3 // namespace: System.Text.RegularExpressions
6 // author: Dan Lewis (dlewis@gmx.co.uk)
10 // Permission is hereby granted, free of charge, to any person obtaining
11 // a copy of this software and associated documentation files (the
12 // "Software"), to deal in the Software without restriction, including
13 // without limitation the rights to use, copy, modify, merge, publish,
14 // distribute, sublicense, and/or sell copies of the Software, and to
15 // permit persons to whom the Software is furnished to do so, subject to
16 // the following conditions:
18 // The above copyright notice and this permission notice shall be
19 // included in all copies or substantial portions of the Software.
21 // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
22 // EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
23 // MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
24 // NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
25 // LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
26 // OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
27 // WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
32 using System.Collections;
33 using System.Reflection;
34 using System.Reflection.Emit;
35 using System.Runtime.Serialization;
37 using RegularExpression = System.Text.RegularExpressions.Syntax.RegularExpression;
38 using Parser = System.Text.RegularExpressions.Syntax.Parser;
40 using System.Diagnostics;
43 namespace System.Text.RegularExpressions {
46 public class Regex : ISerializable {
49 private static int cache_size = 15;
53 public static void CompileToAssembly (RegexCompilationInfo [] regexes, AssemblyName aname)
55 Regex.CompileToAssembly(regexes, aname, new CustomAttributeBuilder [] {}, null);
59 public static void CompileToAssembly (RegexCompilationInfo [] regexes, AssemblyName aname,
60 CustomAttributeBuilder [] attribs)
62 Regex.CompileToAssembly(regexes, aname, attribs, null);
66 public static void CompileToAssembly (RegexCompilationInfo [] regexes, AssemblyName aname,
67 CustomAttributeBuilder [] attribs, string resourceFile)
69 throw new NotImplementedException ();
70 // TODO : Make use of attribs and resourceFile parameters
72 AssemblyBuilder asmBuilder = AppDomain.CurrentDomain.DefineDynamicAssembly (aname, AssemblyBuilderAccess.RunAndSave);
73 ModuleBuilder modBuilder = asmBuilder.DefineDynamicModule("InnerRegexModule",aname.Name);
74 Parser psr = new Parser ();
76 System.Console.WriteLine("CompileToAssembly");
78 for(int i=0; i < regexes.Length; i++)
80 System.Console.WriteLine("Compiling expression :" + regexes[i].Pattern);
81 RegularExpression re = psr.ParseRegularExpression (regexes[i].Pattern, regexes[i].Options);
85 CILCompiler cmp = new CILCompiler (modBuilder, i);
86 bool reverse = (regexes[i].Options & RegexOptions.RightToLeft) !=0;
87 re.Compile (cmp, reverse);
93 // Define a runtime class with specified name and attributes.
94 TypeBuilder builder = modBuilder.DefineType("ITest");
96 asmBuilder.Save(aname.Name);
101 public static string Escape (string str)
103 return Parser.Escape (str);
106 public static string Unescape (string str)
108 return Parser.Unescape (str);
111 public static bool IsMatch (string input, string pattern)
113 return IsMatch (input, pattern, RegexOptions.None);
116 public static bool IsMatch (string input, string pattern, RegexOptions options)
118 Regex re = new Regex (pattern, options);
119 return re.IsMatch (input);
122 public static Match Match (string input, string pattern)
124 return Regex.Match (input, pattern, RegexOptions.None);
127 public static Match Match (string input, string pattern, RegexOptions options)
129 Regex re = new Regex (pattern, options);
130 return re.Match (input);
133 public static MatchCollection Matches (string input, string pattern)
135 return Matches (input, pattern, RegexOptions.None);
138 public static MatchCollection Matches (string input, string pattern, RegexOptions options)
140 Regex re = new Regex (pattern, options);
141 return re.Matches (input);
144 public static string Replace (string input, string pattern, MatchEvaluator evaluator)
146 return Regex.Replace (input, pattern, evaluator, RegexOptions.None);
149 public static string Replace (string input, string pattern, MatchEvaluator evaluator,
150 RegexOptions options)
152 Regex re = new Regex (pattern, options);
153 return re.Replace (input, evaluator);
156 public static string Replace (string input, string pattern, string replacement)
158 return Regex.Replace (input, pattern, replacement, RegexOptions.None);
161 public static string Replace (string input, string pattern, string replacement,
162 RegexOptions options)
164 Regex re = new Regex (pattern, options);
165 return re.Replace (input, replacement);
168 public static string [] Split (string input, string pattern)
170 return Regex.Split (input, pattern, RegexOptions.None);
173 public static string [] Split (string input, string pattern, RegexOptions options)
175 Regex re = new Regex (pattern, options);
176 return re.Split (input);
180 [MonoTODO ("should be used somewhere ? FactoryCache ?")]
