How to replace a not existing value by 0? - bash

Question

So I have a subject folder, located in /home/subject, which contains subjects such as :
Geography
Math
These subjects are files.
And each one of them contain, the name of a student with his mark.
So for example it will be,
For Geography
Mattew 15
Elena 14
And Math :
Matthew 10
Elena 19
I also have a student folder, located in /home/student, which is empty for now.
And the purpose of this folder is to put inside of it :
The name of the student as the name of the file;
The marks of all subjets this student received.
So here is my code :
rm /home/student/*
for subjectFile in /home/subject/*; do
awk -v subject="$(basename "$subjectFile")" '{ print subject, $2 >>"/home/student/" $1 }' "$subjectFile"
done
This loop iterates over all the subject files, inside of the subject folder.
The $subjectFile value is something like :
/home/subject/Math
/home/subject/Geograpy
/home/subject/(MySubject)
Etc.. => Depending on what subjects are available.
I then, get the basename of each of these subject files :
Math
Geography
(...)
And then print the result of the second column, the mark number, inside of my student name that I get through the first column of the subject file : so for this example, for the subject Geography, I'll get Matthew.
I also didn't want to simply infinite append the result, but overwrite the previous results each time I run this script, so I typed : rm /home/student/* , to erase any student files before to proceed to the append.
This works great.
But then I have a request,
How can I make it to replace by 0 the subject mark of a student, if this one is undefined?
A student that did not received any mark for a specific subject, while others did?
As for example :
For Geography
Mattew 15
And Math :
Matthew 10
Elena 19
So for Elena,
This shall create an Elena file, inside of the student folder with :
Geography 0
Math 19

