Merge pull request #2095 from bimbashrestha/grep

[bugs] zstdgrep/grep inconsistencies

programs/zstdgrep

@@ -28,8 +28,6 @@ zcat=${ZCAT:-zstdcat}
 endofopts=0
 pattern_found=0
 grep_args=""
-hyphen=0
-silent=0
 
 prog=${0##*/}
 
@@ -43,92 +41,158 @@ esac
 # skip all options and pass them on to grep taking care of options
 # with arguments, and if -e was supplied
 
+escape='
+  s/'\''/'\''\\'\'''\''/g
+  $s/$/'\''/
+'
+
+# We might want to create a c program in the future
+# and replace this file with that if these
+# unsupported options become necessary
+usage="Usage: $0 [OPTION]... [-e] PATTERN [FILE]...
+OPTIONs are the same as grep with the exception of
+the following unsupported options:
+    --dereference-recursive (-R)
+    --directories (-d)
+    --exclude
+    --exclude-from
+    --exclude-dir
+    --include
+    --null (-Z),
+    --null-data (-z)
+    --recursive (-r)
+grep --help below:
+"
+
+operands=
+files_with_matches=0
+files_without_matches=0
+no_filename=0
+with_filename=0
+
 while [ "$#" -gt 0 ] && [ "${endofopts}" -eq 0 ]; do
-    case "$1" in
+    option=$1
-    # from GNU grep-2.5.1 -- keep in sync!
+    shift
-        -[ABCDXdefm])
+    optarg=
-            if [ "$#" -lt 2 ]; then
+
-                printf '%s: missing argument for %s flag\n' "${prog}" "$1" >&2
+    case $option in
-                exit 1
+        (-[0123456789EFGHIKLPRTUVZabchilnoqrsuvwxyz]?*)
-            fi
+            arg2=-\'$(expr "X$option" : 'X-.[0-9]*\(.*\)' | sed "$escape")
-            case "$1" in
+            eval "set -- $arg2 "'${1+"$@"}'
-                -e)
+            option=$(expr "X$option" : 'X\(-.[0-9]*\)');;
-                    pattern="$2"
+        (--binary-*=* | --[lm]a*=* | --reg*=*) ;;
-                    pattern_found=1
+        (-[ABCDXdefm] | binary-* | --file | --[lm]a* | --reg*)
-                    shift 2
+            case ${1?"$option option requires an argument"} in
-                    break
+                (*\'*) optarg=" '"$(printf '%s\n' "$1" | sed "$escape");;
-                    ;;
+                (*) optarg=" '$1'";;
-                -f)
-                    pattern_found=2
-                    ;;
-                *)
-                    ;;
             esac
-            grep_args="${grep_args} $1 $2"
+            shift;;
-            shift 2
+        (-f?*\'*) optarg=" '"$(expr "X$option" : 'X-f\(.*\)' | sed "$escape"); option=-f;;
-            ;;
+        (-f?*) optarg=" '"$(expr "X$option" : 'X-f\(.*\)')\'; option=-f;;
-        --)
+        (--file=*\'*) optarg=" '"$(expr "X$option" : 'X--file=\(.*\)' | sed "$escape"); option=--file;;
-            shift
+        (--file=*) optarg=" '"$(expr "X$option" : 'X--file=\(.*\)')\'; option=--file;;
-            endofopts=1
+        (--) endofopts=1; break;;
-            ;;
+        (-?*) ;;
-        -)
+        (*)
-            hyphen=1
+            case $option in
-            shift
+                (*\'*) operands="$operands '"$(printf '%s\n' "$option" | sed "$escape");;
-            ;;
+                (*) operands="$operands '$option'";;
-        -h)
-            silent=1
-            shift
-            ;;
-        -*)
-            grep_args="${grep_args} $1"
-            shift
-            ;;
-        *)
-            # pattern to grep for
-            endofopts=1
-            ;;
             esac
+            ${POSIXLY_CORRECT+break}
+            endofopts=1
+            continue;;
+        esac
+
+        case $option in
+            (-[drRzZ] | --di* | --exc* | --inc* | --rec* | --nu*)
+                printf >&2 '%s: %s: option not supported\n' "$0" "$option"
+                exit 2;;
+            (-e* | --reg*) pattern_found=1;;
+            (-f | --file)
+                case $optarg in
+                    (" '-'" | " '/dev/stdin'" | " '/dev/fd/0'")
+                        option=-e
+                        optarg=" '"$(sed "$escape") || exit 2;;
+                esac
+                pattern_found=1;;
+            (--h | --he | --hel | --help) echo "$usage"; eval "$grep --help" || exit 2; exit;;
+            (-H | --wi | --wit | --with | --with- | --with-f | --with-fi \
+            | --with-fil | --with-file | --with-filen | --with-filena | --with-filenam \
+            | --with-filename)
+                with_filename=1
+                continue;;
+            (-l | --files-with-*) files_with_matches=1;;
+            (-L | --files-witho*) files_without_matches=1;;
+            (-h | --no-f*) no_filename=1;;
+        esac
+
+        case $option in
+            (*\'?*) option=\'$(printf '%s\n' "$option" | sed "$escape");;
+            (*) option="'$option'";;
+        esac
+
+        grep_args="$option$optarg"
+        grep="$grep $grep_args"
     done
 
