Good speaking style [closed]

 use utf8; sub correct{ my $string = shift; $string =~ s/(\w+)/\L$1/g; $string =~ s/(\w+)\s*([.,!?:]+)\s*/$1$2 /g; $string =~ s/(\A|[.!?]+)(\s*)(\w+)/$1$2\u$3/g; return $string; } print correct "раз ,два, три , четыри, ПЯТЬ - вышел ZaIcHiK погулять!ТУТ охотник выбегает ...прямо в зайчика СТРЕЛЯЕТ ? КОНЕЦ ."; 

Result: One, two, three, four, five - zaichik went out for a walk! Then the hunter runs out ... Shoots right at the bunny? The end.

PS borrowed an example for the test from Ilya Pirogov , I hope he will not be offended :) PS This is a solution in Perl

 string Corrector(string innerText, UserInfo userInfo) { var reg1 = new Regex(@"[\.\,\:\!\?][^\s]"); //после знака препинания нет пробела или переноса var reg2 = new Regex(@"[\.]\ [^А-ЯA-Z]"); //после точки и пробела не заглавная буква var reg3 = new Regex(@"\b[^А-ЯA-Z]"); //после переноса строки не заглавная буква var reg4 = new Regex(@"\ [\.\,\:\!\?]"); //пробел перед знаком препинания if(reg1.IsMatch(innerText) || reg2.IsMatch(innerText) || reg3.IsMatch(innerText) || reg4.IsMatch(innerText)) { return("Прошу прощения, товарищи. К сожалению, я безнадежно безграмотная соволочь! В связи с этим, не имею возможности в письменном виде высказать вам свою мысль."); } return innerText; } 

In PHP:

 function corrector($string){ $string=strtolower($string); $patterns = array(); $patterns[0] = '/ *\. */'; $patterns[1] = '/ *\, */'; $patterns[2] = '/ *\! */'; $patterns[3] = '/ *\? */'; $patterns[4] = '/ *\: */'; $replacements = array(); $replacements[0] = '. '; $replacements[1] = ', '; $replacements[2] = '! '; $replacements[3] = '? '; $replacements[4] = ': '; $string=preg_replace($patterns, $replacements, $string); $string=preg_replace('/\. \. \. /', '... ', $string); $string=preg_replace_callback( '/([\.\!\?]|(\.\.\.)) [a-zA-Zа-яА-Я]/', create_function( '$matches', 'return strtoupper($matches[0]);' ), $string ); $string=preg_replace_callback( '/^ *[a-zA-Zа-яА-Я]/', create_function( '$matches', 'return strtoupper($matches[0]);' ), $string ); return $string; } $string=" test . ..test ?teSt ."; echo corrector($string); 

That's what I got on java + oop without regexp . Enforcing different formatting for hyphens and dashes was simply not enough patience = /

 input string (для наглядности строки в кавычках): ' hello DUMMY-user ! a lot of thanks ,- this is a "Hello World" Text ... just to check ' result string: 'Hello dummy - user! A lot of thanks, - this is a "hello world" text... just to check.' 

USE_CUSTOM_FORMATTERS - a switch of additional formatters that are needed, but were not explicitly specified in the statement of the problem

