Major fix for localization: finalizing German, French, Spanish, Italian,

Greek, Slovak, Ukrainian, Turkish and Russian. Also added full support for Polish language. All servlets are now fully supported by those languages.
2026-10-25 05:54:15 +00:00 · 2026-07-07 21:36:59 +02:00
1 parent 2ca403a1fc
commit 57d188e69a
@@ -44,14 +44,16 @@
            <input type="radio" name="language" value="default" id="lang_en" onchange="this.form.submit()" #(lang_en)#::checked="checked"#(/lang_en)# /><label for="lang_en" #(active_en)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_en)#>English</label>&nbsp;    
            <input type="radio" name="language" value="de" id="lang_de" onchange="this.form.submit()" #(lang_de)#::checked="checked"#(/lang_de)# /><label for="lang_de" #(active_de)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_de)#>Deutsch</label>&nbsp;
            <input type="radio" name="language" value="fr" id="lang_fr" onchange="this.form.submit()" #(lang_fr)#::checked="checked"#(/lang_fr)# /><label for="lang_fr" #(active_fr)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_fr)#>Fran&ccedil;ais</label>&nbsp;
-            <input type="radio" name="language" value="zh" id="lang_zh" onchange="this.form.submit()" #(lang_zh)#::checked="checked"#(/lang_zh)# /><label for="lang_zh" #(active_zh)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_zh)#>&#20013;&#25991</label>
-            <input type="radio" name="language" value="ru" id="lang_ru" onchange="this.form.submit()" #(lang_ru)#::checked="checked"#(/lang_ru)# /><label for="lang_ru" #(active_ru)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_ru)#>&#1056;&#1091;&#1089;&#1089;&#1082;&#1080;&#1081;</label>
-            <input type="radio" name="language" value="uk" id="lang_uk" onchange="this.form.submit()" #(lang_uk)#::checked="checked"#(/lang_uk)# /><label for="lang_uk" #(active_uk)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_uk)#>&#1059;&#1082;&#1088;&#1072;&#1111;&#1085;&#1089;&#1100;&#1082;&#1072;</label>
-            <input type="radio" name="language" value="hi" id="lang_hi" onchange="this.form.submit()" #(lang_hi)#::checked="checked"#(/lang_hi)# /><label for="lang_hi" #(active_hi)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_hi)#>&#2361;&#2367;&#2344;&#2381;&#2342;&#2368;</label>
-            <input type="radio" name="language" value="ja" id="lang_ja" onchange="this.form.submit()" #(lang_ja)#::checked="checked"#(/lang_ja)# /><label for="lang_ja" #(active_ja)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_ja)#>&#26085;&#26412;&#35486;</label>
+            <input type="radio" name="language" value="pl" id="lang_pl" onchange="this.form.submit()" #(lang_pl)#::checked="checked"#(/lang_pl)# /><label for="lang_pl" #(active_pl)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_pl)#>Polski</label>&nbsp;
            <input type="radio" name="language" value="el" id="lang_el" onchange="this.form.submit()" #(lang_el)#::checked="checked"#(/lang_el)# /><label for="lang_el" #(active_el)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_el)#>Greek</label>
            <input type="radio" name="language" value="it" id="lang_it" onchange="this.form.submit()" #(lang_it)#::checked="checked"#(/lang_it)# /><label for="lang_it" #(active_el)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_el)#>Italiano</label>            
-              <input type="radio" name="language" value="es" id="lang_es" onchange="this.form.submit()" #(lang_es)#::checked="checked"#(/lang_es)# /><label for="lang_es" #(active_es)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_es)#>Español</label>
+            <input type="radio" name="language" value="es" id="lang_es" onchange="this.form.submit()" #(lang_es)#::checked="checked"#(/lang_es)# /><label for="lang_es" #(active_es)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_es)#>Español</label>
+            <input type="radio" name="language" value="uk" id="lang_uk" onchange="this.form.submit()" #(lang_uk)#::checked="checked"#(/lang_uk)# /><label for="lang_uk" #(active_uk)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_uk)#>&#1059;&#1082;&#1088;&#1072;&#1111;&#1085;&#1089;&#1100;&#1082;&#1072;</label>
+            <input type="radio" name="language" value="ru" id="lang_ru" onchange="this.form.submit()" #(lang_ru)#::checked="checked"#(/lang_ru)# /><label for="lang_ru" #(active_ru)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_ru)#>&#1056;&#1091;&#1089;&#1089;&#1082;&#1080;&#1081;</label>
+            <input type="radio" name="language" value="zh" id="lang_zh" onchange="this.form.submit()" #(lang_zh)#::checked="checked"#(/lang_zh)# /><label for="lang_zh" #(active_zh)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_zh)#>&#20013;&#25991</label>
+            <input type="radio" name="language" value="hi" id="lang_hi" onchange="this.form.submit()" #(lang_hi)#::checked="checked"#(/lang_hi)# /><label for="lang_hi" #(active_hi)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_hi)#>&#2361;&#2367;&#2344;&#2381;&#2342;&#2368;</label>
+            <input type="radio" name="language" value="ja" id="lang_ja" onchange="this.form.submit()" #(lang_ja)#::checked="checked"#(/lang_ja)# /><label for="lang_ja" #(active_ja)#::title="Click to generate translated pages"::class="label-success" title="Active : translated pages are available"#(/active_ja)#>&#26085;&#26412;&#35486;</label>
+            
          </fieldset>
        </li>
        <!-- take care that no other items are changed, but also change the former if no js is enabled -->
@@ -70,12 +70,18 @@ Wichtig:
  englische Text auf mehreren Seiten vorkommt, braucht **jede** Seite ihren
  eigenen Abschnitt mit diesem Eintrag (Übersetzung ist pro Datei-Abschnitt).

