The intelligent textile sensors based on fiber (1D) and fabric (2D) are the ideal candidates for wearable devices. Their flexible weaving and unique structure endow them with flexibility, lightweight, good air permeability, and feasible integration with garments. In view of the spring-up of novel textile-based strain sensors, the novel materials and fabrication approaches were elaborated from spatial perspectives, i.e., 1D fibers/yarn and 2D fabric. The intrinsic sensing mechanism is the primary factor affecting sensor sensitivity, and the variation trend of the sensing signal is closely related to it. Although existing studies have involved various sensing mechanisms, there is still lacking systematic classification and discussion. Hence, the sensing mechanisms of textile-based sensors were elaborated from spatial perspectives. Considering that strain sensors were mostly based on resistance variation, the sensing mechanisms of resistive textile-based strain sensors were mainly focused, mainly including fiber deformation, tunneling effect, crack propagation, fabric deformation, electrical contact and bridge connection. Meanwhile, the corresponding resistance prediction models, usually used as important data fitting methodology, were also comprehensively discussed, which can reproduce the resistance trend and provide guidance for the sensor performance. Finally, the multifunctionality of textile-based strain sensors was summarized, namely multi-mode signal detection, visual interaction, energy collection, thermal management and medical treatment were discussed. It was expected to provide research insights into the multifunctional integration of textile sensors.