 public class FormatStr { // debug mode switcher - dump results after each formatter private static boolean DEBUG = false; // additional formatters switcher - use additional formatters private static boolean USE_CUSTOM_FORMATTERS = true; public static void main(String[] args) { // create list List<ABaseFormatter> formatters = new LinkedList<ABaseFormatter>(); // add extra formatters if (USE_CUSTOM_FORMATTERS) { formatters.add(new DashFormatter()); formatters.add(new TrimStringFormatter()); formatters.add(new RemoveExtraWhitespacesFormatter()); } // add required formatters formatters.add(new LastDotFormatter()); formatters.add(new StringToLowerCaseFormatter()); formatters.add(new FirstCharToCapitalFormatter()); formatters.add(new SentenceFromUpperCaseCharFormatter()); formatters.add(new BeforePunctMarkFormatter()); formatters.add(new AfterPunctMarkFormatter()); // input str String str = " hello DUMMY-user ! a lot of thanks ,- this is a \"Hello World\" Text ... just \nto check "; dump("input string:", str); String result = str; for (ABaseFormatter sf : formatters) { result = sf.format(result); if (DEBUG) { dump("applied formatter " + sf.getClass().getName(), result); } } dump("result string:", result); } private static void dump(final String purpose, final String str) { System.out.println(purpose + "\n'" + str + "'\n"); } } abstract class ABaseFormatter { // private Set<Character> set = new HashSet<Character>(); public static Set<Character> SENTENCE_SEPARATOR_CHARS = new HashSet<Character>(); public static Set<Character> PUNCTUATION_CHARS = new HashSet<Character>(); static { Collections.addAll(SENTENCE_SEPARATOR_CHARS, '.', '!', '?'); Collections.addAll(PUNCTUATION_CHARS, ',', '.', '!', '?', ':'); } /** * Format string * * @param string */ public final String format(String string) { if (null == string) { return null; } return formatStr(string); } /** * Override to apply formatting to whole string * * @param str */ protected String formatStr(String str) { StringBuilder buff = new StringBuilder(str); // default behavior - apply per character formatting for (int i = 0; i < buff.length();) { i = formatChar(buff, i, buff.charAt(i)); } return buff.toString(); } /** * Override to apply formatting per char * * @param buff * @param pos * @param ch */ protected int formatChar(StringBuilder buff, int pos, char ch) { return pos + 1; } } class SentenceFromUpperCaseCharFormatter extends ABaseFormatter { @Override protected int formatChar(StringBuilder buff, int pos, char ch) { if (pos == 0) { // find first letter if possible while (pos < buff.length()) { if (Character.isLetter(buff.charAt(pos))) { buff.setCharAt(pos, Character.toUpperCase(buff.charAt(pos))); break; } pos++; } } return pos + 1; } } class AfterPunctMarkFormatter extends ABaseFormatter { @Override protected int formatChar(StringBuilder buff, int pos, char ch) { // add whitespace after punctuation mark if (PUNCTUATION_CHARS.contains(ch)) { // if dots are coming just move to next one if ('.' != ch) { // add whitespace between punct mark and word ch = buff.charAt(pos + 1); if (' ' != ch && '\n' != ch) { buff.insert(pos + 1, ' '); } } } return pos + 1; } } class BeforePunctMarkFormatter extends ABaseFormatter { @Override protected int formatChar(StringBuilder str, int pos, char ch) { // delete leading whitespaces before punctuation marks if (PUNCTUATION_CHARS.contains(ch)) { // count leading whitespaces to the punctuation mark; for (int start = pos; start > 1; start--) { char prevCh = str.charAt(start - 1); if (' ' != prevCh && '\n' != prevCh) { str.replace(start, pos, ""); pos = start + 1; break; } } // new sentence should start from upper case letter if (SENTENCE_SEPARATOR_CHARS.contains(ch) && (pos > 0 && str.charAt(pos - 1) != '.')) { // find first letter if possible while (pos < str.length()) { if (Character.isLetter(str.charAt(pos))) { str.setCharAt(pos, Character.toUpperCase(str.charAt(pos))); break; } pos++; } } } return pos + 1; } } class LastDotFormatter extends ABaseFormatter { @Override protected String formatStr(String str) { // default behavior - apply per character formatting for (int i = str.length() - 1; i >= 0; i--) { char ch = str.charAt(i); if (SENTENCE_SEPARATOR_CHARS.contains(ch)) { break; } // add end of sentence char to the last character // of the last sentence if (Character.isLetter(ch)) { return str + '.'; } } return str; } } class FirstCharToCapitalFormatter extends ABaseFormatter { @Override public String formatStr(String string) { return Character.toUpperCase(string.charAt(0)) + string.substring(1); } } class RemoveExtraWhitespacesFormatter extends ABaseFormatter { @Override public String formatStr(String string) { while (-1 != string.indexOf(" ")) { string = string.replaceAll(" ", " "); } return string; } } class StringToLowerCaseFormatter extends ABaseFormatter { @Override public String formatStr(String string) { return string.toLowerCase(); } } class TrimStringFormatter extends ABaseFormatter { @Override public String formatStr(String string) { return string.trim(); } } class DashFormatter extends ABaseFormatter { @Override protected int formatChar(StringBuilder buff, int pos, char ch) { // delete leading whitespaces before dashes if ('-' == ch) { // add white space before dash ch = buff.charAt(pos - 1); if (pos > 1 && ' ' != ch && '\n' != ch) { buff.insert(pos, ' '); } // add whitespace after pos = pos + 1; ch = buff.charAt(pos + 1); if (pos + 1 < buff.length() && (' ' != ch && '\n' != ch)) { buff.insert(pos + 1, ' '); } } return pos + 1; } } 

