Repository navigation
vdb-cache-less-experiment
Directory actions
More options
Directory actions
More options
vdb-cache-less-experiment
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
parent directory.. | ||||
-------------------------------------------------------------------
do not include in the build-system
-------------------------------------------------------------------
the files here are the result of an investigation into the possibility
to produce SAM-output from a cSRA without a vdb-cache-file in approx.
the same time as sam-dump with a vdb-cache-file.
The answer is yes, it is possible with the help of some lookup-files.
files:
--------------------------------------------------------------------
full.sh ... execute all 4 steps ( 2 of them in parallel )
this is to be executed to perform the equivalent
of sam-dump ( but without headers, and only aligned data )
step[1|2|3|4].sh ... bash-script for the each of the 4 steps
step[1|2|3|4].cpp ... source-code for the binary in each step
tools.hpp ... common code used in the .cpp files
mmap.hpp ... --- " ---
rr_lookup.hpp ... --- " ---
extract_row_count.cpp ... does what it's name says, used by step1.sh
explain_flags.cpp ... helper binary to explain the bits of a decimal SAM-flag value
sam-cmp.cpp ... helper to compare 1 column between 2 SAM-files
sam-format-computation.* ... graphical explanation of the algorithm to
produce SAM with lookup-tables
acc.sh ... helper script to download accession if not present
prefetch.sh ... helper script to locate the prefetch-binary
vdb-dump.sh ... helper script to locate the vdb-dump-binary
cleanup.sh ... deletes all files, that should not be in git
--------------------------------------------------------------------
how to execute:
1) run "./acc.sh" ... to download the accession to be used
( to not include the download-time into the measurement )
2) run "time ./full.sh" ... performs the conversion into SAM
( output-file: SRR_TST.SAM )
The renaming into SRRXXXXXXa, and the deletion of the vdb-cache-file
is neccessary to avoid triggering the schema-functions into using a vdb-cache-file.