Subversion
svn_string.h
Go to the documentation of this file.
00001 /**
00002  * @copyright
00003  * ====================================================================
00004  *    Licensed to the Apache Software Foundation (ASF) under one
00005  *    or more contributor license agreements.  See the NOTICE file
00006  *    distributed with this work for additional information
00007  *    regarding copyright ownership.  The ASF licenses this file
00008  *    to you under the Apache License, Version 2.0 (the
00009  *    "License"); you may not use this file except in compliance
00010  *    with the License.  You may obtain a copy of the License at
00011  *
00012  *      http://www.apache.org/licenses/LICENSE-2.0
00013  *
00014  *    Unless required by applicable law or agreed to in writing,
00015  *    software distributed under the License is distributed on an
00016  *    "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
00017  *    KIND, either express or implied.  See the License for the
00018  *    specific language governing permissions and limitations
00019  *    under the License.
00020  * ====================================================================
00021  * @endcopyright
00022  *
00023  * @file svn_string.h
00024  * @brief Counted-length strings for Subversion, plus some C string goodies.
00025  *
00026  * There are two string datatypes: @c svn_string_t and @c svn_stringbuf_t.
00027  * The former is a simple pointer/length pair useful for passing around
00028  * strings (or arbitrary bytes) with a counted length. @c svn_stringbuf_t is
00029  * buffered to enable efficient appending of strings without an allocation
00030  * and copy for each append operation.
00031  *
00032  * @c svn_string_t contains a <tt>const char *</tt> for its data, so it is
00033  * most appropriate for constant data and for functions which expect constant,
00034  * counted data. Functions should generally use <tt>const @c svn_string_t
00035  * *</tt> as their parameter to indicate they are expecting a constant,
00036  * counted string.
00037  *
00038  * @c svn_stringbuf_t uses a plain <tt>char *</tt> for its data, so it is
00039  * most appropriate for modifiable data.
00040  *
00041  * <h3>Invariants</h3>
00042  *
00043  *   1. Null termination:
00044  *
00045  *      Both structures maintain a significant invariant:
00046  *
00047  *         <tt>s->data[s->len] == '\\0'</tt>
00048  *
00049  *      The functions defined within this header file will maintain
00050  *      the invariant (which does imply that memory is
00051  *      allocated/defined as @c len+1 bytes).  If code outside of the
00052  *      @c svn_string.h functions manually builds these structures,
00053  *      then they must enforce this invariant.
00054  *
00055  *      Note that an @c svn_string(buf)_t may contain binary data,
00056  *      which means that strlen(s->data) does not have to equal @c
00057  *      s->len. The NULL terminator is provided to make it easier to
00058  *      pass @c s->data to C string interfaces.
00059  *
00060  *
00061  *   2. Non-NULL input:
00062  *
00063  *      All the functions assume their input data is non-NULL,
00064  *      unless otherwise documented, and may seg fault if passed
00065  *      NULL.  The input data may *contain* null bytes, of course, just
00066  *      the data pointer itself must not be NULL.
00067  *
00068  * <h3>Memory allocation</h3>
00069  *
00070  *   All the functions make a deep copy of all input data, and never store
00071  *   a pointer to the original input data.
00072  */
00073 
00074 
00075 #ifndef SVN_STRING_H
00076 #define SVN_STRING_H
00077 
00078 #include <apr.h>          /* for apr_size_t */
00079 #include <apr_pools.h>    /* for apr_pool_t */
00080 #include <apr_tables.h>   /* for apr_array_header_t */
00081 
00082 #include "svn_types.h"    /* for svn_boolean_t, svn_error_t */
00083 
00084 #ifdef __cplusplus
00085 extern "C" {
00086 #endif /* __cplusplus */
00087 
00088 /**
00089  * @defgroup svn_string String handling
00090  * @{
00091  */
00092 
00093 
00094 
00095 /** A simple counted string. */
00096 typedef struct svn_string_t
00097 {
00098   const char *data; /**< pointer to the bytestring */
00099   apr_size_t len;   /**< length of bytestring */
00100 } svn_string_t;
00101 
00102 /** A buffered string, capable of appending without an allocation and copy
00103  * for each append. */
00104 typedef struct svn_stringbuf_t
00105 {
00106   /** a pool from which this string was originally allocated, and is not
00107    * necessarily specific to this string.  This is used only for allocating
00108    * more memory from when the string needs to grow.
