Shell string processing, getting file names and suffix names

"http://www.lichaozheng.info/2012/03/20/shell-get filename and suffix name/code: file=" Thisfile.txt "echo" FileName: ${file%.*} "echo" Extension: ${file##*.} "Output: Filename:thisfile extension: txt: Bash string processing matching $OLD/$NEW} based on pattern ${str/substring Replace the first one. ${str//$OLD/$NEW} replaces all. Note: You cannot use regular expressions, you can only use the shell extensions of? *. can only use shell wildcard characters such as *?  [list] [!list] [A-z]. ${str/# $OLD/$NEW} replaces the beginning. If Str starts with an old string, it is replaced. ${str/% $OLD/$NEW} replaces the end. If Str ends with an old string, it is replaced.   [user@laptop ~]# str= ' Hello World '   [User@laptop ~]# Echo ${str/o/o}  Hello World [User@laptop ~]# echo $ {str//o/o}  Hello World [user@laptop ~]# str= ' Hello World '   [User@laptop ~]# echo ${str/#He/he}  Hello worl d [User@laptop ~]# echo ${str/#o/he}  Hello World [user@laptop ~]# Echo ${str/%he/he}  Hello World [user@laptop ~]# Echo ${str/%ld/ld}  Hello World   If the string contains/characters are replaced, then escape, write \/.   [User@laptop ~]# filename= "/user/admin/monitoring/process.sh"   [User@laptop ~]# Echo ${filename/#\/user/\ /tmp} /TMP/ADMIn/monitoring/process.sh [User@laptop ~]# echo ${filename/%.*/.ksh} /user/admin/monitoring/process.ksh [ User@laptop ~]#   Separate directories for the environment variable path, one for each row. ECHO-E ${path/:/\n}   [User@laptop ctmw]# echo $PATH  /usr/kerberos/sbin:/usr/kerberos/bin:/usr/apache/ Apache-ant-1.7.1/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/user/bin [User@laptop ctmw]# Echo-e ${path//:/' \ n '} /usr/kerberos/sbin/usr/kerberos/bin/usr/apache/apache-ant-1.7.1/bin/usr/local/sbin Usr/local/bin/sbin/bin/usr/sbin/usr/bin/user/bin [User@laptop ctmw]# echo ' ${path//:/$ ' \ n '} '  /usr/kerberos/ Sbin/usr/kerberos/bin/usr/apache/apache-ant-1.7.1/bin/usr/local/sbin/usr/local/bin/sbin/bin/usr/sbin/usr/bin/ User/bin   SUBSTRING deletion based on pattern matching deletion is a special kind of substitution ${str/$SUB} deletes the first sub substring in str ${str//$SUB} Deletes all sub substrings in Str ${str#$ PREFIX} Go to the head, remove the shortest matching prefix from the beginning ${str## $PREFIX} go to the head, remove the longest matching prefix from the beginning ${str% $SUFFIX} go tail, remove the shortest matching suffix from the end ${str%% $SUFFIX} go tail, remove the longest matching suffix from the end Note: Often remember the meaning of # and%, there is a way to help memory look at the keyboard, #Before $,% after $, just know # go head,% go tail. Note: You cannot use regular expressions, you can only use the shell extensions of? *.   [user@laptop ~]# str= ' Hello world '   [User@laptop ~]# Echo ${str#he}  llo World [User@laptop ~]# Echo ${st r#he*o}  World [User@laptop ~]# echo ${str# #He *o}  rld [user@laptop ~]# prefix= "*o"   [User@laptop ~]# Echo ${str# $PREFIX}  World [User@laptop ~]# echo ${str## $PREFIX}  rld [user@laptop ~]# echo ${str%o*}  Hello W [ User@laptop ~]# echo ${str%%o*}  Hell [user@laptop ~]# suffix= "o*"   [User@laptop ~]# echo ${str% $SUFFIX}  H Ello W [User@laptop ~]# echo ${str%% $SUFFIX}  Hell   Typical application: Get file extension [user@laptop ~]# file=hello.jpg  [user@l Aptop ~]# echo ${file##*.}   jpg   using the SED command to implement regular expression substitution use the SED command to make regular expression substitutions. echo "$STR" | Sed "s/$OLD/$NEW/" replaces the old substring in str with NEW.   [User@laptop ~]# str= 123456789″  [User@laptop ~]# echo "$STR" | Sed s/345/ok/  12ok6789 [user@laptop ~]# old=345  [user@laptop ~]# new=ok  [User@laptop ~]# echo "$STR" | Sed "s/$OLD/$NEW/"  12ok6789   using the TR command to implement the substitution of character sets using the TR command, you can replace a character, and it can be a replacement from a batch of characters to another. For example, lowercase letters become uppercase letters, or vice versa.   [User@laptop ~]# echo "Bash" | TR "[A-z]" [b-z]   CBTI The above command is to replace a in the original string with a B, replaced by C, and so on.   Online question: Is there a command in Linux that can replace the + or-or-or +   that is "+"-->-" "-"-->" +   For example gmt+8-9 into gmt-8+9   [User@laptop ~]# echo "gmt+8-9″| Sed ' s/-/#/g ' | Sed ' s/+/-/g ' | Sed ' s/#/+/g '   gmt-8+9 above are the answers provided online. If you do it in TR, it's simpler.   [User@laptop ~]# echo "gmt+8-9″| TR "+-" "-+"   gmt-8+9   path string processing dirname ${fullpath} Fetch directory section. BaseName ${fullpath} takes the filename part. BaseName ${fullpath} ${ext} takes the filename part and removes the specified extension. [User@laptop ~]# fullpath=/user/work/project/backup.tar.gz  [user@laptop ~]# dirname "$FULLPATH"  /user/ Work/project [user@laptop ~]# basename "$FULLPATH"     backup.tar.gz [user@laptop ~]# basename "$FULLPATH". GZ&N Bsp Backup.tar [user@laptop ~]# basename "$FULLPATH" .tar  backup.tar.gz [user@laptop ~]# "$FULLPATH". basename  Backup directory section: ${fullpath%/*}     (similar to DirName "$FULLPATH") take file name: file=${fullpath##*/} (similar to basename "$FULLP ATH ") Take the shortest basic name: ${file%%.*} take the longest base name: ${file%.*} take the shortest extension: ${file##*.}   or  ${fullpath##*.} take the longest extension: ${file#*.}   or  ${fullpath#*.}   [User@laptop ~]# fullpath=/user/work/project/backup.tar.gz  [user@laptop ~] # echo ${fullpath%/*} /user/work/project [user@laptop ~]# dirname ' $FULLPATH '  /user/work/project [ User@laptop ~]# file=${fullpath##*/}  [User@laptop ~]# echo $FILE   backup.tar.gz [user@laptop ~]# basename] $FU Llpath "  backup.tar.gz [user@laptop ~]# echo ${file%%.*}  backup [User@laptop ~]# echo ${file%.*}  backup. tar [User@laptop ~]# echo ${file##*.}   GZ [User@laptop ~]# echo ${file#*.}   tar.gz [User@laptop ~]#