WebSVN – HelenOS – Diff – /branches/dd/kernel/generic/src/lib/string.c

 #include <cpu.h>
 #include <arch/asm.h>
 #include <arch.h>
 #include <console/kconsole.h>
+char invalch = '?';
+/** Byte mask consisting of lowest @n bits (out of eight). */
+#define LO_MASK_8(n) ((uint8_t)((1 << (n)) - 1))
+/** Byte mask consisting of lowest @n bits (out of 32). */
+#define LO_MASK_32(n) ((uint32_t)((1 << (n)) - 1))
+/** Byte mask consisting of highest @n bits (out of eight). */
+#define HI_MASK_8(n) (~LO_MASK_8(8 - (n)))
+/** Number of data bits in a UTF-8 continuation byte. */
+#define CONT_BITS 6
-/** Return number of characters in a string.
+/** Decode a single character from a substring.
+ *
+ * Decode a single character from a substring of size @a sz. Decoding starts
+ * at @a offset and this offset is moved to the beginning of the next
+ * character. In case of decoding error, offset generally advances at least
+ * by one. However, offset is never moved beyond (str + sz).
+ *
+ * @param str   String (not necessarily NULL-terminated).
+ * @param index Index (counted in plain characters) where to start
+ *              the decoding.
+ * @param limit Size of the substring.
+ *
+ * @return  Value of decoded character or '?' on decoding error.
+ *
+ */
+wchar_t chr_decode(const char *str, size_t *offset, size_t sz)
+{
+    uint8_t b0, b;          /* Bytes read from str. */
+    wchar_t ch;
+    int b0_bits;        /* Data bits in first byte. */
+    int cbytes;     /* Number of continuation bytes. */
+    if (*offset + 1 > sz)
+        return invalch;
+    b0 = (uint8_t) str[(*offset)++];
+    /* Determine code length. */
+    if ((b0 & 0x80) == 0) {
+        /* 0xxxxxxx (Plain ASCII) */
+        b0_bits = 7;
+        cbytes = 0;
+    } else if ((b0 & 0xe0) == 0xc0) {
+        /* 110xxxxx 10xxxxxx */
+        b0_bits = 5;
+        cbytes = 1;
+    } else if ((b0 & 0xf0) == 0xe0) {
+        /* 1110xxxx 10xxxxxx 10xxxxxx */
+        b0_bits = 4;
+        cbytes = 2;
+    } else if ((b0 & 0xf8) == 0xf0) {
+        /* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
+        b0_bits = 3;
+        cbytes = 3;
+    } else {
+        /* 10xxxxxx -- unexpected continuation byte. */
+        return invalch;
+    }
+    if (*offset + cbytes > sz) {
+        return invalch;
+    }
+    ch = b0 & LO_MASK_8(b0_bits);
+    /* Decode continuation bytes. */
+    while (cbytes > 0) {
+        b = (uint8_t) str[(*offset)++];
+        /* Must be 10xxxxxx. */
+        if ((b & 0xc0) != 0x80) {
+            return invalch;
+        }
+        /* Shift data bits to ch. */
+        ch = (ch << CONT_BITS) | (wchar_t) (b & LO_MASK_8(CONT_BITS));
+        --cbytes;
+    }
+    return ch;
+}
+/** Encode a single character to string representation.
+ *
+ * Encode a single character to string representation (i.e. UTF-8) and store
+ * it into a buffer at @a offset. Encoding starts at @a offset and this offset
+ * is moved to the position where the next character can be written to.
+ *
+ * @param ch        Input character.
+ * @param str       Output buffer.
+ * @param offset    Offset (in bytes) where to start writing.
+ * @param sz        Size of the output buffer.
+ *
+ * @return True if the character was encoded successfully or false if there
+ *     was not enough space in the output buffer or the character code
+ *     was invalid.
+ */
+bool chr_encode(const wchar_t ch, char *str, size_t *offset, size_t sz)
+{
+    uint32_t cc;        /* Unsigned version of ch. */
+    int cbytes;     /* Number of continuation bytes. */
+    int b0_bits;        /* Number of data bits in first byte. */
+    int i;
