return iter(filenames)
 
+  def subtree_size(self, prefix:str = "") -> int:
+    layer_path = self.get_path_to_file("")        
+
+    remove = layer_path
+    if len(remove) and remove[-1] != '/':
+      remove += '/'
+
+    total_bytes = 0
+    for filename, binary in self._data.items():
+      f_prefix = f.removeprefix(remove)[:len(prefix)]
+      if f_prefix == prefix:
+        total_bytes += len(binary)
+
+    return total_bytes
+
 class GoogleCloudStorageInterface(StorageInterface):
   exists_batch_size = Batch._MAX_BATCH_SIZE
   delete_batch_size = Batch._MAX_BATCH_SIZE
@@ -816,6 +844,8 @@ class GoogleCloudStorageInterface(StorageInterface):
     blobs = self._bucket.list_blobs(
       prefix=path, 
       delimiter=delimiter,
+      page_size=2500,
+      fields="items(name),nextPageToken",
     )
 
     for page in blobs.pages:
@@ -835,6 +865,24 @@ class GoogleCloudStorageInterface(StorageInterface):
           yield filename
 
 
+  @retry
+  def subtree_size(self, prefix:str = "") -> int:
+    layer_path = self.get_path_to_file("")        
+    path = posixpath.join(layer_path, prefix)
+
+    blobs = self._bucket.list_blobs(
+      prefix=path,
+      page_size=5000,
+      fields="items(name,size),nextPageToken",
+    )
+
+    total_bytes = 0
+    for page in blobs.pages:
+      for blob in page:
+        total_bytes += blob.size
+
+    return total_bytes
+
   def release_connection(self):
     global GC_POOL
     with GCS_BUCKET_POOL_LOCK:
@@ -882,6 +930,8 @@ class HttpInterface(StorageInterface):
     key = self.get_path_to_file(file_path)
     headers = self.default_headers()
     with self.session.head(key, headers=headers) as resp:
+      if resp.status_code in (404, 403):
+        return None
       resp.raise_for_status()
       return resp.headers
 
@@ -889,6 +939,9 @@ class HttpInterface(StorageInterface):
     headers = self.head(file_path)
     return int(headers["Content-Length"])
 
+  def subtree_size(self, prefix:str = "") -> int:
+    raise NotImplementedError()
+
   @retry
   def get_file(self, file_path, start=None, end=None, part_size=None):
     key = self.get_path_to_file(file_path)
@@ -899,24 +952,20 @@ class HttpInterface(StorageInterface):
       end = int(end - 1) if end is not None else ''
       headers["Range"] = f"bytes={start}-{end}"
     
-    resp = self.session.get(key, headers=headers)
-    
-    if resp.status_code in (404, 403):
-      return (None, None, None, None)
-    resp.close()
-    resp.raise_for_status()
+    with self.session.get(key, headers=headers, stream=True) as resp:    
+      if resp.status_code in (404, 403):
+        return (None, None, None, None)
+      resp.raise_for_status()
+      resp.raw.decode_content = False
+      content = resp.raw.read()
+      content_encoding = resp.headers.get('Content-Encoding', None)  
 
     # Don't check MD5 for http because the etag can come in many
     # forms from either GCS, S3 or another service entirely. We
     # probably won't figure out how to decode it right.
     # etag = resp.headers.get('etag', None)
-    content_encoding = resp.headers.get('Content-Encoding', None)
-
-    # requests automatically decodes these
-    if content_encoding in (None, '', 'gzip', 'deflate', 'br'):
-      content_encoding = None
     
-    return (resp.content, content_encoding, None, None)
+    return (content, content_encoding, None, None)
 
   @retry
   def save_file(self, src, dest, resumable) -> tuple[bool, int]:
@@ -1017,7 +1066,6 @@ class HttpInterface(StorageInterface):
         )
 
       for res in results.get("items", []):
-        print(res["name"])
         yield res["name"].removeprefix(strip)
       
       token = results.get("nextPageToken", None)
@@ -1490,6 +1538,44 @@ class S3Interface(StorageInterface):
       for filename in iterate(resp):
         yield filename
 
+  def subtree_size(self, prefix:str = "") -> int:
+    layer_path = self.get_path_to_file("")        
+    path = posixpath.join(layer_path, prefix)
+
+    @retry
+    def s3lst(path, continuation_token=None):
+      kwargs = {
+        'Bucket': self._path.bucket,
+        'Prefix': path,
+        **self._additional_attrs
+      }
+
+      if continuation_token:
+        kwargs['ContinuationToken'] = continuation_token
+
+      return self._conn.list_objects_v2(**kwargs)
+
+    resp = s3lst(path)
+    
+    def iterate(resp):
+      if 'Contents' not in resp.keys():
+        resp['Contents'] = []
+
+      for item in resp['Contents']:
+        yield item.get('Size', 0)
+
+    total_bytes = 0
+    for num_bytes in iterate(resp):
+      total_bytes += num_bytes
+
+    while resp['IsTruncated'] and resp['NextContinuationToken']:
+      resp = s3lst(path, resp['NextContinuationToken'])
+
+      for num_bytes in iterate(resp):
+        total_bytes += num_bytes
+
+    return total_bytes
+
   def release_connection(self):
     global S3_POOL
     service = self._path.alias or 's3'

cloudfiles_cli/cloudfiles_cli.py

@@ -809,7 +809,10 @@ def du(paths, grand_total, summarize, human_readable):
     npath = normalize_path(path)
     if ispathdir(path):
       cf = CloudFiles(npath)
-      results.append(cf.size(cf.list()))
+      if summarize:
+        results.append(cf.subtree_size())
+      else:
+        results.append(cf.size(cf.list()))
     else:
       cf = CloudFiles(os.path.dirname(npath))
       sz = cf.size(os.path.basename(npath))
@@ -839,7 +842,10 @@ def du(paths, grand_total, summarize, human_readable):
 
   summary = {}
   for path, res in zip(paths, results):
-    summary[path] = sum(res.values())
+    if isinstance(res, int):
+      summary[path] = res
+    else:
+      summary[path] = sum(res.values())
     if summarize:
       print(f"{SI(summary[path])}\t{path}")
 

cloud_files-5.8.2.dist-info/pbr.json

@@ -1 +0,0 @@
-{"git_version": "99528f8", "is_release": true}