From 65f41a742e025487f8ec7f2e7ca2a3af3283fc96 Mon Sep 17 00:00:00 2001
From: Sergey Poznyakoff <gray@gnu.org>
Date: Mon, 13 May 2019 15:20:24 +0300
Subject: wordsplit: optionally disable splitting of unexpandable variable and
 command refs

* include/wordsplit.h (WRDSO_NOVARSPLIT)
(WRDSO_NOCMDSPLIT): New options.
* src/wordsplit.c (scan_word): Treat any variable reference, even
containing whitespace, as a single word if WRDSO_NOVARSPLIT is set.
Ditto for commands and WRDSO_NOCMDSPLIT.
* tests/wordsplit.at: Add new tests.
* tests/wsp.c: Recognize novarsplit and nocmdsplit
options.
For future use: recognize bskeep_words, bskeep_quote, bskeep.
---
 include/wordsplit.h | 11 ++++++++---
 src/wordsplit.c     |  6 ++++--
 tests/wordsplit.at  | 18 ++++++++++++++++++
 tests/wsp.c         | 11 ++++++++---
 4 files changed, 38 insertions(+), 8 deletions(-)

diff --git a/include/wordsplit.h b/include/wordsplit.h
index 3a7ab25..a175275 100644
--- a/include/wordsplit.h
+++ b/include/wordsplit.h
@@ -201,9 +201,7 @@ struct wordsplit
 #define WRDSO_FAILGLOB        0x00000002
 /* Allow a leading period to be matched by metacharacters. */
 #define WRDSO_DOTGLOB         0x00000004
-#if 0 /* Unused value */
-#define WRDSO_ARGV            0x00000008
-#endif
+/* Unused value:              0x00000008 */
 /* Keep backslash in unrecognized escape sequences in words */
 #define WRDSO_BSKEEP_WORD     0x00000010
 /* Handle octal escapes in words */
@@ -220,6 +218,13 @@ struct wordsplit
 #define WRDSO_OESC_QUOTE      0x00000200
 /* Handle hex escapes in quoted strings */
 #define WRDSO_XESC_QUOTE      0x00000400
+/* Unused: 0x00000800 */
+/* Don't split variable references, even if they contain whitespace
+   (e.g. ${VAR:-foo bar}) */
+#define WRDSO_NOVARSPLIT     0x00001000
+/* Don't split commands, even containing whitespace, e.g.
+   $(echo foo bar) */
+#define WRDSO_NOCMDSPLIT     0x00002000
 
 #define WRDSO_BSKEEP          WRDSO_BSKEEP_WORD     
 #define WRDSO_OESC            WRDSO_OESC_WORD       
diff --git a/src/wordsplit.c b/src/wordsplit.c
index e979f27..521a1eb 100644
--- a/src/wordsplit.c
+++ b/src/wordsplit.c
@@ -2058,11 +2058,13 @@ scan_word (struct wordsplit *wsp, size_t start, int consume_all)
 
 	  if (command[i] == '$')
 	    {
-	      if (!(wsp->ws_flags & WRDSF_NOVAR)
+	      if ((!(wsp->ws_flags & WRDSF_NOVAR)
+		   || (wsp->ws_options & WRDSO_NOVARSPLIT))
 		  && command[i+1] == '{'
 		  && find_closing_paren (command, i + 2, len, &i, "{}") == 0)
 		continue;
-	      if (!(wsp->ws_flags & WRDSF_NOCMD)
+	      if ((!(wsp->ws_flags & WRDSF_NOCMD)
+		   || (wsp->ws_options & WRDSO_NOCMDSPLIT))
 		  && command[i+1] == '('
 		  && find_closing_paren (command, i + 2, len, &i, "()") == 0)
 		continue;
diff --git a/tests/wordsplit.at b/tests/wordsplit.at
index e3af703..1f2e80d 100644
--- a/tests/wordsplit.at
+++ b/tests/wordsplit.at
@@ -950,6 +950,24 @@ TOTAL: 3
 [input exhausted
 ]))
 
+TESTWSP([variable nosplit],[],[novar novarsplit],
+[begin ${VAR:- a b} end],
+[NF: 3
+0: begin
+1: "${VAR:- a b}"
+2: end
+TOTAL: 3
+])
+
+TESTWSP([command nosplit],[],[nocmd nocmdsplit],
+[begin $(words a b) end],
+[NF: 3
+0: begin
+1: "$(words a b)"
+2: end
+TOTAL: 3
+])
+
 m4_popdef([TESTWSP])
 m4_popdef([wspnum])
 m4_popdef([wspid])
diff --git a/tests/wsp.c b/tests/wsp.c
index cca3a36..bd13e63 100644
--- a/tests/wsp.c
+++ b/tests/wsp.c
@@ -61,9 +61,14 @@ struct kwd bool_keytab[] = {
 };
 
 struct kwd opt_keytab[] = {
-  { "nullglob", WRDSO_NULLGLOB },
-  { "failglob", WRDSO_FAILGLOB },
-  { "dotglob",  WRDSO_DOTGLOB },
+  { "nullglob",     WRDSO_NULLGLOB },
+  { "failglob",     WRDSO_FAILGLOB },
+  { "dotglob",      WRDSO_DOTGLOB },
+  { "bskeep_words", WRDSO_BSKEEP_WORD },
+  { "bskeep_quote", WRDSO_BSKEEP_QUOTE },
+  { "bskeep",       WRDSO_BSKEEP_WORD|WRDSO_BSKEEP_QUOTE },
+  { "novarsplit",   WRDSO_NOVARSPLIT },
+  { "nocmdsplit",   WRDSO_NOCMDSPLIT },
   { NULL, 0 }
 };
 
-- 
cgit v1.2.1