dcicutils 8.8.4.1b15__py3-none-any.whl → 8.8.4.1b17__py3-none-any.whl

dcicutils/file_utils.py

@@ -1,4 +1,6 @@
 import glob
+import hashlib
+import io
 import os
 import pathlib
 from datetime import datetime
@@ -103,6 +105,76 @@ def are_files_equal(filea: str, fileb: str) -> bool:
         return False
 
 
+def compute_file_md5(file: str) -> str:
+    """
+    Returns the md5 checksum for the given file.
+    """
+    if not isinstance(file, str):
+        return ""
+    try:
+        md5 = hashlib.md5()
+        with open(file, "rb") as file:
+            for chunk in iter(lambda: file.read(4096), b""):
+                md5.update(chunk)
+        return md5.hexdigest()
+    except Exception:
+        return ""
+
+
+def compute_file_etag(file: str) -> Optional[str]:
+    """
+    Returns the AWS S3 "etag" for the given file; this value is md5-like but
+    not the same as a normal md5. We use this to compare that a file in S3
+    appears to be the exact the same file as a local file.
+    """
+    try:
+        with io.open(file, "rb") as f:
+            return _compute_file_etag(f)
+    except Exception:
+        return None
+
+
+def _compute_file_etag(f: io.BufferedReader) -> str:
+    # See: https://stackoverflow.com/questions/75723647/calculate-md5-from-aws-s3-etag
+    MULTIPART_THRESHOLD = 8388608
+    MULTIPART_CHUNKSIZE = 8388608
+    # BUFFER_SIZE = 1048576
+    # Verify some assumptions are correct
+    # assert(MULTIPART_CHUNKSIZE >= MULTIPART_THRESHOLD)
+    # assert((MULTIPART_THRESHOLD % BUFFER_SIZE) == 0)
+    # assert((MULTIPART_CHUNKSIZE % BUFFER_SIZE) == 0)
+    hash = hashlib.md5()
+    read = 0
+    chunks = None
+    while True:
+        # Read some from stdin, if we're at the end, stop reading
+        bits = f.read(1048576)
+        if len(bits) == 0:
+            break
+        read += len(bits)
+        hash.update(bits)
+        if chunks is None:
+            # We're handling a multi-part upload, so switch to calculating
+            # hashes of each chunk
+            if read >= MULTIPART_THRESHOLD:
+                chunks = b''
+        if chunks is not None:
+            if (read % MULTIPART_CHUNKSIZE) == 0:
+                # Dont with a chunk, add it to the list of hashes to hash later
+                chunks += hash.digest()
+                hash = hashlib.md5()
+    if chunks is None:
+        # Normal upload, just output the MD5 hash
+        etag = hash.hexdigest()
+    else:
+        # Multipart upload, need to output the hash of the hashes
+        if (read % MULTIPART_CHUNKSIZE) != 0:
+            # Add the last part if we have a partial chunk
+            chunks += hash.digest()
+        etag = hashlib.md5(chunks).hexdigest() + "-" + str(len(chunks) // 16)
+    return etag
+
+
 def create_random_file(file: Optional[str] = None, prefix: Optional[str] = None, suffix: Optional[str] = None,
                        nbytes: int = 1024, binary: bool = False, line_length: Optional[int] = None) -> str:
     """

dcicutils/http_utils.py

@@ -1,20 +1,31 @@
 from contextlib import contextmanager
 import requests
-from typing import Optional
+from typing import Callable, Optional
 from dcicutils.tmpfile_utils import temporary_file
 
 
 @contextmanager
-def download(url: str, suffix: Optional[str] = None, binary: bool = True) -> Optional[str]:
+def download(url: str, suffix: Optional[str] = None, binary: bool = True,
+             progress: Optional[Callable] = None) -> Optional[str]:
     """
     Context manager to ownload the given URL into a temporary file and yields the file
     path to it. An optional file suffix may be specified. Defaults to binary file mode;
     if this is not desired then pass False as the binary argument.
     """
+    if not callable(progress):
+        progress = None
     with temporary_file(suffix=suffix) as file:
         response = requests.get(url, stream=True)
+        if progress:
+            nbytes = 0
+            nbytes_total = None
+            if isinstance(content_length := response.headers.get("Content-Length"), str) and content_length.isdigit():
+                nbytes_total = int(content_length)
         with open(file, "wb" if binary is True else "w") as f:
             for chunk in response.iter_content(chunk_size=8192):
                 if chunk:
                     f.write(chunk)
+                if progress:
+                    nbytes += len(chunk)
+                    progress(nbytes, nbytes_total)
         yield file

