Fixed bug #35655 (whitespace following end of heredoc is lost).

5 files changed, 71 insertions, 48 deletions

diff --git a/NEWS b/NEWS
index 61745e176e..8e5f792988 100644
--- a/NEWS
+++ b/NEWS
@@ -26,6 +26,7 @@ PHP                                                                        NEWS
 - Fixed many bugs in OCI8. (Tony)
 - Fixed crash and leak in mysqli when using 4.1.x client libraries and
   connecting to 5.x server. (Andrey)
+- Fixed bug #35655 (whitespace following end of heredoc is lost). (Ilia)
 - Fixed bug #35630 (strtotime() crashes on certain relative identifiers).
   (Ilia)
 - Fixed bug #35629 (crash in http:// wrapper on multiple redirects). (Ilia)
diff --git a/Zend/tests/bug35655.phpt b/Zend/tests/bug35655.phpt
new file mode 100755
index 0000000000..30f00cb32e
--- /dev/null
+++ b/Zend/tests/bug35655.phpt
@@ -0,0 +1,18 @@
+--TEST--
+Bug #35655 (whitespace following end of heredoc is lost)
+--FILE--
+<?php
+$code = '
+<? 
+  $x = <<<EOT
+some string    
+EOT
+  $y = 2;
+?>';
+highlight_string($code);
+?>
+--EXPECT--
+<code><span style="color: #000000">
+<br /><span style="color: #0000BB">&lt;?&nbsp;<br />&nbsp;&nbsp;$x&nbsp;</span><span style="color: #007700">=&nbsp;&lt;&lt;&lt;EOT<br /></span><span style="color: #0000BB">some&nbsp;string&nbsp;&nbsp;&nbsp;&nbsp;<br /></span><span style="color: #007700">EOT<br />&nbsp;&nbsp;</span><span style="color: #0000BB">$y&nbsp;</span><span style="color: #007700">=&nbsp;</span><span style="color: #0000BB">2</span><span style="color: #007700">;<br /></span><span style="color: #0000BB">?&gt;</span>
+</span>
+</code>
diff --git a/Zend/zend_highlight.c b/Zend/zend_highlight.c
index d32b79c7b5..a6151717d0 100644
--- a/Zend/zend_highlight.c
+++ b/Zend/zend_highlight.c
@@ -94,7 +94,7 @@ ZEND_API void zend_highlight(zend_syntax_highlighter_ini *syntax_highlighter_ini
 	int token_type;
 	char *last_color = syntax_highlighter_ini->highlight_html;
 	char *next_color;
-	int in_string=0, post_heredoc = 0;
+	int in_string=0;
 
 	zend_printf("<code>");
 	zend_printf("<span style=\"color: %s\">\n", last_color);
@@ -151,14 +151,9 @@ ZEND_API void zend_highlight(zend_syntax_highlighter_ini *syntax_highlighter_ini
 		switch (token_type) {
 			case T_END_HEREDOC:
 				zend_html_puts(token.value.str.val, token.value.str.len TSRMLS_CC);
-				post_heredoc = 1;
 				break;
 			default:
 				zend_html_puts(LANG_SCNG(yy_text), LANG_SCNG(yy_leng) TSRMLS_CC);
-				if (post_heredoc) {
-					 zend_html_putc('\n');
-					 post_heredoc = 0;
-				}
 				break;
 		}
 
@@ -215,19 +210,18 @@ ZEND_API void zend_strip(TSRMLS_D)
 			case EOF:
 				return;
 			
