static void Callback(RhetoricalFigures type)
{
var s = string.Empty;
s += "Finding " + type + "....";
Console.WriteLine(s);
}
public RhetoricalFigure(Subsequence subsequence, RhetoricalFigures type, int windowId)
{
Tokens = subsequence;
Type = type;
WindowId = windowId;
}
// Methods:
public static List<RhetoricalFigure> MergeFigures(List<List<Subsequence>> subsequences, RhetoricalFigures type, bool multiWindow = true, string demarcation = FigureComponentsSeparator)
{
// Some figures may be out of order WRT the start of the text; reorder them here.
subsequences = subsequences.OrderBy(s => s[0].SentenceId).ThenBy(s => s[0][0].Left).ToList();
var deepSubsequences = subsequences;
// Merge multi-window figures.
if (multiWindow) {
for (int i = 0; i < deepSubsequences.Count - 1; ++i) {
for (int j = i + 1; j < deepSubsequences.Count; ++j) {
var intersection = deepSubsequences[i].Intersect(deepSubsequences[j]);
if (intersection.Any()) {
var intersectionList = intersection.ToList();
if (deepSubsequences[i].Last() == intersectionList.Last() && deepSubsequences[j].First() == intersectionList.First()) {
var merger = deepSubsequences[i].Union(deepSubsequences[j]).ToList();
deepSubsequences[i] = merger;
deepSubsequences.RemoveAt(j);
i -= 1;
break;
}
}
}
}
}
// At this point, no subsequence component of any figure should contain part of any other subsequence component.
for (int i = 0; i < deepSubsequences.Count; ++i)
deepSubsequences[i] = deepSubsequences[i].Distinct().ToList();
// N.B. V. the following for a discussion of how 'Distinct()' results are ordered:
// http://stackoverflow.com/questions/4734852/does-c-sharp-distinct-method-keep-original-ordering-of-sequence-intact
// Flatten out subsequence lists.
var flatSubsequences = new List<Subsequence>();
foreach (var s in deepSubsequences)
flatSubsequences.Add(new Subsequence(s.SelectMany(x => x), s[0].WindowId));
// At this point, no subsequence component of any figure should contain part of any other subsequence component.
//for (int i = 0; i < flatSubsequences.Count; ++i) {
// if (flatSubsequences[i].Distinct().Count() < flatSubsequences[i].Count) {
// flatSubsequences.RemoveAt(i);
// deepSubsequences.RemoveAt(i);
// i -= 1;
// continue;
// }
//}
// Remove duplicate list instances and merge those contained in others.
for (int i = 0; i < flatSubsequences.Count - 1; ++i) {
for (int j = i + 1; j < flatSubsequences.Count; ++j) {
if (flatSubsequences[i].IsSupersetOf(flatSubsequences[j])) {
flatSubsequences.RemoveAt(j);
deepSubsequences.RemoveAt(j);
i -= 1;
break;
}
else if (flatSubsequences[j].IsSupersetOf(flatSubsequences[i])) {
flatSubsequences[i] = flatSubsequences[j];
deepSubsequences[i] = new List<Subsequence>(deepSubsequences[j]);
i -= 1;
break;
}
}
}
// Remove any duplicate subsequences within each figure.
for (int i = 0; i < deepSubsequences.Count; ++i)
deepSubsequences[i] = deepSubsequences[i].Distinct().ToList();
// Make sure figure constituents are properly ordered.
for (int i = 0; i < deepSubsequences.Count; ++i) {
for (int j = 0; j < deepSubsequences[i].Count; ++j)
deepSubsequences[i][j].OrderBy(s => s.SentenceId).ThenBy(s => s.Left).ToList();
}
for (int i = 0; i < deepSubsequences.Count; ++i) {
deepSubsequences[i] = deepSubsequences[i].OrderBy(s => s[0].SentenceId).ToList();
for (int j = 0; j < deepSubsequences[i].Count; ++j) {
var dsij = deepSubsequences[i][j];
deepSubsequences[i][j] = new Subsequence(dsij.OrderBy(s => s.Left), dsij.ContainingSentence, dsij.ContainingSubsequence, dsij.WindowId);
if (demarcation != null)
deepSubsequences[i][j].Add(new SubsequenceToken(new Token(demarcation, "", 0)));
}
}
var figures = new List<RhetoricalFigure>();
foreach (var deepSubsequence in deepSubsequences) {
var d = deepSubsequence.OrderBy(x => x.SentenceId).ThenBy(x => x[0].Left).ToList(); // Sort figure constituents so leftmost in text appears first, etc.
