emacs: src/coding.c comparison

comparison src/coding.c @ 22226:557fac086b1b

(ascii_skip_code): New variable. (detect_coding_mask): Skip ASCII codes at the head according to ascii_skip_code. (coding_priorities): New variable. (Fset_coding_priority_internal): New function. (detect_coding): Call set_coding_mask with coding_priorities as arg CODING. (init_coding_once): Initialize ascii_skip_code. (syms_of_coding): Declare set-coding-priority-internal as a Lisp function.

author	Kenichi Handa <handa@m17n.org>
date	Mon, 25 May 1998 08:08:07 +0000
parents	fc4aaf1b1772
children	1deb7d79af00

comparison

equal deleted inserted replaced

-:6f56af1aab96
+:557fac086b1b
 "coding-category-big5",
 "coding-category-raw-text",
 "coding-category-binary"
 };
-/* Table pointers to coding systems corresponding to each coding
+/* Table of pointers to coding systems corresponding to each coding
 categories.  */
 struct coding_system *coding_system_table[CODING_CATEGORY_IDX_MAX];
+/* Table of coding category masks.  Nth element is a mask for a coding
+cateogry of which priority is Nth.  */
+static
+int coding_priorities[CODING_CATEGORY_IDX_MAX];
 /* Flag to tell if we look up translation table on character code
 conversion.  */
 Lisp_Object Venable_character_translation;
 /* Standard translation table to look up on decoding (reading).  */
 highest priority.  Priorities of categories are also specified by a
 user in a Lisp variable `coding-category-list'.
 */
+static
+int ascii_skip_code[256];
 /* Detect how a text of length SRC_BYTES pointed by SOURCE is encoded.
 If it detects possible coding systems, return an integer in which
 appropriate flag bits are set.  Flag bits are defined by macros
 CODING_CATEGORY_MASK_XXX in `coding.h'.
 unsigned char *source;
 int src_bytes, *priorities, *skip;
 {
 register unsigned char c;
 unsigned char *src = source, *src_end = source + src_bytes;
-unsigned int mask = (CODING_CATEGORY_MASK_ISO_7BIT
+unsigned int mask;
-		       | CODING_CATEGORY_MASK_ISO_SHIFT);
 int i;
 /* At first, skip all ASCII characters and control characters except
 for three ISO2022 specific control characters.  */
+ascii_skip_code[ISO_CODE_SO] = 0;
+ascii_skip_code[ISO_CODE_SI] = 0;
+ascii_skip_code[ISO_CODE_ESC] = 0;
 label_loop_detect_coding:
-while (src < src_end)
+while (src < src_end && ascii_skip_code[*src]) src++;
-{
-c = *src;
-if (c >= 0x80
-	  || ((mask & CODING_CATEGORY_MASK_ISO_7BIT)
-	      && c == ISO_CODE_ESC)
-	  || ((mask & CODING_CATEGORY_MASK_ISO_SHIFT)
-	      && (c == ISO_CODE_SI || c == ISO_CODE_SO)))
-	break;
-src++;
-}
 *skip = src - source;
 if (src >= src_end)
 /* We found nothing other than ASCII.  There's nothing to do.  */
 return 0;
 mask = detect_coding_iso2022 (src, src_end);
 if (mask == 0)
 	{
 	  /* No valid ISO2022 code follows C.  Try again.  */
 	  src++;
-	  mask = (c != ISO_CODE_ESC
+	  if (c == ISO_CODE_ESC)
-		  ? CODING_CATEGORY_MASK_ISO_7BIT
+	    ascii_skip_code[ISO_CODE_ESC] = 1;
-		  : CODING_CATEGORY_MASK_ISO_SHIFT);
+	  else
+	    ascii_skip_code[ISO_CODE_SO] = ascii_skip_code[ISO_CODE_SI] = 1;
 	  goto label_loop_detect_coding;
 	}
 if (priorities)
 	goto label_return_highest_only;
 }
 unsigned char *src;
 int src_bytes;
 {
 unsigned int idx;
 int skip, mask, i;
-int priorities[CODING_CATEGORY_IDX_MAX];
 Lisp_Object val = Vcoding_category_list;
-i = 0;
+mask = detect_coding_mask (src, src_bytes, coding_priorities, &skip);
-while (CONSP (val) && i < CODING_CATEGORY_IDX_MAX)
-{
-if (! SYMBOLP (XCONS (val)->car))
-	break;
-idx = XFASTINT (Fget (XCONS (val)->car, Qcoding_category_index));
-if (idx >= CODING_CATEGORY_IDX_MAX)
-	break;
-priorities[i++] = (1 << idx);
-val = XCONS (val)->cdr;
-}
-/* If coding-category-list is valid and contains all coding
-categories, `i' should be CODING_CATEGORY_IDX_MAX now.  If not,
-the following code saves Emacs from craching.  */
-while (i < CODING_CATEGORY_IDX_MAX)
-priorities[i++] = CODING_CATEGORY_MASK_RAW_TEXT;
-mask = detect_coding_mask (src, src_bytes, priorities, &skip);
 coding->heading_ascii = skip;
 if (!mask) return;
 /* We found a single coding system of the highest priority in MASK.  */
 	 coding_system_table[i]);
 }
 return Qnil;
 }
+DEFUN ("set-coding-priority-internal", Fset_coding_priority_internal,
+Sset_coding_priority_internal, 0, 0, 0,
+"Update internal database for the current value of `coding-category-list'.\n\
+This function is internal use only.")
+()
+{
+int i = 0, idx;
+Lisp_Object val = Vcoding_category_list;
+while (CONSP (val) && i < CODING_CATEGORY_IDX_MAX)
+{
+if (! SYMBOLP (XCONS (val)->car))
+	break;
+idx = XFASTINT (Fget (XCONS (val)->car, Qcoding_category_index));
+if (idx >= CODING_CATEGORY_IDX_MAX)
+	break;
+coding_priorities[i++] = (1 << idx);
+val = XCONS (val)->cdr;
+}
+/* If coding-category-list is valid and contains all coding
+categories, `i' should be CODING_CATEGORY_IDX_MAX now.  If not,
+the following code saves Emacs from craching.  */
+while (i < CODING_CATEGORY_IDX_MAX)
+coding_priorities[i++] = CODING_CATEGORY_MASK_RAW_TEXT;
+return Qnil;
+}
 #endif /* emacs */
 /*** 8. Post-amble ***/
 setup_coding_system (Qnil, &keyboard_coding);
 setup_coding_system (Qnil, &terminal_coding);
 setup_coding_system (Qnil, &safe_terminal_coding);
 bzero (coding_system_table, sizeof coding_system_table);
+bzero (ascii_skip_code, sizeof ascii_skip_code);
+for (i = 0; i < 128; i++)
+ascii_skip_code[i] = 1;
 #if defined (MSDOS) || defined (WINDOWSNT)
 system_eol_type = CODING_EOL_CRLF;
 #else
 system_eol_type = CODING_EOL_LF;
 defsubr (&Sterminal_coding_system);
 defsubr (&Sset_keyboard_coding_system_internal);
 defsubr (&Skeyboard_coding_system);
 defsubr (&Sfind_operation_coding_system);
 defsubr (&Supdate_iso_coding_systems);
+defsubr (&Sset_coding_priority_internal);
 DEFVAR_LISP ("coding-system-list", &Vcoding_system_list,
 "List of coding systems.\n\
 \n\
 Do not alter the value of this variable manually.  This variable should be\n\

Mercurial > emacs

comparison src/coding.c @ 22226:557fac086b1b