/vendor/pcre/doc/html/pcreapi.html

http://github.com/feyeleanor/RubyGoLightly · HTML · 1985 lines · 1985 code · 0 blank · 0 comment · 0 complexity · 2a29757d96575a3e7260f91b4efec45a MD5 · raw file 

    

  1. <html>
    2. 

  2. <head>
    3. 

  3. <title>pcreapi specification</title>
    4. 

  4. </head>
    5. 

  5. <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
    6. 

  6. <h1>pcreapi man page</h1>
    7. 

  7. <p>
    8. 

  8. Return to the <a href="index.html">PCRE index page</a>.
    9. 

  9. </p>
    10. 

  10. <p>
    11. 

  11. This page is part of the PCRE HTML documentation. It was generated automatically
    12. 

  12. from the original man page. If there is any nonsense in it, please consult the
    13. 

  13. man page, in case the conversion went wrong.
    14. 

  14. <br>
    15. 

  15. <ul>
    16. 

  16. <li><a name="TOC1" href="#SEC1">PCRE NATIVE API</a>
    17. 

  17. <li><a name="TOC2" href="#SEC2">PCRE API OVERVIEW</a>
    18. 

  18. <li><a name="TOC3" href="#SEC3">NEWLINES</a>
    19. 

  19. <li><a name="TOC4" href="#SEC4">MULTITHREADING</a>
    20. 

  20. <li><a name="TOC5" href="#SEC5">SAVING PRECOMPILED PATTERNS FOR LATER USE</a>
    21. 

  21. <li><a name="TOC6" href="#SEC6">CHECKING BUILD-TIME OPTIONS</a>
    22. 

  22. <li><a name="TOC7" href="#SEC7">COMPILING A PATTERN</a>
    23. 

  23. <li><a name="TOC8" href="#SEC8">COMPILATION ERROR CODES</a>
    24. 

  24. <li><a name="TOC9" href="#SEC9">STUDYING A PATTERN</a>
    25. 

  25. <li><a name="TOC10" href="#SEC10">LOCALE SUPPORT</a>
    26. 

  26. <li><a name="TOC11" href="#SEC11">INFORMATION ABOUT A PATTERN</a>
    27. 

  27. <li><a name="TOC12" href="#SEC12">OBSOLETE INFO FUNCTION</a>
    28. 

  28. <li><a name="TOC13" href="#SEC13">REFERENCE COUNTS</a>
    29. 

  29. <li><a name="TOC14" href="#SEC14">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a>
    30. 

  30. <li><a name="TOC15" href="#SEC15">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a>
    31. 

  31. <li><a name="TOC16" href="#SEC16">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a>
    32. 

  32. <li><a name="TOC17" href="#SEC17">DUPLICATE SUBPATTERN NAMES</a>
    33. 

  33. <li><a name="TOC18" href="#SEC18">FINDING ALL POSSIBLE MATCHES</a>
    34. 

  34. <li><a name="TOC19" href="#SEC19">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a>
    35. 

  35. <li><a name="TOC20" href="#SEC20">SEE ALSO</a>
    36. 

  36. <li><a name="TOC21" href="#SEC21">AUTHOR</a>
    37. 

  37. <li><a name="TOC22" href="#SEC22">REVISION</a>
    38. 

  38. </ul>
    39. 

  39. <br><a name="SEC1" href="#TOC1">PCRE NATIVE API</a><br>
    40. 

  40. <P>
    41. 

  41. <b>#include &#60;pcre.h&#62;</b>
    42. 

  42. </P>
    43. 

  43. <P>
    44. 

  44. <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
    45. 

  45. <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
    46. 

  46. <b>const unsigned char *<i>tableptr</i>);</b>
    47. 

  47. </P>
    48. 

  48. <P>
    49. 

  49. <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
    50. 

  50. <b>int *<i>errorcodeptr</i>,</b>
    51. 

  51. <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
    52. 

  52. <b>const unsigned char *<i>tableptr</i>);</b>
    53. 

  53. </P>
    54. 

  54. <P>
    55. 

  55. <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i>,</b>
    56. 

  56. <b>const char **<i>errptr</i>);</b>
    57. 

  57. </P>
    58. 

  58. <P>
    59. 

  59. <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
    60. 

  60. <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
    61. 

  61. <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
    62. 

  62. </P>
    63. 

  63. <P>
    64. 

  64. <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
    65. 

  65. <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
    66. 

  66. <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
    67. 

  67. <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
    68. 

  68. </P>
    69. 

  69. <P>
    70. 

  70. <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
    71. 

  71. <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
    72. 

  72. <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
    73. 

  73. <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
    74. 

  74. </P>
    75. 

  75. <P>
    76. 

  76. <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
    77. 

  77. <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
    78. 

  78. <b>int <i>buffersize</i>);</b>
    79. 

  79. </P>
    80. 

  80. <P>
    81. 

  81. <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
    82. 

  82. <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
    83. 

  83. <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
    84. 

  84. <b>const char **<i>stringptr</i>);</b>
    85. 

  85. </P>
    86. 

  86. <P>
    87. 

  87. <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
    88. 

  88. <b>const char *<i>name</i>);</b>
    89. 

  89. </P>
    90. 

  90. <P>
    91. 

  91. <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
    92. 

  92. <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
    93. 

  93. </P>
    94. 

  94. <P>
    95. 

  95. <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
    96. 

  96. <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
    97. 

  97. <b>const char **<i>stringptr</i>);</b>
    98. 

  98. </P>
    99. 

  99. <P>
    100. 

  100. <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
    101. 

  101. <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
    102. 

  102. </P>
    103. 

  103. <P>
    104. 

  104. <b>void pcre_free_substring(const char *<i>stringptr</i>);</b>
    105. 

  105. </P>
    106. 

  106. <P>
    107. 

  107. <b>void pcre_free_substring_list(const char **<i>stringptr</i>);</b>
    108. 

  108. </P>
    109. 

  109. <P>
    110. 

  110. <b>const unsigned char *pcre_maketables(void);</b>
    111. 

  111. </P>
    112. 

  112. <P>
    113. 

  113. <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
    114. 

  114. <b>int <i>what</i>, void *<i>where</i>);</b>
    115. 

  115. </P>
    116. 

  116. <P>
    117. 

  117. <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
    118. 

  118. <b>*<i>firstcharptr</i>);</b>
    119. 

  119. </P>
    120. 

  120. <P>
    121. 

  121. <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
    122. 

  122. </P>
    123. 

  123. <P>
    124. 

  124. <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
    125. 

  125. </P>
    126. 

  126. <P>
    127. 

  127. <b>char *pcre_version(void);</b>
    128. 

  128. </P>
    129. 

  129. <P>
    130. 

  130. <b>void *(*pcre_malloc)(size_t);</b>
    131. 

  131. </P>
    132. 

  132. <P>
    133. 

  133. <b>void (*pcre_free)(void *);</b>
    134. 

  134. </P>
    135. 

  135. <P>
    136. 

  136. <b>void *(*pcre_stack_malloc)(size_t);</b>
    137. 

  137. </P>
    138. 

  138. <P>
    139. 

  139. <b>void (*pcre_stack_free)(void *);</b>
    140. 

  140. </P>
    141. 

  141. <P>
    142. 

  142. <b>int (*pcre_callout)(pcre_callout_block *);</b>
    143. 

  143. </P>
    144. 

  144. <br><a name="SEC2" href="#TOC1">PCRE API OVERVIEW</a><br>
    145. 

  145. <P>
    146. 

  146. PCRE has its own native API, which is described in this document. There are
    147. 

  147. also some wrapper functions that correspond to the POSIX regular expression
    148. 

  148. API. These are described in the
    149. 

  149. <a href="pcreposix.html"><b>pcreposix</b></a>
    150. 

  150. documentation. Both of these APIs define a set of C function calls. A C++
    151. 

  151. wrapper is distributed with PCRE. It is documented in the
    152. 

  152. <a href="pcrecpp.html"><b>pcrecpp</b></a>
    153. 

  153. page.
    154. 

  154. </P>
    155. 

  155. <P>
    156. 

  156. The native API C function prototypes are defined in the header file
    157. 

  157. <b>pcre.h</b>, and on Unix systems the library itself is called <b>libpcre</b>.
    158. 

  158. It can normally be accessed by adding <b>-lpcre</b> to the command for linking
    159. 

  159. an application that uses PCRE. The header file defines the macros PCRE_MAJOR
    160. 

  160. and PCRE_MINOR to contain the major and minor release numbers for the library.
    161. 

  161. Applications can use these to include support for different releases of PCRE.
    162. 

  162. </P>
    163. 

  163. <P>
    164. 

  164. The functions <b>pcre_compile()</b>, <b>pcre_compile2()</b>, <b>pcre_study()</b>,
    165. 

  165. and <b>pcre_exec()</b> are used for compiling and matching regular expressions
    166. 

  166. in a Perl-compatible manner. A sample program that demonstrates the simplest
    167. 

  167. way of using them is provided in the file called <i>pcredemo.c</i> in the source
    168. 

  168. distribution. The
    169. 

  169. <a href="pcresample.html"><b>pcresample</b></a>
    170. 

  170. documentation describes how to compile and run it.
    171. 

  171. </P>
    172. 

  172. <P>
    173. 

  173. A second matching function, <b>pcre_dfa_exec()</b>, which is not
    174. 

  174. Perl-compatible, is also provided. This uses a different algorithm for the
    175. 

  175. matching. The alternative algorithm finds all possible matches (at a given
    176. 

  176. point in the subject), and scans the subject just once. However, this algorithm
    177. 

  177. does not return captured substrings. A description of the two matching
    178. 

  178. algorithms and their advantages and disadvantages is given in the
    179. 

  179. <a href="pcrematching.html"><b>pcrematching</b></a>
    180. 

  180. documentation.
    181. 

  181. </P>
    182. 

  182. <P>
    183. 

  183. In addition to the main compiling and matching functions, there are convenience
    184. 

  184. functions for extracting captured substrings from a subject string that is
    185. 

  185. matched by <b>pcre_exec()</b>. They are:
    186. 

  186. <pre>
    187. 

  187.   <b>pcre_copy_substring()</b>
    188. 

  188.   <b>pcre_copy_named_substring()</b>
    189. 

  189.   <b>pcre_get_substring()</b>
    190. 

  190.   <b>pcre_get_named_substring()</b>
    191. 

  191.   <b>pcre_get_substring_list()</b>
    192. 

  192.   <b>pcre_get_stringnumber()</b>
    193. 

  193.   <b>pcre_get_stringtable_entries()</b>
    194. 

  194. </pre>
    195. 

  195. <b>pcre_free_substring()</b> and <b>pcre_free_substring_list()</b> are also
    196. 

  196. provided, to free the memory used for extracted strings.
    197. 

  197. </P>
    198. 

  198. <P>
    199. 

  199. The function <b>pcre_maketables()</b> is used to build a set of character tables
    200. 

  200. in the current locale for passing to <b>pcre_compile()</b>, <b>pcre_exec()</b>,
    201. 

  201. or <b>pcre_dfa_exec()</b>. This is an optional facility that is provided for
    202. 

  202. specialist use. Most commonly, no special tables are passed, in which case
    203. 

  203. internal tables that are generated when PCRE is built are used.
    204. 

  204. </P>
    205. 

  205. <P>
    206. 

  206. The function <b>pcre_fullinfo()</b> is used to find out information about a
    207. 

  207. compiled pattern; <b>pcre_info()</b> is an obsolete version that returns only
    208. 

  208. some of the available information, but is retained for backwards compatibility.
    209. 

  209. The function <b>pcre_version()</b> returns a pointer to a string containing the
    210. 

  210. version of PCRE and its date of release.
    211. 

  211. </P>
    212. 

  212. <P>
    213. 

  213. The function <b>pcre_refcount()</b> maintains a reference count in a data block
    214. 

  214. containing a compiled pattern. This is provided for the benefit of
    215. 

  215. object-oriented applications.
    216. 

  216. </P>
    217. 

  217. <P>
    218. 

  218. The global variables <b>pcre_malloc</b> and <b>pcre_free</b> initially contain
    219. 

  219. the entry points of the standard <b>malloc()</b> and <b>free()</b> functions,
    220. 

  220. respectively. PCRE calls the memory management functions via these variables,
    221. 

  221. so a calling program can replace them if it wishes to intercept the calls. This
    222. 

  222. should be done before calling any PCRE functions.
    223. 

  223. </P>
    224. 

  224. <P>
    225. 

  225. The global variables <b>pcre_stack_malloc</b> and <b>pcre_stack_free</b> are also
    226. 

  226. indirections to memory management functions. These special functions are used
    227. 

  227. only when PCRE is compiled to use the heap for remembering data, instead of
    228. 

  228. recursive function calls, when running the <b>pcre_exec()</b> function. See the
    229. 

  229. <a href="pcrebuild.html"><b>pcrebuild</b></a>
    230. 

  230. documentation for details of how to do this. It is a non-standard way of
    231. 

  231. building PCRE, for use in environments that have limited stacks. Because of the
    232. 

  232. greater use of memory management, it runs more slowly. Separate functions are
    233. 

  233. provided so that special-purpose external code can be used for this case. When
    234. 

  234. used, these functions are always called in a stack-like manner (last obtained,
    235. 

  235. first freed), and always for memory blocks of the same size. There is a
    236. 

  236. discussion about PCRE's stack usage in the
    237. 

  237. <a href="pcrestack.html"><b>pcrestack</b></a>
    238. 

  238. documentation.
    239. 

  239. </P>
    240. 

  240. <P>
    241. 

  241. The global variable <b>pcre_callout</b> initially contains NULL. It can be set
    242. 

  242. by the caller to a "callout" function, which PCRE will then call at specified
    243. 

  243. points during a matching operation. Details are given in the
    244. 

  244. <a href="pcrecallout.html"><b>pcrecallout</b></a>
    245. 

  245. documentation.
    246. 

  246. <a name="newlines"></a></P>
    247. 

  247. <br><a name="SEC3" href="#TOC1">NEWLINES</a><br>
    248. 

  248. <P>
    249. 

  249. PCRE supports five different conventions for indicating line breaks in
    250. 

  250. strings: a single CR (carriage return) character, a single LF (linefeed)
    251. 

  251. character, the two-character sequence CRLF, any of the three preceding, or any
    252. 

  252. Unicode newline sequence. The Unicode newline sequences are the three just
    253. 

  253. mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
    254. 

  254. U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
    255. 

  255. (paragraph separator, U+2029).
    256. 

  256. </P>
    257. 

  257. <P>
    258. 

  258. Each of the first three conventions is used by at least one operating system as
    259. 

  259. its standard newline sequence. When PCRE is built, a default can be specified.
    260. 

  260. The default default is LF, which is the Unix standard. When PCRE is run, the
    261. 

  261. default can be overridden, either when a pattern is compiled, or when it is
    262. 

  262. matched.
    263. 

  263. </P>
    264. 

  264. <P>
    265. 

  265. At compile time, the newline convention can be specified by the <i>options</i>
    266. 

  266. argument of <b>pcre_compile()</b>, or it can be specified by special text at the
    267. 

  267. start of the pattern itself; this overrides any other settings. See the
    268. 

