The rapid expansion of IoT devices has led to the generation of vast amounts of time-series data, presenting both opportunities and challenges, particularly in the demand for efficient data storage and transmission solutions. Although mainstream lossless floating-point compression algorithms have achieved impressive compression ratios, there is a continuous need for further optimization to handle increasingly large datasets. This paper presents Monkey, a novel streaming lossless compression algorithm that enhances the adaptive-length encoding scheme to achieve superior compression performance. Building on the erasing-based method, Monkey employs a segmentation strategy to dynamically locate processed consecutive data points during decoding and removes the flags traditionally used for segmentation. Extensive experiments on widely adopted datasets demonstrate that Monkey substantially advances many XOR-based state-of-the-art lossless compression algorithms: It consistently outperforms Elf+ across all tested datasets with up to 11.5% improvement in compression ratio.