Importing from a folder in a SMB share fails #59

New Issue

2023-01-23T16:35:56Z

rishighan commented

2023-01-23 16:35:56 +00:00

(Migrated from github.com)

Assuming that you have a:

folder mounted as a SAMBA share,
you perform a clean import in ThreeTwo from the Import page, and,
browser tab remains open, internet connection unabated,

the import fails cryptically.

Assuming that you have a: 1. folder mounted as a SAMBA share, 2. you perform a clean import in ThreeTwo from the `Import` page, and, 3. browser tab remains open, internet connection unabated, the import fails cryptically.

rishighan commented

2023-03-15 15:52:54 +00:00

(Migrated from github.com)

Troubleshooting this further, specifically around the files @nstephenh shared with me:

The use of sanitize method here was causing illegal characters to be removed from the filename, subsequently breaking functionality that attempts to extract the said file into the covers folder.
A .cbr file, which in reality was a .cbz file tripped up the uncompression code, failing out and terminating the queue.

Troubleshooting this further, specifically around the files @nstephenh shared with me: 1. The use of `sanitize` method [here](https://github.com/rishighan/threetwo-core-service/blob/8b814c0ac5f60a2a575b92f02eb217635bb1ed31/utils/uncompression.utils.ts#L165) was causing _illegal_ characters to be removed from the filename, subsequently breaking functionality that attempts to extract the said file into the `covers` folder. 2. A `.cbr` file, which in reality was a `.cbz` file tripped up the uncompression code, failing out and terminating the queue.

rishighan commented

2023-03-15 15:59:44 +00:00

(Migrated from github.com)

The aforementioned findings are still not severe enough to warrant terminating the import queue. Ideally:

The uncompression code should not rely on filename extensions, and use MIMEtype instead to infer if the file is a .zip or a .rar
Failures to uncompress, parse the files should throw an error which gets pushed into an array of issues and then the final array should be caught. @nstephenh Please feel free to weigh in.

The aforementioned findings are still not severe enough to warrant terminating the import queue. Ideally: 1. The uncompression code should not rely on filename extensions, and use MIMEtype instead to infer if the file is a `.zip` or a `.rar` 2. Failures to uncompress, parse the files should throw an error which gets pushed into an `array` of issues and then the final array should be caught. @nstephenh Please feel free to weigh in.

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: frishi/threetwo#59