ECS: How to handle Arabic characters in object names

ECS and Arabic Characters: Elastic Cloud Storage (ECS) does support UTF-8 encoding, which can represent Arabic letters and other Unicode characters.

1. S3 Encoding and Decoding:

• When Arabic object names are uploaded over S3, the names are encoded into UTF-8 before being stored.
• This ensures that the object names, which are in Arabic or any other non-ASCII characters, are safely transmitted and stored in ECS.

2. Data Representation in dataheadsvc-access.log:

• In the logs, the object name is shown in its encoded form (for example, %D8%AD%D9%85%D8%AF for محمد), which is a URL-encoded representation of the UTF-8 encoding.
• This is standard practice for representing non-ASCII characters in logs to ensure compatibility and avoid data corruption.

3. Restoring Original Names:

• To view the object names in their original Arabic form, the encoded values (for example, %D8%AD%D9%85%D8%AF) must be decoded back into UTF-8, which will render the text as seen in the application (for example, S3 Browser).

• The behavior is expected and aligns with how modern systems handle non-ASCII characters in object names.
• If we plan to list these objects with Arabic names directly from ECS logs. Decoding the encoded names programmatically (for example, using a script in Python or another language that supports URL decoding) is required.
• For user-facing systems, it is good to ensure that the encoding and decoding processes are seamless, so the Arabic names appear as expected without manual intervention.