https://github.com/emilio-berti/bash_scripts
https://github.com/emilio-berti/bash_scripts
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/emilio-berti/bash_scripts
- Owner: emilio-berti
- Created: 2020-02-14T13:10:09.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2022-11-14T15:27:44.000Z (over 2 years ago)
- Last Synced: 2025-01-21T04:28:09.322Z (4 months ago)
- Language: Shell
- Size: 39.1 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Less known bash features
To split output and error streams:
`command 1> message.mssg 2> error.err`.
The tee pipe permits to save the output to a file *while* passing it to the next command:
`command | tee message.log | wc -l`.
If you want to exit the shell when an error occurres *set* will do the trick: `set -e`.
If you want to run a child process that doesn't terminate when you exit the shell, use *nohup*: `nohup command`.
To remove with prompt files containing a string:
```bash
rm -i $(grep -l 'This file is obsolete' * )
```
## select
This is a built-in function that creates menus and perform operations on choices.
```bash
#!/bin/bash
if (( $# == 0 ))
then
ext=".txt$"
elif (( $# > 1 ))
then
printf "%b" "Error. Only one (optional) argument can be passed.\n"
exit 1
else
ext=$1
fi
cols=$(tput cols)
files="Exit $(ls | grep $ext)"
PS3='File to print? ' #select prompt
until [ "$file" == "Exit" ]
do
printf "%b" "\a\nSelect a file to print:\n" >&2
select file in $files
do
if [ "$file" == "Exit" ]
then
echo "Exiting..."
break
elif [ -n "$file" ]
then
cols=$(tput cols)
printf "%b" "Printing $file...\n\n"
for (( i=1 ; i < $cols; i++ )); do printf "%b" "=" ; done
printf "%b" "\b\n"
head $file
for (( i=1 ; i < $cols; i++ )); do printf "%b" "=" ; done
printf "%b" "\b\n"
break #necessary to go back to menu
else
echo "Invalid selection!"
fi
done
done
```
The above will prompt a menu asking for a file to be _head_-ed.
## awk
awk is a complete programming language for string manipulation.
awk has a default variable *NF* that stores the number of fields (integer); *$NF* is instead the last field (string).
To print the *n* word in a line:
```bash
awk '{print $1}' < file.txt
```
To reverse the words in a line:
```bash
awk '{for (i=NF; i>=0; i--) {printf "%s ", $i;} printf "\n"}' filename
```
awk can take a command that is ran at the beginning or at the end using the keywords *BEGIN* and *END*. For example, to count the sum of all sizes of files:
```bash
ls -l | awk '{sum += $5}; END {print sum}'
```
To skip a line of text depending on how it starts:
```bash
ls -l | awk '/^total/{next} {sum += 5}; END {print sum}'
```
This skip (*next*) every line that starts with 't-o-t-a-l', but execute the sum for every one else.
Print a paragraph with a keyphrase (paragraphs are separated by blank lines):
```bash
/keyphrase/ { flag=1 }
flag == 1 { print }
/^$/ { flag = 0 }
```
and `awk -f para.awk -v keyphrase < myfile`
## logic and arithmetic
Use *$(( ))* or *let* for arithmetic expressions:
`select` provides a way to select an element from a list:
```bash
select x in $(ls)
do
echo "You selected '$x'"
done
```
```bash
COUNT=$((COUNT + 5**POW + MAX * 2))
let COUNT+='5+MAX*2'
```
# If conditions
General if elif else:
```bash
if ...
then
...
elif ...
then
...
else
...
fi
```
Condition statements:
```bash
if [ -d $DIR ] # $DIR is a directory
if [ -e $FILE ] # $FILE exists
if [ -z $VAR ] # $VAR is empty
if [ -n $VAR ] # $VAR has length > 0
if [ $str1 = $str2 ] # string $str1 is equal to string $str2
if [ $n1 -eq $n2 ] # number $n1 is equal to number $n2
if [ $n1 -ne $n2 ] # number $n1 is **not** equal to number $n2
if [ file1 -ot -file2 ] # file1 is older than file2
if [ file1 -nt -file2 ] # file1 is newer than file2
if [ -d $DIR -a -e $FILE ] # $DIR is a directory **and** (-a) $FILE exists
if [ -d $DIR -o -d $FILE ] # $DIR is a directory **or** (-o) $FILE is a directory
if [[ "${MYFILENAME}" == *.jpg ]] # pattern matching, requires two [[ ]]
```
# While loops
General while loop:
```bash
while (( COUNT < MAX ))
do
...
let COUNT++
done
```
# For loops
General for loop:
```bash
for (( i=0; i < 10; i++ ))
do
...
done
```
For floating values:
```bash
for float in $(seq 1.0 .01 1.1) ...
```
# Dates
ISO 8601 is a good standard to display dates. Set it up with, or alias it with:
```bash
ISO_8601='%Y-%m-%d %H:%M:%S %Z'
date "+$ISO_8601" #don't put the + in the variable itself
alias isodate="date '+%Y-%m-%d %H:%M:%S %Z'"
```
```
$ isodate
2022-10-22 08:43:35 CEST
```
Date to Epoch seconds: `date "+%s".
_printf -v var comm_ assign to variable _var_ the output of _comm_. This can be useful to make logs with date:
```bash
$ printf -v today '%(%F_%T)T'
$ printf $today
2022-10-22_09:03:39
```
The '%(%F_%T)T' stands for today's date (%F) and time (%T).
# gdrive (use at your own risk)
https://github.com/prasmussen/gdrive