glcpp: prevent accidental token pasting

Explicit token pasting is handled specially, but the token
printing code can accidentally paste identifiers. Check for
this accidental pasting and prevent it by inserting a space
if necessary.

Supersedes a similar but much more limited test to prevent pasting
`+` and `-` to make `++` and `--`.

A test is provided in glcpp/tests/150-token-accidental-concatenation.
The accidental pasting test in 068-accidental-pasting is also updated
to cover some of the other cases.

Signed-off-by: Eric R. Smith <eric.smith@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37214>

src/compiler/glsl/glcpp/glcpp-parse.y

@@ -1450,6 +1450,21 @@ _token_paste(glcpp_parser_t *parser, token_t *token, token_t *other)
    return token;
 }
 
+/*
+ * Check to see if we need a space in between two tokens, for example to
+ * keep "+ +" from collapsing to "++"
+ */
+static bool
+need_space_between(int token1, int token2)
+{
+   if ( (token1 == '+' || token1 == '-') && token2 == token1 )
+      return true;
+   if (token1 == IDENTIFIER &&
+       (token2 == IDENTIFIER || token2 == INTEGER))
+      return true;
+   return false;
+}
+
 static void
 _token_list_print(glcpp_parser_t *parser, token_list_t *list)
 {
@@ -1458,8 +1473,13 @@ _token_list_print(glcpp_parser_t *parser, token_list_t *list)
    if (list == NULL)
       return;
 
-   for (node = list->head; node; node = node->next)
+   for (node = list->head; node; node = node->next) {
       _token_print(parser->output, node->token);
+      /* avoid accidental token concatenation */
+      if (node->next &&
+          need_space_between(node->token->type, node->next->token->type))
+         _mesa_string_buffer_append_char(parser->output, ' ');
+   }
 }
 
 void
@@ -2030,21 +2050,6 @@ _glcpp_parser_expand_node(glcpp_parser_t *parser, token_node_t *node,
 
       replacement = _token_list_copy(parser, macro->replacements);
 
-      /* If needed insert space in front of replacements to isolate them from
-       * the code they will be inserted into. For example:
-       *
-       *    #define VALUE -1.0
-       *    int a = -VALUE;
-       *
-       * Should be evaluated to int a = - -1.0; not int a = --1.0;
-       */
-      if (node_prev &&
-          (node_prev->token->type == '-' || node_prev->token->type == '+') &&
-          node_prev->token->type == replacement->head->token->type) {
-         token_t *new_token = _token_create_ival(parser, SPACE, SPACE);
-         _token_list_prepend(parser, replacement, new_token);
-      }
-
       _glcpp_parser_apply_pastes(parser, replacement);
       return replacement;
    }

src/compiler/glsl/glcpp/tests/068-accidental-pasting.c

@@ -1,3 +1,6 @@
+#define VALUE1 -1.0
+#define VALUE2 +2
+#define ONEPLUS(x) 1+x
 #define empty
 <empty<
 <empty=
@@ -9,3 +12,6 @@
 |empty|
 +empty+
 -empty-
+-VALUE1
++VALUE2
+ONEPLUS(+1)

src/compiler/glsl/glcpp/tests/068-accidental-pasting.c.expected

@@ -1,4 +1,7 @@
 
+
+
+
 < <
 < =
 > >
@@ -9,3 +12,6 @@
 | |
 + +
 - -
+- -1.0
++ +2
+1+ +1

src/compiler/glsl/glcpp/tests/150-token-accidental-concatenation.c

@@ -0,0 +1,16 @@
+/*
+ * avoid accidental concatenation of strings
+ * but make sure deliberate concatenation works
+ */
+#define a xa
+#define b xb
+#define A(x, y) x y
+#define B(x, y) x ## y
+#define C(x) x
+
+A(r, s)
+B(r, s)
+C(r)s
+A(a, b)
+//B(a, b) // not working yet; gcc and clang give `ab`, we give `xaxb`
+C(a)b

src/compiler/glsl/glcpp/tests/150-token-accidental-concatenation.c.expected

@@ -0,0 +1,16 @@
+ 
+
+
+
+
+
+
+
+
+
+r s
+rs
+r s
+xa xb
+
+xa xb