+    if (*offset >= sz)
+        return false;
+    if (ch < 0)
+        return false;
+    /* Bit operations should only be done on unsigned numbers. */
+    cc = (uint32_t) ch;
+    /* Determine how many continuation bytes are needed. */
+    if ((cc & ~LO_MASK_32(7)) == 0) {
+        b0_bits = 7;
+        cbytes = 0;
+    } else if ((cc & ~LO_MASK_32(11)) == 0) {
+        b0_bits = 5;
+        cbytes = 1;
+    } else if ((cc & ~LO_MASK_32(16)) == 0) {
+        b0_bits = 4;
+        cbytes = 2;
+    } else if ((cc & ~LO_MASK_32(21)) == 0) {
+        b0_bits = 3;
+        cbytes = 3;
+    } else {
+        /* Codes longer than 21 bits are not supported. */
+        return false;
+    }
+    /* Check for available space in buffer. */
+    if (*offset + cbytes >= sz)
+        return false;
+    /* Encode continuation bytes. */
+    for (i = cbytes; i > 0; --i) {
+        str[*offset + i] = 0x80 | (cc & LO_MASK_32(CONT_BITS));
+        cc = cc >> CONT_BITS;
+    }
+    /* Encode first byte. */
+    str[*offset] = (cc & LO_MASK_32(b0_bits)) | HI_MASK_8(8 - b0_bits - 1);
+    /* Advance offset. */
+    *offset += (1 + cbytes);
+    return true;
+}
+/** Get bytes used by UTF-8 characters.
+ *
+ * Get the number of bytes (count of plain characters) which
+ * are used by a given count of UTF-8 characters in a string.
+ * As UTF-8 encoding is multibyte, there is no constant
+ * correspondence between number of characters and used bytes.
+ *
+ * @param str   UTF-8 string to consider.
+ * @param count Number of UTF-8 characters to count.
+ *
+ * @return Number of bytes used by the characters.
+ *
+ */
+size_t utf8_count_bytes(const char *str, count_t count)
+{
+    size_t size = 0;
+    index_t index = 0;
+    index_t iprev;
+    wchar_t ch;
+    while (true) {
+        iprev = index;
+        if (size >= count)
+            break;
+        ch = chr_decode(str, &index, UTF8_NO_LIMIT);
+        if (ch == '\0') break;
+        size++;
+    }
+    return iprev;
+}
+/** Check whether character is plain ASCII.
+ *
+ * @return True if character is plain ASCII.
+ *
+ */
+bool ascii_check(const wchar_t ch)
+{
+    if ((ch >= 0) && (ch <= 127))
+        return true;
+    return false;
+}
+/** Check whether character is Unicode.
+ *
+ * @return True if character is valid Unicode code point.
+ *
+ */
+bool unicode_check(const wchar_t ch)
+{
+    if ((ch >= 0) && (ch <= 1114111))
+        return true;
+    return false;
+}
+/** Return number of plain characters in a string.
+ *
- * @param str NULL terminated string.
+ * @param str NULL-terminated string.
+ *
  * @return Number of characters in str.
+ *
  */
 size_t strlen(const char *str)
+{
-    int i;
+    size_t size;
+    for (size = 0; str[size]; size++);
+    return size;
+}
+/** Return number of UTF-8 characters in a string.
+ *
+ * @param str NULL-terminated UTF-8 string.
+ *
+ * @return Number of UTF-8 characters in str.
+ *
+ */
+size_t strlen_utf8(const char *str)
+{
+    size_t size = 0;
+    index_t index = 0;
+    while (chr_decode(str, &index, UTF8_NO_LIMIT) != 0) {
+        size++;
+    }
+    return size;
+}
+/** Return number of UTF-32 characters in a string.
+ *
+ * @param str NULL-terminated UTF-32 string.
+ *
+ * @return Number of UTF-32 characters in str.
+ *
+ */
+size_t strlen_utf32(const wchar_t *str)
+{
+    size_t size;
-    for (i = 0; str[i]; i++);
+    for (size = 0; str[size]; size++);
-    return i;
+    return size;
+}
 /** Compare two NULL terminated strings
+ *
  * Do a char-by-char comparison of two NULL terminated strings.

Subversion Repositories HelenOS

(root)/branches/dd/kernel/generic/src/lib/string.c – Rev 4055 → 4201