Answer 1

Simple distribution using pure bash
Let's try: This will create a little tree in /tmp:
cd /tmp && tar -zxvf <(base64 -d <<eof
H4sIAFUFUFwAA+3XXU6EMBQF4D6ziu7A/l2uLsBHE7eAYwUMghlKnNm9EKfEEHUyBpg4nu+lJJDQ
5HBK226KpqmuxJJUj4mGUTOpz2MktHXG9RdEWiitrWUhadFZHXRtyLZSiidflbsfnjt2/49qP/Jv
u4dnvwnLfAcn5K9JuT5/w0zIfw2T/F+yUMz+jiHg1Lnv87c09j9l0+dvU3JCqtln8oV/nv9dFkLh
36RWyW3l60zqm+S+KCspNSfnnhwsbtJ/X+dV2c68BBzvfzr2nw33/XeWUvR/DWP/WcYFgOICcI0F
4OJN+p/7Jt9mr8V+znec8P8/7P8cK4v+r2HsP8X6u1h/Qv0vX+x/6B59ff7znyFNw/5fGZz/AQAA
AAAAAAAAAAB+7R1PsalnACgAAA==
eof
)
This will create (and print in terminal, because of -v flag in tar command):
school/
school/subject/
school/subject/math
school/subject/english
school/subject/geography
school/student/
Quick overview:
cd /tmp/school/subject/
grep . * | sort -t: -k2
will render:
geography:Elena 14
english:Elena 15
math:Elena 19
math:Matthew 10
geography:Matthew 15
english:Matthew 17
geography:Phil 15
math:Phil 17
english:Phil 18
Building student stat files:
cd /tmp/school
rm student/*
for file in subject/*;do
subj=${file##*/}
while read student note ;do
echo >>student/${student,,} ${subj^} $note
done < $file
done
Nota: This use ${VARNAME^} to upper first char and ${VARNAME,,} to lower all string. So filenames are lowercaps and subject become capitalized in student files.
Then now:
ls -l student
total 12
-rw-r--r-- 1 user user 32 jan 29 08:57 elena
-rw-r--r-- 1 user user 32 jan 29 08:57 matthew
-rw-r--r-- 1 user user 32 jan 29 08:57 phil
and
cat student/phil
English 18
Geography 15
Math 17
Then now: searching for missing notation
for file in student/*;do
for subj in subject/*;do
subj=${subj##*/}
grep -q ^${subj^}\ $file || echo ${subj^} 0 >> $file
done
done
This could be tested (This will randomely drop 0 or 1 mark in all files):
for file in subject/*;do
((val=1+(RANDOM%4)))
((val<4)) && sed ${val}d -i $file
done
Then run:
cd /tmp/school
rm student/*
for file in subject/*;do
subj=${file##*/}
while read student note ;do
echo >>student/${student,,} ${subj^} $note
done < $file
done
for file in student/*;do
for subj in subject/*;do
subj=${subj##*/}
grep -q ^${subj^}\ $file || echo ${subj^} 0 >> $file
done
done
Ok, now:
grep ' 0$' student/*
student/matthew:Geography 0
Nota: As I'v been used $RANDOM, result may differ in your tests;-)
Another aproach: two steps again but
First step: building student list, then student files imediately with 0 notation:
cd /tmp/school
rm student/*
declare -A students
for file in subject/* ;do
while read student mark ;do
[ "$student" ] && students[$student]=
done <$file
done
for file in subject/*;do
class=(${!students[#]})
while read student mark ;do
subj=${file##*/}
echo >> student/${student,,} ${subj^} $mark
class=(${class[#]/$student})
done <$file
for student in ${class[#]};do
echo >> student/${student,,} ${subj^} 0
done
done
Statistic tool
For fun, with a lot of bashisms and without file creation, there is a pretty dump tool:
#!/bin/bash
declare -A students
declare subjects=() sublen=0 stdlen=0
for file in subject/* ;do # read all subject files
subj=${file##*/}
subjects+=($subj) # Add subject to array
sublen=$(( ${#subj} > sublen ? ${#subj} : sublen )) # Max subject string len
declare -A mark_$subj # Create subject's associative array
while read student mark ;do
stdlen=$(( ${#student} > $stdlen ? ${#student} : stdlen ))
[ "$student" ] && { # Skip empty lines
((students[$student]++)) # Count student's marks
printf -v mark_$subj[$student] "%d" $mark # Store student's mark
}
done <$file
done
printf -v formatstr %${#subjects[#]}s; # prepare format string for all subjects
formatstr="%-${stdlen}s %2s ${formatstr// / %${sublen}s}"
printf -v headline "$formatstr" Student Qt "${subjects[#]}"
echo "$headline" # print head line
echo "${headline//[^ ]/-}" # underscore head line
for student in ${!students[#]};do # Now one line by student...
marks=() # Clear marks
for subject in ${subjects[#]};do
eval "marks+=(\${mark_$subject[\$student]:-0})" # Add subject mark or 0
done
printf "$formatstr\n" $student ${students[$student]} ${marks[#]}
done
This may print out something like:
Student Qt english geography math
------- -- ------- --------- ----
Phil 2 18 15 0
Matthew 3 17 15 10
Elena 2 0 14 19
Nota
This script was built for bash v4.4.12 and tested under bash v5.0.
More
You could download bigger demo script: scholl-averages-demo.sh (view in browser as text .txt).
Always pure bash without forks, but with
average by student, average by subject and overall average, in pseudo float
subject and student sorted alphabeticaly
support UTF-8 in student names
.
Student Qt art biology english geography history math Average
------- -- --- ------- ------- --------- ------- ---- -------
Elena 5 12 0 15 14 17 19 12.83
Iñacio 6 12 15 19 18 12 14 15.00
Matthew 5 19 18 17 15 17 0 14.33
Phil 5 15 19 18 0 13 17 13.67
Renée 6 14 19 18 17 18 15 16.83
Theresa 5 17 14 0 12 17 18 13.00
William 6 17 17 15 15 13 14 15.17
------- -- --- ------- ------- --------- ------- ---- -------
Avgs 7 15.14 14.57 14.57 13.00 15.28 13.86 14.40

Answer 2

I would do it something like this. First make empty files for every student:
cat /home/subject/* | cut -d' ' -f1 | sort -u | while read student_name; do > /home/students/$student ; done
Then I would go through each one and add the marks:
for student in `ls /home/students` ; do
for file in /home/subjects/* ; do
subject="`basename $file`"
mark="`egrep "^$student [0-9]+" $file | cut -d' ' -f2`"
if [ -z "$mark" ]; then
echo "$subject 0" >> /home/students/$student
else
echo "$subject $mark" >> /home/students/$student
fi
done
done
something like that anyway