docs: add HF Space metadata and app metadata

README.md

@@ -1,7 +1,7 @@
 ---
-title: Phonemizer (EUS–ESP)
+title: Itzultzaile fonetikoa — Phonetic Translator (EUS–ESP)
 emoji: "🔤"
-colorFrom: blue
+colorFrom: orange
 colorTo: indigo
 sdk: gradio
 app_file: app.py
@@ -9,19 +9,25 @@ pinned: false
 ---
  
 
-# Phonemizer (EUS–ESP)
+# Itzultzaile fonetikoa — Phonetic Translator (EUS–ESP)
 
-# Phonemizer — Gradio demo (Hugging Face Space)
+# Itzultzaile fonetikoa — Phonetic Translator (EUS–ESP) — Gradio demo (Hugging Face Space)
 
-This Space provides a small web UI to phonemize Basque (eu) and Spanish (es) text.
+This Space provides a small web UI to produce phonetic transcriptions for Basque (eu) and Spanish (es) text.
 
 How to use
 - Input text: paste text into the main box or upload a `.txt` file.
 - Language: select `eu` (Basque) or `es` (Spanish).
 - Symbols: choose `sampa` (default) or `ipa` for the phoneme output format.
-- Separate phonemes: toggle whether phonemes are separated by spaces to make easier to see multi-character phonemes.
-- Submit: press `Submit` to run normalization + phonemization.
-- Download: use the download buttons to get the phonemes or normalized text as `.txt` files.
+- Separate phonos: toggle whether phonos are separated by spaces to make it easier to see multi-character phonos.
+- Submit: press `Submit` to run normalization + phonetic transcription.
+- Download: use the download buttons to get the phonos or normalized text as `.txt` files.
+
+## Notes / Notas
+
+- EN: Diacritics are NOT used.
+
+- EN: Diphthongs: the semiconsonant is always `w`/`j` (aire -> 'a j ɾ e; causó -> k a w s 'o)
 
 Privacy
 - This Space does not store user inputs beyond temporary files used to serve downloads. Do not upload sensitive data.

gradio_phonemizer.py

@@ -101,16 +101,17 @@ def _read_uploaded_file(file_obj) -> str:
 
 
 def process(text: str,
-            uploaded_file, 
-            language: str,
-            symbol: str,
-            separate_phonemes: bool) -> Tuple[str, Optional[str], str, Optional[str]]:
-    """Process either text input or uploaded txt file and return (text_output, download_file_path)
-
-    If the user uploaded a file, the function will return the path to a tmp file
-    suitable for download as the second return value and an empty text output.
-    If the user provided text in the box, the function will return the phonemes
-    as text and also a downloadable txt file containing the same output.
+        uploaded_file,
+        language: str,
+        symbol: str,
+        separate_phonemes: bool) -> Tuple[str, Optional[str], str, Optional[str], Optional[str], Optional[str]]:
+    """Process either text input or uploaded txt file and return a 6-tuple:
+
+    (result_text, ph_file_path, normalized_text, norm_file_path, ph_path, norm_path)
+
+    If the user uploaded a file, the function will return the path to a file
+    suitable for download and also populate the inline text results. When the
+    user provided text, both inline text and downloadable files are produced.
     """
     # Prefer uploaded file if present
     source_text = ""
@@ -126,15 +127,15 @@ def process(text: str,
         phon = Phonemizer(language=language, symbol=symbol)
     except PhonemizerError as e:
         if language == 'eu':
-            err = f"Ezin izan da fonemizadorea hasi: {e}\nEgiaztatu 'modulo1y2' eta 'dict' karpetak."
+            err = f"Ezin izan da itzultzaile fonetikoa hasi: {e}\nEgiaztatu 'modulo1y2' eta 'dict' karpetak."
         else:
-            err = f"No se pudo inicializar el fonemizador: {e}\nComprueba las carpetas 'modulo1y2' y 'dict'."
+            err = f"No se pudo inicializar el traductor fonético: {e}\nComprueba las carpetas 'modulo1y2' y 'dict'."
         # Return 6 outputs matching the UI: result text, file, normalized text, norm file, ph_path, norm_path
         return err, None, "", None, "", ""
     except Exception as e:
         if language == 'eu':
             return f"Hasieratze errore ezezaguna: {e}", None, "", None, "", ""
-        return f"Error inesperado al inicializar: {e}", None
+        return f"Error inesperado al inicializar: {e}", None, "", None, "", ""
         
