Regroup a Jenkins build list under the previous sections¶
What it does: stack sections reads a Jenkins build list and writes it
again under the #### Title #### headers of a previous list. Each
entry keeps its flags and whether it is commented out. A name no
previous section files goes under --unfiled.
Why you’d use it: a solved build list is one flat file. The site’s previous list already files each package under MPI, Chemistry, and the other headings the operators read. The regrouped list is that order.
Regroup the recorded lists¶
tests/jenkins_sections/previous.list has MPI and Chemistry.
tests/jenkins_sections/next.list is the next generation, with CP2K
commented out and not named by either previous section.
eb-stack stack sections tests/jenkins_sections/next.list \
--from tests/jenkins_sections/previous.list \
--out /tmp/grouped.list
That command printed:
list=/tmp/grouped.list sections=3 of=3
/tmp/grouped.list is:
# next generation
#### MPI ####
# OpenMPI-5.0.10-GCC-15.2.0.eb --hooks=/sw/h.py
#### Chemistry ####
# GROMACS-2025.0-foss-2025b.eb
#### New in this generation ####
# CP2K-2026.1-foss-2025b.eb
of=3 is the previous list’s two sections plus the unfiled header.
The default unfiled title is New in this generation. The comment
block before the first entry stays on top. OpenMPI keeps
--hooks=/sw/h.py. eb_stack_sections is the same command. See
the CLI reference.