[Qemacs-commit] qemacs charset.c hex.c

[Charlie Gordon]

CVSROOT:        /sources/qemacs
Module name:    qemacs
Changes by:     Charlie Gordon <chqrlie>        17/04/25 12:56:26

Modified files:
        .              : charset.c hex.c 

Log message:
        charset: improve eol and binary detection
        - accept extra CR in badly encoded DOS files
        - ignore ^Z in DOS files, do not detect them as binary

CVSWeb URLs:
http://cvs.savannah.gnu.org/viewcvs/qemacs/charset.c?cvsroot=qemacs&r1=1.51&r2=1.52
http://cvs.savannah.gnu.org/viewcvs/qemacs/hex.c?cvsroot=qemacs&r1=1.52&r2=1.53

Patches:
Index: charset.c
===================================================================
RCS file: /sources/qemacs/qemacs/charset.c,v
retrieving revision 1.51
retrieving revision 1.52
diff -u -b -r1.51 -r1.52
--- charset.c   25 Apr 2017 16:46:41 -0000      1.51
+++ charset.c   25 Apr 2017 16:56:26 -0000      1.52
@@ -138,7 +138,8 @@
 {
     const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                            (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                           (1U << 0x1f);
     const u8 *p = buf;
     const u8 *p_end = p + size;
     uint32_t c;
@@ -363,7 +364,8 @@
 {
     const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                            (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                           (1U << 0x1f);
     const u8 *p = buf;
     const u8 *p_end = p + size;
     uint32_t c;
@@ -562,7 +564,8 @@
 {
     const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                            (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                           (1U << 0x1f);
     const u8 *p = buf;
     const u8 *p_end = p + (size & ~1);
     uint32_t c;
@@ -693,7 +696,8 @@
 {
     const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                            (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                           (1U << 0x1f);
     const u8 *p = buf;
     const u8 *p_end = p + (size & ~1);
     uint32_t c;
@@ -837,7 +841,8 @@
 {
     const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                            (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                           (1U << 0x1f);
     const u8 *p = buf;
     const u8 *p_end = p + (size & ~3);
     uint32_t c;
@@ -967,7 +972,8 @@
 {
     const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                            (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                           (1U << 0x1f);
     const u8 *p = buf;
     const u8 *p_end = p + (size & ~3);
     uint32_t c;
@@ -1219,6 +1225,11 @@
     while (p < p1) {
         c = *p++;
         if (c == '\r') {
+            if (*p == '\r') {
+                /* possibly spurious extra ^M in DOS file: ignore it,
+                 * next iteration will determine if encoding is MAC or DOS
+                 */
+            } else
             if (*p == '\n') {
                 p++;
                 eol_bits |= 1 << EOL_DOS;
@@ -1502,7 +1513,8 @@
     {
         const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                                (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                               (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                               (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                               (1U << 0x1f);
 
         for (i = 0; i < size; i++) {
             c = buf[i];

Index: hex.c
===================================================================
RCS file: /sources/qemacs/qemacs/hex.c,v
retrieving revision 1.52
retrieving revision 1.53
diff -u -b -r1.52 -r1.53
--- hex.c       16 Apr 2017 21:51:20 -0000      1.52
+++ hex.c       25 Apr 2017 16:56:26 -0000      1.53
@@ -197,7 +197,8 @@
 {
     const uint32_t magic = (1U << '\b') | (1U << '\t') | (1U << '\f') |
                            (1U << '\n') | (1U << '\r') | (1U << '\033') |
-                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1f);
+                           (1U << 0x0e) | (1U << 0x0f) | (1U << 0x1a) |
+                           (1U << 0x1f);
     int i, c;
 
     for (i = 0; i < size; i++) {