static void Main()
{
var rankings = new Dictionary <string, string>();
foreach (var line in System.IO.File.ReadAllLines("../../DraftTierList_01-11-2018_Overall.txt"))
{
var entry = line.Split(';');
rankings.Add(entry[0], entry[1]);
}
string path = @"../../EternalCardName_Corpus2.txt";
if (!SymSpell.CreateDictionary(path, ""))
{
Console.Error.WriteLine("File not found: " + System.IO.Path.GetFullPath(path));
}
//verbosity=Top=0: the suggestion with the highest term frequency of the suggestions of smallest edit distance found
//verbosity=Closest=1: all suggestions of smallest edit distance found, the suggestions are ordered by term frequency
//verbosity=All=2: all suggestions <= maxEditDistance, the suggestions are ordered by edit distance, then by term frequency (slower, no early termination)
SymSpell.verbose = 2;
SymSpell.editDistanceMax = 3;
//SymSpell.lp = 7;
Application.EnableVisualStyles();
Application.SetCompatibleTextRenderingDefault(false);
Application.Run(new Overlay(rankings));
}
static void Main(string[] args)
{
int initialCapacity = 82765;
string dictionaryPath = @"D:\sbwce.txt";
int prefixLength = 7;
int maxEditDistanceDictionary = 2;
var symSpell = new SymSpell(initialCapacity, maxEditDistanceDictionary, prefixLength);
symSpell.CreateDictionary(dictionaryPath);
Type typecontroller = typeof(SymSpell);
System.Reflection.FieldInfo finfo = typecontroller.GetField("words", System.Reflection.BindingFlags.Instance | System.Reflection.BindingFlags.NonPublic | System.Reflection.BindingFlags.GetField);
System.Collections.Generic.Dictionary <string, System.Int64> collection = null;
if (finfo != null)
{
collection = (System.Collections.Generic.Dictionary <string, System.Int64>)finfo.GetValue(symSpell);
}
using (System.IO.StreamWriter fileWriter = new System.IO.StreamWriter(@"D:\Personal\Master\Materias\TFM SLN\DictionaryFiles\load4.log"))
{
foreach (System.Collections.Generic.KeyValuePair <string, System.Int64> kvPair in collection)
{
if (kvPair.Value > 50 && !Regex.IsMatch(kvPair.Key, @"^-?\d+$"))
{
fileWriter.WriteLine("{0} {1}", kvPair.Key, kvPair.Value);
}
}
fileWriter.Close();
};
}
static void Main(string[] args)
{
if (args.Length > 2)
{
Console.Error.Write("Creating dictionary ...");
long memSize = GC.GetTotalMemory(true);
Stopwatch stopWatch = new Stopwatch();
stopWatch.Start();
//parameters
int initialCapacity = 82765;
int termIndex = 0; //column of the term in the dictionary text file
int countIndex = 1; //column of the term frequency in the dictionary text file
//dictionaryType
string dictionaryType = args[0].ToLower();
if ("load.create".IndexOf(dictionaryType) == -1)
{
Console.Error.WriteLine("Error in parameter 1"); return;
}
//dictionaryPath
string dictionaryPath = AppDomain.CurrentDomain.BaseDirectory + args[1];
//prefix length (optional parameter)
int offset = 0;
string lookupType = "";
int prefixLength = 7;
if (!int.TryParse(args[2], out prefixLength))
{
prefixLength = 7;
}
else
{
offset = 1;
}
//lookupType
if (args.Length > 2 + offset)
{
lookupType = args[2 + offset].ToLower();
if ("lookup.lookupcompound.wordsegment".IndexOf(lookupType) == -1)
{
Console.Error.WriteLine("Error in parameter " + (3 + offset).ToString()); return;
}
}
//maxEditDistance
int maxEditDistanceDictionary = 2; //maximum edit distance per dictionary precalculation
if (args.Length > 3 + offset)
{
if (!int.TryParse(args[3 + offset], out maxEditDistanceDictionary))
{
Console.Error.WriteLine("Error in parameter " + (4 + offset).ToString()); return;
}
}
//output stats
bool outputStats = false;//false, true
if (args.Length > 4 + offset)
{
if (!bool.TryParse(args[4 + offset], out outputStats))
{
Console.Error.WriteLine("Error in parameter " + (5 + offset).ToString()); return;
}
}
//verbosity
var suggestionVerbosity = SymSpell.Verbosity.Top; //Top, Closest, All
