speaksy 0.1.0__py3-none-any.whl → 0.1.1__py3-none-any.whl

speaksy/core.py

@@ -3,6 +3,7 @@
 import io
 import logging
 import os
+import re
 import signal
 import subprocess
 import sys
@@ -198,52 +199,53 @@ def route_transcription(audio_buf, groq, local, config):
 # Text Cleanup (LLM post-processing)
 # ---------------------------------------------------------------------------
 
-CLEANUP_PROMPT = (
-    "Clean up this voice transcription. Fix grammar, punctuation, and "
-    "capitalization. Remove filler words (um, uh, like, you know, so, "
-    "basically, actually). Do NOT change the meaning, add new content, "
-    "or remove meaningful words. If the text is already clean, return it "
-    "unchanged. Return ONLY the cleaned text, nothing else."
-)
+# Filler words to remove (with word boundaries)
+FILLER_PATTERNS = [
+    r"\b(um+|uh+|er+|ah+)\b",
+    r"\b(like,?\s+)(?=\w)",  # "like" as filler, not "I like pizza"
+    r"\b(you know,?\s*)",
+    r"\b(basically,?\s*)",
+    r"\b(actually,?\s*)(?![\w])",  # "actually" as filler
+    r"\b(so,?\s+)(?=[a-z])",  # "so" at start as filler
+    r"\b(i mean,?\s*)",
+    r"\b(kind of|kinda)\s+",
+    r"\b(sort of|sorta)\s+",
+]
 
 
 class TextCleaner:
-    """Post-process transcribed text through a fast LLM."""
+    """Clean up transcribed text using simple regex rules."""
 
-    CHAT_URL = "https://api.groq.com/openai/v1/chat/completions"
-
-    def __init__(self, api_key, model="llama-3.1-8b-instant"):
-        self.api_key = api_key
-        self.model = model
+    def __init__(self, api_key=None, model=None):
+        # API key and model not used - kept for backward compatibility
+        pass
 
     def clean(self, text):
-        if not self.api_key or not text:
+        if not text:
             return text
 
-        try:
-            resp = httpx.post(
-                self.CHAT_URL,
-                headers={"Authorization": f"Bearer {self.api_key}"},
-                json={
-                    "model": self.model,
-                    "messages": [
-                        {"role": "system", "content": CLEANUP_PROMPT},
-                        {"role": "user", "content": text},
-                    ],
-                    "temperature": 0,
-                    "max_tokens": len(text) * 2,
-                },
-                timeout=10.0,
-            )
-            resp.raise_for_status()
-            cleaned = resp.json()["choices"][0]["message"]["content"].strip()
-            if cleaned:
-                log.info(f'[Cleanup] "{text}" -> "{cleaned}"')
-                return cleaned
-        except Exception as e:
-            log.warning(f"Text cleanup failed ({e}), using raw transcription")
+        original = text
+        cleaned = text
+
+        # Remove filler words
+        for pattern in FILLER_PATTERNS:
+            cleaned = re.sub(pattern, "", cleaned, flags=re.IGNORECASE)
+
+        # Clean up multiple spaces
+        cleaned = re.sub(r"\s+", " ", cleaned).strip()
+
+        # Capitalize first letter
+        if cleaned:
+            cleaned = cleaned[0].upper() + cleaned[1:]
+
+        # Add period if no ending punctuation
+        if cleaned and cleaned[-1] not in ".!?":
+            cleaned += "."
+
+        if cleaned != original:
+            log.info(f'[Cleanup] "{original}" -> "{cleaned}"')
 
-        return text
+        return cleaned
 
 
 # ---------------------------------------------------------------------------

{speaksy-0.1.0.dist-info → speaksy-0.1.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: speaksy
-Version: 0.1.0
+Version: 0.1.1
 Summary: Voice typing for Linux. Talk it. Type it. Ship it.
 Project-URL: Homepage, https://github.com/oneKn8/speaksy
 Project-URL: Repository, https://github.com/oneKn8/speaksy

{speaksy-0.1.0.dist-info → speaksy-0.1.1.dist-info}/RECORD

@@ -2,12 +2,12 @@ speaksy/__init__.py,sha256=eG5kdi4U1tayyWwhBUByiHrdFcZpCJ6bninHJclA4VU,63
 speaksy/__main__.py,sha256=Y5c4dtERDNv0x1C-HcP1nauSAbssfyuP7MlyN8l1M28,110
 speaksy/cli.py,sha256=Fd-g5g11RMFyKgpdYGtvQAuYIVY6BcL7nyOwFa5GMHM,10225
 speaksy/config.py,sha256=bGiATdV79E27pu7TQTm2VxDDdAKYdzjQrrIbu5GCz2o,4163
-speaksy/core.py,sha256=xkE0KtS3paGzZSnc481R47bECnns3_wSXkysSW66Uzs,16987
+speaksy/core.py,sha256=NxbGWjKWkTXJdoS9pEL_qLchNttKF4R_HkIW19VdEkY,16695
 speaksy/runner.py,sha256=l5aiczErHVLBc3Ir3XIfqxZUAiq4tN6T8OCVqTWwnEU,643
 speaksy/service.py,sha256=EwQs3yB3Ys-WwvP0UI2K_U5_zJqJCJLkK29_MTTcVTE,5449
 speaksy/setup_wizard.py,sha256=-Glu74R-ZKL097Ji8vOmEDC7salaSk9kueONxWs0t4c,7033
-speaksy-0.1.0.dist-info/METADATA,sha256=JwjfdFK_CSimqUGGZC0VwJxW1lvWmrCENw9B56cFCBw,6926
-speaksy-0.1.0.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
-speaksy-0.1.0.dist-info/entry_points.txt,sha256=LFch1KPmgD4hHNVXY_DSJ24fPnB8GaVKF4aKUrqDjzs,45
-speaksy-0.1.0.dist-info/licenses/LICENSE,sha256=KgmDIQPh17s8aGNha9ebeUXZHi533ew6VyCLcY7IJE4,1066
-speaksy-0.1.0.dist-info/RECORD,,
+speaksy-0.1.1.dist-info/METADATA,sha256=LnGcQYfq7p-p3ojsCxTuiuwJjxOvNui2e83yUfYwKQE,6926
+speaksy-0.1.1.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+speaksy-0.1.1.dist-info/entry_points.txt,sha256=LFch1KPmgD4hHNVXY_DSJ24fPnB8GaVKF4aKUrqDjzs,45
+speaksy-0.1.1.dist-info/licenses/LICENSE,sha256=KgmDIQPh17s8aGNha9ebeUXZHi533ew6VyCLcY7IJE4,1066
+speaksy-0.1.1.dist-info/RECORD,,