REGEX from array, returns no matches.

dr_strangelove · 06-23-2010, 05:37 PM

Hello,

I'm attempting to use REGEX to isolate media (video and audio) essence that relate to alias/reference Quicktime movies. The Quicktime movie filename (excluding ".mov" can always be found in the media-essence filename.

Stages of my process are:

I cache the contents of a text file into an array ($file). This text file/array contains a list of Quicktime movies and media essence. Essence for each Quicktime movie is contained in this text file. An example text file is below:

Code:

MOVIE1.mov
media.dir/MOVIE1.m2v
media.dir/MOVIE1.wav
media.dir/MOVIE1_1.wav
media.dir/MOVIE1_2.wav
media.dir/MOVIE1_3.wav
MOVIE2.mov
media.dir/MOVIE2.m2v
media.dir/MOVIE2.wav
media.dir/MOVIE2_1.wav
media.dir/MOVIE2_2.wav
media.dir/MOVIE2_3.wav
MOVIE3.mov
media.dir/MOVIE3.m2v
media.dir/MOVIE3.wav
media.dir/MOVIE3_1.wav
media.dir/MOVIE3_2.wav
media.dir/MOVIE3_3.wav

I isolate which items in the list are Quicktime movies (simply REGEX matching ".mov" and cache these in another array for later use in the script.
Each item within the Quicktime-movie array ($mov_array) is used as search criteria when searching the remainder of the file list.
I plan on using all items which match the Quicktime mov filename (pre ".mov") to generate ffmpeg tasks.

My problem is that the following statement

Code:

if [[ $essence =~ $id ]]

returns no results.

If I manually intervene and state that

Code:

#id="SOMETHINGOROTHER"

then I get results.

Extracted code is below:

Code:

#!/bin/bash

#This script loops through a text file and creates an ffmpeg string to wrap this essence into a single container.

number_mov=0
array=()
mov_array=${#array[*]}
regex1="mov"
topdir="/Users/Dev/1107613"
subdir="/media.dir/"

cd $topdir

IFS='
'
file=( $( < MOVIES.txt ) )									# Here we grab the contents of the text file within the folder
for (( i = 0 ; i < ${#file[@]} ; i++ ))						# Here we loop through its contents and check for the presence of each line
	do
	a=${file[$i]}
	if [[ $a =~ $regex1 ]]; then							# If we match regex1 (normally, is it a mov?) then ...
    mov_array=(${mov_array[@]} "$a")						# ... add this to the array of movs
	number_mov=$((number_mov+1))
	fi
done

for mov in ${mov_array[@]}									# Loop through our MOV array ...
do
	id=$(echo $mov|sed 's/.mov//g')							# ... remove the file extension, remember, only movs should be in this array!
	echo 'Media ID -' $id
	for essence in ${file[@]}								# Loop through our MOV array ...
		do	
			#id="MOVIE1"
			echo "Checking"
			echo $id
			echo $essence
			if [[ $essence =~ $id ]]
		 		then
		  		echo "Match"
			else
				echo "Unlucky"
			fi
			echo
		done
done

Many thanks in advance and apologies if I'm missing something bleedingly obvious.

dr_strangelove

NB* This script is being developed on Mac OS X 10.5.8, for eventual deployment on Linux version 2.6.27.7-9-pae (geeko@buildhost) (gcc version 4.3.2 [gcc-4_3-branch revision 141291] (SUSE Linux) )

rweaver · 06-23-2010, 05:54 PM

Code:

IFS='
'
file=( $( < tt.txt.fil ) )                                                                      # Here we grab the contents of the text file within the folder
for (( i = 0 ; i < ${#file[@]} ; i++ ))                                         # Here we loop through its contents and check for the presence of each line
        do
        a=${file[$i]}
        if [[ $a =~ $regex1 ]]; then                                                    # If we match regex1 (normally, is it a mov?) then ...
        mov_array=(${mov_array[@]} "$a")                                                # ... add this to the array of movs
        number_mov=$((number_mov+1))
        fi
done

for mov in ${mov_array[@]}                                                                      # Loop through our MOV array ...
do
        id=$(echo $mov|sed 's/.mov//g')                                                 # ... remove the file extension, remember, only movs should be in this array!
        #echo 'Media ID -' $id
        for essence in ${file[@]}                                                               # Loop through our MOV array ...
                do
                        echo "id:$id essence:$essence mov:$mov\n"
                        if [[ $essence =~ $id ]]
                                then
                                echo "essence: $essence\n"
                        else
                                w00t=0
                        fi
                done
done

Which returns...

Code:

id:MOVIE1 essence:MOVIE1.mov mov:MOVIE1.mov
essence: MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE1.m2v mov:MOVIE1.mov
essence: media.dir/MOVIE1.m2v
id:MOVIE1 essence:media.dir/MOVIE1.wav mov:MOVIE1.mov
essence: media.dir/MOVIE1.wav
id:MOVIE1 essence:media.dir/MOVIE1_1.wav mov:MOVIE1.mov
essence: media.dir/MOVIE1_1.wav
id:MOVIE1 essence:media.dir/MOVIE1_2.wav mov:MOVIE1.mov
essence: media.dir/MOVIE1_2.wav
id:MOVIE1 essence:media.dir/MOVIE1_3.wav mov:MOVIE1.mov
essence: media.dir/MOVIE1_3.wav
id:MOVIE1 essence:MOVIE2.mov mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE2.m2v mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE2.wav mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE2_1.wav mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE2_2.wav mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE2_3.wav mov:MOVIE1.mov
id:MOVIE1 essence:MOVIE3.mov mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE3.m2v mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE3.wav mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE3_1.wav mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE3_2.wav mov:MOVIE1.mov
id:MOVIE1 essence:media.dir/MOVIE3_3.wav mov:MOVIE1.mov
id:media.dir/MOVIE1.m2v essence:MOVIE1.mov mov:media.dir/MOVIE1.m2v
id:media.dir/MOVIE1.m2v essence:media.dir/MOVIE1.m2v mov:media.dir/MOVIE1.m2v
essence: media.dir/MOVIE1.m2v
<SNIP>

So that being said, im unsure what exactly your problem is.

dr_strangelove · 06-23-2010, 06:30 PM

Thanks for your super swift response rweaver

The text file example fortuitously, includes media essence and Quicktime movies in sequential order - this is not guaranteed. The ultimate aim is to be able to associate each $id to all the $essence that it relates. Whilst my script is ultimately flawed somehow, I'd like the output to be as below, where a 'Match' is indicated where (emboldened) and non-matched and also highlighted (italicised) ... I will later include a little logic to use these 'switches' to create ffmpeg commands.

Code:

Media ID - MOVIE1
Checking
MOVIE1
MOVIE1.mov
Match

Checking
MOVIE1
media.dir/MOVIE1.m2v
Match

Checking
MOVIE1
media.dir/MOVIE1.wav
Match

Checking
MOVIE1
media.dir/MOVIE1_1.wav
Match

Checking
MOVIE1
media.dir/MOVIE1_2.wav
Match

Checking
MOVIE1
media.dir/MOVIE1_3.wav
Match

Checking
MOVIE1
MOVIE2.mov
Unlucky

Checking
MOVIE1
media.dir/MOVIE2.m2v
Unlucky

... ETC ...

Thanks for your continuing patience.

dr_strangelove