Add support for multiple files

mohd-akram · Jan 6, 2024 · 9bf1908 · 9bf1908
1 parent 85d7a02
commit 9bf1908
Show file tree

Hide file tree

Showing 4 changed files with 59 additions and 23 deletions.
diff --git a/jawk b/jawk
@@ -5,7 +5,7 @@ set -eu
 name=`basename $0`
 
 usage() {
-	echo "usage: $name [-v var=value] [-f progfile | 'prog'] [file]" >&2
+	echo >&2 "usage: $name [-v var=value] [-f progfile | 'prog'] [file ...]"
 	exit 1
 }
 
@@ -22,7 +22,7 @@ done
 i=0
 skip=
 f=
-file=
+files=
 for arg do
 	i=$((i+1))
 	shift
@@ -34,19 +34,26 @@ for arg do
 		fi
 	else
 		if [ ! "$p" ]; then prog=$arg; p=1
-		elif [ ! "$f" ]; then file=$arg; f=1
-		else usage; fi
+		else
+			if [ ! "$f" ]; then files="$arg"
+			else files=$(printf '%s\t%s' "$files" "$arg"); fi
+			f=$((f+1));
+		fi
 		continue
 	fi
 	set -- "$@" "$arg"
 done
 
 if [ ! "$p" ]; then usage; fi
-if [ ! "$f" ]; then file=-; fi
+if [ ! "$f" ]; then files=-; fi
 
 jawk='
-BEGIN { JSON="\1"; TYPE="\2"; __KEYS="\3"; FS="\n"; __jawk__init() }
-{ __parse_value($0); $0 = _[JSON]; NR = ++__NR; FILENAME = __FILENAME }
+BEGIN {
+	ARGC=__ARGC=split(__ARGV, ARGV, "\t"); RS=FS="\n"
+	JSON="\1"; TYPE="\2"; __KEYS="\3"; __jawk__init()
+}
+/^---/ { FILENAME=substr($0, 4); __FNR=0; next }
+{ ARGC=__ARGC; RS=FS="\n"; __parse_value($0); $0=_[JSON]; NR=++__NR; FNR=++__FNR }
 
 function __jawk__init(i) {
 	__CHAR[0]   =   "\0"; __CHAR[1]   =   "\1"; __CHAR[2]   =   "\2"
@@ -306,9 +313,15 @@ STRING="\"$CHAR*($ESCAPE$CHAR*)*\""
 NUMBER='-?(0|[1-9][0-9]*)([.][0-9]*)?([eE][+-]?[0-9]*)?'
 KEYWORD='null|false|true'
 SPACE='[[:space:]]+'
+JSON="$STRING|$NUMBER|$KEYWORD|[][{}:,]"
 
 : ${AWK=$(command -v gawk || echo awk)}
 : ${EGREP=$(command -v ugrep || echo 'grep -E')}
 
-$EGREP -o "$STRING|$NUMBER|$KEYWORD|[][{}:,]" "$file" |
-$AWK -v __FILENAME="$file" "$@" "$jawk$prog"
+IFS=$(printf '\t')
+for file in $files; do
+	printf -- '---%s\n' "$file"
+	$EGREP -o "$JSON" "$file" 2>/dev/null
+done | $AWK -v __ARGV="$files" "$@" "$jawk$prog
+BEGIN { ARGC=1; RS=FS=\"\n\" }
+"
diff --git a/src/jawk.awk b/src/jawk.awk
@@ -1,5 +1,9 @@
-BEGIN { JSON="\1"; TYPE="\2"; __KEYS="\3"; FS="\n"; __jawk__init() }
-{ __parse_value($0); $0 = _[JSON]; NR = ++__NR; FILENAME = __FILENAME }
+BEGIN {
+	ARGC=__ARGC=split(__ARGV, ARGV, "\t"); RS=FS="\n"
+	JSON="\1"; TYPE="\2"; __KEYS="\3"; __jawk__init()
+}
+/^---/ { FILENAME=substr($0, 4); __FNR=0; next }
+{ ARGC=__ARGC; RS=FS="\n"; __parse_value($0); $0=_[JSON]; NR=++__NR; FNR=++__FNR }
 