181 public static int CacheSize {
182 get { return cache_size; }
185 throw new ArgumentOutOfRangeException ("CacheSize");
193 private static FactoryCache cache = new FactoryCache (200); // TODO put some meaningful number here
197 // This constructor is used by compiled regular expressions that are
198 // classes derived from Regex class. No initialization required.
203 public Regex (string pattern) : this (pattern, RegexOptions.None)
207 public Regex (string pattern, RegexOptions options)
209 this.pattern = pattern;
210 this.roptions = options;
216 this.machineFactory = cache.Lookup (this.pattern, this.roptions);
218 if (this.machineFactory == null) {
219 // parse and install group mapping
221 Parser psr = new Parser ();
222 RegularExpression re = psr.ParseRegularExpression (this.pattern, this.roptions);
223 this.group_count = re.GroupCount;
224 this.mapping = psr.GetMapping ();
229 //if ((this.roptions & RegexOptions.Compiled) != 0)
230 // //throw new Exception ("Not implemented.");
231 // cmp = new CILCompiler ();
233 cmp = new PatternCompiler ();
235 re.Compile (cmp, RightToLeft);
237 // install machine factory and add to pattern cache
239 this.machineFactory = cmp.GetMachineFactory ();
240 this.machineFactory.Mapping = mapping;
241 cache.Add (this.pattern, this.roptions, this.machineFactory);
243 this.group_count = this.machineFactory.GroupCount;
244 this.mapping = this.machineFactory.Mapping;
253 Regex (SerializationInfo info, StreamingContext context) :
254 this (info.GetString ("pattern"),
255 (RegexOptions) info.GetValue ("options", typeof (RegexOptions)))
259 #if ONLY_1_1 && !TARGET_JVM
260 // fixes public API signature
265 // public instance properties
267 public RegexOptions Options {
268 get { return roptions; }
271 public bool RightToLeft {
272 get { return (roptions & RegexOptions.RightToLeft) != 0; }
275 // public instance methods
277 public string [] GetGroupNames ()
279 string [] names = new string [mapping.Count];
280 mapping.Keys.CopyTo (names, 0);
285 public int[] GetGroupNumbers ()
287 int[] numbers = new int [mapping.Count];
288 mapping.Values.CopyTo (numbers, 0);
293 public string GroupNameFromNumber (int i)
298 foreach (string name in mapping.Keys) {
299 if ((int) mapping [name] == i)
306 public int GroupNumberFromName (string name)
308 if (mapping.Contains (name))
309 return (int) mapping [name];
316 public bool IsMatch (string input)
318 return IsMatch (input, RightToLeft ? input.Length : 0);
321 public bool IsMatch (string input, int startat)
323 return Match (input, startat).Success;
326 public Match Match (string input)
328 return Match (input, RightToLeft ? input.Length : 0);
331 public Match Match (string input, int startat)
333 return CreateMachine ().Scan (this, input, startat, input.Length);
336 public Match Match (string input, int startat, int length)
338 return CreateMachine ().Scan (this, input, startat, startat + length);
341 public MatchCollection Matches (string input)
343 return Matches (input, RightToLeft ? input.Length : 0);
346 public MatchCollection Matches (string input, int startat)
348 Match m = Match (input, startat);
349 return new MatchCollection (m);
354 public string Replace (string input, MatchEvaluator evaluator)
356 return Replace (input, evaluator, Int32.MaxValue, RightToLeft ? input.Length : 0);
359 public string Replace (string input, MatchEvaluator evaluator, int count)
361 return Replace (input, evaluator, count, RightToLeft ? input.Length : 0);
366 public Adapter (MatchEvaluator ev) { this.ev = ev; }
367 public void Evaluate (Match m, StringBuilder sb) { sb.Append (ev (m)); }
370 delegate void MatchAppendEvaluator (Match match, StringBuilder sb);
372 public string Replace (string input, MatchEvaluator evaluator, int count, int startat)
