auto-coder 0.1.44__tar.gz → 0.1.45__tar.gz

{auto-coder-0.1.44 → auto-coder-0.1.45}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.44
+Version: 0.1.45
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
@@ -42,12 +42,13 @@ Requires-Dist: sqlmodel
 Requires-Dist: wudao
 Requires-Dist: SwissArmyTransformer
 Requires-Dist: jieba
-Requires-Dist: byzerllm>=0.1.73
+Requires-Dist: byzerllm>=0.1.75
 Requires-Dist: GitPython
 Requires-Dist: openai>=1.14.3
 Requires-Dist: anthropic
 Requires-Dist: google-generativeai
 Requires-Dist: protobuf
+Requires-Dist: azure-cognitiveservices-speech
 
 <p align="center">
   <picture>    
@@ -68,7 +69,7 @@ Auto-Coder (powered by Byzer-LLM)
 
 *Latest News* 🔥
 
-- [2024/04] Release Auto-Coder 0.1.42
+- [2024/04] Release Auto-Coder 0.1.45
 - [2024/03] Release Auto-Coder 0.1.25
 - [2024/03] Release Auto-Coder 0.1.24
 

{auto-coder-0.1.44 → auto-coder-0.1.45}/README.md

@@ -17,7 +17,7 @@ Auto-Coder (powered by Byzer-LLM)
 
 *Latest News* 🔥
 
-- [2024/04] Release Auto-Coder 0.1.42
+- [2024/04] Release Auto-Coder 0.1.45
 - [2024/03] Release Auto-Coder 0.1.25
 - [2024/03] Release Auto-Coder 0.1.24
 

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/auto_coder.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.44
+Version: 0.1.45
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
@@ -42,12 +42,13 @@ Requires-Dist: sqlmodel
 Requires-Dist: wudao
 Requires-Dist: SwissArmyTransformer
 Requires-Dist: jieba
-Requires-Dist: byzerllm>=0.1.73
+Requires-Dist: byzerllm>=0.1.75
 Requires-Dist: GitPython
 Requires-Dist: openai>=1.14.3
 Requires-Dist: anthropic
 Requires-Dist: google-generativeai
 Requires-Dist: protobuf
+Requires-Dist: azure-cognitiveservices-speech
 
 <p align="center">
   <picture>    
@@ -68,7 +69,7 @@ Auto-Coder (powered by Byzer-LLM)
 
 *Latest News* 🔥
 
-- [2024/04] Release Auto-Coder 0.1.42
+- [2024/04] Release Auto-Coder 0.1.45
 - [2024/03] Release Auto-Coder 0.1.25
 - [2024/03] Release Auto-Coder 0.1.24
 

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/auto_coder.egg-info/SOURCES.txt

@@ -16,6 +16,7 @@ src/autocoder/agent/coder.py
 src/autocoder/common/JupyterClient.py
 src/autocoder/common/ShellClient.py
 src/autocoder/common/__init__.py
+src/autocoder/common/audio.py
 src/autocoder/common/cleaner.py
 src/autocoder/common/code_auto_execute.py
 src/autocoder/common/code_auto_generate.py

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/auto_coder.egg-info/requires.txt

@@ -33,9 +33,10 @@ sqlmodel
 wudao
 SwissArmyTransformer
 jieba
-byzerllm>=0.1.73
+byzerllm>=0.1.75
 GitPython
 openai>=1.14.3
 anthropic
 google-generativeai
 protobuf
+azure-cognitiveservices-speech

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/autocoder/agent/coder.py

@@ -4,12 +4,10 @@ from pydantic import BaseModel,Field
 class ReactThink(BaseModel):
     thoughts:str = Field(...,description="Thoughts on current situation, reflect on how you should proceed to fulfill the user requirement")
     state: bool = Field(...,description="Decide whether you need to take more actions to complete the user requirement. Return true if you think so. Return false if you think the requirement has been completely fulfilled.")
-
         
-
 class Coder:
     def __init__(self,llm:byzerllm.ByzerLLM) -> None:
-        self.llm = llm
+        self.llm = llm 
         self.memory = []
 
     @byzerllm.prompt(llm=lambda self:self.llm)
@@ -17,10 +15,10 @@ class Coder:
         '''
         # User Requirement
         {user_requirement}
+        
         # Context
-        {context}                
+        {context}
         '''
-
+    
     def run(self,with_message):
-        pass    
-
+        pass

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/autocoder/auto_coder.py

@@ -56,7 +56,7 @@ def main():
     print("-" * 50)
     for arg, value in vars(args).items():
         print(f"{arg:20}: {value}")
-    print("-" * 50)
+    print("-" * 50)    
 
     # init store
     store = Store(os.path.join(args.source_dir,".auto-coder","metadata.db"))
@@ -178,10 +178,10 @@ def main():
             response,contexts = rag.stream_search(args.query)
             
             s = ""
-            print("\n\n=============RESPONSE==================\n\n")            
+            print("\n\n=============RESPONSE==================\n\n")                                
             for res in response:                
                 print(res,end="")  
-                s  += res                
+                s  += res                          
             
             print("\n\n=============CONTEXTS==================")
         