-# if no -e option was found, take next argument as grep-pattern
+eval "set -- $operands "'${1+"$@"}'
-if [ "${pattern_found}" -lt 1 ]; then
+
-    if [ "$#" -ge 1 ]; then
+if test $pattern_found -eq 0; then
-        pattern="$1"
+    case ${1?"missing pattern; try \`$0 --help' for help"} in
+        (*\'*) grep="$grep -- '"$(printf '%s\n' "$1" | sed "$escape");;
+        (*) grep="$grep -- '$1'";;
+    esac
     shift
-    elif [ "${hyphen}" -gt 0 ]; then
-        pattern="-"
-    else
-        printf '%s: missing pattern\n' "${prog}" >&2
-        exit 1
-    fi
 fi
 
-EXIT_CODE=0
+if test $# -eq 0; then
-# call grep ...
+    set -- -
-if [ "$#" -lt 1 ]; then
-    # ... on stdin
-    set -f # Disable file name generation (globbing).
-    # shellcheck disable=SC2086
-    "${zcat}" - | "${grep}" ${grep_args} -- "${pattern}" -
-    EXIT_CODE=$?
-    set +f
-else
-    # ... on all files given on the command line
-    if [ "${silent}" -lt 1 ] && [ "$#" -gt 1 ]; then
-        grep_args="-H ${grep_args}"
 fi
-    set -f
+
-    while [ "$#" -gt 0 ]; do
+exec 3>&1
-        # shellcheck disable=SC2086
+res=0
-        if [ $pattern_found -eq 2 ]; then
+
-            "${zcat}" -- "$1" | "${grep}" --label="${1}" ${grep_args} -- -
+for i do
-        else
+    zcat_status=$(
-            "${zcat}" -- "$1" | "${grep}" --label="${1}" ${grep_args} -- "${pattern}" -
+        exec 5>&1
+        ($zcat -- "$i" 5>&-; echo $? >&5) 3>&- |
+        if test $files_with_matches -eq 1; then
+            eval "$grep" >/dev/null && { printf '%s\n' "$i" || exit 2; }
+        elif test $files_without_matches -eq 1; then
+            eval "$grep" >/dev/null || {
+                r=$?
+                if test $r -eq 1; then
+                printf '%s\n' "$i" || r=2
                 fi
-        [ "$?" -ne 0 ] && EXIT_CODE=1
+                exit $r
-        shift
+            }
+        elif test $with_filename -eq 0 && { test $# -eq 1 || test $no_filename -eq 1; }; then
+            eval "$grep"
+        else
+            case $i in
+            (*'
+        '* | *'&'* | *'\'* | *'|'*)
+                i=$(printf '%s\n' "$i" |
+                    sed '
+                    $!N
+                    $s/[&\|]/\\&/g
+                    $s/\n/\\n/g
+                    ');;
+            esac
+            sed_script="s|^|$i:|"
+
+            # Fail if grep or sed fails.
+            r=$(
+                exec 4>&1
+                (eval "$grep" 4>&-; echo $? >&4) 3>&- | sed "$sed_script" >&3 4>&-
+            ) && exit $r
+            r=$?
+            test 1 -lt $r && exit $r || exit 2
+        fi >&3 5>&-
+    )
+    r=$?
+    test 128 -lt $r && exit $r
+    test "$zcat_status" -eq 0 || test "$zcat_status" -eq 2 || r=2
+    test $res -lt $r && res=$r
 done
-    set +f
+exit $res
-fi
-
-exit "${EXIT_CODE}"

tests/playTests.sh

@@ -254,6 +254,38 @@ ZCAT=./zstdcat $ZSTDGREP 2>&1 "1234" tmp_grep_bad.zst && die "Should have failed
 ZCAT=./zstdcat $ZSTDGREP 2>&1 "1234" tmp_grep_bad.zst | grep "No such file or directory" || true
 rm -f tmp_grep*
 
+println "\n===> zstdgrep pipe in with -f "
+echo "start" > tmp_grep
+echo "stop" >> tmp_grep
+echo "start" | ZCAT=./zstdcat $ZSTDGREP -f - tmp_grep > tmp_grep_out1
+echo "start" | grep -f - tmp_grep > tmp_grep_out2
+$DIFF tmp_grep_out1 tmp_grep_out2
+rm -f tmp_grep*
+
+println "\n===> zstdgrep --regexp= multiple"
+echo "start" > tmp_grep
+echo "stop" >> tmp_grep
+ZCAT=./zstdcat $ZSTDGREP --regexp=start --regexp=stop tmp_grep > tmp_grep_out1
+grep -e start -e stop tmp_grep > tmp_grep_out2
+$DIFF tmp_grep_out1 tmp_grep_out2
+rm -f tmp_grep*
+
+println "\n===> zstdgrep multiple -e"
+echo "start" > tmp_grep
+echo "stop" >> tmp_grep
+ZCAT=./zstdcat $ZSTDGREP -e start -e stop tmp_grep > tmp_grep_out1
+grep -e start -e stop tmp_grep > tmp_grep_out2
+$DIFF tmp_grep_out1 tmp_grep_out2
+rm -f tmp_grep*
+
+println "\n===> zstdgrep multiple --regexp"
+echo "start" > tmp_grep
+echo "stop" >> tmp_grep
+ZCAT=./zstdcat $ZSTDGREP --regexp start --regexp stop tmp_grep > tmp_grep_out1
+grep -e start -e stop tmp_grep > tmp_grep_out2
+$DIFF tmp_grep_out1 tmp_grep_out2
+rm -f tmp_grep*
+
 println "\n===>  --exclude-compressed flag"
 rm -rf precompressedFilterTestDir
 mkdir -p precompressedFilterTestDir