-### Lokalisierter Abschnitts-Header (`#Dosya:`)
-Ausnahme: `tr.lng` verwendet für viele Abschnitte das türkische Schlüsselwort
-**`#Dosya:`** statt `#File:`. Werkzeuge, die Abschnitte parsen, müssen **beide**
-erkennen, z. B. per Regex `^#(?:File|Dosya):\s*(.*)$`. Andere `#Xxx:`-Präfixe
-(`#YaCy:`, `#Subject:`, `#URL:` …) sind **keine** Header, sondern Übersetzungs-
-schlüssel, die zufällig mit `#` beginnen.
+### Abschnitts-Header: nur `#File:`
+Der Runtime-Loader (`Translator.loadTranslationsLists`) erkennt derzeit **nur**
+`#File:` als Abschnitts-Header. Zeilen mit anderen Präfixen, z. B. `#Dosya:`,
+werden wie normale Kommentarzeilen behandelt; die folgenden Einträge landen dann
+nicht im beabsichtigten Dateiabschnitt und greifen zur Laufzeit nicht korrekt.
+
+Hinweis: Ältere oder importierte Dateien können lokalisierte Header wie
+`#Dosya:` enthalten. Solche Header müssen vor einem Runtime-Test nach `#File:`
+normalisiert werden (oder der Java-Loader muss explizit erweitert werden).
+Andere `#Xxx:`-Präfixe (`#YaCy:`, `#Subject:`, `#URL:` …) sind ebenfalls keine
+Header. Da alle Zeilen mit `#` ignoriert werden, sind auch `#key==wert`-Zeilen
+auskommentierte Einträge und keine aktiven Übersetzungen.

 ---

@@ -137,6 +143,17 @@ Ablauf (`Translator.translate` / `translateFilesRecursive`):
 Es gibt **keinen** automatischen Extraktor, der „übersetzbare Strings“
 erkennt — die Schlüssel werden von Hand gepflegt.

+Wichtig: Beim Erzeugen der lokalisierten Dateien werden nur Quelldateien
+geschrieben, für die in der Sprachdatei ein passender `#File:`-Abschnitt
+existiert. Fehlt der Abschnitt, wird diese Datei nicht als lokalisierte Kopie
+erzeugt.
+
+Die erzeugten Seiten liegen zur Laufzeit unter `DATA/LOCALE/htroot/<sprache>`.
+Beim Umschalten oder automatischen Refresh einer Sprache muss dieser Ordner
+vorher gelöscht werden; sonst können nicht mehr erzeugte Altdateien weiterhin
+ausgeliefert werden. Aktuelle YaCy-Versionen erledigen das beim Sprachwechsel
+und beim versionsbedingten Startup-Refresh automatisch.
+
 ---

 ## 5. Template-Markup in den Quelldateien
@@ -206,21 +223,70 @@ Leerzeichen, bis zum `#[path]#`), **nicht** die ganze Zeile inkl. `#[path]#`.
  JSON-Snippets) — nur die umgebende Prosa übersetzen.
 - **Test-/Demo-Dateien** und rein technische Bezeichner (Feldnamen wie
  `num_ctx`, `max_tokens`, Rollennamen wie `search-query`).
+- **Technische Link-Ziele, Pfade, Servlets und URLs** bleiben literal. Nicht
+  übersetzen oder durch Leerzeichen beschädigen: `Network.html` bleibt
+  `Network.html`, `sharedBlacklist.html` bleibt `sharedBlacklist.html`,
+  `share.json` bleibt `share.json`, `styles/prosilver/template/overall_header.html`
+  bleibt unverändert und URLs wie `http://localhost:8090/proxy.html?...`
+  dürfen nicht lokalisiert werden.
+
+Prüfung aus dem Repository-Root:
+
+```bash
+python3 locales/validate-locale-links.py --exclude pl.lng
+```
+
+`--exclude` ist nützlich, wenn eine Sprache parallel in einem anderen Arbeitszweig
+bearbeitet wird. Für einzelne Sprachen kann `--include de.lng --include fr.lng`
+verwendet werden. Ein sauberer Lauf endet mit `OK: ... no link target issues found.`

 ---

 ## 8. `master.lng.xlf`

-`master.lng.xlf` ist eine XLIFF-Datei, die den **Gesamtbestand** aller
-übersetzbaren Strings pro Datei als `<source>`-Elemente führt.
+`master.lng.xlf` ist die source-basierte XLIFF-Referenz der übersetzbaren
+Strings pro Datei. Die Wahrheit für diesen Master liegt in den Quellen unter
+`htroot`, nicht in bereits vorhandenen `.lng`-Dateien.

- Sie wird **generiert** (`TranslationManager.createMasterTranslationLists`):
-  Grundlage sind die vorhandenen `.lng`-Schlüssel, gefiltert danach, ob sie noch
-  als Teilstring in der jeweiligen Quelldatei vorkommen (`content.indexOf >= 0`).
+- Sie wird mit `GenerateSourceMasterXliff` aus sichtbaren Textknoten und
+  ausgewählten UI-Attributen (`alt`, `title`, `placeholder`, `aria-label`,
+  Button-`value`) erzeugt.
+- Jeder Kandidat wird gegen den Roh-Quelltext und die Runtime-Wortgrenzen der
+  Übersetzung geprüft. Nicht darstellbare `.lng`-Keys, z. B. Keys mit `==` oder
+  einem abschließenden `=`, werden verworfen.
 - **Nicht** von Hand mit Hash-/Zeilen-IDs pflegen — nach Änderungen an Quellen
-  oder `.lng`-Dateien besser über das YaCy-Tooling neu erzeugen. Ein kompletter
-  `<file>…</file>`-Block darf jedoch sauber entfernt werden (z. B. wenn die
-  zugehörige Seite gelöscht wurde).
+  über das YaCy-Tooling neu erzeugen und das Delta prüfen.
+- Beim Refresh wird die Zieldatei ersetzt. Stale Master-Einträge fallen dadurch
+  weg, auch wenn sie noch in alten `.lng`-Dateien stehen.
+
+Refresh aus dem Repository-Root:
+
+```bash
+java -cp 'build/classes/java/main:lib/*' \
+  net.yacy.utils.translation.GenerateSourceMasterXliff \
+  htroot locales/master.lng.xlf
+```
+
+Falls die Klassen noch nicht kompiliert sind, vorher `ant compile` ausführen.
+Das zweite Argument ist wichtig: ohne `locales/master.lng.xlf` schreibt das Tool
+standardmäßig nach `./source-master.lng.xlf` im Repository-Root. Existiert die
+Zieldatei bereits, wird sie ersetzt.
+
+### Legacy: bestandbasierter Master
+
+`GenerateMasterXliff` erzeugt nur einen bestandbasierten Master aus vorhandenen
+`.lng`-Schlüsseln, gefiltert danach, ob sie noch als Teilstring in der
+jeweiligen Quelldatei vorkommen (`content.indexOf >= 0`):
+
+```bash
+java -cp 'build/classes/java/main:lib/*' \
+  net.yacy.utils.translation.GenerateMasterXliff \
+  locales /tmp/master-from-lng.lng.xlf
+```
+
+Dieses Tool ist nützlich zur Diagnose von Altbestand, aber nicht als
+Vollständigkeitsreferenz: englische UI-Texte, die noch in keiner `.lng`-Datei als
+Schlüssel vorkommen, erscheinen dort nicht.

