updated for version 7.4.241v7.4.241

Problem:    The string returned by submatch() does not distinguish between a
            NL from a line break and a NL that stands for a NUL character.
Solution:   Add a second argument to return a list. (ZyX)

9 files changed, 128 insertions, 6 deletions

diff --git a/runtime/doc/eval.txt b/runtime/doc/eval.txt
index 007c7f7469..cdf8d72bb2 100644
--- a/runtime/doc/eval.txt
+++ b/runtime/doc/eval.txt
@@ -1990,7 +1990,8 @@ strridx( {haystack}, {needle} [, {start}])
 				Number	last index of {needle} in {haystack}
 strtrans( {expr})		String	translate string to make it printable
 strwidth( {expr})		Number	display cell length of the String {expr}
-submatch( {nr})			String	specific match in ":s" or substitute()
+submatch( {nr}[, {list}])	String or List
+					specific match in ":s" or substitute()
 substitute( {expr}, {pat}, {sub}, {flags})
 				String	all {pat} in {expr} replaced with {sub}
 synID( {lnum}, {col}, {trans})	Number	syntax ID at {lnum} and {col}
@@ -5797,12 +5798,23 @@ strwidth({expr})					*strwidth()*
 		Ambiguous, this function's return value depends on 'ambiwidth'.
 		Also see |strlen()|, |strdisplaywidth()| and |strchars()|.
 
-submatch({nr})						*submatch()*
+submatch({nr}[, {list}])				*submatch()*
 		Only for an expression in a |:substitute| command or
 		substitute() function.
 		Returns the {nr}'th submatch of the matched text.  When {nr}
 		is 0 the whole matched text is returned.
+		Note that a NL in the string can stand for a line break of a
+		multi-line match or a NUL character in the text.
 		Also see |sub-replace-expression|.
+
+		If {list} is present and non-zero then submatch() returns 
+		a list of strings, similar to |getline()| with two arguments. 
+		NL characters in the text represent NUL characters in the
+		text.
+		Only returns more than one item for |:substitute|, inside
+		|substitute()| this list will always contain one or zero
+		items, since there are no real line breaks.
+
 		Example: >
 			:s/\d\+/\=submatch(0) + 1/
 <		This finds the first number in the line and adds one to it.
diff --git a/src/eval.c b/src/eval.c
index 05f91b303e..7b682572cf 100644
--- a/src/eval.c
+++ b/src/eval.c
@@ -8129,7 +8129,7 @@ static struct fst
     {"strridx",		2, 3, f_strridx},
     {"strtrans",	1, 1, f_strtrans},
     {"strwidth",	1, 1, f_strwidth},
-    {"submatch",	1, 1, f_submatch},
+    {"submatch",	1, 2, f_submatch},
     {"substitute",	4, 4, f_substitute},
     {"synID",		3, 3, f_synID},
     {"synIDattr",	2, 3, f_synIDattr},
@@ -17890,9 +17890,32 @@ f_submatch(argvars, rettv)
     typval_T	*argvars;
     typval_T	*rettv;
 {
-    rettv->v_type = VAR_STRING;
-    rettv->vval.v_string =
-		    reg_submatch((int)get_tv_number_chk(&argvars[0], NULL));
+    int		error = FALSE;
+    char_u	**match;
+    char_u	**s;
+    listitem_T	*li;
+    int		no;
+    int		retList = 0;
+
+    no = (int)get_tv_number_chk(&argvars[0], &error);
+    if (error)
+	return;
+    error = FALSE;
+    if (argvars[1].v_type != VAR_UNKNOWN)
+	retList = get_tv_number_chk(&argvars[1], &error);
+    if (error)
+	return;
+
+    if (retList == 0)
+    {
+	rettv->v_type = VAR_STRING;
+	rettv->vval.v_string = reg_submatch(no);
+    }
+    else
+    {
+	rettv->v_type = VAR_LIST;
+	rettv->vval.v_list = reg_submatch_list(no);
+    }
 }
 
 /*
diff --git a/src/proto/regexp.pro b/src/proto/regexp.pro
index 38c9c33c4e..a42c5944a8 100644
--- a/src/proto/regexp.pro
+++ b/src/proto/regexp.pro
@@ -10,6 +10,7 @@ char_u *regtilde __ARGS((char_u *source, int magic));
 int vim_regsub __ARGS((regmatch_T *rmp, char_u *source, char_u *dest, int copy, int magic, int backslash));
 int vim_regsub_multi __ARGS((regmmatch_T *rmp, linenr_T lnum, char_u *source, char_u *dest, int copy, int magic, int backslash));
 char_u *reg_submatch __ARGS((int no));
+list_T *reg_submatch_list __ARGS((int no));
 regprog_T *vim_regcomp __ARGS((char_u *expr_arg, int re_flags));
 void vim_regfree __ARGS((regprog_T *prog));
 int vim_regexec __ARGS((regmatch_T *rmp, char_u *line, colnr_T col));
diff --git a/src/regexp.c b/src/regexp.c
index 32a51884b2..0a590f66d7 100644
--- a/src/regexp.c
+++ b/src/regexp.c
@@ -7897,6 +7897,85 @@ reg_submatch(no)
 