var figure = new RhetoricalFigure(new Subsequence(d.SelectMany(x => x)), type, d[0].WindowId);
//figure.Tokens = figure.Tokens.OrderBy(s => s.SentenceId).ThenBy(s => s.Left).ToList(); // This sort could cause problems with the collapsed 'figure'. Stick to the one just above.
figures.Add(figure);
}
return figures;
}
protected string DoStuff(string[] args)
{
string rv = string.Empty;
// Put test methods here:
string[] pathParts =
{
//Repository.LocalTextPath,
Repository.NlpTextsPath,
"sonnets.txt"
//"Washington - Inaugural Address (1789).txt"
//"Obama - Inaugural Address (2009).txt"
//"Obama - Inaugural Address (excerpt, 2009).txt"
//"Churchill - We Shall Fight on the Beaches (1940).txt"
//"Churchill - We Shall Fight on the Beaches (excerpt, 1940).txt"
//"Test Sentences.txt"
//"test.txt"
//"epizeuxis_test.txt" // and ploce
//"polysyndeton_test.txt"
//"anaphora_test.txt" // and epistrophe
//"epistrophe_test.txt"
//"symploce_test.txt"
//"epanalepsis_test.txt"
//"anadiplosis_test.txt"
//"antimetabole_test.txt"
//"polyptoton_test.txt"
//"isocolon_test.txt"
//"chiasmus_test.txt"
//"oxymoron_test.txt"
//"Stevens - Farewell to Florida.txt"
};
var path = Path.Combine(pathParts);
if (args.Count() > 0)
{
var args0 = args[0].Trim();
if (args0 != string.Empty)
{
if (File.Exists(args0))
{
path = args0;
}
else if (File.Exists(Repository.NlpTextsPath + args0))
{
path = Repository.NlpTextsPath + args0;
}
}
}
//var result = Miscellaneous.GetPermutationTree<string>("root", new List<string>() { "antonym", "synonym", "derived" }, 3);
AnalyzerOptions options = AnalyzerOptions.OmitPunctuationTokens | AnalyzerOptions.OmitFalseDuplicatePhrases | AnalyzerOptions.UsePunctuationDelimitedPhrases;
string ignore = "";
Analyzer a = new Analyzer(path, ignore: ignore, options: options);
TimeSpan begin = Process.GetCurrentProcess().TotalProcessorTime;
if (args.Count() > 1) // Deserialize JSON
{
var args1 = args[1].Trim();
var all = false;
if (args1 == string.Empty)
{
args1 = "{ All: {} }";
}
var rhetoricalFigureParameters = JsonConvert.DeserializeObject <Dictionary <string, RhetoricalFigureParameters> >(args1);
RhetoricalFigures exclusions = RhetoricalFigures.None;
foreach (var rfp in rhetoricalFigureParameters)
{
var key = rfp.Key;
RhetoricalFigures rhetoricalFigure;
if (!Enum.TryParse(key, out rhetoricalFigure))
{
continue;
}
if (rhetoricalFigure == RhetoricalFigures.All)
{
all = true;
continue;
}
var windowSize = rfp.Value.windowSize;
var extra = rfp.Value.extra;
exclusions |= rhetoricalFigure;
a.FindRhetoricalFigures(rhetoricalFigure, windowSize, extra, Callback);
}
if (all)
{
a.FindRhetoricalFigures(RhetoricalFigures.All, callback: Callback, exclusions: exclusions);
}
}
else
{
