CODEWITHSUNDEEP
regexbashperl
firedev
firedev

Reputation: 83

How can you use perl and regex to append character at end of each new line between match

I am using bash perl and sed to take GraphQL queries and convert them to TypeScript types. I have it all figured out except one piece - I cannot figure out how to insert a comma at the end of each line within a regex match.

This is the raw *.graphqls text that I'm successfully matching:

enum ContentType {
    LESSON
    WARMUP
    QUIZ
    PROJECT
}

The Regex that makes the match (selects everything between and including enum and } ):

s/(?s)(?=enum).*?(})

I can easily replace this entire segment with bash and perl:

ENUM_MATCH="s/(?s)(?=enum).*?(})/replacement/"

perl -i -0pe "${ENUM_MATCH}" "${FULL_PATH}"

But what I need to do is replace this enum text with the same text, except with a comma on each enum line item - end result would look like this:

enum ContentType {
    LESSON,
    WARMUP,
    QUIZ,
    PROJECT
}

Upvotes: 0

Views: 373

Answers (4)

zdim
zdim

Reputation: 66881

If a trailing comma (after PROJECT) is allowed, which is legal as of C99 and C++11

perl -i -pe'if ($r = /^\s*enum/ .. /}\s*$/) { s/$/,/ if $r!=1 && $r!~/E0$/ }' file

This uses range operator, and its return with which to exclude its first and last line.

Upvotes: 1

ikegami
ikegami

Reputation: 385897

With trailing comma:

s{enum[^{]*\{\K([^}]*)}{ $1 =~ s/\n/,\n/gr }seg

Without:

s{enum[^{]*\{\K([^}]*)}{ $1 =~ s/\n/,\n/gr =~ s/,(?=[^,]*\z// }seg

Upvotes: 0

PesaThe
PesaThe

Reputation: 7499

If , is allowed after PROJECT:

awk '$1=="}"{f=0} f{$0=$0 ","} $1=="enum"{f=1} 1' data

Upvotes: 0

Kent
Kent

Reputation: 195089

an awk one-liner works for your example:

 awk '/enum .*\{/{print;next}p{print p (/\}/?"":",")}{p=$0}END{print p}' file

Upvotes: 0

Related Questions