00109    */
00110   apr_pool_t *pool;
00111 
00112   /** pointer to the bytestring */
00113   char *data;
00114 
00115   /** length of bytestring */
00116   apr_size_t len;
00117 
00118   /** total size of buffer allocated */
00119   apr_size_t blocksize;
00120 } svn_stringbuf_t;
00121 
00122 
00123 /**
00124  * @defgroup svn_string_svn_string_t svn_string_t functions
00125  * @{
00126  */
00127 
00128 /** Create a new bytestring containing a C string (NULL-terminated). */
00129 svn_string_t *
00130 svn_string_create(const char *cstring, apr_pool_t *pool);
00131 
00132 /** Create a new bytestring containing a generic string of bytes
00133  * (NOT NULL-terminated) */
00134 svn_string_t *
00135 svn_string_ncreate(const char *bytes, apr_size_t size, apr_pool_t *pool);
00136 
00137 /** Create a new string with the contents of the given stringbuf */
00138 svn_string_t *
00139 svn_string_create_from_buf(const svn_stringbuf_t *strbuf, apr_pool_t *pool);
00140 
00141 /** Create a new bytestring by formatting @a cstring (NULL-terminated)
00142  * from varargs, which are as appropriate for apr_psprintf().
00143  */
00144 svn_string_t *
00145 svn_string_createf(apr_pool_t *pool, const char *fmt, ...)
00146   __attribute__((format(printf, 2, 3)));
00147 
00148 /** Create a new bytestring by formatting @a cstring (NULL-terminated)
00149  * from a @c va_list (see svn_stringbuf_createf()).
00150  */
00151 svn_string_t *
00152 svn_string_createv(apr_pool_t *pool, const char *fmt, va_list ap)
00153   __attribute__((format(printf, 2, 0)));
00154 
00155 /** Return TRUE if a bytestring is empty (has length zero). */
00156 svn_boolean_t
00157 svn_string_isempty(const svn_string_t *str);
00158 
00159 /** Return a duplicate of @a original_string. */
00160 svn_string_t *
00161 svn_string_dup(const svn_string_t *original_string, apr_pool_t *pool);
00162 
00163 /** Return @c TRUE iff @a str1 and @a str2 have identical length and data. */
00164 svn_boolean_t
00165 svn_string_compare(const svn_string_t *str1, const svn_string_t *str2);
00166 
00167 /** Return offset of first non-whitespace character in @a str, or return
00168  * @a str->len if none.
00169  */
00170 apr_size_t
00171 svn_string_first_non_whitespace(const svn_string_t *str);
00172 
00173 /** Return position of last occurrence of @a ch in @a str, or return
00174  * @a str->len if no occurrence.
00175  */
00176 apr_size_t
00177 svn_string_find_char_backward(const svn_string_t *str, char ch);
00178 
00179 /** @} */
00180 
00181 
00182 /**
00183  * @defgroup svn_string_svn_stringbuf_t svn_stringbuf_t functions
00184  * @{
00185  */
00186 
00187 /** Create a new bytestring containing a C string (NULL-terminated). */
00188 svn_stringbuf_t *
00189 svn_stringbuf_create(const char *cstring, apr_pool_t *pool);
00190 
00191 /** Create a new bytestring containing a generic string of bytes
00192  * (NON-NULL-terminated)
00193  */
00194 svn_stringbuf_t *
00195 svn_stringbuf_ncreate(const char *bytes, apr_size_t size, apr_pool_t *pool);
00196 
00197 /** Create a new empty bytestring with at least @a minimum_size bytes of
00198  * space available in the memory block.