374 Adapter a = new Adapter (evaluator);
375 return Replace (input, new MatchAppendEvaluator (a.Evaluate), count, startat);
378 string Replace (string input, MatchAppendEvaluator evaluator, int count, int startat)
380 StringBuilder result = new StringBuilder ();
384 result.Append (input, 0, ptr);
386 Match m = Match (input, startat);
392 throw new SystemException ("how");
393 result.Append (input, ptr, m.Index - ptr);
394 evaluator (m, result);
396 ptr = m.Index + m.Length;
403 result.Append (input, ptr, input.Length - ptr);
405 return result.ToString ();
408 public string Replace (string input, string replacement)
410 return Replace (input, replacement, Int32.MaxValue, RightToLeft ? input.Length : 0);
413 public string Replace (string input, string replacement, int count)
415 return Replace (input, replacement, count, RightToLeft ? input.Length : 0);
418 public string Replace (string input, string replacement, int count, int startat)
420 ReplacementEvaluator ev = new ReplacementEvaluator (this, replacement);
421 return Replace (input, new MatchAppendEvaluator (ev.EvaluateAppend), count, startat);
426 public string [] Split (string input)
428 return Split (input, Int32.MaxValue, RightToLeft ? input.Length : 0);
431 public string [] Split (string input, int count)
433 return Split (input, count, RightToLeft ? input.Length : 0);
436 public string [] Split (string input, int count, int startat)
438 ArrayList splits = new ArrayList ();
440 count = Int32.MaxValue;
444 while (--count > 0) {
448 m = Match (input, ptr);
454 splits.Add (input.Substring (m.Index + m.Length, ptr - m.Index - m.Length));
456 splits.Add (input.Substring (ptr, m.Index - ptr));
458 int gcount = m.Groups.Count;
459 for (int gindex = 1; gindex < gcount; gindex++) {
460 Group grp = m.Groups [gindex];
461 splits.Add (input.Substring (grp.Index, grp.Length));
467 ptr = m.Index + m.Length;
471 if (RightToLeft && ptr >= 0)
472 splits.Add (input.Substring (0, ptr));
473 if (!RightToLeft && ptr <= input.Length)
474 splits.Add (input.Substring (ptr));
476 return (string []) splits.ToArray (typeof (string));
479 // This method is called at the end of the constructor of compiled
480 // regular expression classes to do internal initialization.
481 protected void InitializeReferences ()
484 throw new NotSupportedException ("This operation is only allowed once per object.");
486 refsInitialized = true;
488 // Compile pattern that results in performance loss as existing
489 // CIL code is ignored but provides support for regular
490 // expressions compiled to assemblies.
494 protected bool UseOptionC ()
496 return ((roptions & RegexOptions.Compiled) != 0);
499 protected bool UseOptionR ()
501 return ((roptions & RegexOptions.RightToLeft) != 0);
506 public override string ToString ()
511 // ISerializable interface
512 void ISerializable.GetObjectData (SerializationInfo info, StreamingContext context)
514 info.AddValue ("pattern", this.ToString (), typeof (string));
515 info.AddValue ("options", this.Options, typeof (RegexOptions));
520 internal int GroupCount {
521 get { return group_count; }
526 private IMachine CreateMachine ()
528 return machineFactory.NewInstance ();
531 private IMachineFactory machineFactory;
532 private IDictionary mapping;
533 private int group_count;
534 private bool refsInitialized;
539 protected internal string pattern;
540 protected internal RegexOptions roptions;
542 // MS undocumented members
544 protected internal System.Collections.Hashtable capnames;
546 protected internal System.Collections.Hashtable caps;
548 protected internal int capsize;
550 protected internal string [] capslist;
552 protected internal RegexRunnerFactory factory;