|
1
|
- Get a CLUE with Linux;
Or, how to save 182
manhours
analyzing test results by
using
LINUX regular expression
tools
at the command line
rather than in Microsoft
Excel.
- (spreadsheets are amazing tools!)
|
|
2
|
- Regular Expressions in Linux saved over 182 Manhours during a 19 test
cycle with 10 reports for each cycle.
- Linux saved 13 hours per test cycle of sorting and comparing in Excel by
using regular expression commands in a Linux shell, taking roughly 15
minutes.
- A BASH shell script was written to automate the process;
it runs in less than FIVE SECONDS.
- 13 hours to 15 minutes to less than 5 seconds.
- First an overview of Linux; then the script will be explained.
|
|
3
|
|
|
4
|
|
|
5
|
|
|
6
|
|
|
7
|
- Setting up your SHELL environment will help productivity, visibility and
create custom tools:
see also #1 of the lesser tenets: allow the user to tailor the
environment...
http://linuxmeister.net/Intro-to-Linux/bashrc-n-history-details.html
http://linuxmeister.net/Notes/bashrc-basic.html
|
|
8
|
|
|
9
|
|
|
10
|
|
|
11
|
|
|
12
|
- awk – Aho, Alfred V.; Weinberger, Peter J.; and Kernighan Brian W. (1977)
https://en.wikipedia.org/wiki/AWK#Versions_and_implementations
http://linuxmeister.net/Commands/awk-example.html http://linuxmeister.net/Notes/tabs-convert-to-spaces.html
- sed – stream editor (also used with Perl)
http://linuxmeister.net/Notes/using-sed-to-count-executable-files.html
ls `env | grep PATH | grep -v XNLSPATH | sed -e 's/^PATH=//' |
sed -e 's/\:/ /g'` 2>/dev/null | wc -l
- grep – global regular expression search
http://linuxmeister.net/Notes/using-find-grep-xargs.html http://linuxmeister.net/Commands/use-of-grep.html
find . -type f -name *.html | xargs grep johnmeister.com/linux |
perl -pi -e 's$http://johnmeister.com/linux$http://linuxmeister.net$g'
|
|
13
|
- A computer system under test in the early development stages monitored
an ARINC bus and reported statuses. Some of the notifications were
considered nuisance messages because of intial settings.
Some might be related to a sensor with too narrow a hysteresis, a
parameter incorrectly set in the s/w,
an out of adjustment gadget, or less likely, an actual defective
component.
note: the content of these reports and the specific systems
involved are not included, "dummy data is used“
- 10 reports were generated for each test cycle of the system.
- 19 test runs were made under different conditions to ensure full
compliance.
- a USB device is used to copy the text files from the system.
- The 10 reports were copied into Excel
- each report would be sorted in Excel by ATA chapter
- copying and sorting the reports would take an hour.
- after all 10 reports were sorted, the engineer compared them, looking
for common errors.
- It took another 3 hours to compare those 10 sorted reports
- The engineer had 13 hours invested in this analysis for just ONE test
cycle, there were 19.
- The engineer then examined each of the common errors in the original
reports.
- Once he determined the issue, he'd resolve that one and move to the
next.
Analyzing his process and test results, we determined the plain
text files had a consistent format
with an easily spotted "key field", i.e. the ATA
Maintenance Chapter.
The reports had names unique to the product under test. This info
was also in each report.
To protect the original reports and to make handling the sorts
easier,
We copied the originals with new names from 01.txt to 10.txt,
with a "txt" suffix for interoperability.
|
|
14
|
|
|
15
|
|
|
16
|
|
|
17
|
|
|
18
|
|
|
19
|
|
|
20
|
|
|
21
|
|
Notes
