add support for reading in UTF-8 sequences to some of the low-level

input functions

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@1989 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

diff --git a/ChangeLog b/ChangeLog
index 977ec9498650d3f399794304ab085f719382c263..d772a97bb0223d41a74a8830304bf55e5b3f83ae 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -71,6 +71,9 @@ CVS code -
          suggested by Bill Soudan)
        - Remove the -D/--dos and -M/--mac command line options, as they
          aren't much use with the new file format autodetection. (DLR)
+       - Add support for reading in UTF-8 sequences to some of the
+         low-level input routines.  Changes to get_kbinput() and
+         get_translated_kbinput(). (DLR)
 - files.c:
   do_insertfile()
        - Readd the NANO_SMALL #ifdef around the start_again: label to

diff --git a/src/nano.h b/src/nano.h
index 769410a64613a580e25e3a7a0db3f5fa671dab64..6bc2a0a1e4a76700041e3545145299b6b39f50f0 100644 (file)
--- a/src/nano.h
+++ b/src/nano.h
@@ -492,6 +492,10 @@ typedef enum {
     TOP, CENTER, NONE
 } topmidnone;
 
+typedef enum {
+    NO_SEQ, ESCAPE_SEQ, UTF8_SEQ
+} seq_type;
+
 /* Minimum editor window rows required for nano to work correctly. */
 #define MIN_EDITOR_ROWS 3
 

diff --git a/src/proto.h b/src/proto.h
index eb0653755e1eeccd337ae1aaeea4863126f86d84..5f789d3fc360f8bd0ec0155e5a96f2da96cfd6e7 100644 (file)
--- a/src/proto.h
+++ b/src/proto.h
@@ -505,7 +505,7 @@ void reset_kbinput(void);
 #endif
 void unget_kbinput(int kbinput, bool meta_key);
 int get_kbinput(WINDOW *win, bool *meta_key, bool *func_key);
-int get_translated_kbinput(int kbinput, bool *func_key, bool *es
+int get_translated_kbinput(int kbinput, seq_type *seq
 #ifndef NANO_SMALL
        , bool reset
 #endif

diff --git a/src/winio.c b/src/winio.c
index 935197c1a6ecfba81b0c71fe1c8217d2216b0d95..632b35399d2fbaf6719db69804b6a44785874577 100644 (file)
--- a/src/winio.c
+++ b/src/winio.c
@@ -96,7 +96,7 @@ static bool resetstatuspos = FALSE;
 /* Reset all the input routines that rely on character sequences. */
 void reset_kbinput(void)
 {
-    get_translated_kbinput(0, NULL, NULL, TRUE);
+    get_translated_kbinput(0, NULL, TRUE);
     get_ascii_kbinput(0, 0, TRUE);
     get_untranslated_kbinput(0, 0, FALSE, TRUE);
 }
@@ -123,7 +123,7 @@ void unget_kbinput(int kbinput, bool meta_key)
 int get_kbinput(WINDOW *win, bool *meta_key, bool *func_key)
 {
     int kbinput, retval = ERR;
-    bool es;
+    seq_type seq;
 
 #ifndef NANO_SMALL
     allow_pending_sigwinch(TRUE);
@@ -138,49 +138,77 @@ int get_kbinput(WINDOW *win, bool *meta_key, bool *func_key)
         * to get_translated_kbinput().  Continue until we get a
         * complete sequence. */
        kbinput = wgetch(win);
-       retval = get_translated_kbinput(kbinput, func_key, &es
+       retval = get_translated_kbinput(kbinput, &seq
 #ifndef NANO_SMALL
                , FALSE
 #endif
                );
 
-       /* If we got an escape sequence, read it in, including the
-        * initial non-escape, as verbatim input. */
-       if (es) {
-           int *escape_seq = NULL;
-           size_t es_len;
-
-           /* Read in the complete escape sequence, putting the initial
-            * non-escape at the beginning of it. */
-           escape_seq = get_verbatim_kbinput(win, kbinput, escape_seq,
-               &es_len, FALSE);
-
-           /* If the escape sequence is one character long, set
-            * meta_key to TRUE, make the non-escape character
-            * lowercase, and save that as the result. */
-           if (es_len == 1) {
-               *meta_key = TRUE;
-               retval = tolower(kbinput);
-           /* If the escape sequence is more than one character
-            * long, set meta_key to FALSE, translate the escape
-            * sequence into the corresponding key value, and save
-            * that as the result. */
-           } else if (es_len > 1) {
-               bool ignore_seq;
-
-               *meta_key = FALSE;
-               retval = get_escape_seq_kbinput(escape_seq, es_len,
+       /* If we got a one-character sequence and it's outside the ASCII
+        * range, set func_key to TRUE. */
+       if (seq == NO_SEQ) {
+           if (retval > 255)
+               *func_key = TRUE;
+       /* If we got a multi-character sequence, read it in, including
+        * the initial character, as verbatim input. */
+       } else {
+           int *sequence = NULL;
+           size_t seq_len;
+
+           sequence = get_verbatim_kbinput(win, kbinput, sequence,
+               &seq_len, FALSE);
+
+           /* Handle escape sequences. */
+           if (seq == ESCAPE_SEQ) {
+               /* If the escape sequence is one character long, set
+                * meta_key to TRUE, make the sequence character
+                * lowercase, and save that as the result. */
+               if (seq_len == 1) {
+                   *meta_key = TRUE;
+                   retval = tolower(kbinput);
+               /* If the escape sequence is more than one character
+                * long, set meta_key to FALSE, translate the escape
+                * sequence into the corresponding key value, and save
+                * that as the result. */
+               } else if (seq_len > 1) {
+                   bool ignore_seq;
+
+                   *meta_key = FALSE;
+                   retval = get_escape_seq_kbinput(sequence, seq_len,
                        &ignore_seq);
 