 ---

@@ -240,7 +306,14 @@ Leerzeichen, bis zum `#[path]#`), **nicht** die ganze Zeile inkl. `#[path]#`.
 5. **Einordnen:** neuen `#File:`-Abschnitt anlegen; Einträge längster-zuerst
   sortieren (Abschnitt 6, Regel 5). Neue Abschnitte können am Dateiende
   angehängt werden (die Datei ist nicht streng sortiert).
-6. **Zeilenenden beachten** (Abschnitt 10).
+6. **Vollständigkeit beidseitig prüfen:**
+   - `master.lng.xlf -> <sprache>.lng`: fehlende Source-Schlüssel ergänzen.
+   - `<sprache>.lng -> master.lng.xlf`: Extras prüfen und in der Regel entfernen;
+     sie sind stale oder stammen aus einem nicht frisch generierten Master.
+   - Für jeden aktiven Sprach-Key prüfen: `key in htroot/<#File>`.
+   - Doppelte Schlüssel und auskommentierte `#...==...`-Einträge entfernen oder
+     bewusst reaktivieren.
+7. **Zeilenenden beachten** (Abschnitt 10).

 ---

@@ -0,0 +1,163 @@
+#!/usr/bin/env python3
+"""Validate that locale translations preserve technical link targets.
+
+The .lng format is plain text replacement. File names, servlet endpoints,
+paths and URLs inside a source string must therefore remain literal in the
+translation. This script detects translated or damaged targets such as
+Network.html -> Netzwerk.html, servletshare.json, or URLs with inserted spaces.
+"""
+
+from __future__ import annotations
+
+import argparse
+import re
+import sys
+from pathlib import Path
+
+
+TARGET_EXTENSIONS = (
+    "html",
+    "inc",
+    "json",
+    "xml",
+    "rss",
+    "css",
+    "js",
+    "pac",
+)
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Validate literal link targets in YaCy .lng locale files.",
+    )
+    parser.add_argument(
+        "--locales",
+        type=Path,
+        default=Path("locales"),
+        help="Locale directory containing .lng files (default: locales).",
+    )
+    parser.add_argument(
+        "--source",
+        type=Path,
+        default=Path("htroot"),
+        help="YaCy htroot directory used to identify valid source targets (default: htroot).",
+    )
+    parser.add_argument(
+        "--include",
+        action="append",
+        default=[],
+        metavar="FILE",
+        help="Only validate this .lng file name. Can be used multiple times.",
+    )
+    parser.add_argument(
+        "--exclude",
+        action="append",
+        default=[],
+        metavar="FILE",
+        help="Skip this .lng file name. Can be used multiple times.",
+    )
+    return parser.parse_args()
+
+
+def collect_known_targets(source_dir: Path, locale_files: list[Path]) -> set[str]:
+    known: set[str] = set()
+
+    if source_dir.exists():
+        suffixes = {f".{ext}" for ext in TARGET_EXTENSIONS}
+        for path in source_dir.rglob("*"):
+            if path.is_file() and path.suffix.lower() in suffixes:
+                relative = path.relative_to(source_dir).as_posix()
+                known.add(relative)
+                known.add(path.name)
+
+    for locale_file in locale_files:
+        for line in locale_file.read_text(encoding="utf-8", errors="ignore").splitlines():
+            if line.startswith("#File:"):
+                target = line[6:].strip()
+                known.add(target)
+                known.add(Path(target).name)
+
+    return known
+
+
+def locale_files(locales_dir: Path, include: list[str], exclude: list[str]) -> list[Path]:
+    include_set = set(include)
+    exclude_set = set(exclude)
+    files = sorted(locales_dir.glob("*.lng"))
+    if include_set:
+        files = [path for path in files if path.name in include_set]
+    if exclude_set:
+        files = [path for path in files if path.name not in exclude_set]
+    return files
+
+
+def main() -> int:
+    args = parse_args()
+    files = locale_files(args.locales, args.include, args.exclude)
+    known_targets = collect_known_targets(args.source, files)
+
+    extensions = "|".join(TARGET_EXTENSIONS)
+    file_token = re.compile(
+        rf"(?<![\w./:-])([A-Za-z0-9_./-]+\.(?:{extensions}))(?![\w./-])",
+        re.IGNORECASE,
+    )
+    value_token = re.compile(
+        rf"(?<![\w./:-])([^\s\"'<>(),;]+\.(?:{extensions}))(?![\w./-])",
+        re.IGNORECASE,
+    )
+    url_token = re.compile(r"https?://[^\s\"'<>),]*")
+
+    failures = 0
+    for locale_file in files:
+        current_section: str | None = None
+        missing_key_targets: list[tuple[int, str | None, list[str], str]] = []
+        unknown_value_targets: list[tuple[int, str | None, str, str]] = []
+
+        for line_number, line in enumerate(
+            locale_file.read_text(encoding="utf-8", errors="ignore").splitlines(),
+            1,
+        ):
+            if line.startswith("#File:"):
+                current_section = line[6:].strip()
+                continue
+            if "==" not in line or line.startswith("#"):
+                continue
+
+            source, target = line.split("==", 1)
+            source_targets = set(file_token.findall(source)) | set(url_token.findall(source))
+            target_targets = set(file_token.findall(target)) | set(url_token.findall(target))
+
+            missing = sorted(token for token in source_targets if token not in target_targets)
+            if missing:
+                missing_key_targets.append((line_number, current_section, missing, line))
+
+            source_file_targets = set(value_token.findall(source))
+            for token in value_token.findall(target):
+                clean = token.strip(".,:;!?")
+                if clean in known_targets or clean in source_file_targets:
+                    continue
+                if clean.startswith(("http://", "https://")):
+                    continue
+                unknown_value_targets.append((line_number, current_section, clean, line))
+
+        if missing_key_targets or unknown_value_targets:
+            failures += len(missing_key_targets) + len(unknown_value_targets)
+            print(f"\n## {locale_file.name}")
+            for line_number, section, missing, line in missing_key_targets:
+                print(f"{line_number}: missing technical target(s) {missing} in {section}")
+                print(f"  {line}")
+            for line_number, section, token, line in unknown_value_targets:
+                print(f"{line_number}: unknown translated/damaged target {token!r} in {section}")
+                print(f"  {line}")
+
+    if failures:
+        print(f"\nFAILED: {failures} locale link target issue(s) found.")
+        return 1
+
+    print(f"OK: checked {len(files)} locale file(s), no link target issues found.")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())
@@ -47,6 +47,7 @@ import java.util.List;
 import java.util.Map;
 import java.util.Map.Entry;
 import java.util.Set;
+import java.util.regex.Pattern;

