segment: Refine treatment of start of macro body.

Previously, if the first line of the macro body (the same line as the
closing parenthesis in the DEFINE) was blank, we reported it as a blank
line to the lexer.  The parser for DEFINE could check for that (by seeing
whether the first line of macro body was empty or all-spaces) but it seems
more elegant to do it in the segmenter.  This implements that.

diff --git a/src/language/lexer/segment.c b/src/language/lexer/segment.c
index 1a8e8a6b5fb54c479ba988855777aaf3834b0432..a4fea0b213118559d474b94a2bc4efa4008ff0d0 100644 (file)
--- a/src/language/lexer/segment.c
+++ b/src/language/lexer/segment.c
@@ -1512,7 +1512,10 @@ segmenter_parse_define_2__ (struct segmenter *s,
     {
       s->nest--;
       if (!s->nest)
-        s->state = S_DEFINE_3;
+        {
+          s->state = S_DEFINE_3;
+          s->substate = 0;
+        }
       return ofs;
     }
 
@@ -1554,9 +1557,15 @@ segmenter_parse_define_3__ (struct segmenter *s,
 
          The line might be blank, whether completely empty or just spaces and
          comments.  That's OK: we need to report blank lines because they can
-         have significance. */
-      *type = SEG_MACRO_BODY;
+         have significance.
+
+         However, if the first line of the macro body (the same line as the
+         closing parenthesis in the argument definition) is blank, we just
+         report it as spaces because it's not significant. */
+      *type = (s->substate == 0 && is_all_spaces (input, ofs)
+               ? SEG_SPACES : SEG_MACRO_BODY);
       s->state = S_DEFINE_4;
+      s->substate = 1;
       return ofs;
     }
   else

diff --git a/tests/language/lexer/scan.at b/tests/language/lexer/scan.at
index d778d07fa2097af516b0086748c14ee582997202..d1fb66fcbeb5dd90deefefc81c62a32f851149e4 100644 (file)
--- a/tests/language/lexer/scan.at
+++ b/tests/language/lexer/scan.at
@@ -894,7 +894,7 @@ SKIP
 MACRO_ID "!macro1"
 LPAREN
 RPAREN
-STRING ""
+SKIP
 SKIP
 STRING "var1 var2 var3"
 SKIP
@@ -940,7 +940,7 @@ SKIP
 MACRO_ID "!macro1"
 LPAREN
 RPAREN
-STRING ""
+SKIP
 SKIP
 STRING "var1 var2 var3"
 MACRO_ID "!enddefine"
@@ -983,6 +983,34 @@ SKIP
 MACRO_ID "!macro1"
 LPAREN
 RPAREN
+SKIP
+SKIP
+MACRO_ID "!enddefine"
+ENDCMD
+-SKIP
+STOP
+])
+PSPP_CHECK_SCAN([-i])
+AT_CLEANUP
+\f
+AT_SETUP([DEFINE command - blank lines])
+AT_KEYWORDS([scan])
+AT_DATA([input], [dnl
+define !macro1()
+
+
+!enddefine.
+])
+AT_DATA([expout-base], [dnl
+ID "define"
+SKIP
+MACRO_ID "!macro1"
+LPAREN
+RPAREN
+SKIP
+SKIP
+STRING ""
+SKIP
 STRING ""
 SKIP
 MACRO_ID "!enddefine"
@@ -1018,7 +1046,7 @@ ID "c"
 LPAREN
 RPAREN
 RPAREN
-STRING ""
+SKIP
 SKIP
 MACRO_ID "!enddefine"
 ENDCMD
@@ -1063,7 +1091,7 @@ LPAREN
 RPAREN
 SKIP
 RPAREN
-STRING ""
+SKIP
 SKIP
 MACRO_ID "!enddefine"
 ENDCMD
@@ -1096,7 +1124,7 @@ COMMA
 ID "z"
 SKIP
 RPAREN
-STRING ""
+SKIP
 SKIP
 STRING "content 1"
 SKIP
@@ -1241,7 +1269,7 @@ SKIP
 MACRO_ID "!macro1"
 LPAREN
 RPAREN
-STRING ""
+SKIP
 SKIP
 STRING "content line 1"
 SKIP

diff --git a/tests/language/lexer/segment.at b/tests/language/lexer/segment.at
index 65a271f69865d59b83305c63224985ad1f196c4a..67e647d2214ec8c22daa1bf753ed2c7fabe99114 100644 (file)
--- a/tests/language/lexer/segment.at
+++ b/tests/language/lexer/segment.at
@@ -1066,7 +1066,7 @@ identifier      define    space
 macro_id        !macro1
 punct           (
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_body      var1_var2_var3
@@ -1115,7 +1115,7 @@ identifier      define    space
 macro_id        !macro1
 punct           (
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_body      var1_var2_var3
@@ -1159,7 +1159,7 @@ identifier      define    space
 macro_id        !macro1
 punct           (
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_id        !enddefine
@@ -1184,7 +1184,7 @@ identifier      define    space
 macro_id        !macro1
 punct           (
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_body
@@ -1224,7 +1224,7 @@ identifier      c
 punct           (
 punct           )
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_id        !enddefine
@@ -1273,7 +1273,7 @@ punct           )
 newline         \n (later)
 
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_id        !enddefine
@@ -1309,7 +1309,7 @@ identifier      z
 newline         \n (later)
 
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_body      content_1
@@ -1451,7 +1451,7 @@ identifier      define    space
 macro_id        !macro1
 punct           (
 punct           )
-macro_body
+spaces
 newline         \n (DEFINE)
 
 macro_body      content_line_1