I have the following vectors that getting different entries based on another array.

• PIP_ID
• SAMPLE_TYPE_IDS
• MPX_KIT_IDS

Depending on the content of those I am creating different files using if statements/rules. For example:

if [ $PIP_ID == 8 ]; then   
    EXTENSION="ATLAS_"
fi
if [[ $SAMPLE_TYPE_IDS == 81 ]] || [[ $SAMPLE_TYPE_IDS == 86 ]]; then
    EXTENSION="SmartSeq3_"
fi
if [[ $SAMPLE_TYPE_IDS == 61 ]] || [[ $SAMPLE_TYPE_IDS == 72 ]] || [[ $SAMPLE_TYPE_IDS == 102]] || [[ $SAMPLE_TYPE_IDS == 84 ]] ; then  
    EXTENSION="scRNAv3_"
fi
if ([[ $MPX_KIT_IDS == 89 ]] || [[ $MPX_KIT_IDS == 91 ]] || [[ $MPX_KIT_IDS ==  92 ]] || [[ $MPX_KIT_IDS == 93 ]]) && [[ SAMPLE_TYPE_IDS != 105 ]] ; then
    EXTENSION="AGENTtrim_"
fi
if [[ $MPX_KIT_IDS == 17 ]] || [[ $MPX_KIT_IDS == 27 ]] || [[ $MPX_KIT_IDS == 53 ]]; then   
    EXTENSION="miRNA_"
        fi

Note that the one that creates the AGENTtrim_ has multiple OR combined with && statements grouped with parenthesis.

How can I turn those into a more simpler, more elegant solution using case? I do not know how to combine all the 3 variables into one, since the one that creates the AGENTtrim_ gets info from 2 independent variables.

CodePudding user response：

Converting it to a case statement would be ugly. You can simplify it by using extended pattern matching.

if [[ $MPX_KIT_IDS == @(89|91|92|93) && $SAMPLE_TYPE_IDS != 105 ]]; then
    EXTENSION="AGENTtrim_"
elif [[ $MPX_KIT_IDS == @(17|27|53) ]]; then
    EXTENSION="miRNA_"
elif [[ $SAMPLE_TYPE_IDS == @(81|86) ]]; then
    EXTENSION="SmartSeq3_"
elif [[ $SAMPLE_TYPE_IDS == @(61|72|102|84) ]]; then
    EXTENSION="scRNAv3_"
elif [[ $PIP_ID == 8 ]]; then
    EXTENSION="ATLAS_"
fi

CodePudding user response：

To be honest I do not know how to combine all the 3 variables into one

You can just put them together with a separator. For example, assuming there will be no commas in values, just pick a comma as separator and do a string:

shopt -s extglob  

case "$PIP_ID,$SAMPLE_TYPE_IDS,$MPX_KIT_IDS" in
8,*)                           EXTENSION="ATLAS_" ;;
*,81,*|*,86,*)                 EXTENSION="SmartSeq3_" ;;
*,61,*|*,72,*|*,102,*|*,85,*)  EXTENSION="scRNAv3_" ;;
# with extended glob shopt -s extglob
*,!(105),@(89|91))             EXTENSION="AGENTtrim_" ;;
*,17|*,27|*,53)                EXTENSION="miRNA_" ;;
# etc...
esac

CodePudding user response：

This probably helps to make the code clearer (with bash):

Replace

[[ $SAMPLE_TYPE_IDS == 61 ]] || [[ $SAMPLE_TYPE_IDS == 72 ]] || [[ $SAMPLE_TYPE_IDS == 102]] || [[ $SAMPLE_TYPE_IDS == 84 ]]

with

[[ $SAMPLE_TYPE_IDS =~ ^(61|72|102|84)$ ]]

CodePudding user response：

Make use of functions which name can clarify the rule.

atlas_pip() {
  [[ $1 == 8 ]]
}

smart_id() {
  [[ $1 =~ ^(61|72)$ ]]
}

rna_id() {
  [[ $1 =~ ^(61|72|102|84)$ ]]
}

agent_id() {
  [[ $1 =~ ^(89|91|92|93)$ ]]
}

mirna_id() {
  [[ $1 =~ ^(17|27|53)$ ]]
}

atlas_pip $PIP_ID          && EXTENSION="ATLAS_"
smart_id  $SAMPLE_TYPE_IDS && EXTENSION="SmartSeq3_"
rna_id    $SAMPLE_TYPE_IDS && EXTENSION="scRNAv3_"
agent_id  $MPX_KIT_IDS && $SAMPLE_TYPE_IDS != 105 && EXTENSION="AGENTtrim_"
mirna_id  $MPX_KIT_IDS     && EXTENSION="miRNA_"

CodePudding user response：

IF your nonexclusive if statements cannot overwrite because a later condition is also met but has a higher priority, then this nested case structure should should create the same effective logic.

case $PIP_ID in
8) EXTENSION="ATLAS_"       ;;
*) case $SAMPLE_TYPE_IDS in                                 # PIP_ID is NOT 8
          8[16]) EXTENSION="SmartSeq3_" ;;
   61|72|102|84) EXTENSION="scRNAv3_"   ;;
              *) case $MPX_KIT_IDS in
                  [12]7|53) EXTENSION="miRNA_"     ;;
                 89|9[123]) case $SAMPLE_TYPE_IDS in
                            105) : nothing to do        ;;  # sets NO extension if it gets to this
                              *) EXTENSION="AGENTtrim_" ;;
                            esac                   ;;
                 esac                   ;;
   esac                     ;;
esac