-			case T_END_HEREDOC: {
-					char *ptr = LANG_SCNG(yy_text);
-
-					zend_write(ptr, LANG_SCNG(yy_leng) - 1);
-					/* The ensure that we only write one ; and that it followed by the required newline */
-					zend_write("\n", sizeof("\n") - 1);
-					if (ptr[LANG_SCNG(yy_leng) - 1] == ';') {
-						lex_scan(&token TSRMLS_CC);
-					}
-					efree(token.value.str.val);
+			case T_END_HEREDOC:
+				zend_write(LANG_SCNG(yy_text), LANG_SCNG(yy_leng));
+				efree(token.value.str.val);
+				/* read the following character, either newline or ; */
+				if (lex_scan(&token TSRMLS_CC) != T_WHITESPACE) {
+					zend_write(LANG_SCNG(yy_text), LANG_SCNG(yy_leng));
 				}
-				break;
-			
+				zend_write("\n", sizeof("\n") - 1);
+				prev_space = 1;
+				token.type = 0;
+				continue;
+
 			default:
 				zend_write(LANG_SCNG(yy_text), LANG_SCNG(yy_leng));
 				break;
diff --git a/Zend/zend_language_scanner.l b/Zend/zend_language_scanner.l
index af4059fef7..830fae6c7f 100644
--- a/Zend/zend_language_scanner.l
+++ b/Zend/zend_language_scanner.l
@@ -1733,7 +1733,6 @@ NEWLINE ("\r"|"\n"|"\r\n")
 
 <ST_HEREDOC>^{LABEL}(";")?{NEWLINE} {
 	int label_len;
-	unsigned char unput_semicolon;
 
 	CG(zend_lineno)++;
 	if (yytext[yyleng-2]=='\r') {
@@ -1744,17 +1743,12 @@ NEWLINE ("\r"|"\n"|"\r\n")
 
 	if (yytext[label_len-1]==';') {
 		label_len--;
-		unput_semicolon=1;
-	} else{
-		unput_semicolon=0;
 	}
 
 	if (label_len==CG(heredoc_len) && !memcmp(yytext, CG(heredoc), label_len)) {
 		zendlval->value.str.val = estrndup(yytext, label_len); /* unput destroys yytext */
 		zendlval->value.str.len = label_len;
-		if (unput_semicolon) {
-			unput(';');
-		}
+		yyless(yyleng - (yyleng - label_len));
 		efree(CG(heredoc));
 		CG(heredoc)=NULL;
 		CG(heredoc_len)=0;
diff --git a/ext/tokenizer/tests/bug26463.phpt b/ext/tokenizer/tests/bug26463.phpt
index 7480aa2f59..679ffe3c37 100644
--- a/ext/tokenizer/tests/bug26463.phpt
+++ b/ext/tokenizer/tests/bug26463.phpt
@@ -14,11 +14,11 @@ DDDD;
 var_dump(token_get_all($str));
 ?>
 --EXPECTF--
-array(17) {
+array(19) {
   [0]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(367)
     [1]=>
     string(6) "<?php
 "
@@ -26,7 +26,7 @@ array(17) {
   [1]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(309)
     [1]=>
     string(2) "$x"
   }
@@ -35,7 +35,7 @@ array(17) {
   [3]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(371)
     [1]=>
     string(6) "<<<DD
 "
@@ -43,7 +43,7 @@ array(17) {
   [4]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(307)
     [1]=>
     string(13) "jhdsjkfhjdsh
 "
@@ -51,67 +51,83 @@ array(17) {
   [5]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(372)
     [1]=>
     string(2) "DD"
   }
   [6]=>
-  string(1) "."
+  array(2) {
+    [0]=>
+    int(370)
+    [1]=>
+    string(1) "
+"
+  }
   [7]=>
+  string(1) "."
+  [8]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(315)
     [1]=>
     string(2) """"
   }
-  [8]=>
-  string(1) ";"
   [9]=>
+  string(1) ";"
+  [10]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(370)
     [1]=>
     string(1) "
 "
   }
-  [10]=>
+  [11]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(309)
     [1]=>
     string(2) "$a"
   }
-  [11]=>
-  string(1) "="
   [12]=>
+  string(1) "="
+  [13]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(371)
     [1]=>
     string(8) "<<<DDDD
 "
   }
-  [13]=>
+  [14]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(307)
     [1]=>
     string(13) "jhdsjkfhjdsh
 "
   }
-  [14]=>
+  [15]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(372)
     [1]=>
     string(4) "DDDD"
   }
-  [15]=>
-  string(1) ";"
   [16]=>
+  string(1) ";"
+  [17]=>
+  array(2) {
+    [0]=>
+    int(370)
+    [1]=>
+    string(1) "
+"
+  }
+  [18]=>
   array(2) {
     [0]=>
-    int(%d)
+    int(369)
     [1]=>
     string(2) "?>"
   }