if (args.Length > 5 + offset)
{
if (!Enum.TryParse(args[5 + offset], true, out suggestionVerbosity))
{
Console.Error.WriteLine("Error in parameter " + (6 + offset).ToString()); return;
}
}
//create object
var symSpell = new SymSpell(initialCapacity, maxEditDistanceDictionary, prefixLength);
//load dictionary
switch (dictionaryType)
{
case "load":
if (!symSpell.LoadDictionary(dictionaryPath, termIndex, countIndex))
{
Console.Error.WriteLine("File not found!");
return;
}
break;
case "create":
if (!symSpell.CreateDictionary(dictionaryPath))
{
Console.Error.WriteLine("File not found!");
return;
}
break;
default:
break;
}
stopWatch.Stop();
long memDelta = GC.GetTotalMemory(true) - memSize;
//not to stdout, but to Console.Error: status info will alway be on console, but not redirected or piped
Console.Error.WriteLine("\rDictionary: " + symSpell.WordCount.ToString("N0") + " words, "
+ symSpell.EntryCount.ToString("N0") + " entries, edit distance=" + symSpell.MaxDictionaryEditDistance.ToString()
+ " in " + stopWatch.Elapsed.TotalMilliseconds.ToString("0.0") + "ms "
+ (memDelta / 1024 / 1024.0).ToString("N0") + " MB");
//warm up
var result = symSpell.Lookup("warmup", SymSpell.Verbosity.All);
//lookup suggestions for single-word input strings
string inputTerm;
while (!string.IsNullOrEmpty(inputTerm = (Console.ReadLine() ?? "").Trim()))
{
switch (lookupType)
{
case "lookup":
var suggestions = symSpell.Lookup(inputTerm, suggestionVerbosity, maxEditDistanceDictionary, true);
//display suggestions, edit distance and term frequency
foreach (var suggestion in suggestions)
{
if (outputStats)
{
Console.WriteLine(suggestion.term + " " + suggestion.distance.ToString() + " " + suggestion.count.ToString("N0"));
}
else
{
Console.WriteLine(suggestion.term);
}
}
break;
case "lookupcompound":
var suggestions2 = symSpell.LookupCompound(inputTerm);
//display suggestions, edit distance and term frequency
foreach (var suggestion in suggestions2)
{
if (outputStats)
{
Console.WriteLine(suggestion.term + " " + suggestion.distance.ToString() + " " + suggestion.count.ToString("N0"));
}
else
{
Console.WriteLine(suggestion.term);
}
}
break;
case "wordsegment":
var suggestions3 = symSpell.WordSegmentation(inputTerm);
//display suggestions, edit distance and term frequency
foreach (var suggestion in suggestions3)
{
if (outputStats)
{
Console.WriteLine(suggestion.correctedString + " " + suggestion.distanceSum.ToString("N0") + " " + suggestion.probabilityLogSum.ToString());
}
else
{
Console.WriteLine(suggestion.correctedString);
}
}
break;
default:
break;
}
}
}
else
{
//PrefixLength is number
//help
Console.WriteLine("SymSpell.CommandLine DictionaryType DictionaryPath [PrefixLength] LookupType [MaxEditDistance] [OutputStats] [Verbosity]");
Console.WriteLine();
Console.WriteLine("DictionaryType=load|create");
Console.WriteLine(" load: load dictionary from dictionary file");
Console.WriteLine(" create: create dictionary from text corpus");
Console.WriteLine("DictionaryPath: path to dictionary/corpus file");
Console.WriteLine("PrefixLength: default=7 (speed/memory consumption trade-off)"); //dictionary param
Console.WriteLine(" 5: low memory, slow lookup");
Console.WriteLine(" 6: medium memory, medium lookup");
Console.WriteLine(" 7: high memory, fast lookup");
//lookup intended for correction of single word
//lookupcompound intended for correction of multiple words, it can insert only a single space per token, faster than wordsegmentation
//wordsegmentation intended for segmentation and correction of multiple words, it can insert multiple spaces per token, slower than lookupcompound
Console.WriteLine("LookupType=lookup|lookupcompound|wordsegment");
Console.WriteLine(" lookup: correct single word");