 import net.yacy.cora.util.CommonPattern;
 import net.yacy.cora.util.ConcurrentLog;
@@ -68,6 +69,10 @@ public class Translator {

    public final static String LANG_FILENAME_FILTER = "^.*\\.lng$";

+    private static final Pattern TECHNICAL_TARGET_PATTERN = Pattern.compile(
+            "^(?:https?://.*|[A-Za-z0-9_./:%?#=&;,+~@!$'()\\[\\]-]+\\.(?:html|inc|json|xml|rss|css|js|pac)(?:[?#].*)?)$",
+            Pattern.CASE_INSENSITIVE);
+
    /**
     * Translate source using entries in translationTable
     * @param source text to translate. Must be non null.
@@ -112,8 +117,12 @@ public class Translator {
                        }

                        if (boundary) { // boundary check ok -> translate
-                            builder.replace(index, index + key.length(), translation);
-                            index = builder.indexOf(key, index + translation.length());
+                            if (isInsideTechnicalTarget(builder, index, key.length())) {
+                                index = builder.indexOf(key, index + key.length());
+                            } else {
+                                builder.replace(index, index + key.length(), translation);
+                                index = builder.indexOf(key, index + translation.length());
+                            }
                        } else { // otherwise just skip to next occurence
                            index = builder.indexOf(key, index + key.length());
                        }
@@ -124,6 +133,35 @@ public class Translator {
        return builder.toString();
    }

+    private static boolean isInsideTechnicalTarget(final CharSequence content, final int index, final int length) {
+        int start = index;
+        while (start > 0 && isTechnicalTargetCharacter(content.charAt(start - 1))) {
+            start--;
+        }
+
+        int end = index + length;
+        while (end < content.length() && isTechnicalTargetCharacter(content.charAt(end))) {
+            end++;
+        }
+
+        if (start == index && end == index + length) {
+            return false;
+        }
+
+        final String token = content.subSequence(start, end).toString();
+        return TECHNICAL_TARGET_PATTERN.matcher(token).matches();
+    }
+
+    private static boolean isTechnicalTargetCharacter(final char c) {
+        return Character.isLetterOrDigit(c)
+                || c == '_' || c == '-' || c == '.' || c == '/'
+                || c == ':' || c == '%' || c == '?' || c == '#'
+                || c == '=' || c == '&' || c == ';' || c == ','
+                || c == '+' || c == '~' || c == '@' || c == '!'
+                || c == '$' || c == '\'' || c == '(' || c == ')'
+                || c == '[' || c == ']';
+    }
+
    /**
     * Load multiple translationLists from one File. Each List starts with #File: relative/path/to/file
     * (within each file section translation is done in order of the language file entries, on conflicts
@@ -300,6 +338,7 @@ public class Translator {
            final File destDir = new File(env.getDataPath("locale.translated_html", "DATA/LOCALE/htroot"), lang.substring(0, lang.length() - 4));// cut .lng
            final File translationFile = new File(langPath, lang);