  268. <a href="pcrepattern.html"><b>pcrepattern</b></a>
    269. 

  269. page for details of the special character sequences.
    270. 

  270. </P>
    271. 

  271. <P>
    272. 

  272. In the PCRE documentation the word "newline" is used to mean "the character or
    273. 

  273. pair of characters that indicate a line break". The choice of newline
    274. 

  274. convention affects the handling of the dot, circumflex, and dollar
    275. 

  275. metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
    276. 

  276. recognized line ending sequence, the match position advancement for a
    277. 

  277. non-anchored pattern. There is more detail about this in the
    278. 

  278. <a href="#execoptions">section on <b>pcre_exec()</b> options</a>
    279. 

  279. below.
    280. 

  280. </P>
    281. 

  281. <P>
    282. 

  282. The choice of newline convention does not affect the interpretation of
    283. 

  283. the \n or \r escape sequences, nor does it affect what \R matches, which is
    284. 

  284. controlled in a similar way, but by separate options.
    285. 

  285. </P>
    286. 

  286. <br><a name="SEC4" href="#TOC1">MULTITHREADING</a><br>
    287. 

  287. <P>
    288. 

  288. The PCRE functions can be used in multi-threading applications, with the
    289. 

  289. proviso that the memory management functions pointed to by <b>pcre_malloc</b>,
    290. 

  290. <b>pcre_free</b>, <b>pcre_stack_malloc</b>, and <b>pcre_stack_free</b>, and the
    291. 

  291. callout function pointed to by <b>pcre_callout</b>, are shared by all threads.
    292. 

  292. </P>
    293. 

  293. <P>
    294. 

  294. The compiled form of a regular expression is not altered during matching, so
    295. 

  295. the same compiled pattern can safely be used by several threads at once.
    296. 

  296. </P>
    297. 

  297. <br><a name="SEC5" href="#TOC1">SAVING PRECOMPILED PATTERNS FOR LATER USE</a><br>
    298. 

  298. <P>
    299. 

  299. The compiled form of a regular expression can be saved and re-used at a later
    300. 

  300. time, possibly by a different program, and even on a host other than the one on
    301. 

  301. which it was compiled. Details are given in the
    302. 

  302. <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
    303. 

  303. documentation. However, compiling a regular expression with one version of PCRE
    304. 

  304. for use with a different version is not guaranteed to work and may cause
    305. 

  305. crashes.
    306. 

  306. </P>
    307. 

  307. <br><a name="SEC6" href="#TOC1">CHECKING BUILD-TIME OPTIONS</a><br>
    308. 

  308. <P>
    309. 

  309. <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
    310. 

  310. </P>
    311. 

  311. <P>
    312. 

  312. The function <b>pcre_config()</b> makes it possible for a PCRE client to
    313. 

  313. discover which optional features have been compiled into the PCRE library. The
    314. 

  314. <a href="pcrebuild.html"><b>pcrebuild</b></a>
    315. 

  315. documentation has more details about these optional features.
    316. 

  316. </P>
    317. 

  317. <P>
    318. 

  318. The first argument for <b>pcre_config()</b> is an integer, specifying which
    319. 

  319. information is required; the second argument is a pointer to a variable into
    320. 

  320. which the information is placed. The following information is available:
    321. 

  321. <pre>
    322. 

  322.   PCRE_CONFIG_UTF8
    323. 

  323. </pre>
    324. 

  324. The output is an integer that is set to one if UTF-8 support is available;
    325. 

  325. otherwise it is set to zero.
    326. 

  326. <pre>
    327. 

  327.   PCRE_CONFIG_UNICODE_PROPERTIES
    328. 

  328. </pre>
    329. 

  329. The output is an integer that is set to one if support for Unicode character
    330. 

  330. properties is available; otherwise it is set to zero.
    331. 

  331. <pre>
    332. 

  332.   PCRE_CONFIG_NEWLINE
    333. 

  333. </pre>
    334. 

  334. The output is an integer whose value specifies the default character sequence
    335. 

  335. that is recognized as meaning "newline". The four values that are supported
    336. 

  336. are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY. The
    337. 

  337. default should normally be the standard sequence for your operating system.
    338. 

  338. <pre>
    339. 

  339.   PCRE_CONFIG_BSR
    340. 

  340. </pre>
    341. 

  341. The output is an integer whose value indicates what character sequences the \R
    342. 

  342. escape sequence matches by default. A value of 0 means that \R matches any
    343. 

  343. Unicode line ending sequence; a value of 1 means that \R matches only CR, LF,
    344. 

  344. or CRLF. The default can be overridden when a pattern is compiled or matched.
    345. 

  345. <pre>
    346. 

  346.   PCRE_CONFIG_LINK_SIZE
    347. 

  347. </pre>
    348. 

  348. The output is an integer that contains the number of bytes used for internal
    349. 

  349. linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
    350. 

  350. allow larger regular expressions to be compiled, at the expense of slower
    351. 

  351. matching. The default value of 2 is sufficient for all but the most massive
    352. 

  352. patterns, since it allows the compiled pattern to be up to 64K in size.
    353. 

  353. <pre>
    354. 

  354.   PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
    355. 

  355. </pre>
    356. 

  356. The output is an integer that contains the threshold above which the POSIX
    357. 

  357. interface uses <b>malloc()</b> for output vectors. Further details are given in
    358. 

  358. the
    359. 

  359. <a href="pcreposix.html"><b>pcreposix</b></a>
    360. 

  360. documentation.
    361. 

  361. <pre>
    362. 

  362.   PCRE_CONFIG_MATCH_LIMIT
    363. 

  363. </pre>
    364. 

  364. The output is an integer that gives the default limit for the number of
    365. 

  365. internal matching function calls in a <b>pcre_exec()</b> execution. Further
    366. 

  366. details are given with <b>pcre_exec()</b> below.
    367. 

  367. <pre>
    368. 

  368.   PCRE_CONFIG_MATCH_LIMIT_RECURSION
    369. 

  369. </pre>
    370. 

  370. The output is an integer that gives the default limit for the depth of
    371. 

  371. recursion when calling the internal matching function in a <b>pcre_exec()</b>
    372. 

  372. execution. Further details are given with <b>pcre_exec()</b> below.
    373. 

  373. <pre>
    374. 

  374.   PCRE_CONFIG_STACKRECURSE
    375. 

  375. </pre>
    376. 

  376. The output is an integer that is set to one if internal recursion when running
    377. 

  377. <b>pcre_exec()</b> is implemented by recursive function calls that use the stack
    378. 

  378. to remember their state. This is the usual way that PCRE is compiled. The
    379. 

  379. output is zero if PCRE was compiled to use blocks of data on the heap instead
    380. 

  380. of recursive function calls. In this case, <b>pcre_stack_malloc</b> and
    381. 

  381. <b>pcre_stack_free</b> are called to manage memory blocks on the heap, thus
    382. 

  382. avoiding the use of the stack.
    383. 

  383. </P>
    384. 

  384. <br><a name="SEC7" href="#TOC1">COMPILING A PATTERN</a><br>
    385. 

  385. <P>
    386. 

  386. <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
    387. 

  387. <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
    388. 

  388. <b>const unsigned char *<i>tableptr</i>);</b>
    389. 

  389. <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
    390. 

  390. <b>int *<i>errorcodeptr</i>,</b>
    391. 

  391. <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
    392. 

  392. <b>const unsigned char *<i>tableptr</i>);</b>
    393. 

  393. </P>
    394. 

  394. <P>
    395. 

  395. Either of the functions <b>pcre_compile()</b> or <b>pcre_compile2()</b> can be
    396. 

  396. called to compile a pattern into an internal form. The only difference between
    397. 

  397. the two interfaces is that <b>pcre_compile2()</b> has an additional argument,
    398. 

  398. <i>errorcodeptr</i>, via which a numerical error code can be returned.
    399. 

  399. </P>
    400. 

  400. <P>
    401. 

  401. The pattern is a C string terminated by a binary zero, and is passed in the
    402. 

  402. <i>pattern</i> argument. A pointer to a single block of memory that is obtained
    403. 

  403. via <b>pcre_malloc</b> is returned. This contains the compiled code and related
    404. 

  404. data. The <b>pcre</b> type is defined for the returned block; this is a typedef
    405. 

  405. for a structure whose contents are not externally defined. It is up to the
    406. 

  406. caller to free the memory (via <b>pcre_free</b>) when it is no longer required.
    407. 

  407. </P>
    408. 

  408. <P>
    409. 

  409. Although the compiled code of a PCRE regex is relocatable, that is, it does not
    410. 

  410. depend on memory location, the complete <b>pcre</b> data block is not
    411. 

  411. fully relocatable, because it may contain a copy of the <i>tableptr</i>
    412. 

  412. argument, which is an address (see below).
    413. 

  413. </P>
    414. 

  414. <P>
    415. 

  415. The <i>options</i> argument contains various bit settings that affect the
    416. 

  416. compilation. It should be zero if no options are required. The available
    417. 

  417. options are described below. Some of them, in particular, those that are
    418. 

  418. compatible with Perl, can also be set and unset from within the pattern (see
    419. 

  419. the detailed description in the
    420. 

  420. <a href="pcrepattern.html"><b>pcrepattern</b></a>
    421. 

  421. documentation). For these options, the contents of the <i>options</i> argument
    422. 

  422. specifies their initial settings at the start of compilation and execution. The
    423. 

  423. PCRE_ANCHORED and PCRE_NEWLINE_<i>xxx</i> options can be set at the time of
    424. 

  424. matching as well as at compile time.
    425. 

  425. </P>
    426. 

  426. <P>
    427. 

  427. If <i>errptr</i> is NULL, <b>pcre_compile()</b> returns NULL immediately.
    428. 

  428. Otherwise, if compilation of a pattern fails, <b>pcre_compile()</b> returns
    429. 

  429. NULL, and sets the variable pointed to by <i>errptr</i> to point to a textual
    430. 

  430. error message. This is a static string that is part of the library. You must
    431. 

  431. not try to free it. The offset from the start of the pattern to the character
    432. 

  432. where the error was discovered is placed in the variable pointed to by
    433. 

  433. <i>erroffset</i>, which must not be NULL. If it is, an immediate error is given.
    434. 

  434. </P>
    435. 

  435. <P>
    436. 

  436. If <b>pcre_compile2()</b> is used instead of <b>pcre_compile()</b>, and the
    437. 

  437. <i>errorcodeptr</i> argument is not NULL, a non-zero error code number is
    438. 

  438. returned via this argument in the event of an error. This is in addition to the
    439. 

  439. textual error message. Error codes and messages are listed below.
    440. 

  440. </P>
    441. 

  441. <P>
    442. 

  442. If the final argument, <i>tableptr</i>, is NULL, PCRE uses a default set of
    443. 

  443. character tables that are built when PCRE is compiled, using the default C
    444. 

  444. locale. Otherwise, <i>tableptr</i> must be an address that is the result of a
    445. 

  445. call to <b>pcre_maketables()</b>. This value is stored with the compiled
    446. 

  446. pattern, and used again by <b>pcre_exec()</b>, unless another table pointer is
    447. 

  447. passed to it. For more discussion, see the section on locale support below.
    448. 

  448. </P>
    449. 

  449. <P>
    450. 

  450. This code fragment shows a typical straightforward call to <b>pcre_compile()</b>:
    451. 

  451. <pre>
    452. 

  452.   pcre *re;
    453. 

  453.   const char *error;
    454. 

  454.   int erroffset;
    455. 

  455.   re = pcre_compile(
    456. 

  456.     "^A.*Z",          /* the pattern */
    457. 

  457.     0,                /* default options */
    458. 

  458.     &error,           /* for error message */
    459. 

  459.     &erroffset,       /* for error offset */
    460. 

  460.     NULL);            /* use default character tables */
    461. 

  461. </pre>
    462. 

  462. The following names for option bits are defined in the <b>pcre.h</b> header
    463. 

  463. file:
    464. 

  464. <pre>
    465. 

  465.   PCRE_ANCHORED
    466. 

  466. </pre>
    467. 

  467. If this bit is set, the pattern is forced to be "anchored", that is, it is
    468. 

  468. constrained to match only at the first matching point in the string that is
    469. 

  469. being searched (the "subject string"). This effect can also be achieved by
    470. 

  470. appropriate constructs in the pattern itself, which is the only way to do it in
    471. 

  471. Perl.
    472. 

  472. <pre>
    473. 

  473.   PCRE_AUTO_CALLOUT
    474. 

  474. </pre>
    475. 

  475. If this bit is set, <b>pcre_compile()</b> automatically inserts callout items,
    476. 

  476. all with number 255, before each pattern item. For discussion of the callout
    477. 

  477. facility, see the
    478. 

  478. <a href="pcrecallout.html"><b>pcrecallout</b></a>
    479. 

  479. documentation.
    480. 

  480. <pre>
    481. 

  481.   PCRE_BSR_ANYCRLF
    482. 

  482.   PCRE_BSR_UNICODE
    483. 

  483. </pre>
    484. 

  484. These options (which are mutually exclusive) control what the \R escape
    485. 

  485. sequence matches. The choice is either to match only CR, LF, or CRLF, or to
    486. 

  486. match any Unicode newline sequence. The default is specified when PCRE is
    487. 

  487. built. It can be overridden from within the pattern, or by setting an option
    488. 

  488. when a compiled pattern is matched.
    489. 

  489. <pre>
    490. 

  490.   PCRE_CASELESS
    491. 

  491. </pre>
    492. 

  492. If this bit is set, letters in the pattern match both upper and lower case
    493. 

  493. letters. It is equivalent to Perl's /i option, and it can be changed within a
    494. 

  494. pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
    495. 

  495. concept of case for characters whose values are less than 128, so caseless
    496. 

  496. matching is always possible. For characters with higher values, the concept of
    497. 

  497. case is supported if PCRE is compiled with Unicode property support, but not
    498. 

  498. otherwise. If you want to use caseless matching for characters 128 and above,
    499. 

  499. you must ensure that PCRE is compiled with Unicode property support as well as
    500. 

  500. with UTF-8 support.
    501. 

  501. <pre>
    502. 

  502.   PCRE_DOLLAR_ENDONLY
    503. 

  503. </pre>
    504. 

  504. If this bit is set, a dollar metacharacter in the pattern matches only at the
    505. 

  505. end of the subject string. Without this option, a dollar also matches
    506. 

  506. immediately before a newline at the end of the string (but not before any other
    507. 

  507. newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
    508. 

  508. There is no equivalent to this option in Perl, and no way to set it within a
    509. 

  509. pattern.
    510. 

  510. <pre>
    511. 

  511.   PCRE_DOTALL
    512. 

  512. </pre>
    513. 

  513. If this bit is set, a dot metacharater in the pattern matches all characters,
    514. 

  514. including those that indicate newline. Without it, a dot does not match when
    515. 

  515. the current position is at a newline. This option is equivalent to Perl's /s
    516. 

  516. option, and it can be changed within a pattern by a (?s) option setting. A
    517. 

  517. negative class such as [^a] always matches newline characters, independent of
    518. 

  518. the setting of this option.
    519. 

  519. <pre>
    520. 

  520.   PCRE_DUPNAMES
    521. 

  521. </pre>
    522. 