-               if (retval == ERR && !ignore_seq) {
-                   /* This escape sequence is unrecognized.  Send it
-                    * back. */
-                   for (; es_len > 1; es_len--)
-                       unget_kbinput(escape_seq[es_len - 1], FALSE);
-                   retval = escape_seq[0];
+                   if (retval == ERR && !ignore_seq) {
+                       /* This escape sequence is unrecognized.  Send
+                        * it back. */
+                       for (; seq_len > 1; seq_len--)
+                           unget_kbinput(sequence[seq_len - 1], FALSE);
+                       retval = sequence[0];
+                   }
                }
+           /* Handle UTF-8 sequences. */
+           } else if (seq == UTF8_SEQ) {
+               /* If we have a UTF-8 sequence, set func_key to FALSE,
+                * translate the UTF-8 sequence into the corresponding
+                * wide character value, and save that as the result. */
+               int i = 0;
+               char *s = charalloc(seq_len + 1);
+               wchar_t wc;
+
+               for (; i < seq_len; i++)
+                   s[i] = (char)sequence[i];
+               s[seq_len] = '\0';
+
+               *func_key = FALSE;
+               if (mbtowc(&wc, s, MB_CUR_MAX) == -1) {
+                   /* This UTF-8 sequence is unrecognized.  Send it
+                    * back. */
+                   for (; seq_len > 1; seq_len--)
+                       unget_kbinput(sequence[seq_len - 1], FALSE);
+                   retval = sequence[0];
+               } else
+                   retval = wc;
            }
-           free(escape_seq);
+           free(sequence);
        }
     }
 
@@ -195,11 +223,11 @@ int get_kbinput(WINDOW *win, bool *meta_key, bool *func_key)
     return retval;
 }
 
-/* Translate acceptable ASCII, extended keypad values, and escape
- * sequences into their corresponding key values.  Set func_key to TRUE
- * when we get an extended keypad value, and set es to TRUE when we get
- * an escape sequence.  Assume nodelay(win) is FALSE. */
-int get_translated_kbinput(int kbinput, bool *func_key, bool *es
+/* Translate acceptable ASCII, extended keypad values, and escape and
+ * UTF-8 sequences into their corresponding key values.  Set seq to
+ * ESCAPE_SEQ when we get an escape sequence, or UTF8_SEQ when we get a
+ * UTF-8 sequence.  Assume nodelay(win) is FALSE. */
+int get_translated_kbinput(int kbinput, seq_type *seq
 #ifndef NANO_SMALL
        , bool reset
 #endif
@@ -217,8 +245,7 @@ int get_translated_kbinput(int kbinput, bool *func_key, bool *es
     }
 #endif
 
-    *func_key = FALSE;
-    *es = FALSE;
+    *seq = NO_SEQ;
 
     switch (kbinput) {
        case ERR:
@@ -349,6 +376,12 @@ int get_translated_kbinput(int kbinput, bool *func_key, bool *es
                            break;
 #endif
                        default:
+                           /* A character with its high bit set: UTF-8
+                            * sequence mode.  Set seq to UTF8_SEQ. */
+                           if ((-128 <= kbinput && kbinput < 0) ||
+                               (127 < kbinput && kbinput <= 255))
+                               *seq = UTF8_SEQ;
+
                            retval = kbinput;
                            break;
                    }
@@ -356,9 +389,9 @@ int get_translated_kbinput(int kbinput, bool *func_key, bool *es
                case 1:
                    /* One escape followed by a non-escape: escape
                     * sequence mode.  Reset the escape counter and set
-                    * es to TRUE. */
+                    * seq to ESCAPE_SEQ. */
                    escapes = 0;
-                   *es = TRUE;
+                   *seq = ESCAPE_SEQ;
                    break;
                case 2:
                    switch (kbinput) {
@@ -428,12 +461,8 @@ int get_translated_kbinput(int kbinput, bool *func_key, bool *es
            }
     }
 
-    /* If the result is outside the ASCII range, set func_key to TRUE. */
-    if (retval > 255)
-       *func_key = TRUE;
- 
 #ifdef DEBUG
-    fprintf(stderr, "get_translated_kbinput(): kbinput = %d, func_key = %d, es = %d, escapes = %d, ascii_digits = %lu, retval = %d\n", kbinput, (int)*func_key, (int)*es, escapes, (unsigned long)ascii_digits, retval);
+    fprintf(stderr, "get_translated_kbinput(): kbinput = %d, seq = %d, escapes = %d, ascii_digits = %lu, retval = %d\n", kbinput, (int)*seq, escapes, (unsigned long)ascii_digits, retval);
 #endif
 
     /* Return the result. */