+            FileUtils.deletedelete(destDir);
            if (translateFilesRecursive(sourceDir, destDir, translationFile, "html,template,inc", "locale")) {
                env.setConfig("locale.language", lang.substring(0, lang.length() - 4));
                Formatter.setLocale(env.getConfig("locale.language", "en"));
@@ -39,8 +39,8 @@ public class GenerateMasterXliff {
 	/**
 	 * Read all translation files in YaCy custom lng format from a given folder
 	 * (locales/ as default), and convert them to a single master xliff file
-	 * containing entries to translate. When the ouput master file already exists,
-	 * it is updated with new entries from the lng files.
+	 * containing entries to translate. When the output master file already exists,
+	 * it is replaced with the currently valid entries from the lng files.
 	 * 
 	 * @param args
 	 *            runtime optional arguments<br/>
@@ -74,7 +74,7 @@ public class GenerateMasterXliff {
 				masterXlf = new File("master.lng.xlf");
 			}
 			if (masterXlf.exists()) {
-				System.out.println("Updating master xliff file at " + masterXlf.getAbsolutePath());
+				System.out.println("Replacing master xliff file at " + masterXlf.getAbsolutePath());
 			} else {
 				System.out.println("Generating a new master xliff file at " + masterXlf.getAbsolutePath());
 			}
@@ -0,0 +1,428 @@
+// GenerateSourceMasterXliff.java
+// ---------------------------
+// Copyright 2026 by YaCy contributors
+//
+// This is a part of YaCy, a peer-to-peer based web search engine
+//
+// LICENSE
+//
+// This program is free software; you can redistribute it and/or modify
+// it under the terms of the GNU General Public License as published by
+// the Free Software Foundation; either version 2 of the License, or
+// (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU General Public License
+// along with this program; if not, write to the Free Software
+// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+package net.yacy.utils.translation;
+
+import java.io.File;
+import java.io.IOException;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.util.Arrays;
+import java.util.LinkedHashMap;
+import java.util.LinkedHashSet;
+import java.util.List;
+import java.util.Locale;
+import java.util.Map;
+import java.util.Set;
+import java.util.TreeMap;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import java.util.stream.Stream;
+
+import net.yacy.cora.util.ConcurrentLog;
+import net.yacy.document.SentenceReader;
+
+/**
+ * Generate a source-based XLIFF master candidate by extracting conservative UI
+ * text candidates directly from htroot source files.
+ *
+ * This tool does not use existing *.lng files as input. It is intended to build
+ * and review a source-of-truth candidate before checking individual languages
+ * for completeness.
+ */
+public class GenerateSourceMasterXliff {
+
+    private static final List<String> SOURCE_EXTENSIONS = Arrays.asList(".html", ".template", ".inc");
+
+    private static final Set<String> TEXT_ATTRIBUTES = new LinkedHashSet<>(
+            Arrays.asList("alt", "aria-label", "placeholder", "title", "value"));
+
+    private static final Pattern IGNORED_BLOCK_PATTERN = Pattern.compile(
+            "(?is)<(script|style|pre|code|textarea)\\b[^>]*>.*?</\\1>");
+
+    private static final Pattern HTML_COMMENT_PATTERN = Pattern.compile("(?s)<!--.*?-->");
+
+    private static final Pattern TAG_PATTERN = Pattern.compile("(?s)<[^>]+>");
+
+    private static final Pattern BLOCK_TAG_PATTERN = Pattern.compile(
+            "(?is)</?(?:html|head|body|title|div|p|table|thead|tbody|tfoot|tr|td|th|ul|ol|li|form|fieldset|legend|dl|dt|dd|select|option|button|input|h[1-6]|section|article|nav|header|footer|hr)\\b[^>]*>");
+
+    private static final Pattern LINK_TAG_PATTERN = Pattern.compile("(?is)</?a\\b[^>]*>");
+
+    private static final Pattern UNPRESERVED_TAG_PATTERN = Pattern.compile(
+            "(?is)<(?!/?(?:abbr|b|strong|em|i|br)\\b)[^>]+>");
+
+    private static final Pattern ATTRIBUTE_PATTERN = Pattern.compile(
+            "(?i)\\b([a-zA-Z_:][-a-zA-Z0-9_:.]*)\\s*=\\s*(\"([^\"]*)\"|'([^']*)')");
+
+    private static final Pattern TEMPLATE_BOUNDARY_PATTERN = Pattern.compile(
+            "#\\([^)]*\\)#|#\\(/[^)]*\\)#|#\\{[^}]+\\}#|#\\{/[^}]+\\}#|#%[^%]+%#|::");
+
+    private static final Pattern HTML_SPACE_ENTITY_ONLY_PATTERN = Pattern.compile(
+            "(?i)(?:&nbsp;|&#160;|&ensp;|&emsp;|\\s)+");
+
+    private static final Pattern HTML_SPACE_ENTITY_EDGE_PATTERN = Pattern.compile(
+            "(?i)^(?:&nbsp;|&#160;|&ensp;|&emsp;|\\s)+|(?:&nbsp;|&#160;|&ensp;|&emsp;|\\s)+$");
+
+    private static final Pattern FORMAT_WRAPPER_PATTERN = Pattern.compile(
+            "(?is)^<(strong|b|em|i)>(.*)</\\1>(?:<br\\s*/?>)?$");
+
+    private static final Pattern HTML_LINE_BREAK_EDGE_PATTERN = Pattern.compile(
+            "(?is)^(?:<br\\s*/?>\\s*)+|(?:\\s*<br\\s*/?>)+$");
+
+    private static final Pattern HTML_TAG_ONLY_PATTERN = Pattern.compile("(?is)</?[a-z][^>]*>");
+
+    /**
+     * @param args runtime optional arguments<br/>
+     *            <ul>
+     *            <li>args[0] : source folder path (htroot/ as default)</li>
+     *            <li>args[1] : output xliff master file path
+     *            (source-master.lng.xlf as default)</li>
+     *            </ul>
+     * @throws IOException when a read/write error occurred
+     */
+    public static void main(final String args[]) throws IOException {
+        try {
+            final File sourceFolder = args.length > 0 && args[0] != null ? new File(args[0]) : new File("htroot");
+            if (!sourceFolder.isDirectory()) {
+                System.err.println(sourceFolder.getPath() + " is not a directory");
+                return;
+            }
+
+            final File masterXlf = args.length > 1 && args[1] != null ? new File(args[1]) : new File("source-master.lng.xlf");
+
+            System.out.println("Extracting source text from folder " + sourceFolder.getAbsolutePath());
+            System.out.println((masterXlf.exists() ? "Replacing" : "Generating") + " source master xliff file at "
+                    + masterXlf.getAbsolutePath());
+
+            final ExtractionResult extractionResult = extractSourceMaster(sourceFolder.toPath());
+            final Map<String, Map<String, String>> sourceMaster = extractionResult.sourceMaster;
+            final VerificationResult verificationResult = verifySourceMaster(sourceFolder.toPath(), sourceMaster);
+            if (!verificationResult.isOk()) {
+                throw new IOException("Source master verification failed: " + verificationResult);
+            }
+            new TranslatorXliff().saveAsXliff(null, masterXlf, sourceMaster);
+
+            int unitCount = 0;
+            for (Map<String, String> entries : sourceMaster.values()) {
+                unitCount += entries.size();
+            }
+            System.out.println("Extracted " + unitCount + " source text candidates from " + sourceMaster.size() + " files.");
+            System.out.println("Rejected " + extractionResult.rejectedCandidates
+                    + " extracted candidates without a runtime boundary match.");
+            System.out.println("Verified " + verificationResult.verifiedKeys
+                    + " source text candidates in " + verificationResult.verifiedFiles + " files.");
+        } finally {
+            ConcurrentLog.shutdown();
+        }
+    }
+
+    static ExtractionResult extractSourceMaster(final Path sourceRoot) throws IOException {
+        final ExtractionResult result = new ExtractionResult();
+        try (Stream<Path> files = Files.walk(sourceRoot)) {
+            files.filter(Files::isRegularFile)
+                    .filter(GenerateSourceMasterXliff::hasSupportedExtension)
+                    .forEach(path -> extractFile(sourceRoot, path, result.sourceMaster, result));
+        }
+        return result;
+    }
+
+    private static boolean hasSupportedExtension(final Path path) {
+        final String fileName = path.getFileName().toString().toLowerCase(Locale.ROOT);
+        for (String extension : SOURCE_EXTENSIONS) {
+            if (fileName.endsWith(extension)) {
+                return true;
+            }
+        }
+        return false;
+    }
+
+    private static void extractFile(final Path sourceRoot, final Path sourceFile,
+            final Map<String, Map<String, String>> sourceMaster, final ExtractionResult result) {
+        try {
+            final String content = new String(Files.readAllBytes(sourceFile), StandardCharsets.UTF_8);
+            final LinkedHashSet<String> candidates = extractCandidates(relativeSourcePath(sourceRoot, sourceFile), content);
+            if (!candidates.isEmpty()) {
+                final String relativePath = relativeSourcePath(sourceRoot, sourceFile);
+                final Map<String, String> entries = new LinkedHashMap<>();
+                for (String candidate : candidates) {
+                    if (hasRuntimeBoundaryMatch(content, candidate)) {
+                        entries.put(candidate, null);
+                    } else {
+                        result.rejectedCandidates++;
+                    }
+                }
+                if (!entries.isEmpty()) {
+                    sourceMaster.put(relativePath, entries);
+                }
+            }
+        } catch (IOException e) {
+            ConcurrentLog.warn("TRANSLATOR", "Could not extract source text from " + sourceFile + ": " + e.getMessage());
+        }
+    }
+
+    private static String relativeSourcePath(final Path sourceRoot, final Path sourceFile) {
+        return sourceRoot.relativize(sourceFile).toString().replace(File.separatorChar, '/');
