Gather data

This commit is contained in:
mike
2025-12-28 19:52:25 +01:00
parent f0254f4130
commit f0829abe2f
2 changed files with 230 additions and 1 deletions

View File

@@ -189,7 +189,6 @@ public class ConcurrentWordScorer {
writerThread.join();
// Update hints in the database
updateHintsInDatabase();
System.out.println("\n✓ All endpoints finished!");
}

230
src/puzzle/HintJob.java Normal file
View File

@@ -0,0 +1,230 @@
package puzzle;
import java.io.*;
import java.nio.charset.StandardCharsets;
import java.sql.*;
public class HintJob {
static final String JDBC_URL = "jdbc:postgresql://192.168.1.159:5432/postgres";
static final String JDBC_USER = "puzzle";
static final String JDBC_PASS = "heel-goed-wachtwoord";
static final String OLLAMA_URL = "http://192.168.1.159:8081/v1/chat/completions";
static final String MODEL = "/models/Hadiseh-Mhd/Mixtral-8x7B-Instruct-v0.1-Q4_K_M-GGUF/mixtral-8x7b-instruct-v0.1.Q4_K_M.gguf";
public static void main(String[] args) throws Exception {
Class.forName("org.postgresql.Driver");
var limit = args.length > 0 ? Integer.parseInt(args[0]) : 3000;
try (var c = DriverManager.getConnection(JDBC_URL, JDBC_USER, JDBC_PASS)) {
c.setAutoCommit(false);
try (var sel = c.prepareStatement(
"select ctid::text, woord, hint, hint_score " +
"from export_real_words_with_hints " +
"order by (hint is null or hint = '') desc, updated_at nulls first " +
"limit ? for update skip locked");
var upd = c.prepareStatement(
"update export_real_words_with_hints set hint = ?, hint_score = ?, guessed_word = ?, suggested_hint = ?, updated_at = now() where ctid::text = ?")) {
sel.setInt(1, limit);
var done = 0;
try (var rs = sel.executeQuery()) {
while (rs.next()) {
if (done % 10 == 0) System.out.println("Committed " + done);
var ctid = rs.getString(1);
var woord = rs.getString(2);
var oldHint = rs.getString(3);
var oldScore = rs.getInt(4);
if (rs.wasNull()) oldScore = -1;
var newHint = generateHint(woord);
if (newHint == null || newHint.isBlank()) continue;
newHint = sanitizeHint(newHint);
var scoreRes = scoreHint(newHint, woord);
var newScore = scoreRes.score;
// De gebruiker wil voornamelijk hints toevoegen aan records die er geen hebben.
// En de originele hint behouden omdat de LLM resultaten soms tegenvallen.
if (oldHint != null && !oldHint.isBlank()) {
// Er is al een hint. We genereren nog steeds een suggestie,
// maar we overschrijven de originele 'hint' kolom NIET.
var updSug = c.prepareStatement(
"update export_real_words_with_hints set suggested_hint = ?, hint_score = ?, guessed_word = ?, updated_at = now() where ctid::text = ?");
updSug.setString(1, newHint);
updSug.setInt(2, newScore);
updSug.setString(3, scoreRes.guessedWord);
updSug.setString(4, ctid);
updSug.executeUpdate();
c.commit();
done++;
continue;
}
// Geen bestaande hint, dus we vullen hem nu in
upd.setString(1, newHint);
upd.setInt(2, newScore);
upd.setString(3, scoreRes.guessedWord);
upd.setString(4, newHint);
upd.setString(5, ctid);
upd.executeUpdate();
c.commit();
done++;
}
}
System.out.println("Done. Updated " + done + " rows.");
}
}
}
static String sanitizeHint(String hint) {
if (hint == null) return null;
hint = hint.trim();
if (hint.contains("\n")) {
var lines = hint.split("\n");
hint = lines[lines.length - 1].trim();
}
return hint;
}
record ScoreResult(String guessedWord, int score) {
}
static ScoreResult scoreHint(String hint, String woord) throws Exception {
var prompt =
"Ik geef je een kruiswoordpuzzel hint en het aantal letters van het woord. " +
"Welk Nederlands woord van " + woord.length() + " letters wordt hier gezocht?\n" +
"Hint: " + hint + "\n" +
"Antwoord met alleen het woord, geen uitleg.";
var payload = "{"
