[GSoC 2026] Introduction and Project Interest: Support generating filesystems from manifests

[Sri Lasya Prathipati]

Subject: [GSoC 2026] Introduction and Project Interest: Support
generating filesystems from manifests

Body:

Hello EROFS Community,

My name is Lasya, and I am a 3rd-year Computer Science Engineering
student. I am writing to express my strong interest in the GSoC 2026
project: "Support generating filesystems from manifests."

I have spent the last few days familiarizing myself with the project's
goals and the erofs-utils codebase. I wanted to share my progress to
demonstrate my commitment to this task:

Development Environment: I have successfully configured a WSL2
(Ubuntu) environment and verified that I can compile the project from
source using autogen.sh and make. The local mkfs.erofs binary is
functional on my system.

Code Analysis: I have begun a deep dive into lib/tar.c. I am
specifically studying the tarerofs_parse_tar() function (line 700) to
understand how it iterates through the tar_header and maps metadata to
the erofs_inode structure. I see this as the primary template for the
manifest parser I intend to build.

Initial Research: I am currently comparing the "unix proto" and "BSD
mtree" formats. My goal is to design a manifest parser that is robust
enough to handle various metadata types (UIDs, GIDs, and permissions)
while remaining consistent with the existing importer logic in
erofs-utils.

I am highly motivated to contribute to EROFS and would welcome any
early feedback on whether there are specific manifest formats the
community prioritizes, or any architectural nuances I should keep in
mind while drafting my formal proposal.

Thank you for your time and for this opportunity.

Best regards,

Lasya