Console.WriteLine(" lookupcompound: correct multiple-word string (supports splitting/merging)");
Console.WriteLine(" wordsegment: word segment and correct input string");
Console.WriteLine("MaxEditDistance: default=2 (0: no correction, word segmentation only)");
Console.WriteLine("OutputStats=false|true");
Console.WriteLine(" false: only corrected string");
Console.WriteLine(" true: corrected string, edit distance, word frequency/probability");
Console.WriteLine("Verbosity=top|closest|all"); //no effect for lookupcompound and wordsegment
Console.WriteLine(" top: Top suggestion");
Console.WriteLine(" closest: All suggestions of smallest edit distance found");
Console.WriteLine(" all: All suggestions within maxEditDistance");
Console.WriteLine();
}
}
static void Main(string[] args)
{
if (args.Length >= 2)
{
Console.Error.Write("Creating dictionary ...");
long memSize = GC.GetTotalMemory(true);
Stopwatch stopWatch = new Stopwatch();
stopWatch.Start();
//parameters
int initialCapacity = 82765;
int maxEditDistanceDictionary = 2; //maximum edit distance per dictionary precalculation
if (args.Length > 2)
{
if (!int.TryParse(args[2], out maxEditDistanceDictionary))
{
Console.Error.WriteLine("Error in parameter 3"); return;
}
}
int maxEditDistanceLookup = maxEditDistanceDictionary; //max edit distance per lookup
var suggestionVerbosity = SymSpell.Verbosity.Top; //Top, Closest, All
if (args.Length > 3)
{
if (!Enum.TryParse(args[3], out suggestionVerbosity))
{
Console.Error.WriteLine("Error in parameter 4"); return;
}
}
int prefixLength = 7;
if (args.Length > 4)
{
if (!int.TryParse(args[4], out prefixLength))
{
Console.Error.WriteLine("Error in parameter 5"); return;
}
}
string dictionaryPath = AppDomain.CurrentDomain.BaseDirectory + args[1]; // "../../../../SymSpell/frequency_dictionary_en_82_765.txt";
int termIndex = 0; //column of the term in the dictionary text file
int countIndex = 1; //column of the term frequency in the dictionary text file
//create object
var symSpell = new SymSpell(initialCapacity, maxEditDistanceDictionary, prefixLength);
//load dictionary
switch (args[0].ToLower())
{
case "load":
if (!symSpell.LoadDictionary(dictionaryPath, termIndex, countIndex))
{
Console.Error.WriteLine("File not found!");
return;
}
break;
case "create":
if (!symSpell.CreateDictionary(dictionaryPath))
{
Console.Error.WriteLine("File not found!");
return;
}
break;
default:
break;
}
stopWatch.Stop();
long memDelta = GC.GetTotalMemory(true) - memSize;
//not to stdout, but to Console.Error: status info will alway be on console, but not redirected or piped
Console.Error.WriteLine("\rDictionary: " + symSpell.WordCount.ToString("N0") + " words, "
+ symSpell.EntryCount.ToString("N0") + " entries, edit distance=" + symSpell.MaxDictionaryEditDistance.ToString()
+ " in " + stopWatch.Elapsed.TotalMilliseconds.ToString("0.0") + "ms "
+ (memDelta / 1024 / 1024.0).ToString("N0") + " MB");
//warm up
var result = symSpell.Lookup("warmup", SymSpell.Verbosity.All, 1);
//lookup suggestions for single-word input strings
string inputTerm;
while (!string.IsNullOrEmpty(inputTerm = (Console.ReadLine() ?? "").Trim()))
{
var suggestions = symSpell.Lookup(inputTerm, suggestionVerbosity, maxEditDistanceLookup, true);
//display suggestions, edit distance and term frequency
foreach (var suggestion in suggestions)
{
Console.WriteLine(suggestion.term + " " + suggestion.distance.ToString() + " " + suggestion.count.ToString("N0"));
}
}
}
else
{
//help
Console.WriteLine("SymSpell.CommandLine load Path [MaxEditDistance] [Verbosity] [PrefixLength]");
Console.WriteLine("SymSpell.CommandLine create Path [MaxEditDistance] [Verbosity] [PrefixLength]");
Console.WriteLine();
Console.WriteLine("load: load dictionary from dictionary file");
Console.WriteLine("create: create dictionary from text corpus");