  522. If this bit is set, names used to identify capturing subpatterns need not be
    523. 

  523. unique. This can be helpful for certain types of pattern when it is known that
    524. 

  524. only one instance of the named subpattern can ever be matched. There are more
    525. 

  525. details of named subpatterns below; see also the
    526. 

  526. <a href="pcrepattern.html"><b>pcrepattern</b></a>
    527. 

  527. documentation.
    528. 

  528. <pre>
    529. 

  529.   PCRE_EXTENDED
    530. 

  530. </pre>
    531. 

  531. If this bit is set, whitespace data characters in the pattern are totally
    532. 

  532. ignored except when escaped or inside a character class. Whitespace does not
    533. 

  533. include the VT character (code 11). In addition, characters between an
    534. 

  534. unescaped # outside a character class and the next newline, inclusive, are also
    535. 

  535. ignored. This is equivalent to Perl's /x option, and it can be changed within a
    536. 

  536. pattern by a (?x) option setting.
    537. 

  537. </P>
    538. 

  538. <P>
    539. 

  539. This option makes it possible to include comments inside complicated patterns.
    540. 

  540. Note, however, that this applies only to data characters. Whitespace characters
    541. 

  541. may never appear within special character sequences in a pattern, for example
    542. 

  542. within the sequence (?( which introduces a conditional subpattern.
    543. 

  543. <pre>
    544. 

  544.   PCRE_EXTRA
    545. 

  545. </pre>
    546. 

  546. This option was invented in order to turn on additional functionality of PCRE
    547. 

  547. that is incompatible with Perl, but it is currently of very little use. When
    548. 

  548. set, any backslash in a pattern that is followed by a letter that has no
    549. 

  549. special meaning causes an error, thus reserving these combinations for future
    550. 

  550. expansion. By default, as in Perl, a backslash followed by a letter with no
    551. 

  551. special meaning is treated as a literal. (Perl can, however, be persuaded to
    552. 

  552. give a warning for this.) There are at present no other features controlled by
    553. 

  553. this option. It can also be set by a (?X) option setting within a pattern.
    554. 

  554. <pre>
    555. 

  555.   PCRE_FIRSTLINE
    556. 

  556. </pre>
    557. 

  557. If this option is set, an unanchored pattern is required to match before or at
    558. 

  558. the first newline in the subject string, though the matched text may continue
    559. 

  559. over the newline.
    560. 

  560. <pre>
    561. 

  561.   PCRE_JAVASCRIPT_COMPAT
    562. 

  562. </pre>
    563. 

  563. If this option is set, PCRE's behaviour is changed in some ways so that it is
    564. 

  564. compatible with JavaScript rather than Perl. The changes are as follows:
    565. 

  565. </P>
    566. 

  566. <P>
    567. 

  567. (1) A lone closing square bracket in a pattern causes a compile-time error,
    568. 

  568. because this is illegal in JavaScript (by default it is treated as a data
    569. 

  569. character). Thus, the pattern AB]CD becomes illegal when this option is set.
    570. 

  570. </P>
    571. 

  571. <P>
    572. 

  572. (2) At run time, a back reference to an unset subpattern group matches an empty
    573. 

  573. string (by default this causes the current matching alternative to fail). A
    574. 

  574. pattern such as (\1)(a) succeeds when this option is set (assuming it can find
    575. 

  575. an "a" in the subject), whereas it fails by default, for Perl compatibility.
    576. 

  576. <pre>
    577. 

  577.   PCRE_MULTILINE
    578. 

  578. </pre>
    579. 

  579. By default, PCRE treats the subject string as consisting of a single line of
    580. 

  580. characters (even if it actually contains newlines). The "start of line"
    581. 

  581. metacharacter (^) matches only at the start of the string, while the "end of
    582. 

  582. line" metacharacter ($) matches only at the end of the string, or before a
    583. 

  583. terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
    584. 

  584. Perl.
    585. 

  585. </P>
    586. 

  586. <P>
    587. 

  587. When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
    588. 

  588. match immediately following or immediately before internal newlines in the
    589. 

  589. subject string, respectively, as well as at the very start and end. This is
    590. 

  590. equivalent to Perl's /m option, and it can be changed within a pattern by a
    591. 

  591. (?m) option setting. If there are no newlines in a subject string, or no
    592. 

  592. occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
    593. 

  593. <pre>
    594. 

  594.   PCRE_NEWLINE_CR
    595. 

  595.   PCRE_NEWLINE_LF
    596. 

  596.   PCRE_NEWLINE_CRLF
    597. 

  597.   PCRE_NEWLINE_ANYCRLF
    598. 

  598.   PCRE_NEWLINE_ANY
    599. 

  599. </pre>
    600. 

  600. These options override the default newline definition that was chosen when PCRE
    601. 

  601. was built. Setting the first or the second specifies that a newline is
    602. 

  602. indicated by a single character (CR or LF, respectively). Setting
    603. 

  603. PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
    604. 

  604. CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
    605. 

  605. preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
    606. 

  606. that any Unicode newline sequence should be recognized. The Unicode newline
    607. 

  607. sequences are the three just mentioned, plus the single characters VT (vertical
    608. 

  608. tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
    609. 

  609. separator, U+2028), and PS (paragraph separator, U+2029). The last two are
    610. 

  610. recognized only in UTF-8 mode.
    611. 

  611. </P>
    612. 

  612. <P>
    613. 

  613. The newline setting in the options word uses three bits that are treated
    614. 

  614. as a number, giving eight possibilities. Currently only six are used (default
    615. 

  615. plus the five values above). This means that if you set more than one newline
    616. 

  616. option, the combination may or may not be sensible. For example,
    617. 

  617. PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
    618. 

  618. other combinations may yield unused numbers and cause an error.
    619. 

  619. </P>
    620. 

  620. <P>
    621. 

  621. The only time that a line break is specially recognized when compiling a
    622. 

  622. pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
    623. 

  623. class is encountered. This indicates a comment that lasts until after the next
    624. 

  624. line break sequence. In other circumstances, line break sequences are treated
    625. 

  625. as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
    626. 

  626. as whitespace characters and are therefore ignored.
    627. 

  627. </P>
    628. 

  628. <P>
    629. 

  629. The newline option that is set at compile time becomes the default that is used
    630. 

  630. for <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, but it can be overridden.
    631. 

  631. <pre>
    632. 

  632.   PCRE_NO_AUTO_CAPTURE
    633. 

  633. </pre>
    634. 

  634. If this option is set, it disables the use of numbered capturing parentheses in
    635. 

  635. the pattern. Any opening parenthesis that is not followed by ? behaves as if it
    636. 

  636. were followed by ?: but named parentheses can still be used for capturing (and
    637. 

  637. they acquire numbers in the usual way). There is no equivalent of this option
    638. 

  638. in Perl.
    639. 

  639. <pre>
    640. 

  640.   PCRE_UNGREEDY
    641. 

  641. </pre>
    642. 

  642. This option inverts the "greediness" of the quantifiers so that they are not
    643. 

  643. greedy by default, but become greedy if followed by "?". It is not compatible
    644. 

  644. with Perl. It can also be set by a (?U) option setting within the pattern.
    645. 

  645. <pre>
    646. 

  646.   PCRE_UTF8
    647. 

  647. </pre>
    648. 

  648. This option causes PCRE to regard both the pattern and the subject as strings
    649. 

  649. of UTF-8 characters instead of single-byte character strings. However, it is
    650. 

  650. available only when PCRE is built to include UTF-8 support. If not, the use
    651. 

  651. of this option provokes an error. Details of how this option changes the
    652. 

  652. behaviour of PCRE are given in the
    653. 

  653. <a href="pcre.html#utf8support">section on UTF-8 support</a>
    654. 

  654. in the main
    655. 

  655. <a href="pcre.html"><b>pcre</b></a>
    656. 

  656. page.
    657. 

  657. <pre>
    658. 

  658.   PCRE_NO_UTF8_CHECK
    659. 

  659. </pre>
    660. 

  660. When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
    661. 

  661. automatically checked. There is a discussion about the
    662. 

  662. <a href="pcre.html#utf8strings">validity of UTF-8 strings</a>
    663. 

  663. in the main
    664. 

  664. <a href="pcre.html"><b>pcre</b></a>
    665. 

  665. page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_compile()</b>
    666. 

  666. returns an error. If you already know that your pattern is valid, and you want
    667. 

  667. to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
    668. 

  668. option. When it is set, the effect of passing an invalid UTF-8 string as a
    669. 

  669. pattern is undefined. It may cause your program to crash. Note that this option
    670. 

  670. can also be passed to <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, to suppress
    671. 

  671. the UTF-8 validity checking of subject strings.
    672. 

  672. </P>
    673. 

  673. <br><a name="SEC8" href="#TOC1">COMPILATION ERROR CODES</a><br>
    674. 

  674. <P>
    675. 

  675. The following table lists the error codes than may be returned by
    676. 

  676. <b>pcre_compile2()</b>, along with the error messages that may be returned by
    677. 

  677. both compiling functions. As PCRE has developed, some error codes have fallen
    678. 

  678. out of use. To avoid confusion, they have not been re-used.
    679. 

  679. <pre>
    680. 

  680.    0  no error
    681. 

  681.    1  \ at end of pattern
    682. 

  682.    2  \c at end of pattern
    683. 

  683.    3  unrecognized character follows \
    684. 

  684.    4  numbers out of order in {} quantifier
    685. 

  685.    5  number too big in {} quantifier
    686. 

  686.    6  missing terminating ] for character class
    687. 

  687.    7  invalid escape sequence in character class
    688. 

  688.    8  range out of order in character class
    689. 

  689.    9  nothing to repeat
    690. 

  690.   10  [this code is not in use]
    691. 

  691.   11  internal error: unexpected repeat
    692. 

  692.   12  unrecognized character after (? or (?-
    693. 

  693.   13  POSIX named classes are supported only within a class
    694. 

  694.   14  missing )
    695. 

  695.   15  reference to non-existent subpattern
    696. 

  696.   16  erroffset passed as NULL
    697. 

  697.   17  unknown option bit(s) set
    698. 

  698.   18  missing ) after comment
    699. 

  699.   19  [this code is not in use]
    700. 

  700.   20  regular expression is too large
    701. 

  701.   21  failed to get memory
    702. 

  702.   22  unmatched parentheses
    703. 

  703.   23  internal error: code overflow
    704. 

  704.   24  unrecognized character after (?&#60;
    705. 

  705.   25  lookbehind assertion is not fixed length
    706. 

  706.   26  malformed number or name after (?(
    707. 

  707.   27  conditional group contains more than two branches
    708. 

  708.   28  assertion expected after (?(
    709. 

  709.   29  (?R or (?[+-]digits must be followed by )
    710. 

  710.   30  unknown POSIX class name
    711. 

  711.   31  POSIX collating elements are not supported
    712. 

  712.   32  this version of PCRE is not compiled with PCRE_UTF8 support
    713. 

  713.   33  [this code is not in use]
    714. 

  714.   34  character value in \x{...} sequence is too large
    715. 

  715.   35  invalid condition (?(0)
    716. 

  716.   36  \C not allowed in lookbehind assertion
    717. 

  717.   37  PCRE does not support \L, \l, \N, \U, or \u
    718. 

  718.   38  number after (?C is &#62; 255
    719. 

  719.   39  closing ) for (?C expected
    720. 

  720.   40  recursive call could loop indefinitely
    721. 

  721.   41  unrecognized character after (?P
    722. 

  722.   42  syntax error in subpattern name (missing terminator)
    723. 

  723.   43  two named subpatterns have the same name
    724. 

  724.   44  invalid UTF-8 string
    725. 

  725.   45  support for \P, \p, and \X has not been compiled
    726. 

  726.   46  malformed \P or \p sequence
    727. 

  727.   47  unknown property name after \P or \p
    728. 

  728.   48  subpattern name is too long (maximum 32 characters)
    729. 

  729.   49  too many named subpatterns (maximum 10000)
    730. 

  730.   50  [this code is not in use]
    731. 

  731.   51  octal value is greater than \377 (not in UTF-8 mode)
    732. 

  732.   52  internal error: overran compiling workspace
    733. 

  733.   53  internal error: previously-checked referenced subpattern not found
    734. 

  734.   54  DEFINE group contains more than one branch
    735. 

  735.   55  repeating a DEFINE group is not allowed
    736. 

  736.   56  inconsistent NEWLINE options
    737. 

  737.   57  \g is not followed by a braced, angle-bracketed, or quoted
    738. 

  738.         name/number or by a plain number
    739. 

  739.   58  a numbered reference must not be zero
    740. 

  740.   59  (*VERB) with an argument is not supported
    741. 

  741.   60  (*VERB) not recognized
    742. 

  742.   61  number is too big
    743. 

  743.   62  subpattern name expected
    744. 

  744.   63  digit expected after (?+
    745. 

  745.   64  ] is an invalid data character in JavaScript compatibility mode
    746. 

  746. </pre>
    747. 

  747. The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
    748. 

  748. be used if the limits were changed when PCRE was built.
    749. 

  749. </P>
    750. 

  750. <br><a name="SEC9" href="#TOC1">STUDYING A PATTERN</a><br>
    751. 

  751. <P>
    752. 

  752. <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i></b>
    753. 

  753. <b>const char **<i>errptr</i>);</b>
    754. 

  754. </P>
    755. 

  755. <P>
    756. 

  756. If a compiled pattern is going to be used several times, it is worth spending
    757. 

  757. more time analyzing it in order to speed up the time taken for matching. The
    758. 

  758. function <b>pcre_study()</b> takes a pointer to a compiled pattern as its first
    759. 

  759. argument. If studying the pattern produces additional information that will
    760. 

  760. help speed up matching, <b>pcre_study()</b> returns a pointer to a
    761. 

  761. <b>pcre_extra</b> block, in which the <i>study_data</i> field points to the
    762. 

  762. results of the study.
    763. 

  763. </P>
    764. 

  764. <P>
    765. 

  765. The returned value from <b>pcre_study()</b> can be passed directly to
    766. 

  766. <b>pcre_exec()</b>. However, a <b>pcre_extra</b> block also contains other
    767. 

  767. fields that can be set by the caller before the block is passed; these are
    768. 

  768. described
    769. 

  769. <a href="#extradata">below</a>
    770. 

  770. in the section on matching a pattern.
    771. 

  771. </P>
    772. 

  772. <P>
    773. 

  773. If studying the pattern does not produce any additional information
    774. 

  774. <b>pcre_study()</b> returns NULL. In that circumstance, if the calling program
    775. 

  775. wants to pass any of the other fields to <b>pcre_exec()</b>, it must set up its
    776. 

  776. own <b>pcre_extra</b> block.
    777. 

  777. </P>
    778. 

  778. <P>
    779. 

  779. The second argument of <b>pcre_study()</b> contains option bits. At present, no
    780. 

  780. options are defined, and this argument should always be zero.
    781. 

  781. </P>
    782. 

  782. <P>
    783. 

  783. The third argument for <b>pcre_study()</b> is a pointer for an error message. If
    784. 

  784. studying succeeds (even if no data is returned), the variable it points to is
    785. 

  785. set to NULL. Otherwise it is set to point to a textual error message. This is a
    786. 