//a.FindRhetoricalFigures(RhetoricalFigures.Epizeuxis, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Ploce, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Conduplicatio, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Polysyndeton, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Anaphora, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Epistrophe, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Symploce, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Epanalepsis, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Anadiplosis, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Antimetabole, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Polyptoton, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Isocolon, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Chiasmus, callback: Callback);
//a.FindRhetoricalFigures(RhetoricalFigures.Oxymoron, callback: Callback);
a.FindRhetoricalFigures(RhetoricalFigures.All, callback: Callback);
}
TimeSpan end = Process.GetCurrentProcess().TotalProcessorTime;
Console.WriteLine();
a.Document.WriteLine();
Console.WriteLine();
var figureRows = new List <string>();
var figureColumns = new string[] {
"figure_id",
"token_id",
"type",
"word",
"sentence_id",
"left_edge",
"right_edge",
"tag",
"tag_equiv",
"depth",
"stem"
};
string sep = ",";
var header = String.Join(sep, figureColumns);
figureRows.Add(header);
int i = 0;
foreach (var figure in a.Figures)
{
int j = 0;
foreach (var token in figure.Tokens)
{
if (token.ContainingSentence == null)
{
continue;
}
var rowArray = new object[] {
i, j,
figure.Type,
"\"" + token.Word + "\"",
token.SentenceId, token.Left, token.Right,
"\"" + token.Tag + "\"",
"\"" + token.TagEquivalent + "\"",
token.Depth,
"\"" + token.Stem + "\""
};
var row = String.Join(sep, rowArray);
figureRows.Add(row);
j++;
}
i++;
}
var sentenceRows = new List <string>();
var sentenceColumns = new string[] {
"sentence_id",
"token_id",
"word",
"left_edge",
"right_edge",
"tag",
"tag_equiv",
"depth",
"stem"
};
header = String.Join(sep, sentenceColumns);
sentenceRows.Add(header);
i = 0;
foreach (var sentence in a.Document.Sentences)
{
int j = 0;
foreach (var token in sentence.Tokens)
{
var rowArray = new object[] {
i, j,
"\"" + token.Word + "\"",
token.Left, token.Right,
"\"" + token.Tag + "\"",
"\"" + token.TagEquivalent + "\"",
token.Depth,
"\"" + token.Stem + "\""
};
var row = String.Join(sep, rowArray);
sentenceRows.Add(row);
j++;
}
i++;
}
figureRows.ForEach(x => Console.WriteLine("{0}", x));
Console.WriteLine();
if (args.Count() > 2) // Write CSV representations of figures to file.
{
var args2 = args[2].Trim();
var args2Csv = args2 + ".csv";
var args2Doc = args2 + ".doc.csv";
Console.WriteLine("Writing document: " + args2Csv + Environment.NewLine);
File.WriteAllLines(args2Csv, figureRows);
Console.WriteLine("Writing document: " + args2Doc + Environment.NewLine);
File.WriteAllLines(args2Doc, sentenceRows);
}
foreach (var figure in a.Figures)
{
Console.WriteLine(figure.Text);
}
Console.WriteLine();
Console.WriteLine("Measured time: " + (end - begin).TotalMilliseconds + " ms; " + (end - begin).TotalSeconds + " s; " + (end - begin).TotalMinutes + "m.");
return(rv);
// N.B. This returns to method "Main()", in which a console pause may be commented out; uncomment it for testing.