+ "\"model\":\"" + jsonEscape(MODEL) + "\","
+ "\"messages\":[{\"role\":\"user\",\"content\":\"" + jsonEscape(prompt) + "\"}],"
+ "\"stream\":false,"
+ "\"max_tokens\":3,"
+ "\"temperature\":0.0"
+ "}";
var p = new ProcessBuilder(
"curl", "-sS",
"-H", "Content-Type: application/json",
"-X", "POST", OLLAMA_URL,
"-d", payload
).redirectErrorStream(true).start();
String out;
try (var in = p.getInputStream()) {
out = new String(in.readAllBytes(), StandardCharsets.UTF_8);
}
p.waitFor();
var guessed = jsonGetString(out, "content");
if (guessed != null) guessed = guessed.trim().toUpperCase().replaceAll("[^A-Z]", "");
else guessed = "";
var original = woord.toUpperCase().replaceAll("[^A-Z]", "");
var dist = levenshtein(guessed, original);
// Score: we willen een hoge score voor een goede hint.
// Als de afstand 0 is (exact geraden), is de score maximaal.
// Score = max(0, 100 - dist * 10) - een simpele lineaire schaling
var score = Math.max(0, 100 - (dist * 10));
return new ScoreResult(guessed, score);
}
static int levenshtein(String s1, String s2) {
var dp = new int[s1.length() + 1][s2.length() + 1];
for (var i = 0; i <= s1.length(); i++) dp[i][0] = i;
for (var j = 0; j <= s2.length(); j++) dp[0][j] = j;
for (var i = 1; i <= s1.length(); i++)
for (var j = 1; j <= s2.length(); j++) {
var cost = (s1.charAt(i - 1) == s2.charAt(j - 1)) ? 0 : 1;
dp[i][j] = Math.min(Math.min(dp[i - 1][j] + 1, dp[i][j - 1] + 1), dp[i - 1][j - 1] + cost);
}
return dp[s1.length()][s2.length()];
}
static String generateHint(String woord) throws Exception {
var prompt =
"Geef een korte, duidelijke kruiswoordpuzzel hint in het Nederlands voor het woord: " + woord + ". " +
"Antwoord alleen met de hint. Geen inleiding, geen uitleg, geen aantal letters, geen aanhalingstekens.";
var payload = "{"
+ "\"model\":\"" + jsonEscape(MODEL) + "\","
+ "\"messages\":[{\"role\":\"user\",\"content\":\"" + jsonEscape(prompt) + "\"}],"
+ "\"stream\":false"
+ "}";
var p = new ProcessBuilder(
"curl", "-sS",
"-H", "Content-Type: application/json",
"-X", "POST", OLLAMA_URL,
"-d", payload
).redirectErrorStream(true).start();
String out;
try (var in = p.getInputStream()) {
out = new String(in.readAllBytes(), StandardCharsets.UTF_8);
}
var code = p.waitFor();
if (code != 0) {
System.err.println("curl failed (" + code + "): " + out);
return null;
}
return jsonGetString(out, "content"); // Extract content from {"choices":[{"message":{"content":"..."}}...]}
}
static String jsonEscape(String s) {
var b = new StringBuilder(s.length() + 16);
for (var i = 0; i < s.length(); i++) {
var ch = s.charAt(i);
switch (ch) {
case '\\' -> b.append("\\\\");
case '"' -> b.append("\\\"");
case '\n' -> b.append("\\n");
case '\r' -> b.append("\\r");
case '\t' -> b.append("\\t");
default -> b.append(ch);
}
}
return b.toString();
}
// minimal JSON string extractor for {"key":"value"} with escapes
static String jsonGetString(String json, String key) {
var needle = "\"" + key + "\"";
var i = json.indexOf(needle);
if (i < 0) return null;
i = json.indexOf(':', i + needle.length());
if (i < 0) return null;
i++;
while (i < json.length() && Character.isWhitespace(json.charAt(i))) i++;
if (i >= json.length() || json.charAt(i) != '"') return null;
i++; // after opening quote
var b = new StringBuilder();
var esc = false;
for (; i < json.length(); i++) {
var ch = json.charAt(i);
if (esc) {
switch (ch) {
case 'n' -> b.append('\n');
case 'r' -> b.append('\r');
case 't' -> b.append('\t');
case '"' -> b.append('"');
case '\\' -> b.append('\\');
default -> b.append(ch);
}
esc = false;
} else if (ch == '\\') esc = true;
else if (ch == '"') return b.toString();
else b.append(ch);
}
return null;
}
}