Address PR #19 review comments: naming, test coverage, and documentation

[Copilot]

Applies all actionable feedback from PR #19 review thread #3690473082.

Code Quality

• Naming: wantChunks → wantParts, chunks → parts (consistent with multipart upload terminology)
• Cleanup: Removed commented-out code, unnecessary blank lines
• Error formatting: >%q< → %q (conventional Go error format)
• Spelling: maximun → maximum, terrabytes → terabytes

Test Coverage

Expanded TestOptimalChunkSize from 7 to 15 cases covering all thresholds and boundaries:

• Zero/edge cases: 0 bytes, 1 MB
• Boundary transitions: 100 MB±1B, 1 GB±1B, 10 GB±1B, 100 GB±1B
• Scaled ranges: 5 GB, 500 GB
• Fixed parts calculation for fileSize == 0

Observability

Added logging when git config fails:

func init() {
    v, err := GetLfsCustomTransferInt("lfs.customtransfer.drs.multipart-min-chunk-size", 10)
    if err != nil {
        log.Printf("Warning: Could not read git config for multipart-min-chunk-size, using default (10 MB): %v\n", err)
        MinChunkSize = int64(10) * MB
        return
    }
    MinChunkSize = int64(v) * MB
}

Documentation

Added algorithm description to NEW section matching OLD section structure:

• Calculation logic for each file size range
• Behavior summary (single PUT, fixed chunks, scaled ranges)
• Examples with expected chunk sizes and part counts
• Removed markdown escape artifacts (\- → -)

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[Copilot]

Pull request overview

This PR applies follow-up review feedback from PR #19, refining the OptimalChunkSize algorithm documentation, tightening test coverage and naming around multipart parts calculation, and improving observability and comments in common constants.

Changes:

• Expanded and clarified the engineering note in docs/optimal-chunk-size.md to document the new OptimalChunkSize behavior, test strategy, and cleaned up markdown artifacts.
• Strengthened TestOptimalChunkSize in client/upload/utils_test.go with additional boundary and scaled-range cases, renamed wantChunks → wantParts, and fixed the parts calculation for zero-byte files.
• Updated client/common/constants.go to fix spelling in comments, remove dead code, add logging when git config lookup fails, and normalize error message formatting.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
docs/optimal-chunk-size.md	Documents the new OptimalChunkSize algorithm, its behavior across size ranges, and provides a concrete test plan and examples.
client/upload/utils_test.go	Extends and renames the table-driven tests to validate chunk sizes and part counts across edge cases and scaled ranges, including the 0-byte behavior.
client/upload/utils.go	Minor cleanup in OptimalChunkSize, keeping implementation aligned with the documented algorithm.
client/common/constants.go	Corrects comment spelling, adds a warning log on git config read failure for multipart min chunk size, removes obsolete commented code, and standardizes error message formatting in GetLfsCustomTransferInt.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The test table exercises the 0-byte, exact-threshold, and +1 byte boundary cases, but it does not include the documented -1 byte boundaries or a negative fileSize, even though docs/optimal-chunk-size.md ("Test design notes" #3 and #4) explicitly calls for those cases. To fully match the documented test plan and guard against off-by-one or fallback regressions, consider adding subtests for fileSize < 0 and for each X - 1 byte threshold (e.g., 100 MB - 1 B, 1 GB - 1 B, etc.).