Observed when processing files with a Python script of my own, but likely to be a nuisance to developers of any software, that allows you to store files in arbitrary locations.

A common pattern when modifying files is to first write them to a temporary file in the same directory (ensuring that it is on the same physical drive) and then replacing the original file with the modified file. This pattern gives safety against data corruption, if the program fails while creating the new version of the file.

In Python, that translates in the simplest case to:

tempfile = file.with_suffix(file.suffix + \".tmp\")\ntempfile.write_text(\"hello world\")\ntempfile.replace(file)

On Windows, I find that this pattern frequently fails with an error like

PermissionError: [WinError 32] The process cannot access the file\nbecause it is being used by another process: 'E:\\\\Dropbox\\\\tmp\\\\test.txt'

but only for files in Dropbox. This means that any script I write must allow for retrying the .replace() step.

To give a more complete, read-to use, example I provide the subsequent script, which repeatedly performs the pattern both in %TEMP% and in E:\\Dropbox\\tmp. On my System, Dropbox is on E:\\Dropbox, so that part has to be adjusted for anyone trying the script. For me, the script rarely manages more than 8 repetitions.

import os\nfrom pathlib import Path\n\nTEMPFILE_DIR = Path(os.getenv(\"TEMP\"))\nDROPBOX_ROOT = Path(\"E:/Dropbox\")      # Adjust for you system, usually C:/Users/Username/Dropbox\"\n\ndef main():\n    tempdirfile = TEMPFILE_DIR / \"tmp.txt\"\n    dropboxfile = Path(\"E:/Dropbox/tmp/test.txt\")\n\n    i = 0\n    while True:\n        i += 1\n        print(i)\n        for file in (tempdirfile, dropboxfile):\n            # file.unlink() # would solve in most cases, but still fails every few hundred attempts\n            #               # also defeats the purpose of using the pattern in the first place\n            tempdirfile.parent.mkdir(parents=True, exist_ok=True)\n            tempfile = file.with_suffix(\".txt.tmp\")\n            tempfile.write_text(TEXT)\n            print(\"In {}: {} -> {}\".format(file.parent, file.name, tempfile.name))\n            tempfile.replace(file)\n\nTEXT = \"\"\"\nLorem ipsum dolor sit amet, consectetur adipiscing elit. Mauris justo\npurus, pharetra quis ornare in, tincidunt et odio. Sed pulvinar lorem\nvel justo vestibulum aliquet. Nunc et nisi dui. Cras tempus lacus ac\nquam imperdiet mattis. Maecenas tristique dui vel orci condimentum\nvolutpat. Quisque ut mauris metus. Fusce libero metus, luctus eget\nsagittis a, venenatis et ex. Aenean purus tellus, feugiat et libero\nvel, congue elementum tellus. Cras at interdum nulla. Ut ut nunc id\nnibh lacinia dapibus quis id turpis. Pellentesque habitant morbi\ntristique senectus et netus et malesuada fames ac turpis egestas.\nVestibulum eu iaculis sapien. Fusce lobortis massa at molestie\nposuere.\n\"\"\"\n\nif __name__ == \"__main__\":\n    main()

Observed when processing files with a Python script of my own, but likely to be a nuisance to developers of any software, that allows you to store files in arbitrary locations.

A common pattern when modifying files is to first write them to a temporary file in the same directory (ensuring that it is on the same physical drive) and then replacing the original file with the modified file. This pattern gives safety against data corruption, if the program fails while creating the new version of the file.

In Python, that translates in the simplest case to:

tempfile = file.with_suffix(file.suffix + \".tmp\")\ntempfile.write_text(\"hello world\")\ntempfile.replace(file)

On Windows, I find that this pattern frequently fails with an error like

PermissionError: [WinError 32] The process cannot access the file\nbecause it is being used by another process: 'E:\\\\Dropbox\\\\tmp\\\\test.txt'

but only for files in Dropbox. This means that any script I write must allow for retrying the .replace() step.

To give a more complete, read-to use, example I provide the subsequent script, which repeatedly performs the pattern both in %TEMP% and in E:\\Dropbox\\tmp. On my System, Dropbox is on E:\\Dropbox, so that part has to be adjusted for anyone trying the script. For me, the script rarely manages more than 8 repetitions.

import os\nfrom pathlib import Path\n\nTEMPFILE_DIR = Path(os.getenv(\"TEMP\"))\nDROPBOX_ROOT = Path(\"E:/Dropbox\")      # Adjust for you system, usually C:/Users/Username/Dropbox\"\n\ndef main():\n    tempdirfile = TEMPFILE_DIR / \"tmp.txt\"\n    dropboxfile = Path(\"E:/Dropbox/tmp/test.txt\")\n\n    i = 0\n    while True:\n        i += 1\n        print(i)\n        for file in (tempdirfile, dropboxfile):\n            # file.unlink() # would solve in most cases, but still fails every few hundred attempts\n            #               # also defeats the purpose of using the pattern in the first place\n            tempdirfile.parent.mkdir(parents=True, exist_ok=True)\n            tempfile = file.with_suffix(\".txt.tmp\")\n            tempfile.write_text(TEXT)\n            print(\"In {}: {} -> {}\".format(file.parent, file.name, tempfile.name))\n            tempfile.replace(file)\n\nTEXT = \"\"\"\nLorem ipsum dolor sit amet, consectetur adipiscing elit. Mauris justo\npurus, pharetra quis ornare in, tincidunt et odio. Sed pulvinar lorem\nvel justo vestibulum aliquet. Nunc et nisi dui. Cras tempus lacus ac\nquam imperdiet mattis. Maecenas tristique dui vel orci condimentum\nvolutpat. Quisque ut mauris metus. Fusce libero metus, luctus eget\nsagittis a, venenatis et ex. Aenean purus tellus, feugiat et libero\nvel, congue elementum tellus. Cras at interdum nulla. Ut ut nunc id\nnibh lacinia dapibus quis id turpis. Pellentesque habitant morbi\ntristique senectus et netus et malesuada fames ac turpis egestas.\nVestibulum eu iaculis sapien. Fusce lobortis massa at molestie\nposuere.\n\"\"\"\n\nif __name__ == \"__main__\":\n    main()

If there is a better place to send the bug report, please let me know.

If there is a better place to send the bug report, please let me know.

KlausB wrote:\n

If there is a better place to send the bug report, please let me know.

\n

I've moved it to the API part of the forums for you. 

KlausB wrote:\n

If there is a better place to send the bug report, please let me know.

\n

I've moved it to the API part of the forums for you. 

Hi KlausB 

Apologies for the delay in reviewing this. The issue you're encountering appears to be outside the scope of the Dropbox API. However, it looks like the problem may be related to the Dropbox Desktop Client. To help you get the most relevant support, I’ve moved this post to the Dropbox Client section of the Dropbox Community Forum for further review.

We appreciate your patience and hope someone from the community or our team can assist you further there.

Hi  

Apologies for the delay in reviewing this. The issue you're encountering appears to be outside the scope of the Dropbox API. However, it looks like the problem may be related to the Dropbox Desktop Client. To help you get the most relevant support, I’ve moved this post to the Dropbox Client section of the Dropbox Community Forum for further review.

We appreciate your patience and hope someone from the community or our team can assist you further there.

Hey KlausB - sorry for the late response here, but if you're still having issues with this, you can let us know and we'll reach out via email to investigate further, if needed. 

Hey - sorry for the late response here, but if you're still having issues with this, you can let us know and we'll reach out via email to investigate further, if needed.