  786. static string that is part of the library. You must not try to free it. You
    787. 

  787. should test the error pointer for NULL after calling <b>pcre_study()</b>, to be
    788. 

  788. sure that it has run successfully.
    789. 

  789. </P>
    790. 

  790. <P>
    791. 

  791. This is a typical call to <b>pcre_study</b>():
    792. 

  792. <pre>
    793. 

  793.   pcre_extra *pe;
    794. 

  794.   pe = pcre_study(
    795. 

  795.     re,             /* result of pcre_compile() */
    796. 

  796.     0,              /* no options exist */
    797. 

  797.     &error);        /* set to NULL or points to a message */
    798. 

  798. </pre>
    799. 

  799. At present, studying a pattern is useful only for non-anchored patterns that do
    800. 

  800. not have a single fixed starting character. A bitmap of possible starting
    801. 

  801. bytes is created.
    802. 

  802. <a name="localesupport"></a></P>
    803. 

  803. <br><a name="SEC10" href="#TOC1">LOCALE SUPPORT</a><br>
    804. 

  804. <P>
    805. 

  805. PCRE handles caseless matching, and determines whether characters are letters,
    806. 

  806. digits, or whatever, by reference to a set of tables, indexed by character
    807. 

  807. value. When running in UTF-8 mode, this applies only to characters with codes
    808. 

  808. less than 128. Higher-valued codes never match escapes such as \w or \d, but
    809. 

  809. can be tested with \p if PCRE is built with Unicode character property
    810. 

  810. support. The use of locales with Unicode is discouraged. If you are handling
    811. 

  811. characters with codes greater than 128, you should either use UTF-8 and
    812. 

  812. Unicode, or use locales, but not try to mix the two.
    813. 

  813. </P>
    814. 

  814. <P>
    815. 

  815. PCRE contains an internal set of tables that are used when the final argument
    816. 

  816. of <b>pcre_compile()</b> is NULL. These are sufficient for many applications.
    817. 

  817. Normally, the internal tables recognize only ASCII characters. However, when
    818. 

  818. PCRE is built, it is possible to cause the internal tables to be rebuilt in the
    819. 

  819. default "C" locale of the local system, which may cause them to be different.
    820. 

  820. </P>
    821. 

  821. <P>
    822. 

  822. The internal tables can always be overridden by tables supplied by the
    823. 

  823. application that calls PCRE. These may be created in a different locale from
    824. 

  824. the default. As more and more applications change to using Unicode, the need
    825. 

  825. for this locale support is expected to die away.
    826. 

  826. </P>
    827. 

  827. <P>
    828. 

  828. External tables are built by calling the <b>pcre_maketables()</b> function,
    829. 

  829. which has no arguments, in the relevant locale. The result can then be passed
    830. 

  830. to <b>pcre_compile()</b> or <b>pcre_exec()</b> as often as necessary. For
    831. 

  831. example, to build and use tables that are appropriate for the French locale
    832. 

  832. (where accented characters with values greater than 128 are treated as letters),
    833. 

  833. the following code could be used:
    834. 

  834. <pre>
    835. 

  835.   setlocale(LC_CTYPE, "fr_FR");
    836. 

  836.   tables = pcre_maketables();
    837. 

  837.   re = pcre_compile(..., tables);
    838. 

  838. </pre>
    839. 

  839. The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
    840. 

  840. are using Windows, the name for the French locale is "french".
    841. 

  841. </P>
    842. 

  842. <P>
    843. 

  843. When <b>pcre_maketables()</b> runs, the tables are built in memory that is
    844. 

  844. obtained via <b>pcre_malloc</b>. It is the caller's responsibility to ensure
    845. 

  845. that the memory containing the tables remains available for as long as it is
    846. 

  846. needed.
    847. 

  847. </P>
    848. 

  848. <P>
    849. 

  849. The pointer that is passed to <b>pcre_compile()</b> is saved with the compiled
    850. 

  850. pattern, and the same tables are used via this pointer by <b>pcre_study()</b>
    851. 

  851. and normally also by <b>pcre_exec()</b>. Thus, by default, for any single
    852. 

  852. pattern, compilation, studying and matching all happen in the same locale, but
    853. 

  853. different patterns can be compiled in different locales.
    854. 

  854. </P>
    855. 

  855. <P>
    856. 

  856. It is possible to pass a table pointer or NULL (indicating the use of the
    857. 

  857. internal tables) to <b>pcre_exec()</b>. Although not intended for this purpose,
    858. 

  858. this facility could be used to match a pattern in a different locale from the
    859. 

  859. one in which it was compiled. Passing table pointers at run time is discussed
    860. 

  860. below in the section on matching a pattern.
    861. 

  861. </P>
    862. 

  862. <br><a name="SEC11" href="#TOC1">INFORMATION ABOUT A PATTERN</a><br>
    863. 

  863. <P>
    864. 

  864. <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
    865. 

  865. <b>int <i>what</i>, void *<i>where</i>);</b>
    866. 

  866. </P>
    867. 

  867. <P>
    868. 

  868. The <b>pcre_fullinfo()</b> function returns information about a compiled
    869. 

  869. pattern. It replaces the obsolete <b>pcre_info()</b> function, which is
    870. 

  870. nevertheless retained for backwards compability (and is documented below).
    871. 

  871. </P>
    872. 

  872. <P>
    873. 

  873. The first argument for <b>pcre_fullinfo()</b> is a pointer to the compiled
    874. 

  874. pattern. The second argument is the result of <b>pcre_study()</b>, or NULL if
    875. 

  875. the pattern was not studied. The third argument specifies which piece of
    876. 

  876. information is required, and the fourth argument is a pointer to a variable
    877. 

  877. to receive the data. The yield of the function is zero for success, or one of
    878. 

  878. the following negative numbers:
    879. 

  879. <pre>
    880. 

  880.   PCRE_ERROR_NULL       the argument <i>code</i> was NULL
    881. 

  881.                         the argument <i>where</i> was NULL
    882. 

  882.   PCRE_ERROR_BADMAGIC   the "magic number" was not found
    883. 

  883.   PCRE_ERROR_BADOPTION  the value of <i>what</i> was invalid
    884. 

  884. </pre>
    885. 

  885. The "magic number" is placed at the start of each compiled pattern as an simple
    886. 

  886. check against passing an arbitrary memory pointer. Here is a typical call of
    887. 

  887. <b>pcre_fullinfo()</b>, to obtain the length of the compiled pattern:
    888. 

  888. <pre>
    889. 

  889.   int rc;
    890. 

  890.   size_t length;
    891. 

  891.   rc = pcre_fullinfo(
    892. 

  892.     re,               /* result of pcre_compile() */
    893. 

  893.     pe,               /* result of pcre_study(), or NULL */
    894. 

  894.     PCRE_INFO_SIZE,   /* what is required */
    895. 

  895.     &length);         /* where to put the data */
    896. 

  896. </pre>
    897. 

  897. The possible values for the third argument are defined in <b>pcre.h</b>, and are
    898. 

  898. as follows:
    899. 

  899. <pre>
    900. 

  900.   PCRE_INFO_BACKREFMAX
    901. 

  901. </pre>
    902. 

  902. Return the number of the highest back reference in the pattern. The fourth
    903. 

  903. argument should point to an <b>int</b> variable. Zero is returned if there are
    904. 

  904. no back references.
    905. 

  905. <pre>
    906. 

  906.   PCRE_INFO_CAPTURECOUNT
    907. 

  907. </pre>
    908. 

  908. Return the number of capturing subpatterns in the pattern. The fourth argument
    909. 

  909. should point to an <b>int</b> variable.
    910. 

  910. <pre>
    911. 

  911.   PCRE_INFO_DEFAULT_TABLES
    912. 

  912. </pre>
    913. 

  913. Return a pointer to the internal default character tables within PCRE. The
    914. 

  914. fourth argument should point to an <b>unsigned char *</b> variable. This
    915. 

  915. information call is provided for internal use by the <b>pcre_study()</b>
    916. 

  916. function. External callers can cause PCRE to use its internal tables by passing
    917. 

  917. a NULL table pointer.
    918. 

  918. <pre>
    919. 

  919.   PCRE_INFO_FIRSTBYTE
    920. 

  920. </pre>
    921. 

  921. Return information about the first byte of any matched string, for a
    922. 

  922. non-anchored pattern. The fourth argument should point to an <b>int</b>
    923. 

  923. variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
    924. 

  924. still recognized for backwards compatibility.)
    925. 

  925. </P>
    926. 

  926. <P>
    927. 

  927. If there is a fixed first byte, for example, from a pattern such as
    928. 

  928. (cat|cow|coyote), its value is returned. Otherwise, if either
    929. 

  929. <br>
    930. 

  930. <br>
    931. 

  931. (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
    932. 

  932. starts with "^", or
    933. 

  933. <br>
    934. 

  934. <br>
    935. 

  935. (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
    936. 

  936. (if it were set, the pattern would be anchored),
    937. 

  937. <br>
    938. 

  938. <br>
    939. 

  939. -1 is returned, indicating that the pattern matches only at the start of a
    940. 

  940. subject string or after any newline within the string. Otherwise -2 is
    941. 

  941. returned. For anchored patterns, -2 is returned.
    942. 

  942. <pre>
    943. 

  943.   PCRE_INFO_FIRSTTABLE
    944. 

  944. </pre>
    945. 

  945. If the pattern was studied, and this resulted in the construction of a 256-bit
    946. 

  946. table indicating a fixed set of bytes for the first byte in any matching
    947. 

  947. string, a pointer to the table is returned. Otherwise NULL is returned. The
    948. 

  948. fourth argument should point to an <b>unsigned char *</b> variable.
    949. 

  949. <pre>
    950. 

  950.   PCRE_INFO_HASCRORLF
    951. 

  951. </pre>
    952. 

  952. Return 1 if the pattern contains any explicit matches for CR or LF characters,
    953. 

  953. otherwise 0. The fourth argument should point to an <b>int</b> variable. An
    954. 

  954. explicit match is either a literal CR or LF character, or \r or \n.
    955. 

  955. <pre>
    956. 

  956.   PCRE_INFO_JCHANGED
    957. 

  957. </pre>
    958. 

  958. Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
    959. 

  959. 0. The fourth argument should point to an <b>int</b> variable. (?J) and
    960. 

  960. (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
    961. 

  961. <pre>
    962. 

  962.   PCRE_INFO_LASTLITERAL
    963. 

  963. </pre>
    964. 

  964. Return the value of the rightmost literal byte that must exist in any matched
    965. 

  965. string, other than at its start, if such a byte has been recorded. The fourth
    966. 

  966. argument should point to an <b>int</b> variable. If there is no such byte, -1 is
    967. 

  967. returned. For anchored patterns, a last literal byte is recorded only if it
    968. 

  968. follows something of variable length. For example, for the pattern
    969. 

  969. /^a\d+z\d+/ the returned value is "z", but for /^a\dz\d/ the returned value
    970. 

  970. is -1.
    971. 

  971. <pre>
    972. 

  972.   PCRE_INFO_NAMECOUNT
    973. 

  973.   PCRE_INFO_NAMEENTRYSIZE
    974. 

  974.   PCRE_INFO_NAMETABLE
    975. 

  975. </pre>
    976. 

  976. PCRE supports the use of named as well as numbered capturing parentheses. The
    977. 

  977. names are just an additional way of identifying the parentheses, which still
    978. 

  978. acquire numbers. Several convenience functions such as
    979. 

  979. <b>pcre_get_named_substring()</b> are provided for extracting captured
    980. 

  980. substrings by name. It is also possible to extract the data directly, by first
    981. 

  981. converting the name to a number in order to access the correct pointers in the
    982. 

  982. output vector (described with <b>pcre_exec()</b> below). To do the conversion,
    983. 

  983. you need to use the name-to-number map, which is described by these three
    984. 

  984. values.
    985. 

  985. </P>
    986. 

  986. <P>
    987. 

  987. The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
    988. 

  988. the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
    989. 

  989. entry; both of these return an <b>int</b> value. The entry size depends on the
    990. 

  990. length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
    991. 

  991. entry of the table (a pointer to <b>char</b>). The first two bytes of each entry
    992. 

  992. are the number of the capturing parenthesis, most significant byte first. The
    993. 

  993. rest of the entry is the corresponding name, zero terminated. The names are in
    994. 

  994. alphabetical order. When PCRE_DUPNAMES is set, duplicate names are in order of
    995. 

  995. their parentheses numbers. For example, consider the following pattern (assume
    996. 

  996. PCRE_EXTENDED is set, so white space - including newlines - is ignored):
    997. 

  997. <pre>
    998. 

  998.   (?&#60;date&#62; (?&#60;year&#62;(\d\d)?\d\d) - (?&#60;month&#62;\d\d) - (?&#60;day&#62;\d\d) )
    999. 

  999. </pre>
    1000. 

  1000. There are four named subpatterns, so the table has four entries, and each entry
    1001. 

  1001. in the table is eight bytes long. The table is as follows, with non-printing
    1002. 

  1002. bytes shows in hexadecimal, and undefined bytes shown as ??:
    1003. 

  1003. <pre>
    1004. 

  1004.   00 01 d  a  t  e  00 ??
    1005. 

  1005.   00 05 d  a  y  00 ?? ??
    1006. 

  1006.   00 04 m  o  n  t  h  00
    1007. 

  1007.   00 02 y  e  a  r  00 ??
    1008. 

  1008. </pre>
    1009. 

  1009. When writing code to extract data from named subpatterns using the
    1010. 

  1010. name-to-number map, remember that the length of the entries is likely to be
    1011. 

  1011. different for each compiled pattern.
    1012. 

  1012. <pre>
    1013. 

  1013.   PCRE_INFO_OKPARTIAL
    1014. 

  1014. </pre>
    1015. 

  1015. Return 1 if the pattern can be used for partial matching, otherwise 0. The
    1016. 

  1016. fourth argument should point to an <b>int</b> variable. The
    1017. 

  1017. <a href="pcrepartial.html"><b>pcrepartial</b></a>
    1018. 

  1018. documentation lists the restrictions that apply to patterns when partial
    1019. 

  1019. matching is used.
    1020. 

  1020. <pre>
    1021. 

  1021.   PCRE_INFO_OPTIONS
    1022. 

  1022. </pre>
    1023. 

  1023. Return a copy of the options with which the pattern was compiled. The fourth
    1024. 

  1024. argument should point to an <b>unsigned long int</b> variable. These option bits
    1025. 

  1025. are those specified in the call to <b>pcre_compile()</b>, modified by any
    1026. 

  1026. top-level option settings at the start of the pattern itself. In other words,
    1027. 

  1027. they are the options that will be in force when matching starts. For example,
    1028. 

  1028. if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
    1029. 

  1029. result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
    1030. 

  1030. </P>
    1031. 

  1031. <P>
    1032. 

  1032. A pattern is automatically anchored by PCRE if all of its top-level
    1033. 

  1033. alternatives begin with one of the following:
    1034. 

  1034. <pre>
    1035. 

  1035.   ^     unless PCRE_MULTILINE is set
    1036. 

  1036.   \A    always
    1037. 

  1037.   \G    always
    1038. 

  1038.   .*    if PCRE_DOTALL is set and there are no back references to the subpattern in which .* appears
    1039. 