And I decided to implement it in C ++ . Maybe Hashcode will be useful, who knows ... despite the fact that the task is almost elementary (and, I think, standard). I did everything through classes (well, how else !?), so you can embed it into any system without going into the details of the functioning of this class, knowing only one of its methods. Encapsulation, respectively. Anyway, such things are not reasonable to implement not through classes, since there is a behavior here. Please love and respect:

 #include<iostream> #include<fstream> #include<clocale> #include<string> #include<vector> #include<algorithm> using namespace std; class Corrector // наш класс, реализующий функционал корректора { private: char* filename; // имя файла, где лежит BAD-TEXT. std::string text; // текст из этого файла std::vector<std::string>lines; // все лексемы( Tok = ' ') void CreateText() { for(vector<string>::iterator itr=lines.begin();itr!=lines.end();++itr) text+=*itr+" "; transform(text.begin(),text.end(),text.begin(),tolower); text[0] = toupper(text[0]); Correct(); } public: Corrector(char* fn):filename(fn){setlocale(LC_ALL,"Russian");} void LoadText() // метод загрузки текста { std::string str; ifstream f(filename); while(!f.eof()) { f>>str; lines.push_back(str); } f.close(); CreateText(); } void Correct() // корректор исходного текста { int i = 0; for(string::iterator itr=text.begin();itr!=text.end();++itr) { i++; if(*itr==',' || *itr=='!' || *itr=='?' || *itr=='.' || *itr==';') if(i>0) if(*(itr+1)!=' ') text.insert(itr+1,' '); if(*itr=='!' || *itr=='?' || *itr=='.') if(i>0) if(*(itr-1)==' ') text.erase(itr-1,itr); if(i>2) if(*(itr-2)=='!' || *(itr-2)=='?' || *(itr-2)=='.') *itr = toupper(*itr); } } void SaveText(char* fn) // Сохранение отформатированного текста в Файл "fn" { ofstream f(fn); f<<text<<endl; f.close(); } void Print(){cout<<text<<endl;} // отладочный метод... }; int main() { setlocale(LC_ALL,"Russian"); Corrector corr("badtext.txt"); // корректор corr.LoadText(); // загружаем текст corr.Print(); // отладочный вызов corr.SaveText("formatted.txt"); system("Pause"); } 

Let's say if the source text was like this:

 когда-то я был таким же шалопаем,как и вы,поэтому приношу свои извинения !! совершенно безграмотная сволочь(нет,нет,нет!) но а вы,а !? Так,что,"довай,до свыдааания!" !!! 

So at the output it turns out like this

 Когда-то я был таким же шалопаем, как и вы, поэтому приношу свои извинения!! Совершенно безграмотная сволочь(нет, нет, нет! ) но а вы, а!? Так, что, "довай, до свыдааания! "!!! 

The code is written literally "on the knee", so if errors are found, I will correct them ...

PS I apologize for being cumbersome, but this is C ++ ... nevertheless nothing bad - the code is fully embedded and scalable. Waiting for comments (in style including!) =)))

And Vova for this need to ban