[PATCH v5] regexp-replace and ^, word boundary or look-behind operators (and more).

[Stephane Chazelas <stephane@chazelas.org>]

2024-03-08 15:30:50 +0000, Stephane Chazelas:
[...]
> - it would run in an infinite loop when there's no match in ERE
>   mode.
[...]
> +    [[ -n $subject ]] && break
[...]

D'oh, should be -z instead of -n or || instead of &&.

So, here's a v5 patch (previous one should have been a v4, I
forgot to update the subject):

diff --git a/Functions/Misc/regexp-replace b/Functions/Misc/regexp-replace
index d4408f0f7..db8f63404 100644
--- a/Functions/Misc/regexp-replace
+++ b/Functions/Misc/regexp-replace
@@ -1,91 +1,95 @@
-# Replace all occurrences of a regular expression in a variable.  The
-# variable is modified directly.  Respects the setting of the
-# option RE_MATCH_PCRE.
+# Replace all occurrences of a regular expression in a scalar variable.
+# The variable is modified directly.  Respects the setting of the option
+# RE_MATCH_PCRE, but otherwise sets the zsh emulation mode.
 #
-# First argument: *name* (not contents) of variable.
-# Second argument: regular expression
-# Third argument: replacement string.  This can contain all forms of
-# $ and backtick substitutions; in particular, $MATCH will be replaced
-# by the portion of the string matched by the regular expression.
-
-# we use positional parameters instead of variables to avoid
-# clashing with the user's variable. Make sure we start with 3 and only
-# 3 elements:
-argv=("$1" "$2" "$3")
-
-# $4 records whether pcre is enabled as that information would otherwise
-# be lost after emulate -L zsh
-4=0
-[[ -o re_match_pcre ]] && 4=1
+# Arguments:
+#
+# 1. *name* (not contents) of variable or more generally any lvalue;
+#    expected to be scalar.
+#
+# 2. regular expression
+#
+# 3. replacement string.  This can contain all forms of
+#    $ and backtick substitutions; in particular, $MATCH will be
+#    replaced by the portion of the string matched by the regular
+#    expression. Parsing errors are fatal to the shell process.
+
+if (( $# < 2 || $# > 3 )); then
+  setopt localoptions functionargzero
+  print -ru2 "Usage: $0 <varname> <regexp> [<replacement>]"
+  return 2
+fi
 
-emulate -L zsh
+# ensure variable exists in the caller's scope before referencing it
+# to make sure we don't end up referencing one of our own.
+typeset -g -- "$1" || return 2
+typeset -nu -- var=$1 || return 2
 
+local -i use_pcre=0
+[[ -o re_match_pcre ]] && use_pcre=1
 
-local MATCH MBEGIN MEND
+emulate -L zsh
+
+local regexp=$2 replacement=$3 result MATCH MBEGIN MEND
 local -a match mbegin mend
 
-if (( $4 )); then
+if (( use_pcre )); then
   # if using pcre, we're using pcre_match and a running offset
   # That's needed for ^, \A, \b, and look-behind operators to work
   # properly.
 
   zmodload zsh/pcre || return 2
-  pcre_compile -- "$2" && pcre_study || return 2
+  pcre_compile -- "$regexp" && pcre_study || return 2
+
+  local -i offset=0 start stop
+  local new ZPCRE_OP
+  local -a finds
 
-  # $4 is the current *byte* offset, $5, $6 reserved for later use
-  4=0 6=
+  while pcre_match -b -n $offset -- "$var"; do
+    # we need to perform the evaluation in a scalar assignment so that
+    # if it generates an array, the elements are converted to string (by
+    # joining with the first chararacter of $IFS as usual)
+    new=${(Xe)replacement}
 
-  local ZPCRE_OP
-  while pcre_match -b -n $4 -- "${(P)1}"; do
-    # append offsets and computed replacement to the array
-    # we need to perform the evaluation in a scalar assignment so that if
-    # it generates an array, the elements are converted to string (by
-    # joining with the first character of $IFS as usual)
-    5=${(e)3}
-    argv+=(${(s: :)ZPCRE_OP} "$5")
+    finds+=( ${(s[ ])ZPCRE_OP} "$new" )
 
     # for 0-width matches, increase offset by 1 to avoid
     # infinite loop
-    4=$((argv[-2] + (argv[-3] == argv[-2])))
+    (( offset = finds[-2] + (finds[-3] == finds[-2]) ))
   done
 
-  (($# > 6)) || return # no match
+  (( $#finds )) || return # no match
 
-  set +o multibyte
+  unsetopt multibyte
 
-  # $5 contains the result, $6 the current offset
-  5= 6=1
-  for 2 3 4 in "$@[7,-1]"; do
-    5+=${(P)1[$6,$2]}$4
-    6=$(($3 + 1))
+  offset=1
+  for start stop new in "$finds[@]"; do
+    result+=${var[offset,start]}$new
+    (( offset = stop + 1 ))
   done
-  5+=${(P)1[$6,-1]}
-else
+  result+=${var[offset,-1]}
+
+else # no PCRE
+
   # in ERE, we can't use an offset so ^, (and \<, \b, \B, [[:<:]] where
   # available) won't work properly.
-
-  # $4 is the string to be matched
-  4=${(P)1}
-
-  while [[ -n $4 ]]; do
-    if [[ $4 =~ $2 ]]; then
-      # append initial part and substituted match
-      5+=${4[1,MBEGIN-1]}${(e)3}
-      # truncate remaining string
-      if ((MEND < MBEGIN)); then
-        # zero-width match, skip one character for the next match
-        ((MEND++))
-	5+=${4[1]}
-      fi
-      4=${4[MEND+1,-1]}
-      # indicate we did something
-      6=1
-    else
-      break
+  local subject=$var
+  local -i ok
+  while [[ $subject =~ $regexp ]]; do
+    # append initial part and substituted match
+    result+=$subject[1,MBEGIN-1]${(Xe)replacement}
+    # truncate remaining string
+    if (( MEND < MBEGIN )); then
+      # zero-width match, skip one character for the next match
+      (( MEND++ ))
+      result+=$subject[MBEGIN]
     fi
+    subject=$subject[MEND+1,-1]
+    ok=1
+    [[ -z $subject ]] && break
   done
-  [[ -n $6 ]] || return # no match
-  5+=$4
+  (( ok )) || return
+  result+=$subject
 fi
 
-eval $1=\$5
+var=$result