  1039. </pre>
    1040. 

  1040. For such patterns, the PCRE_ANCHORED bit is set in the options returned by
    1041. 

  1041. <b>pcre_fullinfo()</b>.
    1042. 

  1042. <pre>
    1043. 

  1043.   PCRE_INFO_SIZE
    1044. 

  1044. </pre>
    1045. 

  1045. Return the size of the compiled pattern, that is, the value that was passed as
    1046. 

  1046. the argument to <b>pcre_malloc()</b> when PCRE was getting memory in which to
    1047. 

  1047. place the compiled data. The fourth argument should point to a <b>size_t</b>
    1048. 

  1048. variable.
    1049. 

  1049. <pre>
    1050. 

  1050.   PCRE_INFO_STUDYSIZE
    1051. 

  1051. </pre>
    1052. 

  1052. Return the size of the data block pointed to by the <i>study_data</i> field in
    1053. 

  1053. a <b>pcre_extra</b> block. That is, it is the value that was passed to
    1054. 

  1054. <b>pcre_malloc()</b> when PCRE was getting memory into which to place the data
    1055. 

  1055. created by <b>pcre_study()</b>. The fourth argument should point to a
    1056. 

  1056. <b>size_t</b> variable.
    1057. 

  1057. </P>
    1058. 

  1058. <br><a name="SEC12" href="#TOC1">OBSOLETE INFO FUNCTION</a><br>
    1059. 

  1059. <P>
    1060. 

  1060. <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
    1061. 

  1061. <b>*<i>firstcharptr</i>);</b>
    1062. 

  1062. </P>
    1063. 

  1063. <P>
    1064. 

  1064. The <b>pcre_info()</b> function is now obsolete because its interface is too
    1065. 

  1065. restrictive to return all the available data about a compiled pattern. New
    1066. 

  1066. programs should use <b>pcre_fullinfo()</b> instead. The yield of
    1067. 

  1067. <b>pcre_info()</b> is the number of capturing subpatterns, or one of the
    1068. 

  1068. following negative numbers:
    1069. 

  1069. <pre>
    1070. 

  1070.   PCRE_ERROR_NULL       the argument <i>code</i> was NULL
    1071. 

  1071.   PCRE_ERROR_BADMAGIC   the "magic number" was not found
    1072. 

  1072. </pre>
    1073. 

  1073. If the <i>optptr</i> argument is not NULL, a copy of the options with which the
    1074. 

  1074. pattern was compiled is placed in the integer it points to (see
    1075. 

  1075. PCRE_INFO_OPTIONS above).
    1076. 

  1076. </P>
    1077. 

  1077. <P>
    1078. 

  1078. If the pattern is not anchored and the <i>firstcharptr</i> argument is not NULL,
    1079. 

  1079. it is used to pass back information about the first character of any matched
    1080. 

  1080. string (see PCRE_INFO_FIRSTBYTE above).
    1081. 

  1081. </P>
    1082. 

  1082. <br><a name="SEC13" href="#TOC1">REFERENCE COUNTS</a><br>
    1083. 

  1083. <P>
    1084. 

  1084. <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
    1085. 

  1085. </P>
    1086. 

  1086. <P>
    1087. 

  1087. The <b>pcre_refcount()</b> function is used to maintain a reference count in the
    1088. 

  1088. data block that contains a compiled pattern. It is provided for the benefit of
    1089. 

  1089. applications that operate in an object-oriented manner, where different parts
    1090. 

  1090. of the application may be using the same compiled pattern, but you want to free
    1091. 

  1091. the block when they are all done.
    1092. 

  1092. </P>
    1093. 

  1093. <P>
    1094. 

  1094. When a pattern is compiled, the reference count field is initialized to zero.
    1095. 

  1095. It is changed only by calling this function, whose action is to add the
    1096. 

  1096. <i>adjust</i> value (which may be positive or negative) to it. The yield of the
    1097. 

  1097. function is the new value. However, the value of the count is constrained to
    1098. 

  1098. lie between 0 and 65535, inclusive. If the new value is outside these limits,
    1099. 

  1099. it is forced to the appropriate limit value.
    1100. 

  1100. </P>
    1101. 

  1101. <P>
    1102. 

  1102. Except when it is zero, the reference count is not correctly preserved if a
    1103. 

  1103. pattern is compiled on one host and then transferred to a host whose byte-order
    1104. 

  1104. is different. (This seems a highly unlikely scenario.)
    1105. 

  1105. </P>
    1106. 

  1106. <br><a name="SEC14" href="#TOC1">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a><br>
    1107. 

  1107. <P>
    1108. 

  1108. <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
    1109. 

  1109. <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
    1110. 

  1110. <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
    1111. 

  1111. </P>
    1112. 

  1112. <P>
    1113. 

  1113. The function <b>pcre_exec()</b> is called to match a subject string against a
    1114. 

  1114. compiled pattern, which is passed in the <i>code</i> argument. If the
    1115. 

  1115. pattern has been studied, the result of the study should be passed in the
    1116. 

  1116. <i>extra</i> argument. This function is the main matching facility of the
    1117. 

  1117. library, and it operates in a Perl-like manner. For specialist use there is
    1118. 

  1118. also an alternative matching function, which is described
    1119. 

  1119. <a href="#dfamatch">below</a>
    1120. 

  1120. in the section about the <b>pcre_dfa_exec()</b> function.
    1121. 

  1121. </P>
    1122. 

  1122. <P>
    1123. 

  1123. In most applications, the pattern will have been compiled (and optionally
    1124. 

  1124. studied) in the same process that calls <b>pcre_exec()</b>. However, it is
    1125. 

  1125. possible to save compiled patterns and study data, and then use them later
    1126. 

  1126. in different processes, possibly even on different hosts. For a discussion
    1127. 

  1127. about this, see the
    1128. 

  1128. <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
    1129. 

  1129. documentation.
    1130. 

  1130. </P>
    1131. 

  1131. <P>
    1132. 

  1132. Here is an example of a simple call to <b>pcre_exec()</b>:
    1133. 

  1133. <pre>
    1134. 

  1134.   int rc;
    1135. 

  1135.   int ovector[30];
    1136. 

  1136.   rc = pcre_exec(
    1137. 

  1137.     re,             /* result of pcre_compile() */
    1138. 

  1138.     NULL,           /* we didn't study the pattern */
    1139. 

  1139.     "some string",  /* the subject string */
    1140. 

  1140.     11,             /* the length of the subject string */
    1141. 

  1141.     0,              /* start at offset 0 in the subject */
    1142. 

  1142.     0,              /* default options */
    1143. 

  1143.     ovector,        /* vector of integers for substring information */
    1144. 

  1144.     30);            /* number of elements (NOT size in bytes) */
    1145. 

  1145. <a name="extradata"></a></PRE>
    1146. 

  1146. </P>
    1147. 

  1147. <br><b>
    1148. 

  1148. Extra data for <b>pcre_exec()</b>
    1149. 

  1149. </b><br>
    1150. 

  1150. <P>
    1151. 

  1151. If the <i>extra</i> argument is not NULL, it must point to a <b>pcre_extra</b>
    1152. 

  1152. data block. The <b>pcre_study()</b> function returns such a block (when it
    1153. 

  1153. doesn't return NULL), but you can also create one for yourself, and pass
    1154. 

  1154. additional information in it. The <b>pcre_extra</b> block contains the following
    1155. 

  1155. fields (not necessarily in this order):
    1156. 

  1156. <pre>
    1157. 

  1157.   unsigned long int <i>flags</i>;
    1158. 

  1158.   void *<i>study_data</i>;
    1159. 

  1159.   unsigned long int <i>match_limit</i>;
    1160. 

  1160.   unsigned long int <i>match_limit_recursion</i>;
    1161. 

  1161.   void *<i>callout_data</i>;
    1162. 

  1162.   const unsigned char *<i>tables</i>;
    1163. 

  1163. </pre>
    1164. 

  1164. The <i>flags</i> field is a bitmap that specifies which of the other fields
    1165. 

  1165. are set. The flag bits are:
    1166. 

  1166. <pre>
    1167. 

  1167.   PCRE_EXTRA_STUDY_DATA
    1168. 

  1168.   PCRE_EXTRA_MATCH_LIMIT
    1169. 

  1169.   PCRE_EXTRA_MATCH_LIMIT_RECURSION
    1170. 

  1170.   PCRE_EXTRA_CALLOUT_DATA
    1171. 

  1171.   PCRE_EXTRA_TABLES
    1172. 

  1172. </pre>
    1173. 

  1173. Other flag bits should be set to zero. The <i>study_data</i> field is set in the
    1174. 

  1174. <b>pcre_extra</b> block that is returned by <b>pcre_study()</b>, together with
    1175. 

  1175. the appropriate flag bit. You should not set this yourself, but you may add to
    1176. 

  1176. the block by setting the other fields and their corresponding flag bits.
    1177. 

  1177. </P>
    1178. 

  1178. <P>
    1179. 

  1179. The <i>match_limit</i> field provides a means of preventing PCRE from using up a
    1180. 

  1180. vast amount of resources when running patterns that are not going to match,
    1181. 

  1181. but which have a very large number of possibilities in their search trees. The
    1182. 

  1182. classic example is the use of nested unlimited repeats.
    1183. 

  1183. </P>
    1184. 

  1184. <P>
    1185. 

  1185. Internally, PCRE uses a function called <b>match()</b> which it calls repeatedly
    1186. 

  1186. (sometimes recursively). The limit set by <i>match_limit</i> is imposed on the
    1187. 

  1187. number of times this function is called during a match, which has the effect of
    1188. 

  1188. limiting the amount of backtracking that can take place. For patterns that are
    1189. 

  1189. not anchored, the count restarts from zero for each position in the subject
    1190. 

  1190. string.
    1191. 

  1191. </P>
    1192. 

  1192. <P>
    1193. 

  1193. The default value for the limit can be set when PCRE is built; the default
    1194. 

  1194. default is 10 million, which handles all but the most extreme cases. You can
    1195. 

  1195. override the default by suppling <b>pcre_exec()</b> with a <b>pcre_extra</b>
    1196. 

  1196. block in which <i>match_limit</i> is set, and PCRE_EXTRA_MATCH_LIMIT is set in
    1197. 

  1197. the <i>flags</i> field. If the limit is exceeded, <b>pcre_exec()</b> returns
    1198. 

  1198. PCRE_ERROR_MATCHLIMIT.
    1199. 

  1199. </P>
    1200. 

  1200. <P>
    1201. 

  1201. The <i>match_limit_recursion</i> field is similar to <i>match_limit</i>, but
    1202. 

  1202. instead of limiting the total number of times that <b>match()</b> is called, it
    1203. 

  1203. limits the depth of recursion. The recursion depth is a smaller number than the
    1204. 

  1204. total number of calls, because not all calls to <b>match()</b> are recursive.
    1205. 

  1205. This limit is of use only if it is set smaller than <i>match_limit</i>.
    1206. 

  1206. </P>
    1207. 

  1207. <P>
    1208. 

  1208. Limiting the recursion depth limits the amount of stack that can be used, or,
    1209. 

  1209. when PCRE has been compiled to use memory on the heap instead of the stack, the
    1210. 

  1210. amount of heap memory that can be used.
    1211. 

  1211. </P>
    1212. 

  1212. <P>
    1213. 

  1213. The default value for <i>match_limit_recursion</i> can be set when PCRE is
    1214. 

  1214. built; the default default is the same value as the default for
    1215. 

  1215. <i>match_limit</i>. You can override the default by suppling <b>pcre_exec()</b>
    1216. 

  1216. with a <b>pcre_extra</b> block in which <i>match_limit_recursion</i> is set, and
    1217. 

  1217. PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the <i>flags</i> field. If the limit
    1218. 

  1218. is exceeded, <b>pcre_exec()</b> returns PCRE_ERROR_RECURSIONLIMIT.
    1219. 

  1219. </P>
    1220. 

  1220. <P>
    1221. 

  1221. The <i>pcre_callout</i> field is used in conjunction with the "callout" feature,
    1222. 

  1222. which is described in the
    1223. 

  1223. <a href="pcrecallout.html"><b>pcrecallout</b></a>
    1224. 

  1224. documentation.
    1225. 

  1225. </P>
    1226. 

  1226. <P>
    1227. 

  1227. The <i>tables</i> field is used to pass a character tables pointer to
    1228. 

  1228. <b>pcre_exec()</b>; this overrides the value that is stored with the compiled
    1229. 

  1229. pattern. A non-NULL value is stored with the compiled pattern only if custom
    1230. 

  1230. tables were supplied to <b>pcre_compile()</b> via its <i>tableptr</i> argument.
    1231. 

  1231. If NULL is passed to <b>pcre_exec()</b> using this mechanism, it forces PCRE's
    1232. 

  1232. internal tables to be used. This facility is helpful when re-using patterns
    1233. 

  1233. that have been saved after compiling with an external set of tables, because
    1234. 

  1234. the external tables might be at a different address when <b>pcre_exec()</b> is
    1235. 

  1235. called. See the
    1236. 

  1236. <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
    1237. 

  1237. documentation for a discussion of saving compiled patterns for later use.
    1238. 

  1238. <a name="execoptions"></a></P>
    1239. 

  1239. <br><b>
    1240. 

  1240. Option bits for <b>pcre_exec()</b>
    1241. 

  1241. </b><br>
    1242. 

  1242. <P>
    1243. 

  1243. The unused bits of the <i>options</i> argument for <b>pcre_exec()</b> must be
    1244. 

  1244. zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
    1245. 

  1245. PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NO_UTF8_CHECK and PCRE_PARTIAL.
    1246. 

  1246. <pre>
    1247. 

  1247.   PCRE_ANCHORED
    1248. 

  1248. </pre>
    1249. 

  1249. The PCRE_ANCHORED option limits <b>pcre_exec()</b> to matching at the first
    1250. 

  1250. matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
    1251. 

  1251. to be anchored by virtue of its contents, it cannot be made unachored at
    1252. 

  1252. matching time.
    1253. 

  1253. <pre>
    1254. 

  1254.   PCRE_BSR_ANYCRLF
    1255. 

  1255.   PCRE_BSR_UNICODE
    1256. 

  1256. </pre>
    1257. 

  1257. These options (which are mutually exclusive) control what the \R escape
    1258. 

  1258. sequence matches. The choice is either to match only CR, LF, or CRLF, or to
    1259. 

  1259. match any Unicode newline sequence. These options override the choice that was
    1260. 

  1260. made or defaulted when the pattern was compiled.
    1261. 

  1261. <pre>
    1262. 

  1262.   PCRE_NEWLINE_CR
    1263. 

  1263.   PCRE_NEWLINE_LF
    1264. 

  1264.   PCRE_NEWLINE_CRLF
    1265. 

  1265.   PCRE_NEWLINE_ANYCRLF
    1266. 

  1266.   PCRE_NEWLINE_ANY
    1267. 

  1267. </pre>
    1268. 

  1268. These options override the newline definition that was chosen or defaulted when
    1269. 

  1269. the pattern was compiled. For details, see the description of
    1270. 

  1270. <b>pcre_compile()</b> above. During matching, the newline choice affects the
    1271. 