00199  *
00200  * The allocated string buffer will be one byte larger than @a minimum_size
00201  * to account for a final '\\0'.
00202  *
00203  * @since New in 1.6.
00204  */
00205 svn_stringbuf_t *
00206 svn_stringbuf_create_ensure(apr_size_t minimum_size, apr_pool_t *pool);
00207 
00208 /** Create a new stringbuf with the contents of the given string */
00209 svn_stringbuf_t *
00210 svn_stringbuf_create_from_string(const svn_string_t *str, apr_pool_t *pool);
00211 
00212 /** Create a new bytestring by formatting @a cstring (NULL-terminated)
00213  * from varargs, which are as appropriate for apr_psprintf().
00214  */
00215 svn_stringbuf_t *
00216 svn_stringbuf_createf(apr_pool_t *pool, const char *fmt, ...)
00217   __attribute__((format(printf, 2, 3)));
00218 
00219 /** Create a new bytestring by formatting @a cstring (NULL-terminated)
00220  * from a @c va_list (see svn_stringbuf_createf()).
00221  */
00222 svn_stringbuf_t *
00223 svn_stringbuf_createv(apr_pool_t *pool, const char *fmt, va_list ap)
00224   __attribute__((format(printf, 2, 0)));
00225 
00226 /** Make sure that the string @a str has at least @a minimum_size bytes of
00227  * space available in the memory block.
00228  *
00229  * (@a minimum_size should include space for the terminating NULL character.)
00230  */
00231 void
00232 svn_stringbuf_ensure(svn_stringbuf_t *str, apr_size_t minimum_size);
00233 
00234 /** Set a bytestring @a str to @a value */
00235 void
00236 svn_stringbuf_set(svn_stringbuf_t *str, const char *value);
00237 
00238 /** Set a bytestring @a str to empty (0 length). */
00239 void
00240 svn_stringbuf_setempty(svn_stringbuf_t *str);
00241 
00242 /** Return @c TRUE if a bytestring is empty (has length zero). */
00243 svn_boolean_t
00244 svn_stringbuf_isempty(const svn_stringbuf_t *str);
00245 
00246 /** Chop @a nbytes bytes off end of @a str, but not more than @a str->len. */
00247 void
00248 svn_stringbuf_chop(svn_stringbuf_t *str, apr_size_t nbytes);
00249 
00250 /** Fill bytestring @a str with character @a c. */
00251 void
00252 svn_stringbuf_fillchar(svn_stringbuf_t *str, unsigned char c);
00253 
00254 /** Append a single character @a byte onto @a targetstr.
00255  * This is an optimized version of svn_stringbuf_appendbytes()
00256  * that is much faster to call and execute. Gains vary with the ABI.
00257  * The advantages extend beyond the actual call because the reduced
00258  * register pressure allows for more optimization within the caller.
00259  *
00260  * reallocs if necessary. @a targetstr is affected, nothing else is.
00261  * @since New in 1.7.
00262  */
00263 void
00264 svn_stringbuf_appendbyte(svn_stringbuf_t *targetstr,
00265                          char byte);
00266 
00267 /** Append an array of bytes onto @a targetstr.
00268  *
00269  * reallocs if necessary. @a targetstr is affected, nothing else is.
00270  */
00271 void
00272 svn_stringbuf_appendbytes(svn_stringbuf_t *targetstr,
00273                           const char *bytes,
00274                           apr_size_t count);
00275 
00276 /** Append an @c svn_stringbuf_t onto @a targetstr.
00277  *
00278  * reallocs if necessary. @a targetstr is affected, nothing else is.
00279  */
00280 void
00281 svn_stringbuf_appendstr(svn_stringbuf_t *targetstr,
00282                         const svn_stringbuf_t *appendstr);
00283 
00284 /** Append a C string onto @a targetstr.