 function __jawk__init(i) {
 	__CHAR[0]   =   "\0"; __CHAR[1]   =   "\1"; __CHAR[2]   =   "\2"

diff --git a/src/jawk.sh b/src/jawk.sh
@@ -5,7 +5,7 @@ set -eu
 name=`basename $0`
 
 usage() {
-	echo "usage: $name [-v var=value] [-f progfile | 'prog'] [file]" >&2
+	echo >&2 "usage: $name [-v var=value] [-f progfile | 'prog'] [file ...]"
 	exit 1
 }
 
@@ -22,7 +22,7 @@ done
 i=0
 skip=
 f=
-file=
+files=
 for arg do
 	i=$((i+1))
 	shift
@@ -37,15 +37,18 @@ for arg do
 	# Get prog and file
 	else
 		if [ ! "$p" ]; then prog=$arg; p=1
-		elif [ ! "$f" ]; then file=$arg; f=1
-		else usage; fi
+		else
+			if [ ! "$f" ]; then files="$arg"
+			else files=$(printf '%s\t%s' "$files" "$arg"); fi
+			f=$((f+1));
+		fi
 		continue
 	fi
 	set -- "$@" "$arg"
 done
 
 if [ ! "$p" ]; then usage; fi
-if [ ! "$f" ]; then file=-; fi
+if [ ! "$f" ]; then files=-; fi
 
 jawk=$(cat jawk.awk)
 
@@ -55,9 +58,15 @@ STRING="\"$CHAR*($ESCAPE$CHAR*)*\""
 NUMBER='-?(0|[1-9][0-9]*)([.][0-9]*)?([eE][+-]?[0-9]*)?'
 KEYWORD='null|false|true'
 SPACE='[[:space:]]+'
+JSON="$STRING|$NUMBER|$KEYWORD|[][{}:,]"
 
 : ${AWK=$(command -v gawk || echo awk)}
 : ${EGREP=$(command -v ugrep || echo 'grep -E')}
 
-$EGREP -o "$STRING|$NUMBER|$KEYWORD|[][{}:,]" "$file" |
-$AWK -v __FILENAME="$file" "$@" "$jawk$prog"
+IFS=$(printf '\t')
+for file in $files; do
+	printf -- '---%s\n' "$file"
+	$EGREP -o "$JSON" "$file" 2>/dev/null
+done | $AWK -v __ARGV="$files" "$@" "$jawk$prog
+BEGIN { ARGC=1; RS=FS=\"\n\" }
+"
diff --git a/test/run.sh b/test/run.sh
@@ -167,11 +167,21 @@ test "trailing space in JSON"
 out=$(printf '{"a":1, \n"b":2}' | jawk '{print _["b"]}')
 [ "$out" = '2' ]
 
-test "file"
-echo '{"age":10}' >test/test.json
-out=$(jawk '{print FILENAME;print _["age"]}' test/test.json; rm test/test.json)
-[ "$out" = "$(printf 'test/test.json\n10')" ]
-
 test "NR"
 out=$(printf '{\n"age":10\n}\n{\n"age":12\n}' | jawk '{print NR}')
 [ "$out" = "$(printf '1\n2')" ]
+
+test "files"
+printf '{"age":10}\n{"age":20}' >test/test.json
+printf '{"age":30}\n{"age":40}' >test/test2.json
+out=$(jawk \
+	'BEGIN{print ARGC,ARGV[1],ARGV[2]}{print ARGC,FILENAME,NR,FNR,_["age"]}' \
+		test/test.json test/test2.json
+	rm test/test*.json
+)
+[ "$out" = "$(printf '%s\n' \
+'2 test/test.json test/test2.json' \
+'2 test/test.json 1 1 10' \
+'2 test/test.json 2 2 20' \
+'2 test/test2.json 3 1 30' \
+'2 test/test2.json 4 2 40')" ]