{dcicutils-8.8.4.1b15.dist-info → dcicutils-8.8.4.1b17.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dcicutils
-Version: 8.8.4.1b15
+Version: 8.8.4.1b17
 Summary: Utility package for interacting with the 4DN Data Portal and other 4DN resources
 Home-page: https://github.com/4dn-dcic/utils
 License: MIT

{dcicutils-8.8.4.1b15.dist-info → dcicutils-8.8.4.1b17.dist-info}/RECORD

@@ -28,10 +28,10 @@ dcicutils/es_utils.py,sha256=ZksLh5ei7kRUfiFltk8sd2ZSfh15twbstrMzBr8HNw4,7541
 dcicutils/exceptions.py,sha256=4giQGtpak-omQv7BP6Ckeu91XK5fnDosC8gfdmN_ccA,9931
 dcicutils/ff_mocks.py,sha256=6RKS4eUiu_Wl8yP_8V0CaV75w4ZdWxdCuL1CVlnMrek,36918
 dcicutils/ff_utils.py,sha256=oIhuZPnGtfwj6bWyCc1u23JbMB_6InPp01ZqUOljd8M,73123
-dcicutils/file_utils.py,sha256=vPPFiGK5scq7opVQ_Ne8eeTAjy5OtSxyAgeJUievygQ,6283
+dcicutils/file_utils.py,sha256=msxA3fFTtK09Qc_I3-r9Y5Pp5WVJRPPpLlFYv3Rju-E,8697
 dcicutils/function_cache_decorator.py,sha256=XMyiEGODVr2WoAQ68vcoX_9_Xb9p8pZXdXl7keU8i2g,10026
 dcicutils/glacier_utils.py,sha256=Q4CVXsZCbP-SoZIsZ5NMcawDfelOLzbQnIlQn-GdlTo,34149
-dcicutils/http_utils.py,sha256=RB0x9hRMZM9Xd1x00c5J0iUzUdYzIQR0XKFiQ94HWO0,807
+dcicutils/http_utils.py,sha256=Je5ErNjR5e6lfSXGRncK_lcR_-zP38PIpmHjApy9Wi4,1289
 dcicutils/jh_utils.py,sha256=Gpsxb9XEzggF_-Eq3ukjKvTnuyb9V1SCSUXkXsES4Kg,11502
 dcicutils/kibana/dashboards.json,sha256=wHMB_mpJ8OaYhRRgvpZuihaB2lmSF64ADt_8hkBWgQg,16225
 dcicutils/kibana/readme.md,sha256=3KmHF9FH6A6xwYsNxRFLw27q0XzHYnjZOlYUnn3VkQQ,2164
@@ -73,8 +73,8 @@ dcicutils/trace_utils.py,sha256=g8kwV4ebEy5kXW6oOrEAUsurBcCROvwtZqz9fczsGRE,1769
 dcicutils/validation_utils.py,sha256=cMZIU2cY98FYtzK52z5WUYck7urH6JcqOuz9jkXpqzg,14797
 dcicutils/variant_utils.py,sha256=2H9azNx3xAj-MySg-uZ2SFqbWs4kZvf61JnK6b-h4Qw,4343
 dcicutils/zip_utils.py,sha256=_Y9EmL3D2dUZhxucxHvrtmmlbZmK4FpSsHEb7rGSJLU,3265
-dcicutils-8.8.4.1b15.dist-info/LICENSE.txt,sha256=qnwSmfnEWMl5l78VPDEzAmEbLVrRqQvfUQiHT0ehrOo,1102
-dcicutils-8.8.4.1b15.dist-info/METADATA,sha256=-eKROBZs2uM-7YOnS8nHPkO9U1nHnRz5UJhZYU9svgQ,3440
-dcicutils-8.8.4.1b15.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
-dcicutils-8.8.4.1b15.dist-info/entry_points.txt,sha256=51Q4F_2V10L0282W7HFjP4jdzW4K8lnWDARJQVFy_hw,270
-dcicutils-8.8.4.1b15.dist-info/RECORD,,
+dcicutils-8.8.4.1b17.dist-info/LICENSE.txt,sha256=qnwSmfnEWMl5l78VPDEzAmEbLVrRqQvfUQiHT0ehrOo,1102
+dcicutils-8.8.4.1b17.dist-info/METADATA,sha256=KLVzlNXL2JHkY6n3bL-fbnoVa1kNKAnIbK7bRluuxlM,3440
+dcicutils-8.8.4.1b17.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
+dcicutils-8.8.4.1b17.dist-info/entry_points.txt,sha256=51Q4F_2V10L0282W7HFjP4jdzW4K8lnWDARJQVFy_hw,270
+dcicutils-8.8.4.1b17.dist-info/RECORD,,