TSearch & TFileSearch Version 2.2 -Boyer-Moore-Horspool search algorithm
unit Searches; (*-----------------------------------------------------------------------------*
| Components TSearch & TFileSearch |
| Version: 2.2 |
| Last Update: 10 June 2004 |
| Compilers: Delphi 3 - Delphi 7 |
| Author: Angus Johnson - angusj-AT-myrealbox-DOT-com |
| Copyright: © 2001 -2004 Angus Johnson |
| |
| Description: Delphi implementation of the |
| Boyer-Moore-Horspool search algorithm. |
*-----------------------------------------------------------------------------*) //10.06.: Added support for widestring searches interface uses
windows, sysutils, classes; type TBaseSearch = class(TComponent)
private
fPos : pchar;
fEnd : pchar;
fPattern : string;
fPatLen : integer;
fPatInitialized : boolean;
fCaseSensitive : boolean;
JumpShift : integer;
Shift : array[#..#] of integer;
CaseBlindTable : array[#..#] of char;
procedure InitPattern;
procedure MakeCaseBlindTable;
procedure SetCaseSensitive(CaseSens: boolean);
procedure SetPattern(const Pattern: string);
procedure SetWsPattern(const WsPattern: widestring);
function FindCaseSensitive: integer;
function FindCaseInsensitive: integer;
protected
fStart : pchar;
fDataLength : integer;
procedure ClearData;
procedure SetData(Data: pchar; DataLength: integer); virtual;
public
constructor Create(aowner: tcomponent); override;
destructor Destroy; override;
//The following Find functions return the based offset of Pattern
//else POSITION_EOF (-) if the Pattern is not found ...
function FindFirst: integer;
function FindNext: integer;
function FindFrom(StartPos: integer): integer;
//To simplify searching for widestring patterns -
//assign the WsPattern property instead of the Pattern property
property WsPattern: widestring write SetWsPattern;
property Data: pchar read fStart;
property DataSize: integer read fDataLength;
published
property CaseSensitive: boolean read fCaseSensitive write SetCaseSensitive;
property Pattern: string read fPattern write SetPattern;
end; TSearch = class(TBaseSearch)
public
//Changes visibility of base SetData() method to public ...
//Note: TSearch does NOT own the data. To avoid the overhead of
//copying it, it just gets a pointer to it.
procedure SetData(Data: pchar; DataLength: integer); override;
end; TFileSearch = class(TBaseSearch)
private
fFilename: string;
procedure SetFilename(const Filename: string);
procedure Closefile;
public
destructor Destroy; override;
published
//Assigning 'Filename' creates a memory map of the named file.
//This memory mapping will be closed when either the Filename property is
//assigned to '' or the FileSearch object is destroyed.
property Filename: string read fFilename write SetFilename;
end; procedure Register; const
POSITION_EOF = -; implementation procedure Register;
begin
RegisterComponents('Samples', [TSearch, TFileSearch]);
end; //------------------------------------------------------------------------------
// TBaseSearch methods ...
//------------------------------------------------------------------------------ procedure TBaseSearch.MakeCaseBlindTable;
var
i: char;
begin
for i:= # to # do
CaseBlindTable[i]:= ansilowercase(i)[];
end;
//------------------------------------------------------------------------------ constructor TBaseSearch.Create(AOwner: TComponent);
begin
inherited Create(AOwner);
fStart := nil;
fPattern := '';
fPatLen := ;
MakeCaseBlindTable;
fCaseSensitive := false; //Default to case insensitive searches.
fPatInitialized := false;
end;
//------------------------------------------------------------------------------ destructor TBaseSearch.Destroy;
begin
ClearData;
inherited Destroy;
end;
//------------------------------------------------------------------------------ procedure TBaseSearch.ClearData;
begin
fStart := nil;
fPos := nil;
fEnd := nil;
fDataLength := ;
end;
//------------------------------------------------------------------------------ procedure TBaseSearch.SetPattern(const Pattern: string);
begin
if fPattern = Pattern then exit;
fPattern := Pattern;
fPatLen := length(Pattern);
fPatInitialized := false;
end;
//------------------------------------------------------------------------------ procedure TBaseSearch.SetWsPattern(const WsPattern: widestring);
begin
fPatLen := length(WsPattern)*;
fPatInitialized := false;
if fPatLen = then exit;
SetString(fPattern, pchar(pointer(WsPattern)), fPatLen);
end;
//------------------------------------------------------------------------------ procedure TBaseSearch.SetData(Data: pchar; DataLength: integer);
begin
ClearData;
if (Data = nil) or (DataLength < ) then exit;
fStart := Data;
fDataLength := DataLength;
fEnd := fStart + fDataLength;
end;
//------------------------------------------------------------------------------ procedure TBaseSearch.SetCaseSensitive(CaseSens: boolean);
begin
if fCaseSensitive = CaseSens then exit;
fCaseSensitive := CaseSens;