00285  *
00286  * reallocs if necessary. @a targetstr is affected, nothing else is.
00287  */
00288 void
00289 svn_stringbuf_appendcstr(svn_stringbuf_t *targetstr,
00290                          const char *cstr);
00291 
00292 /** Return a duplicate of @a original_string. */
00293 svn_stringbuf_t *
00294 svn_stringbuf_dup(const svn_stringbuf_t *original_string, apr_pool_t *pool);
00295 
00296 /** Return @c TRUE iff @a str1 and @a str2 have identical length and data. */
00297 svn_boolean_t
00298 svn_stringbuf_compare(const svn_stringbuf_t *str1,
00299                       const svn_stringbuf_t *str2);
00300 
00301 /** Return offset of first non-whitespace character in @a str, or return
00302  * @a str->len if none.
00303  */
00304 apr_size_t
00305 svn_stringbuf_first_non_whitespace(const svn_stringbuf_t *str);
00306 
00307 /** Strip whitespace from both sides of @a str (modified in place). */
00308 void
00309 svn_stringbuf_strip_whitespace(svn_stringbuf_t *str);
00310 
00311 /** Return position of last occurrence of @a ch in @a str, or return
00312  * @a str->len if no occurrence.
00313  */
00314 apr_size_t
00315 svn_stringbuf_find_char_backward(const svn_stringbuf_t *str, char ch);
00316 
00317 /** Return @c TRUE iff @a str1 and @a str2 have identical length and data. */
00318 svn_boolean_t
00319 svn_string_compare_stringbuf(const svn_string_t *str1,
00320                              const svn_stringbuf_t *str2);
00321 
00322 /** @} */
00323 
00324 
00325 /**
00326  * @defgroup svn_string_cstrings C string functions
00327  * @{
00328  */
00329 
00330 /** Divide @a input into substrings along @a sep_chars boundaries, return an
00331  * array of copies of those substrings (plain const char*), allocating both
00332  * the array and the copies in @a pool.
00333  *
00334  * None of the elements added to the array contain any of the
00335  * characters in @a sep_chars, and none of the new elements are empty
00336  * (thus, it is possible that the returned array will have length
00337  * zero).
00338  *
00339  * If @a chop_whitespace is TRUE, then remove leading and trailing
00340  * whitespace from the returned strings.
00341  */
00342 apr_array_header_t *
00343 svn_cstring_split(const char *input,
00344                   const char *sep_chars,
00345                   svn_boolean_t chop_whitespace,
00346                   apr_pool_t *pool);
00347 
00348 /** Like svn_cstring_split(), but append to existing @a array instead of
00349  * creating a new one.  Allocate the copied substrings in @a pool
00350  * (i.e., caller decides whether or not to pass @a array->pool as @a pool).
00351  */
00352 void
00353 svn_cstring_split_append(apr_array_header_t *array,
00354                          const char *input,
00355                          const char *sep_chars,
00356                          svn_boolean_t chop_whitespace,
00357                          apr_pool_t *pool);
00358 
00359 
00360 /** Return @c TRUE iff @a str matches any of the elements of @a list, a list
00361  * of zero or more glob patterns.
00362  */
00363 svn_boolean_t
00364 svn_cstring_match_glob_list(const char *str, const apr_array_header_t *list);
00365 
00366 /** Return @c TRUE iff @a str exactly matches any of the elements of @a list.
00367  *
00368  * @since new in 1.7
00369  */
00370 svn_boolean_t
00371 svn_cstring_match_list(const char *str, const apr_array_header_t *list);
00372 
00373 /**
00374  * Return the number of line breaks in @a msg, allowing any kind of newline
00375  * termination (CR, LF, CRLF, or LFCR), even inconsistent.
00376  *
00377  * @since New in 1.2.