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/autocoder/command_args.py

@@ -53,8 +53,9 @@ def parse_args() -> AutoCoderArgs:
     
     parser.add_argument("--search_engine", default="", help=desc["search_engine"])
     parser.add_argument("--search_engine_token", default="",help=desc["search_engine_token"])
-    parser.add_argument("--enable_rag_search", action='store_true',help=desc["enable_rag_search"])
-    parser.add_argument("--enable_rag_context", action='store_true', default=False, help=desc["enable_rag_context"])
+    
+    parser.add_argument("--enable_rag_search", nargs='?', const=True, default=False, help=desc["enable_rag_search"])
+    parser.add_argument("--enable_rag_context", nargs='?',const=True, default=False, help=desc["enable_rag_context"])    
 
     parser.add_argument("--auto_merge", action='store_true', help=desc["auto_merge"])
 

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/autocoder/common/__init__.py

@@ -4,7 +4,7 @@ import sys
 import subprocess
 import os
 import time
-from typing import List,Dict,Any,Optional
+from typing import List,Dict,Any,Optional,Union
 
 class SourceCode(pydantic.BaseModel):
     module_name: str
@@ -85,8 +85,8 @@ class AutoCoderArgs(pydantic.BaseModel):
 
     search_engine: str = ""
     search_engine_token: str = ""
-    enable_rag_search:bool
-    enable_rag_context:bool
+    enable_rag_search:Union[bool,str] = False
+    enable_rag_context:Union[bool,str] = False
     required_exts:str = ""
     
     auto_merge: bool = pydantic.Field(False, description="Whether to automatically merge the generated code into the existing file")

auto-coder-0.1.45/src/autocoder/common/audio.py

@@ -0,0 +1,92 @@
+import io
+import os
+import queue
+import threading
+import byzerllm
+import time
+import json
+import base64
+from concurrent.futures import ThreadPoolExecutor
+import shutil
+
+def play_wave(filename:str):
+    try:
+        import simpleaudio as sa
+    except ImportError:
+        raise ImportError("The simpleaudio module is required to play audio files. Try running 'pip install simpleaudio'")
+
+    wave_obj = sa.WaveObject.from_wave_file(filename)
+    play_obj = wave_obj.play()
+    play_obj.wait_done()
+
+class PlayStreamAudioFromText:
+    def __init__(self, tts_model_name:str="openai_tts"):
+        self.q = queue.Queue()
+        self.pool = ThreadPoolExecutor(max_workers=5)
+        self.llm = byzerllm.ByzerLLM()
+        self.llm.setup_default_model_name(tts_model_name)
+        self.wav_num = -1
+
+    def text_to_speech(self, text, file_path):
+        print(f"Converting text to speech: {text}")
+        t = self.llm.chat_oai(conversations=[{
+            "role":"user",
+            "content": json.dumps({
+                "input": text,
+                "voice": "echo",
+                "response_format": "wav"
+            }, ensure_ascii=False)
+        }])
+        temp_file_path = file_path + ".tmp"
+        with open(temp_file_path, "wb") as f:
+            f.write(base64.b64decode(t[0].output))
+        shutil.move(temp_file_path, file_path)
+        print(f"Converted successfully: {file_path}")
+        self.q.task_done()
+
+    def play_audio_files(self):
+        idx = 1
+        while True:
+            if self.wav_num == -2:
+                break
+            file_path = f"/tmp/wavs/{idx:03d}.wav"
+            if not os.path.exists(file_path):
+                time.sleep(0.1) # Reduce CPU usage
+                continue
+            play_wave(file_path)
+            idx += 1
+            if idx > self.wav_num:
+                self.wav_num = -2
+
+    def process_texts(self):
+        idx = 1
+        s = ""
+        while True:
+            text = self.q.get()
+            if text is None:
+                self.q.task_done()
+                break
+            s += text
+            if len(s) < 10:
+                continue
+            sentences = s.split("。")
+            for sentence in sentences:
+                if len(sentence) == 0:
+                    continue
+                file_path = f"/tmp/wavs/{idx:03d}.wav"
+                print(f"Processing: {sentence} to {file_path}")
+                self.pool.submit(self.text_to_speech, sentence, file_path)
+                idx += 1
+            s = ""
+        self.wav_num = idx - 1
+
+    def run(self, text_generator):
+        os.makedirs("/tmp/wavs", exist_ok=True)
+        threading.Thread(target=self.play_audio_files).start()
+        threading.Thread(target=self.process_texts)._start()
+        for text in text_generator:
+            self.q.put(text)
+        self.q.put(None)
+        while self.wav_num != -2:
+            time.sleep(0.1)
+

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/autocoder/index/index.py

@@ -154,16 +154,15 @@ class IndexManager:
            if ext in [".md",".html",".txt",".doc",".pdf"]:
                continue
 