fPatInitialized := false;
end;
//------------------------------------------------------------------------------ procedure TBaseSearch.InitPattern;
var
j: integer;
i: char;
begin
if fPatLen = then exit;
for i := # to # do Shift[i]:= fPatLen;
if fCaseSensitive then
begin
for j := to fPatLen- do
Shift[fPattern[j]]:= fPatLen - j;
JumpShift := Shift[fPattern[fPatLen]];
Shift[fPattern[fPatLen]] := ;
end else
begin
for j := to fPatLen- do
Shift[CaseBlindTable[fPattern[j]]]:= fPatLen - j;
JumpShift := Shift[CaseBlindTable[fPattern[fPatLen]]];
Shift[CaseBlindTable[fPattern[fPatLen]]] := ;
end;
fPatInitialized := true;
end;
//------------------------------------------------------------------------------ function TBaseSearch.FindFirst: integer;
begin
fPos := fStart+fPatLen-;
result := FindNext;
end;
//------------------------------------------------------------------------------ function TBaseSearch.FindFrom(StartPos: integer): integer;
begin
if StartPos < fPatLen- then //ie: StartPos must never be less than fPatLen-
fPos := fStart+fPatLen- else
fPos := fStart+StartPos;
result := FindNext;
end;
//------------------------------------------------------------------------------ function TBaseSearch.FindNext: integer;
begin
if not fPatInitialized then InitPattern;
if (fPatLen = ) or (fPatLen >= fDataLength) or (fPos >= fEnd) then
begin
fPos := fEnd;
result := POSITION_EOF;
exit;
end;
if fCaseSensitive then
result := FindCaseSensitive else
result := FindCaseInsensitive;
end;
//------------------------------------------------------------------------------ function TBaseSearch.FindCaseSensitive: integer;
var
i: integer;
j: pchar;
begin
result:= POSITION_EOF;
while fPos < fEnd do
begin
i := Shift[fPos^]; //test last character first
if i <> then //last char does not match
inc(fPos,i)
else
begin //last char matches at least
i := ;
j := fPos - fPatLen;
while (i < fPatLen) and (fPattern[i] = (j+i)^) do inc(i);
if (i = fPatLen) then
begin
result:= fPos-fStart-fPatLen+;
inc(fPos,fPatLen);
break; //FOUND!
end
else
inc(fPos,JumpShift);
end;
end;
end;
//------------------------------------------------------------------------------ function TBaseSearch.FindCaseInsensitive: integer;
var
i: integer;
j: pchar;
begin
result:= POSITION_EOF;
while fPos < fEnd do
begin
i := Shift[CaseBlindTable[fPos^]]; //test last character first
if i <> then //last char does not match
inc(fPos,i)
else
begin //last char matches at least
i := ;
j := fPos - fPatLen;
while (i < fPatLen) and
(CaseBlindTable[fPattern[i]] = CaseBlindTable[(j+i)^]) do inc(i);
if (i = fPatLen) then
begin
result:= fPos-fStart-fPatLen+;
inc(fPos,fPatLen);
break; //FOUND!
end
else
inc(fPos,JumpShift);
end;
end;
end; //------------------------------------------------------------------------------
// TSearch methods ...
//------------------------------------------------------------------------------ procedure TSearch.SetData(Data: pchar; DataLength: integer);
begin
inherited; //changes visibility of base method from protected to public
end; //------------------------------------------------------------------------------
// TFileSearch methods ...
//------------------------------------------------------------------------------ destructor TFileSearch.Destroy;
begin
CloseFile;
inherited Destroy;
end;
//------------------------------------------------------------------------------ procedure TFileSearch.SetFilename(const Filename: string);
var
filehandle: integer;
filemappinghandle: thandle;
size, highsize: integer;
begin
if (csDesigning in ComponentState) then
begin
fFilename := Filename;
exit;
end;
CloseFile;
if (Filename = '') or not FileExists(Filename) then exit;
filehandle := sysutils.FileOpen(Filename, fmopenread or fmsharedenynone);
if filehandle = then exit; //error
size := GetFileSize(filehandle, @highsize);
if (size <= ) or (highsize <> ) then //nb: files > gig not supported
begin
CloseHandle(filehandle);
exit;
end;
filemappinghandle :=
CreateFileMapping(filehandle, nil, page_readonly, , , nil);
if GetLastError = error_already_exists then filemappinghandle := ;
if filemappinghandle <> then
SetData(MapViewOfFile(filemappinghandle,file_map_read,,,),size);
if fStart <> nil then fFilename := Filename;
CloseHandle(filemappinghandle);
CloseHandle(filehandle);
end;
//------------------------------------------------------------------------------ procedure TFileSearch.CloseFile;
begin
if (csDesigning in ComponentState) then exit;
if (fStart <> nil) then UnmapViewOfFile(fStart);
fFilename := '';
ClearData;
end;
//------------------------------------------------------------------------------ end.
TSearch & TFileSearch Version 2.2 -Boyer-Moore-Horspool search algorithm的更多相关文章
- Leetcode OJ : Implement strStr() [ Boyer–Moore string search algorithm ] python solution
class Solution { public: int strStr(char *haystack, char *needle) { , skip[]; char *str = haystack, ...