00378  */
00379 int
00380 svn_cstring_count_newlines(const char *msg);
00381 
00382 /**
00383  * Return a cstring which is the concatenation of @a strings (an array
00384  * of char *) each followed by @a separator (that is, @a separator
00385  * will also end the resulting string).  Allocate the result in @a pool.
00386  * If @a strings is empty, then return the empty string.
00387  *
00388  * @since New in 1.2.
00389  */
00390 char *
00391 svn_cstring_join(const apr_array_header_t *strings,
00392                  const char *separator,
00393                  apr_pool_t *pool);
00394 
00395 /**
00396  * Compare two strings @a atr1 and @a atr2, treating case-equivalent
00397  * unaccented Latin (ASCII subset) letters as equal.
00398  *
00399  * Returns in integer greater than, equal to, or less than 0,
00400  * according to whether @a str1 is considered greater than, equal to,
00401  * or less than @a str2.
00402  *
00403  * @since New in 1.5.
00404  */
00405 int
00406 svn_cstring_casecmp(const char *str1, const char *str2);
00407 
00408 /**
00409  * Parse the C string @a str into a 64 bit number, and return it in @a *n.
00410  * Assume that the number is represented in base @a base.
00411  * Raise an error if conversion fails (e.g. due to overflow), or if the
00412  * converted number is smaller than @a minval or larger than @a maxval.
00413  *
00414  * @since New in 1.7.
00415  */
00416 svn_error_t *
00417 svn_cstring_strtoi64(apr_int64_t *n, const char *str,
00418                      apr_int64_t minval, apr_int64_t maxval,
00419                      int base);
00420 
00421 /**
00422  * Parse the C string @a str into a 64 bit number, and return it in @a *n.
00423  * Assume that the number is represented in base 10.
00424  * Raise an error if conversion fails (e.g. due to overflow).
00425  *
00426  * @since New in 1.7.
00427  */
00428 svn_error_t *
00429 svn_cstring_atoi64(apr_int64_t *n, const char *str);
00430 
00431 /**
00432  * Parse the C string @a str into a 32 bit number, and return it in @a *n.
00433  * Assume that the number is represented in base 10.
00434  * Raise an error if conversion fails (e.g. due to overflow).
00435  *
00436  * @since New in 1.7.
00437  */
00438 svn_error_t *
00439 svn_cstring_atoi(int *n, const char *str);
00440 
00441 /**
00442  * Parse the C string @a str into an unsigned 64 bit number, and return
00443  * it in @a *n. Assume that the number is represented in base @a base.
00444  * Raise an error if conversion fails (e.g. due to overflow), or if the
00445  * converted number is smaller than @a minval or larger than @a maxval.
00446  *
00447  * @since New in 1.7.
00448  */
00449 svn_error_t *
00450 svn_cstring_strtoui64(apr_uint64_t *n, const char *str,
00451                       apr_uint64_t minval, apr_uint64_t maxval,
00452                       int base);
00453 
00454 /**
00455  * Parse the C string @a str into an unsigned 64 bit number, and return
00456  * it in @a *n. Assume that the number is represented in base 10.
00457  * Raise an error if conversion fails (e.g. due to overflow).
00458  *
00459  * @since New in 1.7.
00460  */
00461 svn_error_t *
00462 svn_cstring_atoui64(apr_uint64_t *n, const char *str);
00463 
00464 /**
00465  * Parse the C string @a str into an unsigned 32 bit number, and return
00466  * it in @a *n. Assume that the number is represented in base 10.
00467  * Raise an error if conversion fails (e.g. due to overflow).
00468  *
00469  * @since New in 1.7.
00470  */
00471 svn_error_t *
00472 svn_cstring_atoui(unsigned int *n, const char *str);
00473 
00474 /** @} */
00475 
00476 /** @} */
00477 
00478 
00479 #ifdef __cplusplus
00480 }
00481 #endif /* __cplusplus */
00482 
00483 #endif  /* SVN_STRING_H */
 All Data Structures Files Functions Variables Typedefs Enumerations Enumerator Defines