}
static void Callback(RhetoricalFigures type)
{
var s = string.Empty;
s += "Finding " + type + "....";
Console.WriteLine(s);
}
public void FindRhetoricalFigures(RhetoricalFigures type = RhetoricalFigures.All, int?windowSize = null, object extra = null, FindFiguresCallback callback = null, RhetoricalFigures exclusions = RhetoricalFigures.None)
{
bool anadiplosis = (type.HasFlag(RhetoricalFigures.Anadiplosis) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Anadiplosis);
bool anaphora = (type.HasFlag(RhetoricalFigures.Anaphora) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Anaphora);
bool antimetabole = (type.HasFlag(RhetoricalFigures.Antimetabole) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Antimetabole);
bool chiasmus = (type.HasFlag(RhetoricalFigures.Chiasmus) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Chiasmus);
bool conduplicatio = (type.HasFlag(RhetoricalFigures.Conduplicatio) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Conduplicatio);
bool epanalepsis = (type.HasFlag(RhetoricalFigures.Epanalepsis) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Epanalepsis);
bool epistrophe = (type.HasFlag(RhetoricalFigures.Epistrophe) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Epistrophe);
bool epizeuxis = (type.HasFlag(RhetoricalFigures.Epizeuxis) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Epizeuxis);
bool isocolon = (type.HasFlag(RhetoricalFigures.Isocolon) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Isocolon);
bool oxymoron = (type.HasFlag(RhetoricalFigures.Oxymoron) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Oxymoron);
bool ploce = (type.HasFlag(RhetoricalFigures.Ploce) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Ploce);
bool polyptoton = (type.HasFlag(RhetoricalFigures.Polyptoton) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Polyptoton);
bool polysyndeton = (type.HasFlag(RhetoricalFigures.Polysyndeton) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Polysyndeton);
bool symploce = (type.HasFlag(RhetoricalFigures.Symploce) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Symploce);
if (epizeuxis)
{
if (callback != null)
{
callback(RhetoricalFigures.Epizeuxis);
}
RhetoricalFigure.FindEpizeuxis(this, windowSize);
}
if (ploce)
{
if (callback != null)
{
callback(RhetoricalFigures.Ploce);
}
RhetoricalFigure.FindPloce(this, windowSize);
}
if (conduplicatio)
{
if (callback != null)
{
callback(RhetoricalFigures.Conduplicatio);
}
RhetoricalFigure.FindConduplicatio(this, windowSize, extra);
}
if (polysyndeton)
{
if (callback != null)
{
callback(RhetoricalFigures.Polysyndeton);
}
RhetoricalFigure.FindPolysyndeton(this, windowSize, extra);
}
if (anaphora)
{
if (callback != null)
{
callback(RhetoricalFigures.Anaphora);
}
RhetoricalFigure.FindAnaphora(this, windowSize, extra);
}
if (epistrophe)
{
if (callback != null)
{
callback(RhetoricalFigures.Epistrophe);
}
RhetoricalFigure.FindEpistrophe(this, windowSize, extra);
}
if (symploce)
{
if (callback != null)
{
callback(RhetoricalFigures.Symploce);
}
RhetoricalFigure.FindSymploce(this, windowSize, extra);
}
if (epanalepsis)
{
if (callback != null)
{
callback(RhetoricalFigures.Epanalepsis);
}
RhetoricalFigure.FindEpanalepsis(this, windowSize, extra);
}
if (anadiplosis)
{
if (callback != null)
{
callback(RhetoricalFigures.Anadiplosis);
}
RhetoricalFigure.FindAnadiplosis(this, windowSize, extra);
}
if (antimetabole)
{
if (callback != null)
{
callback(RhetoricalFigures.Antimetabole);
}
RhetoricalFigure.FindAntimetabole(this, windowSize, extra);
}
if (polyptoton)
{
if (callback != null)
{
callback(RhetoricalFigures.Polyptoton);
}
RhetoricalFigure.FindPolyptoton(this, windowSize);
}
if (isocolon)
{
if (callback != null)
{
callback(RhetoricalFigures.Isocolon);
}
RhetoricalFigure.FindIsocolon(this, windowSize, extra);
}
if (chiasmus)
{
if (callback != null)
{
callback(RhetoricalFigures.Chiasmus);
}
RhetoricalFigure.FindChiasmus(this, windowSize, extra);
}
if (oxymoron)
{
if (callback != null)
{
callback(RhetoricalFigures.Oxymoron);