+    }
+
+    static VerificationResult verifySourceMaster(final Path sourceRoot,
+            final Map<String, Map<String, String>> sourceMaster) throws IOException {
+        final VerificationResult result = new VerificationResult();
+        for (Map.Entry<String, Map<String, String>> fileEntry : sourceMaster.entrySet()) {
+            result.verifiedFiles++;
+            final Path sourceFile = sourceRoot.resolve(fileEntry.getKey());
+            if (!Files.isRegularFile(sourceFile)) {
+                result.missingFiles++;
+                continue;
+            }
+
+            final String content = new String(Files.readAllBytes(sourceFile), StandardCharsets.UTF_8);
+            for (String key : fileEntry.getValue().keySet()) {
+                if (key == null || key.isEmpty()) {
+                    result.emptyKeys++;
+                    continue;
+                }
+                if (content.indexOf(key) < 0) {
+                    result.missingRawMatches++;
+                    continue;
+                }
+                if (!hasRuntimeBoundaryMatch(content, key)) {
+                    result.missingRuntimeBoundaryMatches++;
+                    continue;
+                }
+                result.verifiedKeys++;
+            }
+        }
+        return result;
+    }
+
+    private static boolean hasRuntimeBoundaryMatch(final String content, final String key) {
+        if (key == null || key.isEmpty()) {
+            return false;
+        }
+
+        int index = content.indexOf(key);
+        while (index >= 0) {
+            if (hasRuntimeBoundaries(content, key, index)) {
+                return true;
+            }
+            index = content.indexOf(key, index + key.length());
+        }
+        return false;
+    }
+
+    private static boolean hasRuntimeBoundaries(final String content, final String key, final int index) {
+        boolean boundary = index + key.length() >= content.length();
+
+        if (!boundary) {
+            final char c = content.charAt(index + key.length() - 1);
+            final char lc = content.charAt(index + key.length());
+            boundary |= SentenceReader.punctuation(c) || SentenceReader.invisible(c);
+            boundary |= SentenceReader.punctuation(lc) || SentenceReader.invisible(lc);
+        }
+
+        if (boundary && index > 0) {
+            final char c = content.charAt(index - 1);
+            boundary = SentenceReader.punctuation(c) || SentenceReader.invisible(c);
+            final char fc = content.charAt(index);
+            boundary |= SentenceReader.punctuation(fc) || SentenceReader.invisible(fc);
+        }
+        return boundary;
+    }
+
+    static LinkedHashSet<String> extractCandidates(final String content) {
+        return extractCandidates(null, content);
+    }
+
+    static LinkedHashSet<String> extractCandidates(final String sourcePath, final String content) {
+        final LinkedHashSet<String> candidates = new LinkedHashSet<>();
+        String withoutIgnoredBlocks = HTML_COMMENT_PATTERN.matcher(
+                IGNORED_BLOCK_PATTERN.matcher(content).replaceAll("")).replaceAll("");
+        if ("jslicense.html".equals(sourcePath)) {
+            withoutIgnoredBlocks = withoutIgnoredBlocks.replaceAll("(?is)<td\\b[^>]*>.*?</td>", "");
+        }
+
+        Matcher tagMatcher = TAG_PATTERN.matcher(withoutIgnoredBlocks);
+        while (tagMatcher.find()) {
+            addAttributeCandidates(tagMatcher.group(), candidates);
+        }
+
+        final String textSource = UNPRESERVED_TAG_PATTERN.matcher(
+                LINK_TAG_PATTERN.matcher(BLOCK_TAG_PATTERN.matcher(withoutIgnoredBlocks).replaceAll("\n")).replaceAll(""))
+                .replaceAll("\n");
+        addTextCandidates(textSource, candidates);
+        return candidates;
+    }
+
+    private static void addAttributeCandidates(final String tag, final Set<String> candidates) {
+        final String lowerTag = tag.toLowerCase(Locale.ROOT);
+        if (lowerTag.contains("type=\"hidden\"") || lowerTag.contains("type='hidden'")) {
+            return;
+        }
+
+        final Matcher attributeMatcher = ATTRIBUTE_PATTERN.matcher(tag);
+        while (attributeMatcher.find()) {
+            final String attributeName = attributeMatcher.group(1).toLowerCase(Locale.ROOT);
+            if (TEXT_ATTRIBUTES.contains(attributeName)) {
+                if ("value".equals(attributeName) && !isButtonValueAttribute(lowerTag)) {
+                    continue;
+                }
+                final String value = attributeMatcher.group(3) != null ? attributeMatcher.group(3) : attributeMatcher.group(4);
+                final String quote = attributeMatcher.group(3) != null ? "\"" : "'";
+                addCandidate(quote + value + quote, candidates);
+            }
+        }
+    }
+
+    private static boolean isButtonValueAttribute(final String lowerTag) {
+        return lowerTag.startsWith("<button") || lowerTag.contains("type=\"submit\"") || lowerTag.contains("type='submit'")
+                || lowerTag.contains("type=\"button\"") || lowerTag.contains("type='button'")
+                || lowerTag.contains("type=\"reset\"") || lowerTag.contains("type='reset'");
+    }
+
+    private static void addTextCandidates(final String text, final Set<String> candidates) {
+        final String[] templateParts = TEMPLATE_BOUNDARY_PATTERN.split(text);
+        for (String templatePart : templateParts) {
+            final String[] lines = templatePart.split("\\r?\\n");
+            for (String line : lines) {
+                addCandidate(line, candidates);
+            }
+        }
+    }
+
+    private static void addCandidate(final String rawCandidate, final Set<String> candidates) {
+        if (rawCandidate == null) {
+            return;
+        }
+
+        final String candidate = normalizeFormattingWrapper(trimHtmlLineBreaks(
+                trimHtmlSpaceEntities(trimInvisible(rawCandidate))));
+        if (!isUsefulCandidate(candidate)) {
+            return;
+        }
+        candidates.add(candidate);
+    }
+
+    private static String trimInvisible(final String text) {
+        int start = 0;
+        int end = text.length();
+        while (start < end && Character.isWhitespace(text.charAt(start))) {
+            start++;
+        }
+        while (end > start && Character.isWhitespace(text.charAt(end - 1))) {
+            end--;
+        }
+        return text.substring(start, end);
+    }
+
+    private static String trimHtmlSpaceEntities(final String text) {
+        return HTML_SPACE_ENTITY_EDGE_PATTERN.matcher(text).replaceAll("");
+    }
+
+    private static String trimHtmlLineBreaks(final String text) {
+        return HTML_LINE_BREAK_EDGE_PATTERN.matcher(text).replaceAll("");
+    }
+
+    private static String normalizeFormattingWrapper(final String text) {
+        final Matcher matcher = FORMAT_WRAPPER_PATTERN.matcher(text);
+        if (matcher.matches()) {
+            return trimHtmlSpaceEntities(trimInvisible(matcher.group(2)));
+        }
+        return text;
+    }
+
+    private static boolean isUsefulCandidate(final String candidate) {
+        if (candidate.length() < 2) {
+            return false;
+        }
+        if (HTML_SPACE_ENTITY_ONLY_PATTERN.matcher(candidate).matches()) {
+            return false;
+        }
+        if (HTML_TAG_ONLY_PATTERN.matcher(candidate).matches()) {
+            return false;
+        }
+        if (candidate.indexOf("==") >= 0) {
+            return false;
+        }
+        if (candidate.endsWith("=")) {
+            return false;
+        }
+        if (candidate.startsWith("#")) {
+            return false;
+        }
+        if (candidate.startsWith("//") || candidate.startsWith("/*") || candidate.startsWith("*")) {
+            return false;
+        }
+        if (candidate.startsWith(":") || candidate.startsWith(".") || candidate.startsWith("-")) {
+            return false;
+        }
+        if (candidate.indexOf("#[") >= 0 || candidate.indexOf("#(") >= 0 || candidate.indexOf("#{") >= 0
+                || candidate.indexOf("#%") >= 0) {
+            return false;
+        }
+
+        boolean hasLetter = false;
+        for (int i = 0; i < candidate.length(); i++) {
+            if (Character.isLetter(candidate.charAt(i))) {
+                hasLetter = true;
+                break;
+            }
+        }
+        return hasLetter;
+    }
+
+    static final class ExtractionResult {
+
+        private final Map<String, Map<String, String>> sourceMaster = new TreeMap<>();
+
+        private int rejectedCandidates;
+    }
+
+    static final class VerificationResult {
+
+        private int verifiedFiles;
+
+        private int verifiedKeys;
+
+        private int missingFiles;
+
+        private int emptyKeys;
+
+        private int missingRawMatches;
+
+        private int missingRuntimeBoundaryMatches;
+
+        private boolean isOk() {
+            return this.missingFiles == 0 && this.emptyKeys == 0 && this.missingRawMatches == 0
+                    && this.missingRuntimeBoundaryMatches == 0;
+        }
+
+        @Override
+        public String toString() {
+            return "verifiedFiles=" + this.verifiedFiles + ", verifiedKeys=" + this.verifiedKeys
+                    + ", missingFiles=" + this.missingFiles + ", emptyKeys=" + this.emptyKeys
+                    + ", missingRawMatches=" + this.missingRawMatches
+                    + ", missingRuntimeBoundaryMatches=" + this.missingRuntimeBoundaryMatches;
+        }
+    }
+}
@@ -162,17 +162,16 @@ public class TranslationManager extends TranslatorXliff {

