Mycelia Julia Coding Style

When working in the Mycelia repository, follow these strict coding rules.

Import Rules (STRICT)

• NEVER use using - always use import
• Fully qualify all names - use Mycelia.Rhizomorph, DataFrames.DataFrame(), etc.
• No shorthand aliases - avoid R = Mycelia.Rhizomorph; use full qualified names
• Dependencies are imported at top-level in src/Mycelia.jl and available in all source files

Example:

julia
1# WRONG
2using DataFrames
3df = DataFrame()
4
5# WRONG
6import DataFrames: DataFrame
7df = DataFrame()
8
9# CORRECT
10import DataFrames
11df = DataFrames.DataFrame()

Naming Conventions

• Modules: CamelCase (e.g., Rhizomorph, KmerAnalysis)
• Functions/Variables: snake_case (e.g., process_sequences, kmer_count)
• Constants: SCREAMING_SNAKE_CASE (e.g., MAX_KMER_SIZE)
• No emojis in code or comments

Sequence Types (BioSequences)

Always use proper BioSequences types:

• BioSequences.LongDNA{4} for DNA
• BioSequences.LongRNA{4} for RNA
• BioSequences.LongAA for amino acids

Extract from FASTQ:

julia
1FASTX.sequence(BioSequences.LongDNA{4}, record)

Avoid string conversions in k-mer graphs, qualmer graphs, and assembly algorithms. Use string() only for external tool interfaces or final output.

Formatting

• 4-space indentation
• Clear docstrings: """signature..."""
• Prefer pure functions and explicit keyword arguments
• Use joinpath for file paths (portability)
• Avoid type piracy

Default formatter: JuliaFormatter.jl with SciMLStyle or BlueStyle, unless repo indicates Runic.jl.

Architecture Notes

• Main module uses dynamic file inclusion
• All .jl files in src/ are auto-included via Mycelia.jl
• Rhizomorph and Mycelia do NOT export symbols by design
• Tests and callers MUST use fully qualified names