- Moore majority vote algorithm(摩尔投票算法)
Boyer-Moore majority vote algorithm(摩尔投票算法) 简介 Boyer-Moore majority vote algorithm(摩尔投票算法)是一种在线性时间O( ...
- Boyer–Moore (BM)字符串搜索算法
在计算机科学里,Boyer-Moore字符串搜索算法是一种非常高效的字符串搜索算法.它由Bob Boyer和J Strother Moore设计于1977年.此算法仅对搜索目标字符串(关键字)进行预处 ...
- Boyer Moore算法(字符串匹配)
上一篇文章,我介绍了KMP算法. 但是,它并不是效率最高的算法,实际采用并不多.各种文本编辑器的"查找"功能(Ctrl+F),大多采用Boyer-Moore算法. Boyer-Mo ...
- Boyer-Moore 字符串匹配算法
字符串匹配问题的形式定义: 文本(Text)是一个长度为 n 的数组 T[1..n]: 模式(Pattern)是一个长度为 m 且 m≤n 的数组 P[1..m]: T 和 P 中的元素都属于有限的字 ...
- grep之字符串搜索算法Boyer-Moore由浅入深(比KMP快3-5倍)
这篇长文历时近两天终于完成了,前两天帮网站翻译一篇文章“为什么GNU grep如此之快?”,里面提及到grep速度快的一个重要原因是使用了Boyer-Moore算法作为字符串搜索算法,兴趣之下就想了解 ...
- grep之字符串搜索算法Boyer-Moore由浅入深(比KMP快3-5倍)(转)
这篇长文历时近两天终于完成了,前两天帮网站翻译一篇文章“为什么GNU grep如此之快?”,里面提及到grep速度快的一个重要原因是使用了Boyer-Moore算法作为字符串搜索算法,兴趣之下就想了解 ...
- KMP算法简单回顾
前言 虽从事企业应用的设计与开发,闲暇之时,还是偶尔涉猎数学和算法的东西,本篇根据个人角度来写一点关于KMP串匹配的东西,一方面向伟人致敬,另一方面也是练练手,头脑风暴.我在自娱自乐,路过的朋友别太认 ...
- Erlang/Elixir精选-第5期(20200106)
The forgotten ideas in computer science-Joe Armestrong 在2020年的第一期里面,一起回顾2018年Joe的 The forgotten idea ...
随机推荐
- Netty框架入门
一.概述 Netty是由JBOSS提供的一个java开源框架. Netty提供异步的.事件驱动的网络应用程序框架和工具,用以快速开发高性能.高可靠性的网络服务器和客户端程序. 二. ...
- Metro应用Json数据处理
Windows Phone 8 或者 Windows 8 平台对JSON数据的处理方式基本是一致的,需要使用DataContractJsonSerializer类将对象的实例序列化为JSON字符串,并 ...
- java基础4 函数
本文知识点(目录): 1.函数的概述 2.函数的格式 3.自定义函数 4.函数的特点 5.函数的应用 6.函数的重载 1.函数的概述 发现不断进行加法运算,为了提高代码的 ...
- 半小时分组统计个数sql
group by 最后一个时间是多少按多少分组 select count(1), trunc(a.refund_insert_time, 'hh24') + case when to_char(ref ...
- POJ 2752 Seek the Name, Seek the Fame(KMP求公共前后缀)
题目链接:http://poj.org/problem?id=2752 题目大意:给你一串字符串s找到所有的公共前后缀,即既是前缀又是后缀的子串. 解题思路: 如图所示 假设字符串pi与jq为符合条件 ...
- django orm如何作一个优雅一点的filter?
如果有N多fitler条件, 单独放在一个长语句里显然不好看. 还好, django支持字典方式的过滤条件, 写法大约与单独的长语里差不多. 如下: def get_queryset(self): f ...
- 【C#】数据类型(sbyte,byte,short,ushort,int,uint,long,ulong和char。、、、)
C#的数据类型可以分为3类:数值类型,引用类型,指针类型.指针类型仅在不安全代码中使用. 值类型包括简单类型(如字符型,浮点型和整数型等),集合类型和结构型.引用类型包括类类型,接口类型,代表类型和数 ...
- 2017-2018-1 20179202《Linux内核原理与分析》第五周作业
一.系统调用实验: 1.知识准备: (1)系统调用的三层皮:xyz()(API).system_call(所有系统调用的入口) . sys_xyz()(中断服务程序) (2)API是程序员在用户空间下 ...
- 牛客练习赛9 F - 珂朵莉的约数
题目描述 珂朵莉给你一个长为n的序列,有m次查询 每次查询给两个数l,r 设s为区间[l,r]内所有数的乘积 求s的约数个数mod 1000000007 输入描述: 第一行两个正整数n,m第二行一个长 ...
- sql 和xml
一·常用基本功能 表结果: 1. 1 <row> 2 <ID>1035</ID> 3 <Rate>6.12</Rate> 4 <Dat ...