    /**
     * Create a master translation list by reading all translation files
-     * If a masterOutputFile exists, content is preserved (loaded first)
+     * Existing masterOutputFile content is replaced. This keeps the generated
+     * master list in sync with strings that still occur in the current source
+     * files and avoids carrying stale source entries forward.
     *
     * @param localesFolder folder containing *.lng translation files
     * @param masterOutputFile output file (xliff format). Must not be null.
     * @throws IOException
     */
    public void createMasterTranslationLists(final File localesFolder, final File masterOutputFile) throws IOException {
-        if (masterOutputFile.exists()) // if file exists, conserve existing master content (may be updated by external tool)
-            mainTransLists = loadTranslationsListsFromXliff(masterOutputFile);
-        else
-            mainTransLists = new TreeMap<String, Map<String, String>>();
+        mainTransLists = new TreeMap<String, Map<String, String>>();

        List<String> lngFiles = Translator.langFiles(localesFolder);
        for (String filename : lngFiles) {
@@ -227,7 +227,7 @@ public class TranslatorXliff extends Translator {
            final OutputStreamWriter output = new OutputStreamWriter(fileOutStream, StandardCharsets.UTF_8.name());
        ) {
            output.write("<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n");
-            output.write("<xliff version='1.2' xmlns='urn:oasis:names:tc:xliff:document:1.2'> \n");
+            output.write("<xliff version='1.2' xmlns='urn:oasis:names:tc:xliff:document:1.2'>\n");
            for (String afilemap : lng.keySet()) {
                output.write("<file original=\"" + afilemap + "\" " // original required in xliff 1.2
                        + " source-language=\"" + sourceLanguage + "\" "); // required in xliff 1.2
@@ -362,6 +362,7 @@ public final class yacy {
                            try { //is this another version?!
                                final File sourceDir = new File(sb.getConfig(SwitchboardConstants.HTROOT_PATH, SwitchboardConstants.HTROOT_PATH_DEFAULT));
                                final File destDir = new File(sb.getDataPath("locale.translated_html", "DATA/LOCALE/htroot"), tmplang);
+                                FileUtils.deletedelete(destDir);
                                if (new TranslatorXliff().translateFilesRecursive(sourceDir, destDir, new File(locale_source, tmplang + ".lng"), "html,template,inc", "locale")) { //translate it
                                    //write the new Versionnumber
                                    final BufferedWriter bw = new BufferedWriter(new PrintWriter(new FileWriter(new File(destDir, "version"))));
@@ -60,4 +60,21 @@ public class TranslatorTest {
        }
    }

