DNA-based data storage has developed rapidly in recent years and is increasingly viewed as a promising solution for long-term, ultra-dense archival storage. As research moves beyond early demonstrations, encoding has become a central technical challenge that directly affects system reliability, scalability, and practical deployment. This survey presents a structured overview of coding techniques for DNA data storage. We first outline the biochemical and structural constraints of DNA storage, including synthesis and sequencing errors, molecular instability, and context-dependent biases, which strongly influence encoding design. We then review how classical error-correcting codes—such as Reed–Solomon, BCH, convolutional, fountain, and LDPC codes—have been adapted for use in molecular storage systems. Next, we introduce emerging coding approaches designed specifically for DNA storage, including ALD and Damerau distance codes, set-based coding models, and biochemical rule–guided methods such as Yin–Yang coding. Beyond error correction, the survey also examines coding strategies aimed at improving efficiency, with attention to storage rate, biochemical feasibility, and structural robustness. By comparing representative encoding schemes across different channel conditions and application scenarios, this survey summarizes existing theoretical analyses and experimental results and provides practical guidance for selecting suitable codes. Finally, we discuss open challenges and future research directions, highlighting the potential of adaptive, secure, and machine learning–assisted encoding techniques for large-scale DNA data storage systems.