  1271. behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
    1272. 

  1272. the way the match position is advanced after a match failure for an unanchored
    1273. 

  1273. pattern.
    1274. 

  1274. </P>
    1275. 

  1275. <P>
    1276. 

  1276. When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
    1277. 

  1277. match attempt for an unanchored pattern fails when the current position is at a
    1278. 

  1278. CRLF sequence, and the pattern contains no explicit matches for CR or LF
    1279. 

  1279. characters, the match position is advanced by two characters instead of one, in
    1280. 

  1280. other words, to after the CRLF.
    1281. 

  1281. </P>
    1282. 

  1282. <P>
    1283. 

  1283. The above rule is a compromise that makes the most common cases work as
    1284. 

  1284. expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
    1285. 

  1285. set), it does not match the string "\r\nA" because, after failing at the
    1286. 

  1286. start, it skips both the CR and the LF before retrying. However, the pattern
    1287. 

  1287. [\r\n]A does match that string, because it contains an explicit CR or LF
    1288. 

  1288. reference, and so advances only by one character after the first failure.
    1289. 

  1289. </P>
    1290. 

  1290. <P>
    1291. 

  1291. An explicit match for CR of LF is either a literal appearance of one of those
    1292. 

  1292. characters, or one of the \r or \n escape sequences. Implicit matches such as
    1293. 

  1293. [^X] do not count, nor does \s (which includes CR and LF in the characters
    1294. 

  1294. that it matches).
    1295. 

  1295. </P>
    1296. 

  1296. <P>
    1297. 

  1297. Notwithstanding the above, anomalous effects may still occur when CRLF is a
    1298. 

  1298. valid newline sequence and explicit \r or \n escapes appear in the pattern.
    1299. 

  1299. <pre>
    1300. 

  1300.   PCRE_NOTBOL
    1301. 

  1301. </pre>
    1302. 

  1302. This option specifies that first character of the subject string is not the
    1303. 

  1303. beginning of a line, so the circumflex metacharacter should not match before
    1304. 

  1304. it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
    1305. 

  1305. never to match. This option affects only the behaviour of the circumflex
    1306. 

  1306. metacharacter. It does not affect \A.
    1307. 

  1307. <pre>
    1308. 

  1308.   PCRE_NOTEOL
    1309. 

  1309. </pre>
    1310. 

  1310. This option specifies that the end of the subject string is not the end of a
    1311. 

  1311. line, so the dollar metacharacter should not match it nor (except in multiline
    1312. 

  1312. mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
    1313. 

  1313. compile time) causes dollar never to match. This option affects only the
    1314. 

  1314. behaviour of the dollar metacharacter. It does not affect \Z or \z.
    1315. 

  1315. <pre>
    1316. 

  1316.   PCRE_NOTEMPTY
    1317. 

  1317. </pre>
    1318. 

  1318. An empty string is not considered to be a valid match if this option is set. If
    1319. 

  1319. there are alternatives in the pattern, they are tried. If all the alternatives
    1320. 

  1320. match the empty string, the entire match fails. For example, if the pattern
    1321. 

  1321. <pre>
    1322. 

  1322.   a?b?
    1323. 

  1323. </pre>
    1324. 

  1324. is applied to a string not beginning with "a" or "b", it matches the empty
    1325. 

  1325. string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
    1326. 

  1326. valid, so PCRE searches further into the string for occurrences of "a" or "b".
    1327. 

  1327. </P>
    1328. 

  1328. <P>
    1329. 

  1329. Perl has no direct equivalent of PCRE_NOTEMPTY, but it does make a special case
    1330. 

  1330. of a pattern match of the empty string within its <b>split()</b> function, and
    1331. 

  1331. when using the /g modifier. It is possible to emulate Perl's behaviour after
    1332. 

  1332. matching a null string by first trying the match again at the same offset with
    1333. 

  1333. PCRE_NOTEMPTY and PCRE_ANCHORED, and then if that fails by advancing the
    1334. 

  1334. starting offset (see below) and trying an ordinary match again. There is some
    1335. 

  1335. code that demonstrates how to do this in the <i>pcredemo.c</i> sample program.
    1336. 

  1336. <pre>
    1337. 

  1337.   PCRE_NO_UTF8_CHECK
    1338. 

  1338. </pre>
    1339. 

  1339. When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
    1340. 

  1340. string is automatically checked when <b>pcre_exec()</b> is subsequently called.
    1341. 

  1341. The value of <i>startoffset</i> is also checked to ensure that it points to the
    1342. 

  1342. start of a UTF-8 character. There is a discussion about the validity of UTF-8
    1343. 

  1343. strings in the
    1344. 

  1344. <a href="pcre.html#utf8strings">section on UTF-8 support</a>
    1345. 

  1345. in the main
    1346. 

  1346. <a href="pcre.html"><b>pcre</b></a>
    1347. 

  1347. page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_exec()</b> returns
    1348. 

  1348. the error PCRE_ERROR_BADUTF8. If <i>startoffset</i> contains an invalid value,
    1349. 

  1349. PCRE_ERROR_BADUTF8_OFFSET is returned.
    1350. 

  1350. </P>
    1351. 

  1351. <P>
    1352. 

  1352. If you already know that your subject is valid, and you want to skip these
    1353. 

  1353. checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
    1354. 

  1354. calling <b>pcre_exec()</b>. You might want to do this for the second and
    1355. 

  1355. subsequent calls to <b>pcre_exec()</b> if you are making repeated calls to find
    1356. 

  1356. all the matches in a single subject string. However, you should be sure that
    1357. 

  1357. the value of <i>startoffset</i> points to the start of a UTF-8 character. When
    1358. 

  1358. PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
    1359. 

  1359. subject, or a value of <i>startoffset</i> that does not point to the start of a
    1360. 

  1360. UTF-8 character, is undefined. Your program may crash.
    1361. 

  1361. <pre>
    1362. 

  1362.   PCRE_PARTIAL
    1363. 

  1363. </pre>
    1364. 

  1364. This option turns on the partial matching feature. If the subject string fails
    1365. 

  1365. to match the pattern, but at some point during the matching process the end of
    1366. 

  1366. the subject was reached (that is, the subject partially matches the pattern and
    1367. 

  1367. the failure to match occurred only because there were not enough subject
    1368. 

  1368. characters), <b>pcre_exec()</b> returns PCRE_ERROR_PARTIAL instead of
    1369. 

  1369. PCRE_ERROR_NOMATCH. When PCRE_PARTIAL is used, there are restrictions on what
    1370. 

  1370. may appear in the pattern. These are discussed in the
    1371. 

  1371. <a href="pcrepartial.html"><b>pcrepartial</b></a>
    1372. 

  1372. documentation.
    1373. 

  1373. </P>
    1374. 

  1374. <br><b>
    1375. 

  1375. The string to be matched by <b>pcre_exec()</b>
    1376. 

  1376. </b><br>
    1377. 

  1377. <P>
    1378. 

  1378. The subject string is passed to <b>pcre_exec()</b> as a pointer in
    1379. 

  1379. <i>subject</i>, a length (in bytes) in <i>length</i>, and a starting byte offset
    1380. 

  1380. in <i>startoffset</i>. In UTF-8 mode, the byte offset must point to the start of
    1381. 

  1381. a UTF-8 character. Unlike the pattern string, the subject may contain binary
    1382. 

  1382. zero bytes. When the starting offset is zero, the search for a match starts at
    1383. 

  1383. the beginning of the subject, and this is by far the most common case.
    1384. 

  1384. </P>
    1385. 

  1385. <P>
    1386. 

  1386. A non-zero starting offset is useful when searching for another match in the
    1387. 

  1387. same subject by calling <b>pcre_exec()</b> again after a previous success.
    1388. 

  1388. Setting <i>startoffset</i> differs from just passing over a shortened string and
    1389. 

  1389. setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
    1390. 

  1390. lookbehind. For example, consider the pattern
    1391. 

  1391. <pre>
    1392. 

  1392.   \Biss\B
    1393. 

  1393. </pre>
    1394. 

  1394. which finds occurrences of "iss" in the middle of words. (\B matches only if
    1395. 

  1395. the current position in the subject is not a word boundary.) When applied to
    1396. 

  1396. the string "Mississipi" the first call to <b>pcre_exec()</b> finds the first
    1397. 

  1397. occurrence. If <b>pcre_exec()</b> is called again with just the remainder of the
    1398. 

  1398. subject, namely "issipi", it does not match, because \B is always false at the
    1399. 

  1399. start of the subject, which is deemed to be a word boundary. However, if
    1400. 

  1400. <b>pcre_exec()</b> is passed the entire string again, but with <i>startoffset</i>
    1401. 

  1401. set to 4, it finds the second occurrence of "iss" because it is able to look
    1402. 

  1402. behind the starting point to discover that it is preceded by a letter.
    1403. 

  1403. </P>
    1404. 

  1404. <P>
    1405. 

  1405. If a non-zero starting offset is passed when the pattern is anchored, one
    1406. 

  1406. attempt to match at the given offset is made. This can only succeed if the
    1407. 

  1407. pattern does not require the match to be at the start of the subject.
    1408. 

  1408. </P>
    1409. 

  1409. <br><b>
    1410. 

  1410. How <b>pcre_exec()</b> returns captured substrings
    1411. 

  1411. </b><br>
    1412. 

  1412. <P>
    1413. 

  1413. In general, a pattern matches a certain portion of the subject, and in
    1414. 

  1414. addition, further substrings from the subject may be picked out by parts of the
    1415. 

  1415. pattern. Following the usage in Jeffrey Friedl's book, this is called
    1416. 

  1416. "capturing" in what follows, and the phrase "capturing subpattern" is used for
    1417. 

  1417. a fragment of a pattern that picks out a substring. PCRE supports several other
    1418. 

  1418. kinds of parenthesized subpattern that do not cause substrings to be captured.
    1419. 

  1419. </P>
    1420. 

  1420. <P>
    1421. 

  1421. Captured substrings are returned to the caller via a vector of integers whose
    1422. 

  1422. address is passed in <i>ovector</i>. The number of elements in the vector is
    1423. 

  1423. passed in <i>ovecsize</i>, which must be a non-negative number. <b>Note</b>: this
    1424. 

  1424. argument is NOT the size of <i>ovector</i> in bytes.
    1425. 

  1425. </P>
    1426. 

  1426. <P>
    1427. 

  1427. The first two-thirds of the vector is used to pass back captured substrings,
    1428. 

  1428. each substring using a pair of integers. The remaining third of the vector is
    1429. 

  1429. used as workspace by <b>pcre_exec()</b> while matching capturing subpatterns,
    1430. 

  1430. and is not available for passing back information. The number passed in
    1431. 

  1431. <i>ovecsize</i> should always be a multiple of three. If it is not, it is
    1432. 

  1432. rounded down.
    1433. 

  1433. </P>
    1434. 

  1434. <P>
    1435. 

  1435. When a match is successful, information about captured substrings is returned
    1436. 

  1436. in pairs of integers, starting at the beginning of <i>ovector</i>, and
    1437. 

  1437. continuing up to two-thirds of its length at the most. The first element of
    1438. 

  1438. each pair is set to the byte offset of the first character in a substring, and
    1439. 

  1439. the second is set to the byte offset of the first character after the end of a
    1440. 

  1440. substring. <b>Note</b>: these values are always byte offsets, even in UTF-8
    1441. 

  1441. mode. They are not character counts.
    1442. 

  1442. </P>
    1443. 

  1443. <P>
    1444. 

  1444. The first pair of integers, <i>ovector[0]</i> and <i>ovector[1]</i>, identify the
    1445. 

  1445. portion of the subject string matched by the entire pattern. The next pair is
    1446. 

  1446. used for the first capturing subpattern, and so on. The value returned by
    1447. 

  1447. <b>pcre_exec()</b> is one more than the highest numbered pair that has been set.
    1448. 

  1448. For example, if two substrings have been captured, the returned value is 3. If
    1449. 

  1449. there are no capturing subpatterns, the return value from a successful match is
    1450. 

  1450. 1, indicating that just the first pair of offsets has been set.
    1451. 

  1451. </P>
    1452. 

  1452. <P>
    1453. 

  1453. If a capturing subpattern is matched repeatedly, it is the last portion of the
    1454. 

  1454. string that it matched that is returned.
    1455. 

  1455. </P>
    1456. 

  1456. <P>
    1457. 

  1457. If the vector is too small to hold all the captured substring offsets, it is
    1458. 

  1458. used as far as possible (up to two-thirds of its length), and the function
    1459. 

  1459. returns a value of zero. If the substring offsets are not of interest,
    1460. 

  1460. <b>pcre_exec()</b> may be called with <i>ovector</i> passed as NULL and
    1461. 

  1461. <i>ovecsize</i> as zero. However, if the pattern contains back references and
    1462. 

  1462. the <i>ovector</i> is not big enough to remember the related substrings, PCRE
    1463. 

  1463. has to get additional memory for use during matching. Thus it is usually
    1464. 

  1464. advisable to supply an <i>ovector</i>.
    1465. 

  1465. </P>
    1466. 

  1466. <P>
    1467. 

  1467. The <b>pcre_info()</b> function can be used to find out how many capturing
    1468. 

  1468. subpatterns there are in a compiled pattern. The smallest size for
    1469. 

  1469. <i>ovector</i> that will allow for <i>n</i> captured substrings, in addition to
    1470. 

  1470. the offsets of the substring matched by the whole pattern, is (<i>n</i>+1)*3.
    1471. 

  1471. </P>
    1472. 

  1472. <P>
    1473. 

  1473. It is possible for capturing subpattern number <i>n+1</i> to match some part of
    1474. 

  1474. the subject when subpattern <i>n</i> has not been used at all. For example, if
    1475. 

  1475. the string "abc" is matched against the pattern (a|(z))(bc) the return from the
    1476. 

  1476. function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
    1477. 

  1477. happens, both values in the offset pairs corresponding to unused subpatterns
    1478. 

  1478. are set to -1.
    1479. 

  1479. </P>
    1480. 

  1480. <P>
    1481. 

  1481. Offset values that correspond to unused subpatterns at the end of the
    1482. 

  1482. expression are also set to -1. For example, if the string "abc" is matched
    1483. 

  1483. against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
    1484. 

  1484. return from the function is 2, because the highest used capturing subpattern
    1485. 

  1485. number is 1. However, you can refer to the offsets for the second and third
    1486. 

  1486. capturing subpatterns if you wish (assuming the vector is large enough, of
    1487. 

  1487. course).
    1488. 

  1488. </P>
    1489. 

  1489. <P>
    1490. 

  1490. Some convenience functions are provided for extracting the captured substrings
    1491. 

  1491. as separate strings. These are described below.
    1492. 

  1492. <a name="errorlist"></a></P>
    1493. 

  1493. <br><b>
    1494. 

  1494. Error return values from <b>pcre_exec()</b>
    1495. 

  1495. </b><br>
    1496. 

  1496. <P>
    1497. 

  1497. If <b>pcre_exec()</b> fails, it returns a negative number. The following are
    1498. 

  1498. defined in the header file:
    1499. 

  1499. <pre>
    1500. 

  1500.   PCRE_ERROR_NOMATCH        (-1)
    1501. 

  1501. </pre>
    1502. 

