API: Introduced exceptions for common mistakes of the symbol mapper

[libucw.git] / ucw / string.h

diff --git a/ucw/string.h b/ucw/string.h
index f96bd0234012189ecabf268c5d5d5d533273360d..0c64ace57dca303d8f3e6e447fc09cec03b931b0 100644 (file)
--- a/ucw/string.h
+++ b/ucw/string.h
@@ -2,7 +2,7 @@
  *     UCW Library -- String Routines
  *
  *     (c) 2006 Pavel Charvat <pchar@ucw.cz>
- *     (c) 2007--2008 Martin Mares <mj@ucw.cz>
+ *     (c) 2007--2012 Martin Mares <mj@ucw.cz>
  *
  *     This software may be freely distributed and used according to the terms
  *     of the GNU Lesser General Public License.
@@ -11,31 +11,122 @@
 #ifndef _UCW_STRING_H
 #define _UCW_STRING_H
 
+#ifdef CONFIG_UCW_CLEAN_ABI
+#define hex_to_mem ucw_hex_to_mem
+#define mem_to_hex ucw_mem_to_hex
+#define str_count_char ucw_str_count_char
+#define str_format_flags ucw_str_format_flags
+#define str_has_prefix ucw_str_has_prefix
+#define str_has_suffix ucw_str_has_suffix
+#define str_hier_prefix ucw_str_hier_prefix
+#define str_hier_suffix ucw_str_hier_suffix
+#define str_match_pattern ucw_str_match_pattern
+#define str_match_pattern_nocase ucw_str_match_pattern_nocase
+#define str_sepsplit ucw_str_sepsplit
+#define str_unesc ucw_str_unesc
+#define str_wordsplit ucw_str_wordsplit
+#endif
+
 /* string.c */
 
+#ifdef CONFIG_DARWIN
+uns strnlen(const char *str, uns n);   // NOAPI
+#endif
+
+/**
+ * Format a set of flag bits. When the i-th bit of @flags is 1,
+ * set the i-th character of @dest to @fmt[i], otherwise to '-'.
+ **/
 char *str_format_flags(char *dest, const char *fmt, uns flags);
+
+/** Counts occurrences of @chr in @str. **/
 uns str_count_char(const char *str, uns chr);
 
 /* str-esc.c */
 
+/**
+ * Decode a string with backslash escape sequences as in C99 strings.
+ * It is safe to pass @dest equal to @src.
+ **/
 char *str_unesc(char *dest, const char *src);
 
 /* str-split.c */
 
+/**
+ * Split @str to at most @max fields separated by @sep. Occurrences of the
+ * separator are rewritten to string terminators, @rec[i] is set to point
+ * to the i-th field. The total number of fields is returned.
+ *
+ * When there are more than @max fields in @str, the first @max fields
+ * are processed and -1 is returned.
+ **/
 int str_sepsplit(char *str, uns sep, char **rec, uns max);
+
+/**
+ * Split @str to words separated by white-space characters. The spaces
+ * are replaced by string terminators, @rec[i] is set to point to the
+ * i-th field. The total number of fields is returned.
+ *
+ * When there are more than @max fields in @str, the first @max fields
+ * are processed and -1 is returned.
+ *
+ * Fields surrounded by double quotes are also recognized. They can contain
+ * spaces, but no mechanism for escaping embedded quotes is defined.
+ **/
 int str_wordsplit(char *str, char **rec, uns max);
 
 /* str-(i)match.c: Matching of shell patterns */
 
+/**
+ * Test whether the string @str matches the shell-like pattern @patt. Only
+ * "*" and "?" meta-characters are supported.
+ **/
 int str_match_pattern(const char *patt, const char *str);
+
+/** A case-insensitive version of @str_match_pattern(). **/
 int str_match_pattern_nocase(const char *patt, const char *str);
 
 /* str-hex.c */
 
+/**
+ * Create a hexdump of a memory block of @bytes bytes starting at @src.
+ * The @flags contain an optional separator of bytes (0 if bytes should
+ * not be separated), possibly OR-ed with `MEM_TO_HEX_UPCASE` when upper-case
+ * characters should be used.
+ **/
 void mem_to_hex(char *dest, const byte *src, uns bytes, uns flags);
+
+/**
+ * An inverse function to @mem_to_hex(). Takes a hexdump of at most @max_bytes
+ * bytes and stores the bytes to a buffer starting at @dest. Returns a pointer
+ * at the first character after the dump.
+ **/
 const char *hex_to_mem(byte *dest, const char *src, uns max_bytes, uns flags);
 
 // Bottom 8 bits of flags are an optional separator of bytes, the rest is:
 #define MEM_TO_HEX_UPCASE 0x100
 
+/* str-fix.c */
+
+int str_has_prefix(const char *str, const char *prefix); /** Tests if @str starts with @prefix. **/
+int str_has_suffix(const char *str, const char *suffix); /** Tests if @str ends with @suffix. **/
+
+/**
+ * Let @str and @prefix be hierarchical names with components separated by
+ * a character @sep. Returns true if @str starts with @prefix, respecting
+ * component boundaries.
+ *
+ * For example, when @sep is '/' and @str is "/usr/local/bin", then:
+ *
+ * - "/usr/local" is a prefix
+ * - "/usr/local/" is a prefix, too
+ * - "/usr/loc" is not,
+ * - "/usr/local/bin" is a prefix,
+ * - "/usr/local/bin/" is not,
+ * - "/" is a prefix,
+ * - "" is a prefix.
+ **/
+int str_hier_prefix(const char *str, const char *prefix, uns sep);
+int str_hier_suffix(const char *str, const char *suffix, uns sep); /** Like @str_hier_prefix(), but for suffixes. **/
+
 #endif