+    @Test
+    public void testTranslateSkipsTechnicalTargets() {
+        final Translator translator = new Translator();
+        final Map<String, String> translationTable = new HashMap<String, String>();
+        translationTable.put("Network", "Réseau");
+        translationTable.put("share", "partager");
+        translationTable.put("localhost", "hôte-local");
+
+        assertEquals("Réseau", translator.translate(new StringBuilder("Network"), translationTable));
+        assertEquals("<a href=\"Network.html\">Réseau</a>",
+                translator.translate(new StringBuilder("<a href=\"Network.html\">Network</a>"), translationTable));
+        assertEquals("servlet share.json",
+                translator.translate(new StringBuilder("servlet share.json"), translationTable));
+        assertEquals("http://localhost:8090/Network.html",
+                translator.translate(new StringBuilder("http://localhost:8090/Network.html"), translationTable));
+    }
+
 }
@@ -0,0 +1,45 @@
+package net.yacy.utils.translation;
+
+import static org.junit.Assert.assertFalse;
+import static org.junit.Assert.assertTrue;
+
+import java.io.File;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.util.Map;
+
+import org.junit.Rule;
+import org.junit.Test;
+import org.junit.rules.TemporaryFolder;
+
+public class GenerateSourceMasterXliffTest {
+
+    @Rule
+    public TemporaryFolder temporaryFolder = new TemporaryFolder();
+
+    @Test
+    public void testMainExtractsSourceVisibleTextAndVerifiesOutput() throws Exception {
+        final File sourceRoot = this.temporaryFolder.newFolder("htroot");
+        final File sourceFile = new File(sourceRoot, "Network.html");
+        Files.write(sourceFile.toPath(), (
+                "<html><body>\n"
+                + "<a href=\"Network.html\" title=\"Network link\">Network</a>\n"
+                + "<input type=\"submit\" value=\"Save\">\n"
+                + "<input type=\"hidden\" value=\"Internal secret\">\n"
+                + "<script>Ignored script text</script>\n"
+                + "</body></html>\n").getBytes(StandardCharsets.UTF_8));
+
+        final File outputFile = this.temporaryFolder.newFile("source-master.lng.xlf");
+        GenerateSourceMasterXliff.main(new String[] { sourceRoot.getAbsolutePath(), outputFile.getAbsolutePath() });
+
+        final Map<String, Map<String, String>> master = new TranslatorXliff().loadTranslationsListsFromXliff(outputFile);
+        assertTrue(master.containsKey("Network.html"));
+
+        final Map<String, String> entries = master.get("Network.html");
+        assertTrue(entries.containsKey("Network"));
+        assertTrue(entries.containsKey("\"Network link\""));
+        assertTrue(entries.containsKey("\"Save\""));
+        assertFalse(entries.containsKey("\"Internal secret\""));
+        assertFalse(entries.containsKey("Ignored script text"));
+    }
+}