Stars with an external convective envelope have rotation periods that evolve as the square root of the age on the main sequence. This led to empirical "gyrochronology" relations, which should allow us to determine the age of a star from its surface rotation period. Recently, the exquisite Kepler data showed that there is a stalling of the magnetic braking. This was based on the measurement of surface rotation based on the light curve modulation related to the passage of spot on the visible stellar disks. Here, we present the analysis of more than 100,000 of solar-like main-sequence stars (F, G, K, M dwarfs) observed in long cadence for 4 years with the Kepler mission, leading to the largest catalog of rotation periods available for more than 40,000 dwarfs along with their proxies for magnetic activity. We apply our improved rotation pipeline, which combines three different methods (time-frequency analysis, auto-correlation function and composite spectrum) with different calibrations. For K and M dwarfs, we obtained more than 15,000 rotation periods, including new periods for ∼4000 stars. For F and G dwarfs, we implemented a machine learning algorithm to automatically select the reliable rotation periods and reduce visual checks. Finally, we know for the Sun and many solar-like stars, magnetic activity can suppress the amplitude of the acoustic modes. We analyzed ∼1000 solar-like stars observed in short-cadence during the Kepler survey phase where modes were undetected. Surprisingly magnetic activity might not the main reason for not detecting modes in those stars.