feat: allowing human-readable date in presigns() expiration

noteooks/test_general.ipynb

@@ -534,7 +534,7 @@
     {
      "data": {
       "text/plain": [
-       "'https://bucket-external.s3.amazonaws.com/dataset/dataset_qft/moody_qft_danbooru.json?AWSAccessKeyId=AKIAVVWUPRZ2AC6O2S3I&Signature=0z3LgRw9wcdnSqcCmqX7VaqAU7Q%3D&Expires=1727805561'"
+       "'https://bucket-external.s3.amazonaws.com/dataset/dataset_qft/moody_qft_danbooru.json?AWSAccessKeyId=AKIAVVWUPRZ2AC6O2S3I&Signature=sXf%2FvVsHcWcBK8aA%2FBW5i4dseEw%3D&Expires=1759172427'"
       ]
      },
      "execution_count": 1,
@@ -548,7 +548,7 @@
     "\n",
     "uri = \"s3://bucket-external/dataset/dataset_qft/moody_qft_danbooru.json\"\n",
     "\n",
-    "signed = ub.presigns(uri)\n",
+    "signed = ub.presigns(uri, expiration=\"1y\")\n",
     "signed"
    ]
   }

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "unibox"
-version = "0.4.11"
+version = "0.4.12"
 description = "Unibox provides unified interface for common file operations."
 authors = ["yada <trojblue@gmail.com>"]
 license = "MIT"
@@ -20,6 +20,7 @@ pyarrow = ">=10.0.1"
 boto3 = "^1.28.76"
 requests = "^2.31.0"
 orjson = "^3.9.10"
+humanfriendly = "^10.0"
 
 [tool.poetry.extras]
 full = ["ipython"]

unibox/__init__.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+
 from doctest import debug
 from pathlib import Path
 from typing import Union, List, Dict, Any
@@ -140,11 +141,12 @@ def peeks(data: Any, n=3, console_print=False) -> Dict[str, Any]:
     return peeker.peeks(data)
 
 
-def presigns(s3_uri: str, timeout: int = 86400) -> str:
+def presigns(s3_uri: str, expiration: Union[int, str] = "1d") -> str:
     """
     Generate a presigned URL from a given S3 URI.
     :param s3_uri: S3 URI (e.g., 's3://bucket-name/object-key')
-    :param timeout: Time in seconds for the presigned URL to remain valid (default: 1 day)
+        :param expiration: Time in seconds for the presigned URL to remain valid (default: 1 day).
+                        Accepts either an integer (seconds) or human-readable strings like "1d", "1mo", "1y".
     :return: Presigned URL as string. If error, returns None.
     """
-    return s3_client.generate_presigned_uri(s3_uri, timeout)
+    return s3_client.generate_presigned_uri(s3_uri, expiration=expiration)

unibox/utils/s3_client.py

@@ -8,6 +8,12 @@
 from tqdm.auto import tqdm
 import pandas as pd
 
+import logging
+import humanfriendly
+from botocore.exceptions import ClientError
+from typing import Union
+
+
 def parse_s3_url(url: str):
     parsed_url = urlparse(url)
     if parsed_url.scheme != "s3":
@@ -114,19 +120,32 @@ def traverse(self, s3_uri, include_extensions=None, exclude_extensions=None,
 
         return all_entries
 
-    def generate_presigned_uri(self, s3_uri: str, timeout: int = 86400) -> str:
+
+    def generate_presigned_uri(self, s3_uri: str, expiration: Union[int, str] = "1d") -> str:
         """
         Generate a presigned URL from a given S3 URI.
 
         :param s3_uri: S3 URI (e.g., 's3://bucket-name/object-key')
-        :param timeout: Time in seconds for the presigned URL to remain valid (default: 1 day)
+        :param expiration: Time in seconds for the presigned URL to remain valid (default: 1 day).
+                        Accepts either an integer (seconds) or human-readable strings like "1d", "1mo", "1y".
         :return: Presigned URL as string. If error, returns None.
         """
         bucket, key = parse_s3_url(s3_uri)
+
+        # Convert human-readable time to seconds if needed
+        if isinstance(expiration, str):
+            try:
+                expiration = int(humanfriendly.parse_timespan(expiration))
+            except Exception as e:
+                logging.error(f"Invalid time format: {expiration}. Error: {e}")
+                return None
+
         try:
-            response = self.s3.generate_presigned_url('get_object',
-                                                      Params={'Bucket': bucket, 'Key': key},
-                                                      ExpiresIn=timeout)
+            response = self.s3.generate_presigned_url(
+                'get_object',
+                Params={'Bucket': bucket, 'Key': key},
+                ExpiresIn=expiration
+            )
             return response
         except ClientError as e:
             logging.error(f"Failed to generate presigned URL for {s3_uri}: {e}")