BASH Regex - get filename tags, labels and descriptions

Shell Programming and Scripting
1

Hi, I am trying to switch from windows to linux.
I have been using Autohotkey scripts for some little things.
I started writing some bash scripts for NEMO browser in linux mint and I am trying to convert some of AHK scripts but as I am not a programmer. I have a hard time with regex stuff.

What I did is split file path and I got it working ok. -- Directory, FileName, Filename-Without-Extension, Extension

Now every filename may or may not contain [tags], {labels} and (description) in brackets at the end of the filename.
Where I am stuck is regex that works in Autohotkey is not in BASH. I tried sed, awk, but I have no idea what I am doing.

example filename:
multiple.{1}(1)[1] dots.and spaces {labels}(description)[tag1 tag2].txt
filename with only tags [tag1 tag2 tag3].txt

Note:
Tags, labels or description can be any words or numbers, not just what I have in example.
Not every filename will have all 3. Some could have only tags, others only descriptions or labels.

need to split and get this:
shortName = "multiple.{1}(1)[1] dots.and spaces" -- filename without any tags,labels and descriptions
tags = "tag1 tag2" -- tags without brackets
labels = "label" -- labels without brackets
description = "description" -- description without brackets

Original regex from Autohotkey that works in all of my scripts

; NameNoExt is filename without extension
; get filename only without tags[], labels{} or description() 
ShortName:=RegExReplace(NameNoExt, "\s*(\([^()]+\)|{[}]+|\[[^][]+\])+$")
; Get tags
Tags:=RegExReplace(NameNoExt, ".*\[(.*)?\](?!\s).*$", "$1")
; Get labels
Label:= RegExReplace(NameNoExt, ".*\{(.*)?\}(?!\s).*$", "$1")
; Get description
Description := RegExReplace(NameNoExt, ".*\((.*)?\)(?!\s).*$", "$1")

Here is what I have in BASH .. and it is going nowere..
So to mention again, this is where I need help with regex splitting these tags, labels and descriptions.

str="multiple.{1}(1)[1] dots.and spaces {lab}(desc)[tag1 tag2].txt"

IFS='
'

NameNoExt=`echo "${str%.*}"`        # for filename with multiple dots

shortName="???"        # get "multiple.{1}(1)[1] dots"

#regexp=".*\[(.*)?\](?!\s).*$"    # from autohotkey
regexp="\[([^)]+)\]+$"
if [[ $NameNoExt =~ $regexp ]]
then 
  TAG="${BASH_REMATCH[1]}"
else 
  TAG="Nothing"
fi
#regexp=".*\{(.*)?\}(?!\s).*$"    # from autohotkey
regexp="\{([^}]+)\}+$"
if [[ $NameNoExt =~ $regexp ]]
then 
  LAB="${BASH_REMATCH[1]}"
else 
  LAB="Nothing"
fi
#regexp=".*\((.*)?\)(?!\s).*$"    # from autohotkey
regexp="\(([^)]+)\)+$"
if [[ $NameNoExt =~ $regexp ]]
then 
  DESC="${BASH_REMATCH[1]}"
else 
  DESC="Nothing"
fi
zenity --info --text "`printf "Input = \"$str\"\n\nNameNoExt = \"$NameNoExt\"\nShort Name  = \"$shortName\"\nTags  = \"$TAG\"\nLabels  = \"$LAB\"\nDescription  = \"$DESC\""`"

I highly appreciate any help with this.
Thank you all.

2

Hi, try

regexp='.*\[([^)]+)\]'

,

regexp='.*\{([^}]+)\}'

and

regexp='.*\(([^)]+)\)'

Respectively..

3

Hi, thank you.
That works at the given example. But if I am to remove one of the end brackets (like
[*] ), this regex will find other bracket if exist within a filename ( like that [1] in the middle of the name).

But in most cases 99%, filenames will not have any extra brackets within the filename, mostly at the end. Just want to have it so there are no mistakes in case one pops up..

Is there any way to check if brackets are at the tail separated by space from filename (without extension)? And if they are then get contents, if not skippit.
As an example:
some [1] name [tag]
some [1] name <-- does not have brackets at the tail

Anyway, thank you very much for your reply. It helped a lot. I might end up just using that for now until I find a better solution.
Cheers

4

Ok, I think I got this thing working now.
I did not know how to extract content A,B,C from "only" end brackets so I had to create new VAR from the tail brackets (if existed) and then extract contents of A,B and C from that. Otherwise regex would be getting 1,2,3 from brackets in the filename.

Anyway... it is ok as it is now but if there is a better solution I would love to learn it.
note: this is for string testing only, once it is in a real script I would have checking to see if file or directory before getting shortName and extension.

#!/bin/bash
test_str=(
    'test1.{1}(2)[3] dots.spaces {A1}(B1)[C1].ext'
    'test2.[3]{1}(2) dots.spaces [C1](B1){A1}.ext'
    'test3 [3].com [C1].ext'
    'test4 {1}(2)[3]dots.ext'
    'test4 {1} (2) [3] dots.ext'
    'test6 (B1)[C1].ext'
    'test7[C1](B1).ext'
)
IFS='
'
regexp='\s*(\([^)]+\)|\{[^}]+\}|\[[^][]+\])+$'
subst=""
regexTag='.*\[([^]]+)\]'
regexLab='.*\{([^}]+)\}'
regexDes='.*\(([^)]+)\)'


for p in "${test_str[@]}"
do
    NameNoExt=`echo "${p%.*}"`
    if [[ $NameNoExt =~ $regexp ]]; then
        tailBrackets="${BASH_REMATCH[0]}";
        shortName=$(echo "$NameNoExt" | sed -r 's/'$regexp'/'$subst'/g');
    else
        shortName="$NameNoExt"
        tailBrackets=""
    fi
    # if tail bracket exist then extract A,B,C
    if [ $tailBrackets ]; then
        if [[ $tailBrackets =~ $regexTag ]]
        then 
          TAG="${BASH_REMATCH[1]}"
        else 
          TAG=""
        fi
        if [[ $tailBrackets =~ $regexLab ]]
        then 
          LAB="${BASH_REMATCH[1]}"
        else 
          LAB=""
        fi
        if [[ $tailBrackets =~ $regexDes ]]
        then 
          DES="${BASH_REMATCH[1]}"
        else 
          DES=""
        fi
    else
        TAG=""
        LAB=""
        DES=""
    fi
    zenity --info --text "`printf "Filename = \"$p\"\n\nNameNoExt = \"$NameNoExt\"\nShort Name  = \"$shortName\"\nTail Brackets = \"$tailBrackets\"\nA  = \"$LAB\"\nB  = \"$DES\"\nC  = \"$TAG\""`"
done