     return retval;
 }
+
+/*
+ * Used for the submatch() function with the optional non-zero argument: get
+ * the list of strings from the n'th submatch in allocated memory with NULs
+ * represented in NLs.
+ * Returns a list of allocated strings.  Returns NULL when not in a ":s"
+ * command, for a non-existing submatch and for any error.
+ */
+    list_T *
+reg_submatch_list(no)
+    int		no;
+{
+    char_u	*s;
+    linenr_T	slnum;
+    linenr_T	elnum;
+    colnr_T	scol;
+    colnr_T	ecol;
+    int		i;
+    list_T	*list;
+    int		error = FALSE;
+
+    if (!can_f_submatch || no < 0)
+	return NULL;
+
+    if (submatch_match == NULL)
+    {
+	slnum = submatch_mmatch->startpos[no].lnum;
+	elnum = submatch_mmatch->endpos[no].lnum;
+	if (slnum < 0 || elnum < 0)
+	    return NULL;
+
+	scol = submatch_mmatch->startpos[no].col;
+	ecol = submatch_mmatch->endpos[no].col;
+
+	list = list_alloc();
+	if (list == NULL)
+	    return NULL;
+
+	s = reg_getline_submatch(slnum) + scol;
+	if (slnum == elnum)
+	{
+	    if (list_append_string(list, s, ecol - scol) == FAIL)
+		error = TRUE;
+	}
+	else
+	{
+	    if (list_append_string(list, s, -1) == FAIL)
+		error = TRUE;
+	    for (i = 1; i < elnum - slnum; i++)
+	    {
+		s = reg_getline_submatch(slnum + i);
+		if (list_append_string(list, s, -1) == FAIL)
+		    error = TRUE;
+	    }
+	    s = reg_getline_submatch(elnum);
+	    if (list_append_string(list, s, ecol) == FAIL)
+		error = TRUE;
+	}
+    }
+    else
+    {
+	s = submatch_match->startp[no];
+	if (s == NULL || submatch_match->endp[no] == NULL)
+	    return NULL;
+	list = list_alloc();
+	if (list == NULL)
+	    return NULL;
+	if (list_append_string(list, s,
+				 (int)(submatch_match->endp[no] - s)) == FAIL)
+	    error = TRUE;
+    }
+
+    if (error)
+    {
+	list_free(list, TRUE);
+	return NULL;
+    }
+    return list;
+}
 #endif
 
 static regengine_T bt_regengine =
diff --git a/src/testdir/test79.in b/src/testdir/test79.in
index f15ecc0f8d..56955c2318 100644
--- a/src/testdir/test79.in
+++ b/src/testdir/test79.in
diff --git a/src/testdir/test79.ok b/src/testdir/test79.ok
index bb30d14052..0f6ea45451 100644
--- a/src/testdir/test79.ok
+++ b/src/testdir/test79.ok
diff --git a/src/testdir/test80.in b/src/testdir/test80.in
index c62fdc02b2..406fb6dac7 100644
--- a/src/testdir/test80.in
+++ b/src/testdir/test80.in
@@ -117,6 +117,7 @@ STARTTEST
 :set cpo&
 :$put =\"\n\nTEST_5:\"
 :$put =substitute('A123456789', 'A\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)', '\=submatch(0) . submatch(9) . submatch(8) . submatch(7) . submatch(6) . submatch(5) . submatch(4) . submatch(3) . submatch(2) . submatch(1)', '')
+:$put =substitute('A123456789', 'A\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)\(.\)', '\=string([submatch(0, 1), submatch(9, 1), submatch(8, 1), submatch(7, 1), submatch(6, 1), submatch(5, 1), submatch(4, 1), submatch(3, 1), submatch(2, 1), submatch(1, 1)])', '')
 /^TEST_6
 ENDTEST
 
@@ -142,6 +143,7 @@ STARTTEST
 :$put =\"\n\nTEST_7:\"
 :$put =substitute('A
A', 'A.', '\=submatch(0)', '')
 :$put =substitute(\"B\nB\", 'B.', '\=submatch(0)', '')
+:$put =substitute(\"B\nB\", 'B.', '\=string(submatch(0, 1))', '')
 :$put =substitute('-bb', '\zeb', 'a', 'g')
 :$put =substitute('-bb', '\ze', 'c', 'g')
 /^TEST_8
diff --git a/src/testdir/test80.ok b/src/testdir/test80.ok
index 2b79d377a7..b42f604a07 100644
--- a/src/testdir/test80.ok
+++ b/src/testdir/test80.ok
@@ -90,6 +90,7 @@ l
 
 TEST_5:
 A123456789987654321
+[['A123456789'], ['9'], ['8'], ['7'], ['6'], ['5'], ['4'], ['3'], ['2'], ['1']]
 
 
 TEST_6:
@@ -103,6 +104,8 @@ TEST_7:
 A
A
 B
 B
+['B
+']B
 -abab
 c-cbcbc
 
diff --git a/src/version.c b/src/version.c
index 472a7fad03..fbfb5e66a1 100644
--- a/src/version.c
+++ b/src/version.c
@@ -735,6 +735,8 @@ static char *(features[]) =
 static int included_patches[] =
 {   /* Add new patch number below this line */
 /**/
+    241,
+/**/
     240,
 /**/
     239,