General spring-clean of EBCDIC-related issues in the code, which had decayed

over time. Also the documentation. Added one test that can be run in an ASCII
world to do a little testing of EBCDIC-related things. 


git-svn-id: svn://vcs.exim.org/pcre/code/trunk@1033 2f5784b3-3f2a-0410-8824-cb99058d5e15

1 files changed, 52 insertions, 26 deletions

diff --git a/pcre_newline.c b/pcre_newline.c
index a0a13c8..5e257d8 100644
--- a/pcre_newline.c
+++ b/pcre_newline.c
@@ -60,7 +60,7 @@ http://unicode.org/unicode/reports/tr18/. */
 *************************************************/
 
 /* It is guaranteed that the initial value of ptr is less than the end of the
-string that is being processed.
+string that is being processed. 
 
 Arguments:
   ptr          pointer to possible newline
@@ -86,12 +86,14 @@ if (utf)
 else
 #endif  /* SUPPORT_UTF */
   c = *ptr;
+  
+/* Note that this function is called only for ANY or ANYCRLF. */ 
 
 if (type == NLTYPE_ANYCRLF) switch(c)
   {
-  case 0x000a: *lenptr = 1; return TRUE;             /* LF */
-  case 0x000d: *lenptr = (ptr < endptr - 1 && ptr[1] == 0x0a)? 2 : 1;
-               return TRUE;                          /* CR */
+  case CHAR_LF: *lenptr = 1; return TRUE;
+  case CHAR_CR: *lenptr = (ptr < endptr - 1 && ptr[1] == CHAR_LF)? 2 : 1;
+               return TRUE;
   default: return FALSE;
   }
 
@@ -99,20 +101,29 @@ if (type == NLTYPE_ANYCRLF) switch(c)
 
 else switch(c)
   {
-  case 0x000a:                                       /* LF */
-  case 0x000b:                                       /* VT */
-  case 0x000c: *lenptr = 1; return TRUE;             /* FF */
-  case 0x000d: *lenptr = (ptr < endptr - 1 && ptr[1] == 0x0a)? 2 : 1;
-               return TRUE;                          /* CR */
+#ifdef EBCDIC
+  case CHAR_NEL:
+#endif     
+  case CHAR_LF:
+  case CHAR_VT:
+  case CHAR_FF: *lenptr = 1; return TRUE;
+
+  case CHAR_CR: 
+  *lenptr = (ptr < endptr - 1 && ptr[1] == CHAR_LF)? 2 : 1;
+  return TRUE;
+
+#ifndef EBCDIC
 #ifdef COMPILE_PCRE8
-  case 0x0085: *lenptr = utf? 2 : 1; return TRUE;    /* NEL */
+  case CHAR_NEL: *lenptr = utf? 2 : 1; return TRUE;
   case 0x2028:                                       /* LS */
   case 0x2029: *lenptr = 3; return TRUE;             /* PS */
-#else
-  case 0x0085:                                       /* NEL */
+#else   /* 16-bit (can't be EBCDIC) */
+  case CHAR_NEL:
   case 0x2028:                                       /* LS */
   case 0x2029: *lenptr = 1; return TRUE;             /* PS */
-#endif /* COMPILE_PCRE8 */
+#endif  /* COMPILE_PCRE8 */
+#endif  /* Not EBCDIC */
+
   default: return FALSE;
   }
 }
@@ -153,30 +164,45 @@ else
 #endif  /* SUPPORT_UTF */
   c = *ptr;
 
+/* Note that this function is called only for ANY or ANYCRLF. */ 
+
 if (type == NLTYPE_ANYCRLF) switch(c)
   {
-  case 0x000a: *lenptr = (ptr > startptr && ptr[-1] == 0x0d)? 2 : 1;
-               return TRUE;                         /* LF */
-  case 0x000d: *lenptr = 1; return TRUE;            /* CR */
+  case CHAR_LF: 
+  *lenptr = (ptr > startptr && ptr[-1] == CHAR_CR)? 2 : 1;
+  return TRUE;
+   
+  case CHAR_CR: *lenptr = 1; return TRUE;
   default: return FALSE;
   }
 
+/* NLTYPE_ANY */
+
 else switch(c)
   {
-  case 0x000a: *lenptr = (ptr > startptr && ptr[-1] == 0x0d)? 2 : 1;
-               return TRUE;                         /* LF */
-  case 0x000b:                                      /* VT */
-  case 0x000c:                                      /* FF */
-  case 0x000d: *lenptr = 1; return TRUE;            /* CR */
+  case CHAR_LF: 
+  *lenptr = (ptr > startptr && ptr[-1] == CHAR_CR)? 2 : 1;
+  return TRUE;
+                
+#ifdef EBCDIC
+  case CHAR_NEL:
+#endif     
+  case CHAR_VT:
+  case CHAR_FF:
+  case CHAR_CR: *lenptr = 1; return TRUE;
+   
+#ifndef EBCDIC 
 #ifdef COMPILE_PCRE8
-  case 0x0085: *lenptr = utf? 2 : 1; return TRUE;   /* NEL */
-  case 0x2028:                                      /* LS */
-  case 0x2029: *lenptr = 3; return TRUE;            /* PS */
+  case CHAR_NEL: *lenptr = utf? 2 : 1; return TRUE;
+  case 0x2028:                                       /* LS */
+  case 0x2029: *lenptr = 3; return TRUE;             /* PS */
 #else
-  case 0x0085:                                       /* NEL */
+  case CHAR_NEL:
   case 0x2028:                                       /* LS */
   case 0x2029: *lenptr = 1; return TRUE;             /* PS */
-#endif /* COMPILE_PCRE8 */
+#endif  /* COMPILE_PCRE8 */
+#endif  /* NotEBCDIC */
+
   default: return FALSE;
   }
 }