 
     # Normalize then get phonemes. Run normalization per original input line so the
@@ -163,9 +164,9 @@ def process(text: str,
             phonemes = re.sub(r"\s*\|\s*", " ", phonemes)
     except PhonemizerError as e:
         if language == 'eu':
-            msg = f"Fonemizazio errorea: {e}"
+            msg = f"Itzultzaile fonetikoaren errorea: {e}"
         else:
-            msg = f"Error del fonemizador: {e}"
+            msg = f"Error del traductor fonético: {e}"
         return msg, None, "", None, "", ""
     except Exception as e:
         if language == 'eu':
@@ -181,7 +182,7 @@ def process(text: str,
     out_dir.mkdir(parents=True, exist_ok=True)
     from datetime import datetime
     ts = datetime.now().strftime('%Y%m%d_%H%M%S')
-    ph_file = out_dir / f'phonemes_{ts}.txt'
+    ph_file = out_dir / f'phonos_{ts}.txt'
     norm_file = out_dir / f'normalized_{ts}.txt'
     ph_file.write_text(phonemes, encoding='utf-8')
     norm_file.write_text(normalized, encoding='utf-8')
@@ -206,7 +207,7 @@ def download_from_text(text: str) -> Optional[str]:
     out_dir.mkdir(parents=True, exist_ok=True)
     from datetime import datetime
     ts = datetime.now().strftime('%Y%m%d_%H%M%S')
-    filename = f'phonemes_{ts}.txt'
+    filename = f'phonos_{ts}.txt'
     out_path = out_dir / filename
     out_path.write_text(text, encoding='utf-8')
     # Return the path string so Gradio's File component can serve it
@@ -216,7 +217,7 @@ def download_from_text(text: str) -> Optional[str]:
 def build_interface():
     with gr.Blocks(title="Eu/Es Phonemizer") as demo:
         # Simple header (image removed per user preference)
-        header = gr.Markdown("# Fonemizadorea — Euskara (eu) eta Gaztelania (es)")
+        header = gr.Markdown("# Itzultzaile fonetikoa — Euskara (eu) eta Gaztelania (es)")
 
         # Show README instructions in a collapsible panel so users can read
         # usage notes directly inside the app without leaving the UI.
@@ -341,7 +342,7 @@ def build_interface():
                 language = gr.Radio(choices=['eu', 'es'], value='eu', label='Hizkuntza / Idioma')
                 symbol = gr.Radio(choices=['sampa', 'ipa'], value='sampa', label='Sinboloak / Símbolos (Irteera)')
                 # Default checked and Basque-only label; will switch to Spanish when language changes
-                separate_phonemes = gr.Checkbox(label='Banatu fonemak espazioz', value=True)
+                separate_phonemes = gr.Checkbox(label='Banatu fonoak espazioz', value=True)
 
             # Small column to the right of controls that holds the upload box
             with gr.Column(scale=1, elem_id='upload_col'):
@@ -374,8 +375,8 @@ def build_interface():
                 download_norm_btn = gr.DownloadButton('Deskargatu normalizatua', elem_id='download_norm_btn')
 
             with gr.Column(scale=1):
-                result_box = gr.Textbox(lines=12, elem_id='result_box', label='Fonemak', interactive=False)
-                download_ph_btn = gr.DownloadButton('Deskargatu fonemak', elem_id='download_ph_btn')
+                result_box = gr.Textbox(lines=12, elem_id='result_box', label='Fonoak', interactive=False)
+                download_ph_btn = gr.DownloadButton('Deskargatu fonoak', elem_id='download_ph_btn')
 
         # hidden boxes to hold latest generated file paths so download buttons can trigger
         ph_path_box = gr.Textbox(visible=False, elem_id='ph_path_box')
@@ -424,9 +425,9 @@ def build_interface():
             if lang == 'eu':
                 return (
                     gr.update(label='Sinboloak (Irteera)'),                  # symbol
-                    gr.update(label='Banatu fonemak espazioz'),               # separate_phonemes
+                    gr.update(label='Banatu fonoak espazioz'),               # separate_phonemes
                     # keep input/upload labels stable (do not update them to avoid reflow)
-                    gr.update(label='Fonemak'),
+                    gr.update(label='Fonoak'),
                     gr.update(label='Deskargatu irteera (.txt)'),
                     gr.update(label='Normalizatua'),
                     gr.update(label='Deskargatu normalizatua (.txt)'),
@@ -436,9 +437,9 @@ def build_interface():
             else:
                 return (
                     gr.update(label='Símbolos (Salida)'),
-                    gr.update(label='Separar fonemas con espacios'),
+                    gr.update(label='Separar fonos con espacios'),
                     # keep input/upload labels stable (do not update them to avoid reflow)
-                    gr.update(label='Fonemas'),
+                    gr.update(label='Fonos'),
                     gr.update(label='Descargar salida (.txt)'),
                     gr.update(label='Normalizado'),
                     gr.update(label='Descargar normalizado (.txt)'),