Console.WriteLine("MaxEditDistance: default=2");
Console.WriteLine("Verbosity=Top|Closest|All (case-sensitive)");
Console.WriteLine("PrefixLength: default=7 (5:low memory; 7:fast lookup)");
Console.WriteLine();
}
}
//Load a frequency dictionary or create a frequency dictionary from a text corpus
public static void Main(string[] args)
{
// Console.Write("Creating dictionary ...");
long memSize = GC.GetTotalMemory(true);
Stopwatch stopWatch = new Stopwatch();
stopWatch.Start();
//set parameters
const int initialCapacity = 82765;
const int maxEditDistance = 3;
const int prefixLength = 7;
var symSpell = new SymSpell(initialCapacity, maxEditDistance, prefixLength);
symSpell.CreateDictionary(@"C:\Dev\csharp\names.txt");
const SymSpell.Verbosity verbosity = SymSpell.Verbosity.Top;
//var suggestions = symSpell.Lookup("Minuie Coare", verbosity);
// Console.WriteLine(suggestions.Count);
// foreach (var suggestion in suggestions)
// {
// Console.WriteLine(suggestion);
// }
using (Stream corpusStream = File.OpenRead(@"C:\Dev\csharp\keys.txt"))
{
using (StreamReader sr = new StreamReader(corpusStream))
{
String line;
//process a single line at a time only for memory efficiency
while ((line = sr.ReadLine()) != null)
{
var suggestions = symSpell.Lookup(line.Trim(), verbosity);
Console.WriteLine(suggestions.Count);
foreach (var suggestion in suggestions)
{
Console.WriteLine(suggestion);
}
}
}
}
// //Load a frequency dictionary
// //wordfrequency_en.txt ensures high correction quality by combining two data sources:
// //Google Books Ngram data provides representative word frequencies (but contains many entries with spelling errors)
// //SCOWL — Spell Checker Oriented Word Lists which ensures genuine English vocabulary (but contained no word frequencies)
// string path = AppDomain.CurrentDomain.BaseDirectory + "frequency_dictionary_en_82_765.txt"; //path referencing the SymSpell core project
// //string path = "../../frequency_dictionary_en_82_765.txt"; //path when using symspell nuget package (frequency_dictionary_en_82_765.txt is included in nuget package)
// if (!symSpell.LoadDictionary(path, 0, 1)) { Console.Error.WriteLine("\rFile not found: " + Path.GetFullPath(path)); Console.ReadKey(); return; }
// //Alternatively Create the dictionary from a text corpus (e.g. http://norvig.com/big.txt )
// //Make sure the corpus does not contain spelling errors, invalid terms and the word frequency is representative to increase the precision of the spelling correction.
// //You may use SymSpell.CreateDictionaryEntry() to update a (self learning) dictionary incrementally
// //To extend spelling correction beyond single words to phrases (e.g. correcting "unitedkingom" to "united kingdom") simply add those phrases with CreateDictionaryEntry(). or use https://github.com/wolfgarbe/SymSpellCompound
// //string path = "big.txt";
// //if (!symSpell.CreateDictionary(path)) Console.Error.WriteLine("File not found: " + Path.GetFullPath(path));
stopWatch.Stop();
Console.WriteLine(stopWatch.Elapsed.TotalMilliseconds.ToString("0.0") + "ms ");
long memDelta = GC.GetTotalMemory(true) - memSize;
Console.WriteLine((memDelta / 1024 / 1024.0).ToString("N0") + " MB");
// Console.WriteLine("\rDictionary: " + symSpell.WordCount.ToString("N0") + " words, "
// + symSpell.EntryCount.ToString("N0") + " entries, edit distance=" + symSpell.MaxDictionaryEditDistance.ToString()
// + " in " + stopWatch.Elapsed.TotalMilliseconds.ToString("0.0") + "ms "
// + (memDelta / 1024 / 1024.0).ToString("N0") + " MB");
// //warm up
// var result = symSpell.Lookup("warmup", SymSpell.Verbosity.All);
// string input;
// Console.WriteLine("Type a work and hit enter key to get spelling suggestions:");
// while (!string.IsNullOrEmpty(input = (Console.ReadLine() ?? "").Trim()))
// {
// Correct(input, symSpell);
// }
}
