Matching blocks with conditions - regex

Question

I am in the need for some regexp guru help.
I am trying to make a small config system for a home project, but for this it seams that I need a bit more regexp code than my regexp skills can come up with.
I need to be able to extract some info inside blocks based on conditions and actions. For an example.
action1 [condition1 condition2 !condition3] {
Line 1
Line 2
Line 3
}
The conditions are stored in simple variables separated by space. I use these variables to create the regexp used to extract the block info from the file. Most if this is working fine, except that I have no idea how to make the "not matching" part, which basically means that a "word" is not available in the condition variable.
VAR1="condition1 condition2"
VAR2="condition1 condition2 condition3"
When matched against the above, it should match VAR1 but not VAR2.
This is what I have so far
PARAMS="con1 con2 con3"
INPUT_PARAMS="[^!]\\?\\<$(echo $PARAMS | sed 's/ /\\>\\|[^!]\\?\\</g')\\>"
sed -n "/^$ACTION[ \t]*\(\[\($INPUT_PARAMS\)*\]\)\?[ \t]*{/,/}$/p" default.cfg | sed '/^[^{]\+{/d' | sed '/}/d'
Not sure how pretty this is, but it does work, except for not-matching.
EDIT:
Okay I will try to elaborate a bit.
Let's say that I have the below text/config file
action1 [con1 con2 con3] {
Line A
Line B
}
action2 [con1 con2 !con3] {
Line C
}
action3 [con1 con2] {
Line D
}
action4 {
Line E
}
and I have the fallowing conditions to match against
ARG1="con1 con2 con3"
ARG2="con1 con2"
ARG3="con1"
ARG4="con1 con4"
# Matching against ARG1 should print Line A, B, D and E
# Matching against ARG2 should print Line C, D and E
# Matching against ARG3 should print Line E
# Matching against ARG4 should print Line E
Below is a java like example of action2 using normal conditional check. It give a better idea of what I am trying
if (ARG2.contains("con1") && ARG2.contains("con2") && !ARG2.contains("con3")) {
// Print all lines in this block
}

Answer 1

The logic of how you're selecting which records to print lines from isn't clear to me so here's how to create sets of positive and negative conditions using awk:
$ cat tst.awk
BEGIN{
RS = ""; FS = "\n"
# create the set of the positive conditions in the "conds" variable.
n = split(conds,tmp," ")
for (i=1; i<=n; i++)
wanted[tmp[i]]
}
{
# create sets of the positive and negative conditions
# present in the first line of the current record.
delete negPresent # use split("",negPresent) in non-gawk
delete posPresent
n = split($1,tmp,/[][ {]+/)
for (i=2; i<n; i++) {
cond = tmp[i]
sub(/^!/,"",cond) ? negPresent[cond] : posPresent[cond]
}
allPosInWanted = 1
for (cond in posPresent)
if ( !(cond in wanted) )
allPosInWanted = 0
someNegInWanted = 0
for (cond in negPresent)
if (cond in wanted)
someNegInWanted = 1
if (allPosInWanted && !someNegInWanted)
for (i=2;i<NF;i++)
print $i
}
.
$ awk -v conds='con1 con2 con3' -f tst.awk file
Line A
Line B
Line D
Line E
$
$ awk -v conds='con1 con2' -f tst.awk file
Line C
Line D
Line E
$
$ awk -v conds='con1' -f tst.awk file
Line E
$
$ awk -v conds='con1 con4' -f tst.awk file
Line E
$
and now you just have to code whatever logic you like in that final block where the printing is being done to compare the conditions in each of the sets.