}
RhetoricalFigure.FindOxymoron(this, windowSize, extra);
}
}
public void FindRhetoricalFigures(RhetoricalFigures type = RhetoricalFigures.All, int? windowSize = null, object extra = null, FindFiguresCallback callback = null, RhetoricalFigures exclusions = RhetoricalFigures.None)
{
bool anadiplosis = (type.HasFlag(RhetoricalFigures.Anadiplosis) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Anadiplosis);
bool anaphora = (type.HasFlag(RhetoricalFigures.Anaphora) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Anaphora);
bool antimetabole = (type.HasFlag(RhetoricalFigures.Antimetabole) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Antimetabole);
bool chiasmus = (type.HasFlag(RhetoricalFigures.Chiasmus) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Chiasmus);
bool conduplicatio = (type.HasFlag(RhetoricalFigures.Conduplicatio) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Conduplicatio);
bool epanalepsis = (type.HasFlag(RhetoricalFigures.Epanalepsis) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Epanalepsis);
bool epistrophe = (type.HasFlag(RhetoricalFigures.Epistrophe) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Epistrophe);
bool epizeuxis = (type.HasFlag(RhetoricalFigures.Epizeuxis) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Epizeuxis);
bool isocolon = (type.HasFlag(RhetoricalFigures.Isocolon) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Isocolon);
bool oxymoron = (type.HasFlag(RhetoricalFigures.Oxymoron) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Oxymoron);
bool ploce = (type.HasFlag(RhetoricalFigures.Ploce) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Ploce);
bool polyptoton = (type.HasFlag(RhetoricalFigures.Polyptoton) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Polyptoton);
bool polysyndeton = (type.HasFlag(RhetoricalFigures.Polysyndeton) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Polysyndeton);
bool symploce = (type.HasFlag(RhetoricalFigures.Symploce) || type.HasFlag(RhetoricalFigures.All)) && !exclusions.HasFlag(RhetoricalFigures.Symploce);
if (epizeuxis) {
if (callback != null) callback(RhetoricalFigures.Epizeuxis);
RhetoricalFigure.FindEpizeuxis(this, windowSize);
}
if (ploce) {
if (callback != null) callback(RhetoricalFigures.Ploce);
RhetoricalFigure.FindPloce(this, windowSize);
}
if (conduplicatio) {
if (callback != null) callback(RhetoricalFigures.Conduplicatio);
RhetoricalFigure.FindConduplicatio(this, windowSize, extra);
}
if (polysyndeton) {
if (callback != null) callback(RhetoricalFigures.Polysyndeton);
RhetoricalFigure.FindPolysyndeton(this, windowSize, extra);
}
if (anaphora) {
if (callback != null) callback(RhetoricalFigures.Anaphora);
RhetoricalFigure.FindAnaphora(this, windowSize, extra);
}
if (epistrophe) {
if (callback != null) callback(RhetoricalFigures.Epistrophe);
RhetoricalFigure.FindEpistrophe(this, windowSize, extra);
}
if (symploce) {
if (callback != null) callback(RhetoricalFigures.Symploce);
RhetoricalFigure.FindSymploce(this, windowSize, extra);
}
if (epanalepsis) {
if (callback != null) callback(RhetoricalFigures.Epanalepsis);
RhetoricalFigure.FindEpanalepsis(this, windowSize, extra);
}
if (anadiplosis) {
if (callback != null) callback(RhetoricalFigures.Anadiplosis);
RhetoricalFigure.FindAnadiplosis(this, windowSize, extra);
}
if (antimetabole) {
if (callback != null) callback(RhetoricalFigures.Antimetabole);
RhetoricalFigure.FindAntimetabole(this, windowSize, extra);
}
if (polyptoton) {
if (callback != null) callback(RhetoricalFigures.Polyptoton);
RhetoricalFigure.FindPolyptoton(this, windowSize);
}
if (isocolon) {
if (callback != null) callback(RhetoricalFigures.Isocolon);
RhetoricalFigure.FindIsocolon(this, windowSize, extra);
}
if (chiasmus) {
if (callback != null) callback(RhetoricalFigures.Chiasmus);
RhetoricalFigure.FindChiasmus(this, windowSize, extra);
}
if (oxymoron) {
if (callback != null) callback(RhetoricalFigures.Oxymoron);
RhetoricalFigure.FindOxymoron(this, windowSize, extra);
}
}