-           md5 = hashlib.md5(source.source_code.encode('utf-8')).hexdigest()
-           logger.info(f"try to build index for {file_path} md5: {md5}")
+           md5 = hashlib.md5(source.source_code.encode('utf-8')).hexdigest()           
            if source.source_code.strip() == "":
                continue
 
            if source.module_name in index_data and index_data[source.module_name]["md5"] == md5:
                continue
            
-           try:
-               logger.info(f"parse and update index for {file_path} md5: {md5}")
+           try:               
+               start_time = time.monotonic()
                source_code = source.source_code
                if len(source.source_code) > self.max_input_length:
                    logger.warning(f"Warning: The length of source code is too long ({len(source.source_code)}) > model_max_input_length({self.max_input_length}), splitting into chunks...")
@@ -177,7 +176,9 @@ class IndexManager:
                else:
                    symbols = self.get_all_file_symbols(source.module_name, source_code)
                    time.sleep(self.anti_quota_limit)
-               time.sleep(self.anti_quota_limit)
+                
+               logger.info(f"Parse and update index for {file_path} md5: {md5} took {time.monotonic() - start_time:.2f}s") 
+               
            except Exception as e:
                logger.warning(f"Error: {e}")
                continue
@@ -291,18 +292,16 @@ class IndexManager:
         all_results:List[TargetFile] = []
         
         def w():
-            return self._get_meta_str(skip_symbols=True)
+            return self._get_meta_str(skip_symbols=True,max_chunk_size=self.max_input_length-1000)
 
+        logger.info("Find the related files by query according to the files...")
         temp_result = self._query_index_with_thread(query,w)
-        all_results.extend(temp_result)
-            
-
+        all_results.extend(temp_result)            
         
-        if self.args.index_filter_level >= 1:    
-            
-            logger.info("Find the related files by query according to the file and symbols...")
+        if self.args.index_filter_level >= 1:                
+            logger.info("Find the related files by query according to the symbols...")
             def w():
-                return self._get_meta_str(skip_symbols=False)
+                return self._get_meta_str(skip_symbols=False,max_chunk_size=self.max_input_length-1000)
             temp_result = self._query_index_with_thread(query,w)            
             all_results.extend(temp_result)
         
@@ -340,10 +339,13 @@ def build_index_and_filter_files(llm,args:AutoCoderArgs,sources:List[SourceCode]
        for source in sources:
            if source.tag in ["REST","RAG","SEARCH"]:
                 final_files.append(get_file_path(source.module_name))
-
+        
+       logger.info("Building index for all files...")
        index_manager = IndexManager(llm=llm,sources=sources,args=args)
        index_manager.build_index()
        
+       logger.info(f"Finding related files in the index...")
+       start_time = time.monotonic()
        target_files = index_manager.get_target_files_by_query(args.query)
        
        if target_files:
@@ -364,7 +366,9 @@ def build_index_and_filter_files(llm,args:AutoCoderArgs,sources:List[SourceCode]
        
        if not final_files:
            logger.warning("Warning: No related files found, use all files")
-           final_files = [file.module_name for file in sources]                          
+           final_files = [file.module_name for file in sources] 
+       
+       logger.info(f"Find related files took {time.monotonic() - start_time:.2f}s")
    else:
        final_files = [file.module_name for file in sources]
 

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/autocoder/rag/llm_wrapper.py

@@ -6,7 +6,7 @@ from byzerllm.utils.client.types import (
 import pydantic
 from byzerllm import ByzerLLM
 from byzerllm.utils.client import LLMResponse
-from byzerllm.utils import SingleOutputMeta
+from byzerllm.utils.types import SingleOutputMeta
 from autocoder.rag.simple_rag import SimpleRAG
 from loguru import logger
 

{auto-coder-0.1.44 → auto-coder-0.1.45}/src/autocoder/rag/simple_rag.py

@@ -132,13 +132,19 @@ class SimpleRAG:
     
     def search(self,query:str) -> List[SourceCode]:
         if self.args.enable_rag_search:
-            texts,contexts = self.stream_search(query)
+            target_query = query
+            if isinstance(self.args.enable_rag_search,str):
+                target_query = self.args.enable_rag_search
+            texts,contexts = self.stream_search(target_query)
             s = "".join([text for text in texts])
             urls = ",".join(set([context["doc_url"] for context in contexts]))
             ## append RAG: prefix is used to protect avoid the source code is modified by the code auto execute
             return [SourceCode(module_name=f"RAG:{urls}", source_code=s)]
         elif self.args.enable_rag_context:
-            contexts = self.retrieve(query) 
+            target_query = query            
+            if isinstance(self.args.enable_rag_context,str):
+                target_query = self.args.enable_rag_context            
+            contexts = self.retrieve(target_query) 
             for context in contexts:  
                 context["raw_chunk"]
                 try:

auto-coder-0.1.45/src/autocoder/version.py

@@ -0,0 +1 @@
+__version__ = "0.1.45"

auto-coder-0.1.44/src/autocoder/version.py

@@ -1 +0,0 @@
-__version__ = "0.1.44"