  1502. The subject string did not match the pattern.
    1503. 

  1503. <pre>
    1504. 

  1504.   PCRE_ERROR_NULL           (-2)
    1505. 

  1505. </pre>
    1506. 

  1506. Either <i>code</i> or <i>subject</i> was passed as NULL, or <i>ovector</i> was
    1507. 

  1507. NULL and <i>ovecsize</i> was not zero.
    1508. 

  1508. <pre>
    1509. 

  1509.   PCRE_ERROR_BADOPTION      (-3)
    1510. 

  1510. </pre>
    1511. 

  1511. An unrecognized bit was set in the <i>options</i> argument.
    1512. 

  1512. <pre>
    1513. 

  1513.   PCRE_ERROR_BADMAGIC       (-4)
    1514. 

  1514. </pre>
    1515. 

  1515. PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
    1516. 

  1516. the case when it is passed a junk pointer and to detect when a pattern that was
    1517. 

  1517. compiled in an environment of one endianness is run in an environment with the
    1518. 

  1518. other endianness. This is the error that PCRE gives when the magic number is
    1519. 

  1519. not present.
    1520. 

  1520. <pre>
    1521. 

  1521.   PCRE_ERROR_UNKNOWN_OPCODE (-5)
    1522. 

  1522. </pre>
    1523. 

  1523. While running the pattern match, an unknown item was encountered in the
    1524. 

  1524. compiled pattern. This error could be caused by a bug in PCRE or by overwriting
    1525. 

  1525. of the compiled pattern.
    1526. 

  1526. <pre>
    1527. 

  1527.   PCRE_ERROR_NOMEMORY       (-6)
    1528. 

  1528. </pre>
    1529. 

  1529. If a pattern contains back references, but the <i>ovector</i> that is passed to
    1530. 

  1530. <b>pcre_exec()</b> is not big enough to remember the referenced substrings, PCRE
    1531. 

  1531. gets a block of memory at the start of matching to use for this purpose. If the
    1532. 

  1532. call via <b>pcre_malloc()</b> fails, this error is given. The memory is
    1533. 

  1533. automatically freed at the end of matching.
    1534. 

  1534. <pre>
    1535. 

  1535.   PCRE_ERROR_NOSUBSTRING    (-7)
    1536. 

  1536. </pre>
    1537. 

  1537. This error is used by the <b>pcre_copy_substring()</b>,
    1538. 

  1538. <b>pcre_get_substring()</b>, and <b>pcre_get_substring_list()</b> functions (see
    1539. 

  1539. below). It is never returned by <b>pcre_exec()</b>.
    1540. 

  1540. <pre>
    1541. 

  1541.   PCRE_ERROR_MATCHLIMIT     (-8)
    1542. 

  1542. </pre>
    1543. 

  1543. The backtracking limit, as specified by the <i>match_limit</i> field in a
    1544. 

  1544. <b>pcre_extra</b> structure (or defaulted) was reached. See the description
    1545. 

  1545. above.
    1546. 

  1546. <pre>
    1547. 

  1547.   PCRE_ERROR_CALLOUT        (-9)
    1548. 

  1548. </pre>
    1549. 

  1549. This error is never generated by <b>pcre_exec()</b> itself. It is provided for
    1550. 

  1550. use by callout functions that want to yield a distinctive error code. See the
    1551. 

  1551. <a href="pcrecallout.html"><b>pcrecallout</b></a>
    1552. 

  1552. documentation for details.
    1553. 

  1553. <pre>
    1554. 

  1554.   PCRE_ERROR_BADUTF8        (-10)
    1555. 

  1555. </pre>
    1556. 

  1556. A string that contains an invalid UTF-8 byte sequence was passed as a subject.
    1557. 

  1557. <pre>
    1558. 

  1558.   PCRE_ERROR_BADUTF8_OFFSET (-11)
    1559. 

  1559. </pre>
    1560. 

  1560. The UTF-8 byte sequence that was passed as a subject was valid, but the value
    1561. 

  1561. of <i>startoffset</i> did not point to the beginning of a UTF-8 character.
    1562. 

  1562. <pre>
    1563. 

  1563.   PCRE_ERROR_PARTIAL        (-12)
    1564. 

  1564. </pre>
    1565. 

  1565. The subject string did not match, but it did match partially. See the
    1566. 

  1566. <a href="pcrepartial.html"><b>pcrepartial</b></a>
    1567. 

  1567. documentation for details of partial matching.
    1568. 

  1568. <pre>
    1569. 

  1569.   PCRE_ERROR_BADPARTIAL     (-13)
    1570. 

  1570. </pre>
    1571. 

  1571. The PCRE_PARTIAL option was used with a compiled pattern containing items that
    1572. 

  1572. are not supported for partial matching. See the
    1573. 

  1573. <a href="pcrepartial.html"><b>pcrepartial</b></a>
    1574. 

  1574. documentation for details of partial matching.
    1575. 

  1575. <pre>
    1576. 

  1576.   PCRE_ERROR_INTERNAL       (-14)
    1577. 

  1577. </pre>
    1578. 

  1578. An unexpected internal error has occurred. This error could be caused by a bug
    1579. 

  1579. in PCRE or by overwriting of the compiled pattern.
    1580. 

  1580. <pre>
    1581. 

  1581.   PCRE_ERROR_BADCOUNT       (-15)
    1582. 

  1582. </pre>
    1583. 

  1583. This error is given if the value of the <i>ovecsize</i> argument is negative.
    1584. 

  1584. <pre>
    1585. 

  1585.   PCRE_ERROR_RECURSIONLIMIT (-21)
    1586. 

  1586. </pre>
    1587. 

  1587. The internal recursion limit, as specified by the <i>match_limit_recursion</i>
    1588. 

  1588. field in a <b>pcre_extra</b> structure (or defaulted) was reached. See the
    1589. 

  1589. description above.
    1590. 

  1590. <pre>
    1591. 

  1591.   PCRE_ERROR_BADNEWLINE     (-23)
    1592. 

  1592. </pre>
    1593. 

  1593. An invalid combination of PCRE_NEWLINE_<i>xxx</i> options was given.
    1594. 

  1594. </P>
    1595. 

  1595. <P>
    1596. 

  1596. Error numbers -16 to -20 and -22 are not used by <b>pcre_exec()</b>.
    1597. 

  1597. </P>
    1598. 

  1598. <br><a name="SEC15" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a><br>
    1599. 

  1599. <P>
    1600. 

  1600. <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
    1601. 

  1601. <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
    1602. 

  1602. <b>int <i>buffersize</i>);</b>
    1603. 

  1603. </P>
    1604. 

  1604. <P>
    1605. 

  1605. <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
    1606. 

  1606. <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
    1607. 

  1607. <b>const char **<i>stringptr</i>);</b>
    1608. 

  1608. </P>
    1609. 

  1609. <P>
    1610. 

  1610. <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
    1611. 

  1611. <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
    1612. 

  1612. </P>
    1613. 

  1613. <P>
    1614. 

  1614. Captured substrings can be accessed directly by using the offsets returned by
    1615. 

  1615. <b>pcre_exec()</b> in <i>ovector</i>. For convenience, the functions
    1616. 

  1616. <b>pcre_copy_substring()</b>, <b>pcre_get_substring()</b>, and
    1617. 

  1617. <b>pcre_get_substring_list()</b> are provided for extracting captured substrings
    1618. 

  1618. as new, separate, zero-terminated strings. These functions identify substrings
    1619. 

  1619. by number. The next section describes functions for extracting named
    1620. 

  1620. substrings.
    1621. 

  1621. </P>
    1622. 

  1622. <P>
    1623. 

  1623. A substring that contains a binary zero is correctly extracted and has a
    1624. 

  1624. further zero added on the end, but the result is not, of course, a C string.
    1625. 

  1625. However, you can process such a string by referring to the length that is
    1626. 

  1626. returned by <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>.
    1627. 

  1627. Unfortunately, the interface to <b>pcre_get_substring_list()</b> is not adequate
    1628. 

  1628. for handling strings containing binary zeros, because the end of the final
    1629. 

  1629. string is not independently indicated.
    1630. 

  1630. </P>
    1631. 

  1631. <P>
    1632. 

  1632. The first three arguments are the same for all three of these functions:
    1633. 

  1633. <i>subject</i> is the subject string that has just been successfully matched,
    1634. 

  1634. <i>ovector</i> is a pointer to the vector of integer offsets that was passed to
    1635. 

  1635. <b>pcre_exec()</b>, and <i>stringcount</i> is the number of substrings that were
    1636. 

  1636. captured by the match, including the substring that matched the entire regular
    1637. 

  1637. expression. This is the value returned by <b>pcre_exec()</b> if it is greater
    1638. 

  1638. than zero. If <b>pcre_exec()</b> returned zero, indicating that it ran out of
    1639. 

  1639. space in <i>ovector</i>, the value passed as <i>stringcount</i> should be the
    1640. 

  1640. number of elements in the vector divided by three.
    1641. 

  1641. </P>
    1642. 

  1642. <P>
    1643. 

  1643. The functions <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>
    1644. 

  1644. extract a single substring, whose number is given as <i>stringnumber</i>. A
    1645. 

  1645. value of zero extracts the substring that matched the entire pattern, whereas
    1646. 

  1646. higher values extract the captured substrings. For <b>pcre_copy_substring()</b>,
    1647. 

  1647. the string is placed in <i>buffer</i>, whose length is given by
    1648. 

  1648. <i>buffersize</i>, while for <b>pcre_get_substring()</b> a new block of memory is
    1649. 

  1649. obtained via <b>pcre_malloc</b>, and its address is returned via
    1650. 

  1650. <i>stringptr</i>. The yield of the function is the length of the string, not
    1651. 

  1651. including the terminating zero, or one of these error codes:
    1652. 

  1652. <pre>
    1653. 

  1653.   PCRE_ERROR_NOMEMORY       (-6)
    1654. 

  1654. </pre>
    1655. 

  1655. The buffer was too small for <b>pcre_copy_substring()</b>, or the attempt to get
    1656. 

  1656. memory failed for <b>pcre_get_substring()</b>.
    1657. 

  1657. <pre>
    1658. 

  1658.   PCRE_ERROR_NOSUBSTRING    (-7)
    1659. 

  1659. </pre>
    1660. 

  1660. There is no substring whose number is <i>stringnumber</i>.
    1661. 

  1661. </P>
    1662. 

  1662. <P>
    1663. 

  1663. The <b>pcre_get_substring_list()</b> function extracts all available substrings
    1664. 

  1664. and builds a list of pointers to them. All this is done in a single block of
    1665. 

  1665. memory that is obtained via <b>pcre_malloc</b>. The address of the memory block
    1666. 

  1666. is returned via <i>listptr</i>, which is also the start of the list of string
    1667. 

  1667. pointers. The end of the list is marked by a NULL pointer. The yield of the
    1668. 

  1668. function is zero if all went well, or the error code
    1669. 

  1669. <pre>
    1670. 

  1670.   PCRE_ERROR_NOMEMORY       (-6)
    1671. 

  1671. </pre>
    1672. 

  1672. if the attempt to get the memory block failed.
    1673. 

  1673. </P>
    1674. 

  1674. <P>
    1675. 

  1675. When any of these functions encounter a substring that is unset, which can
    1676. 

  1676. happen when capturing subpattern number <i>n+1</i> matches some part of the
    1677. 

  1677. subject, but subpattern <i>n</i> has not been used at all, they return an empty
    1678. 

  1678. string. This can be distinguished from a genuine zero-length substring by
    1679. 

  1679. inspecting the appropriate offset in <i>ovector</i>, which is negative for unset
    1680. 

  1680. substrings.
    1681. 

  1681. </P>
    1682. 

  1682. <P>
    1683. 

  1683. The two convenience functions <b>pcre_free_substring()</b> and
    1684. 

  1684. <b>pcre_free_substring_list()</b> can be used to free the memory returned by
    1685. 

  1685. a previous call of <b>pcre_get_substring()</b> or
    1686. 

  1686. <b>pcre_get_substring_list()</b>, respectively. They do nothing more than call
    1687. 

  1687. the function pointed to by <b>pcre_free</b>, which of course could be called
    1688. 

  1688. directly from a C program. However, PCRE is used in some situations where it is
    1689. 

  1689. linked via a special interface to another programming language that cannot use
    1690. 

  1690. <b>pcre_free</b> directly; it is for these cases that the functions are
    1691. 

  1691. provided.
    1692. 

  1692. </P>
    1693. 

  1693. <br><a name="SEC16" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a><br>
    1694. 

  1694. <P>
    1695. 

  1695. <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
    1696. 

  1696. <b>const char *<i>name</i>);</b>
    1697. 

  1697. </P>
    1698. 

  1698. <P>
    1699. 

  1699. <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
    1700. 

  1700. <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
    1701. 

  1701. <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
    1702. 

  1702. <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
    1703. 

  1703. </P>
    1704. 

  1704. <P>
    1705. 

  1705. <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
    1706. 

  1706. <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
    1707. 

  1707. <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
    1708. 

  1708. <b>const char **<i>stringptr</i>);</b>
    1709. 

  1709. </P>
    1710. 

  1710. <P>
    1711. 

  1711. To extract a substring by name, you first have to find associated number.
    1712. 

  1712. For example, for this pattern
    1713. 

  1713. <pre>
    1714. 

  1714.   (a+)b(?&#60;xxx&#62;\d+)...
    1715. 

  1715. </pre>
    1716. 

  1716. the number of the subpattern called "xxx" is 2. If the name is known to be
    1717. 

  1717. unique (PCRE_DUPNAMES was not set), you can find the number from the name by
    1718. 

  1718. calling <b>pcre_get_stringnumber()</b>. The first argument is the compiled
    1719. 

  1719. pattern, and the second is the name. The yield of the function is the
    1720. 

  1720. subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
    1721. 

  1721. that name.
    1722. 

  1722. </P>
    1723. 

  1723. <P>
    1724. 

  1724. Given the number, you can extract the substring directly, or use one of the
    1725. 

  1725. functions described in the previous section. For convenience, there are also
    1726. 

  1726. two functions that do the whole job.
    1727. 

  1727. </P>
    1728. 

  1728. <P>
    1729. 

  1729. Most of the arguments of <b>pcre_copy_named_substring()</b> and
    1730. 

  1730. <b>pcre_get_named_substring()</b> are the same as those for the similarly named
    1731. 

  1731. functions that extract by number. As these are described in the previous
    1732. 

  1732. section, they are not re-described here. There are just two differences:
    1733. 

  1733. </P>
    1734. 

  1734. <P>
    1735. 

  1735. First, instead of a substring number, a substring name is given. Second, there
    1736. 

  1736. is an extra argument, given at the start, which is a pointer to the compiled
    1737. 

  1737. pattern. This is needed in order to gain access to the name-to-number
    1738. 

  1738. translation table.
    1739. 

  1739. </P>
    1740. 

  1740. <P>
    1741. 

  1741. These functions call <b>pcre_get_stringnumber()</b>, and if it succeeds, they
    1742. 

  1742. then call <b>pcre_copy_substring()</b> or <b>pcre_get_substring()</b>, as
    1743. 

