Experimental knowledge of the partonic structure of mesons (the pion and kaon) is very limited due to the lack of a stable pion target! Our current knowledge of the pion structure function in the valence region is obtained primarily from pionic Drell-Yan scattering, and in the pion sea region at low Bjorken-x, from hard diffractive processes measured on e − p collisions at HERA. These data seem to indicate that the pion sea has approximately one-third of the magnitude of the proton sea, while from the parton model one expects the pion sea to be two-thirds of the proton sea. But...
at large Bjorken-x virtually nothing is known about the contribution of sea quarks and gluons.
The Electron-Ion Collider (EIC) with an acceptance optimized for forward physics has the potential for accessing pion and kaon structure functions over a large kinematic region through the Sullivan process, where one measures the contribution to the electron Deep Inelastic Scattering (DIS) of the meson cloud of a proton target: the nucleon parton distributions contain a component which can be attributed to the meson cloud. 
- Sullivan processes. In these examples, a nucleon’s pion cloud is used to provide access to the pion’s (a) elastic form factor and (b) parton distribution functions. $t = (k − k′)^2$ is a Mandelstam variable and the intermediate pion, $π^∗(P = k − k′)$, $P^2 = t$, is off-shell.
Three informative refs are:
Aitkenhead, et al. "Determination of the pion and kaon structure functions." Physical Review Letters 45
(1980) 157 .
Collins, & Martin, "Hadron reaction mechanisms." Reports on Progress in Physics 45 (1982) 335. Section 6.1.
Aguilar et al, "Pion and kaon structure at the electron-ion collider" European Physical Journal A55 (2019) 1-15.
