dcicutils 8.8.4.1b13__py3-none-any.whl → 8.8.4.1b16__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/misc_utils.py

@@ -19,6 +19,7 @@ import pytz
 import re
 import rfc3986.validators
 import rfc3986.exceptions
+import shortuuid
 import time
 import uuid
 import warnings
@@ -2698,3 +2699,9 @@ def get_cpu_architecture_name() -> str:
         if os_architecture_name == "x86_64": return "amd64"  # noqa
         return os_architecture_name
     return ""
+
+
+def short_uuid(length: Optional[int] = None):
+    if (length is None) or (not isinstance(length, int)) or (length < 1):
+        length = 16
+    return shortuuid.ShortUUID().random(length=length)

{dcicutils-8.8.4.1b13.dist-info → dcicutils-8.8.4.1b16.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dcicutils
-Version: 8.8.4.1b13
+Version: 8.8.4.1b16
 Summary: Utility package for interacting with the 4DN Data Portal and other 4DN resources
 Home-page: https://github.com/4dn-dcic/utils
 License: MIT
@@ -26,8 +26,8 @@ Requires-Dist: PyJWT (>=2.6.0,<3.0.0)
 Requires-Dist: PyYAML (>=6.0.1,<7.0.0)
 Requires-Dist: appdirs (>=1.4.4,<2.0.0)
 Requires-Dist: aws-requests-auth (>=0.4.2,<1)
-Requires-Dist: boto3 (>=1.34.90,<2.0.0)
-Requires-Dist: botocore (>=1.34.90,<2.0.0)
+Requires-Dist: boto3 (>=1.34.93,<2.0.0)
+Requires-Dist: botocore (>=1.34.93,<2.0.0)
 Requires-Dist: chardet (>=5.2.0,<6.0.0)
 Requires-Dist: docker (>=4.4.4,<5.0.0)
 Requires-Dist: elasticsearch (==7.13.4)
@@ -43,6 +43,7 @@ Requires-Dist: pytz (>=2020.4)
 Requires-Dist: redis (>=4.5.1,<5.0.0)
 Requires-Dist: requests (>=2.21.0,<3.0.0)
 Requires-Dist: rfc3986 (>=1.4.0,<2.0.0)
+Requires-Dist: shortuuid (>=1.0.13,<2.0.0)
 Requires-Dist: structlog (>=19.2.0,<20.0.0)
 Requires-Dist: toml (>=0.10.1,<1)
 Requires-Dist: tqdm (>=4.66.2,<5.0.0)

{dcicutils-8.8.4.1b13.dist-info → dcicutils-8.8.4.1b16.dist-info}/RECORD

@@ -28,7 +28,7 @@ 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
@@ -44,7 +44,7 @@ dcicutils/license_policies/park-lab-gpl-pipeline.jsonc,sha256=vLZkwm3Js-kjV44nug
 dcicutils/license_policies/park-lab-pipeline.jsonc,sha256=9qlY0ASy3iUMQlr3gorVcXrSfRHnVGbLhkS427UaRy4,283
 dcicutils/license_utils.py,sha256=d1cq6iwv5Ju-VjdoINi6q7CPNNL7Oz6rcJdLMY38RX0,46978
 dcicutils/log_utils.py,sha256=7pWMc6vyrorUZQf-V-M3YC6zrPgNhuV_fzm9xqTPph0,10883
-dcicutils/misc_utils.py,sha256=Nw47AZs-cSOzGp5TZWrQZftePcahz1yfw6iNwOzUt-s,105530
+dcicutils/misc_utils.py,sha256=eVZ3lEkDebweKCeza2GIo7x3qEqqkj61Ilr17eMFlR0,105744
 dcicutils/obfuscation_utils.py,sha256=fo2jOmDRC6xWpYX49u80bVNisqRRoPskFNX3ymFAmjw,5963
 dcicutils/opensearch_utils.py,sha256=V2exmFYW8Xl2_pGFixF4I2Cc549Opwe4PhFi5twC0M8,1017
 dcicutils/portal_object_utils.py,sha256=gDXRgPsRvqCFwbC8WatsuflAxNiigOnqr0Hi93k3AgE,15422
@@ -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-8.8.4.1b13.dist-info/LICENSE.txt,sha256=qnwSmfnEWMl5l78VPDEzAmEbLVrRqQvfUQiHT0ehrOo,1102
-dcicutils-8.8.4.1b13.dist-info/METADATA,sha256=gPRDYcqVxkYQNNEWUBaES-znrDZQhf-BT2Ie5ooqFwU,3397
-dcicutils-8.8.4.1b13.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
-dcicutils-8.8.4.1b13.dist-info/entry_points.txt,sha256=51Q4F_2V10L0282W7HFjP4jdzW4K8lnWDARJQVFy_hw,270
-dcicutils-8.8.4.1b13.dist-info/RECORD,,
+dcicutils-8.8.4.1b16.dist-info/LICENSE.txt,sha256=qnwSmfnEWMl5l78VPDEzAmEbLVrRqQvfUQiHT0ehrOo,1102
+dcicutils-8.8.4.1b16.dist-info/METADATA,sha256=7R_Eatzjy4Ez8_JufgAKLIQ_O6z0fVvjqHzLEwfE9O0,3440
+dcicutils-8.8.4.1b16.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
+dcicutils-8.8.4.1b16.dist-info/entry_points.txt,sha256=51Q4F_2V10L0282W7HFjP4jdzW4K8lnWDARJQVFy_hw,270
+dcicutils-8.8.4.1b16.dist-info/RECORD,,