  1743. appropriate. <b>NOTE:</b> If PCRE_DUPNAMES is set and there are duplicate names,
    1744. 

  1744. the behaviour may not be what you want (see the next section).
    1745. 

  1745. </P>
    1746. 

  1746. <br><a name="SEC17" href="#TOC1">DUPLICATE SUBPATTERN NAMES</a><br>
    1747. 

  1747. <P>
    1748. 

  1748. <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
    1749. 

  1749. <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
    1750. 

  1750. </P>
    1751. 

  1751. <P>
    1752. 

  1752. When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
    1753. 

  1753. are not required to be unique. Normally, patterns with duplicate names are such
    1754. 

  1754. that in any one match, only one of the named subpatterns participates. An
    1755. 

  1755. example is shown in the
    1756. 

  1756. <a href="pcrepattern.html"><b>pcrepattern</b></a>
    1757. 

  1757. documentation.
    1758. 

  1758. </P>
    1759. 

  1759. <P>
    1760. 

  1760. When duplicates are present, <b>pcre_copy_named_substring()</b> and
    1761. 

  1761. <b>pcre_get_named_substring()</b> return the first substring corresponding to
    1762. 

  1762. the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
    1763. 

  1763. returned; no data is returned. The <b>pcre_get_stringnumber()</b> function
    1764. 

  1764. returns one of the numbers that are associated with the name, but it is not
    1765. 

  1765. defined which it is.
    1766. 

  1766. </P>
    1767. 

  1767. <P>
    1768. 

  1768. If you want to get full details of all captured substrings for a given name,
    1769. 

  1769. you must use the <b>pcre_get_stringtable_entries()</b> function. The first
    1770. 

  1770. argument is the compiled pattern, and the second is the name. The third and
    1771. 

  1771. fourth are pointers to variables which are updated by the function. After it
    1772. 

  1772. has run, they point to the first and last entries in the name-to-number table
    1773. 

  1773. for the given name. The function itself returns the length of each entry, or
    1774. 

  1774. PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
    1775. 

  1775. described above in the section entitled <i>Information about a pattern</i>.
    1776. 

  1776. Given all the relevant entries for the name, you can extract each of their
    1777. 

  1777. numbers, and hence the captured data, if any.
    1778. 

  1778. </P>
    1779. 

  1779. <br><a name="SEC18" href="#TOC1">FINDING ALL POSSIBLE MATCHES</a><br>
    1780. 

  1780. <P>
    1781. 

  1781. The traditional matching function uses a similar algorithm to Perl, which stops
    1782. 

  1782. when it finds the first match, starting at a given point in the subject. If you
    1783. 

  1783. want to find all possible matches, or the longest possible match, consider
    1784. 

  1784. using the alternative matching function (see below) instead. If you cannot use
    1785. 

  1785. the alternative function, but still need to find all possible matches, you
    1786. 

  1786. can kludge it up by making use of the callout facility, which is described in
    1787. 

  1787. the
    1788. 

  1788. <a href="pcrecallout.html"><b>pcrecallout</b></a>
    1789. 

  1789. documentation.
    1790. 

  1790. </P>
    1791. 

  1791. <P>
    1792. 

  1792. What you have to do is to insert a callout right at the end of the pattern.
    1793. 

  1793. When your callout function is called, extract and save the current matched
    1794. 

  1794. substring. Then return 1, which forces <b>pcre_exec()</b> to backtrack and try
    1795. 

  1795. other alternatives. Ultimately, when it runs out of matches, <b>pcre_exec()</b>
    1796. 

  1796. will yield PCRE_ERROR_NOMATCH.
    1797. 

  1797. <a name="dfamatch"></a></P>
    1798. 

  1798. <br><a name="SEC19" href="#TOC1">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a><br>
    1799. 

  1799. <P>
    1800. 

  1800. <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
    1801. 

  1801. <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
    1802. 

  1802. <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
    1803. 

  1803. <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
    1804. 

  1804. </P>
    1805. 

  1805. <P>
    1806. 

  1806. The function <b>pcre_dfa_exec()</b> is called to match a subject string against
    1807. 

  1807. a compiled pattern, using a matching algorithm that scans the subject string
    1808. 

  1808. just once, and does not backtrack. This has different characteristics to the
    1809. 

  1809. normal algorithm, and is not compatible with Perl. Some of the features of PCRE
    1810. 

  1810. patterns are not supported. Nevertheless, there are times when this kind of
    1811. 

  1811. matching can be useful. For a discussion of the two matching algorithms, see
    1812. 

  1812. the
    1813. 

  1813. <a href="pcrematching.html"><b>pcrematching</b></a>
    1814. 

  1814. documentation.
    1815. 

  1815. </P>
    1816. 

  1816. <P>
    1817. 

  1817. The arguments for the <b>pcre_dfa_exec()</b> function are the same as for
    1818. 

  1818. <b>pcre_exec()</b>, plus two extras. The <i>ovector</i> argument is used in a
    1819. 

  1819. different way, and this is described below. The other common arguments are used
    1820. 

  1820. in the same way as for <b>pcre_exec()</b>, so their description is not repeated
    1821. 

  1821. here.
    1822. 

  1822. </P>
    1823. 

  1823. <P>
    1824. 

  1824. The two additional arguments provide workspace for the function. The workspace
    1825. 

  1825. vector should contain at least 20 elements. It is used for keeping track of
    1826. 

  1826. multiple paths through the pattern tree. More workspace will be needed for
    1827. 

  1827. patterns and subjects where there are a lot of potential matches.
    1828. 

  1828. </P>
    1829. 

  1829. <P>
    1830. 

  1830. Here is an example of a simple call to <b>pcre_dfa_exec()</b>:
    1831. 

  1831. <pre>
    1832. 

  1832.   int rc;
    1833. 

  1833.   int ovector[10];
    1834. 

  1834.   int wspace[20];
    1835. 

  1835.   rc = pcre_dfa_exec(
    1836. 

  1836.     re,             /* result of pcre_compile() */
    1837. 

  1837.     NULL,           /* we didn't study the pattern */
    1838. 

  1838.     "some string",  /* the subject string */
    1839. 

  1839.     11,             /* the length of the subject string */
    1840. 

  1840.     0,              /* start at offset 0 in the subject */
    1841. 

  1841.     0,              /* default options */
    1842. 

  1842.     ovector,        /* vector of integers for substring information */
    1843. 

  1843.     10,             /* number of elements (NOT size in bytes) */
    1844. 

  1844.     wspace,         /* working space vector */
    1845. 

  1845.     20);            /* number of elements (NOT size in bytes) */
    1846. 

  1846. </PRE>
    1847. 

  1847. </P>
    1848. 

  1848. <br><b>
    1849. 

  1849. Option bits for <b>pcre_dfa_exec()</b>
    1850. 

  1850. </b><br>
    1851. 

  1851. <P>
    1852. 

  1852. The unused bits of the <i>options</i> argument for <b>pcre_dfa_exec()</b> must be
    1853. 

  1853. zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
    1854. 

  1854. PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL,
    1855. 

  1855. PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART. All but the last three of these are
    1856. 

  1856. the same as for <b>pcre_exec()</b>, so their description is not repeated here.
    1857. 

  1857. <pre>
    1858. 

  1858.   PCRE_PARTIAL
    1859. 

  1859. </pre>
    1860. 

  1860. This has the same general effect as it does for <b>pcre_exec()</b>, but the
    1861. 

  1861. details are slightly different. When PCRE_PARTIAL is set for
    1862. 

  1862. <b>pcre_dfa_exec()</b>, the return code PCRE_ERROR_NOMATCH is converted into
    1863. 

  1863. PCRE_ERROR_PARTIAL if the end of the subject is reached, there have been no
    1864. 

  1864. complete matches, but there is still at least one matching possibility. The
    1865. 

  1865. portion of the string that provided the partial match is set as the first
    1866. 

  1866. matching string.
    1867. 

  1867. <pre>
    1868. 

  1868.   PCRE_DFA_SHORTEST
    1869. 

  1869. </pre>
    1870. 

  1870. Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
    1871. 

  1871. soon as it has found one match. Because of the way the alternative algorithm
    1872. 

  1872. works, this is necessarily the shortest possible match at the first possible
    1873. 

  1873. matching point in the subject string.
    1874. 

  1874. <pre>
    1875. 

  1875.   PCRE_DFA_RESTART
    1876. 

  1876. </pre>
    1877. 

  1877. When <b>pcre_dfa_exec()</b> is called with the PCRE_PARTIAL option, and returns
    1878. 

  1878. a partial match, it is possible to call it again, with additional subject
    1879. 

  1879. characters, and have it continue with the same match. The PCRE_DFA_RESTART
    1880. 

  1880. option requests this action; when it is set, the <i>workspace</i> and
    1881. 

  1881. <i>wscount</i> options must reference the same vector as before because data
    1882. 

  1882. about the match so far is left in them after a partial match. There is more
    1883. 

  1883. discussion of this facility in the
    1884. 

  1884. <a href="pcrepartial.html"><b>pcrepartial</b></a>
    1885. 

  1885. documentation.
    1886. 

  1886. </P>
    1887. 

  1887. <br><b>
    1888. 

  1888. Successful returns from <b>pcre_dfa_exec()</b>
    1889. 

  1889. </b><br>
    1890. 

  1890. <P>
    1891. 

  1891. When <b>pcre_dfa_exec()</b> succeeds, it may have matched more than one
    1892. 

  1892. substring in the subject. Note, however, that all the matches from one run of
    1893. 

  1893. the function start at the same point in the subject. The shorter matches are
    1894. 

  1894. all initial substrings of the longer matches. For example, if the pattern
    1895. 

  1895. <pre>
    1896. 

  1896.   &#60;.*&#62;
    1897. 

  1897. </pre>
    1898. 

  1898. is matched against the string
    1899. 

  1899. <pre>
    1900. 

  1900.   This is &#60;something&#62; &#60;something else&#62; &#60;something further&#62; no more
    1901. 

  1901. </pre>
    1902. 

  1902. the three matched strings are
    1903. 

  1903. <pre>
    1904. 

  1904.   &#60;something&#62;
    1905. 

  1905.   &#60;something&#62; &#60;something else&#62;
    1906. 

  1906.   &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
    1907. 

  1907. </pre>
    1908. 

  1908. On success, the yield of the function is a number greater than zero, which is
    1909. 

  1909. the number of matched substrings. The substrings themselves are returned in
    1910. 

  1910. <i>ovector</i>. Each string uses two elements; the first is the offset to the
    1911. 

  1911. start, and the second is the offset to the end. In fact, all the strings have
    1912. 

  1912. the same start offset. (Space could have been saved by giving this only once,
    1913. 

  1913. but it was decided to retain some compatibility with the way <b>pcre_exec()</b>
    1914. 

  1914. returns data, even though the meaning of the strings is different.)
    1915. 

  1915. </P>
    1916. 

  1916. <P>
    1917. 

  1917. The strings are returned in reverse order of length; that is, the longest
    1918. 

  1918. matching string is given first. If there were too many matches to fit into
    1919. 

  1919. <i>ovector</i>, the yield of the function is zero, and the vector is filled with
    1920. 

  1920. the longest matches.
    1921. 

  1921. </P>
    1922. 

  1922. <br><b>
    1923. 

  1923. Error returns from <b>pcre_dfa_exec()</b>
    1924. 

  1924. </b><br>
    1925. 

  1925. <P>
    1926. 

  1926. The <b>pcre_dfa_exec()</b> function returns a negative number when it fails.
    1927. 

  1927. Many of the errors are the same as for <b>pcre_exec()</b>, and these are
    1928. 

  1928. described
    1929. 

  1929. <a href="#errorlist">above.</a>
    1930. 

  1930. There are in addition the following errors that are specific to
    1931. 

  1931. <b>pcre_dfa_exec()</b>:
    1932. 

  1932. <pre>
    1933. 

  1933.   PCRE_ERROR_DFA_UITEM      (-16)
    1934. 

  1934. </pre>
    1935. 

  1935. This return is given if <b>pcre_dfa_exec()</b> encounters an item in the pattern
    1936. 

  1936. that it does not support, for instance, the use of \C or a back reference.
    1937. 

  1937. <pre>
    1938. 

  1938.   PCRE_ERROR_DFA_UCOND      (-17)
    1939. 

  1939. </pre>
    1940. 

  1940. This return is given if <b>pcre_dfa_exec()</b> encounters a condition item that
    1941. 

  1941. uses a back reference for the condition, or a test for recursion in a specific
    1942. 

  1942. group. These are not supported.
    1943. 

  1943. <pre>
    1944. 

  1944.   PCRE_ERROR_DFA_UMLIMIT    (-18)
    1945. 

  1945. </pre>
    1946. 

  1946. This return is given if <b>pcre_dfa_exec()</b> is called with an <i>extra</i>
    1947. 

  1947. block that contains a setting of the <i>match_limit</i> field. This is not
    1948. 

  1948. supported (it is meaningless).
    1949. 

  1949. <pre>
    1950. 

  1950.   PCRE_ERROR_DFA_WSSIZE     (-19)
    1951. 

  1951. </pre>
    1952. 

  1952. This return is given if <b>pcre_dfa_exec()</b> runs out of space in the
    1953. 

  1953. <i>workspace</i> vector.
    1954. 

  1954. <pre>
    1955. 

  1955.   PCRE_ERROR_DFA_RECURSE    (-20)
    1956. 

  1956. </pre>
    1957. 

  1957. When a recursive subpattern is processed, the matching function calls itself
    1958. 

  1958. recursively, using private vectors for <i>ovector</i> and <i>workspace</i>. This
    1959. 

  1959. error is given if the output vector is not large enough. This should be
    1960. 

  1960. extremely rare, as a vector of size 1000 is used.
    1961. 

  1961. </P>
    1962. 

  1962. <br><a name="SEC20" href="#TOC1">SEE ALSO</a><br>
    1963. 

  1963. <P>
    1964. 

  1964. <b>pcrebuild</b>(3), <b>pcrecallout</b>(3), <b>pcrecpp(3)</b>(3),
    1965. 

  1965. <b>pcrematching</b>(3), <b>pcrepartial</b>(3), <b>pcreposix</b>(3),
    1966. 

  1966. <b>pcreprecompile</b>(3), <b>pcresample</b>(3), <b>pcrestack</b>(3).
    1967. 

  1967. </P>
    1968. 

  1968. <br><a name="SEC21" href="#TOC1">AUTHOR</a><br>
    1969. 

  1969. <P>
    1970. 

  1970. Philip Hazel
    1971. 

  1971. <br>
    1972. 

  1972. University Computing Service
    1973. 

  1973. <br>
    1974. 

  1974. Cambridge CB2 3QH, England.
    1975. 

  1975. <br>
    1976. 

  1976. </P>
    1977. 

  1977. <br><a name="SEC22" href="#TOC1">REVISION</a><br>
    1978. 

  1978. <P>
    1979. 

  1979. Last updated: 24 August 2008
    1980. 

  1980. <br>
    1981. 

  1981. Copyright &copy; 1997-2008 University of Cambridge.
    1982. 

  1982. <br>
    1983. 

  1983. <p>
    1984. 

  1984. Return to the <a href="index.html">PCRE index page</a>.
    1985. 

  1985. </p>
    1986.