Spaces:
Sleeping
Sleeping
Commit ·
f46b6ca
1
Parent(s): a886d4b
code logic completed for uploaded files with qdrant
Browse files- Loaders/generic.py +56 -2
- app.py +34 -16
- config.ini +12 -1
Loaders/generic.py
CHANGED
|
@@ -1,4 +1,6 @@
|
|
| 1 |
#import
|
|
|
|
|
|
|
| 2 |
from langchain_community.document_loaders.generic import GenericLoader
|
| 3 |
from langchain_community.document_loaders.parsers import LanguageParser
|
| 4 |
from langchain_text_splitters import Language
|
|
@@ -10,7 +12,8 @@ class GenericLoaders:
|
|
| 10 |
self.main_config = main_config
|
| 11 |
self.repo_path = repo_path
|
| 12 |
self.rag_path_ext= rag_path_ext
|
| 13 |
-
|
|
|
|
| 14 |
def loaders(self):
|
| 15 |
# Load
|
| 16 |
try :
|
|
@@ -29,10 +32,61 @@ class GenericLoaders:
|
|
| 29 |
)
|
| 30 |
documents = loader.load()
|
| 31 |
print(len(documents))
|
| 32 |
-
|
| 33 |
except Exception as e:
|
| 34 |
print(e)
|
| 35 |
|
| 36 |
return documents
|
| 37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
|
|
|
|
| 1 |
#import
|
| 2 |
+
import os
|
| 3 |
+
import tempfile
|
| 4 |
from langchain_community.document_loaders.generic import GenericLoader
|
| 5 |
from langchain_community.document_loaders.parsers import LanguageParser
|
| 6 |
from langchain_text_splitters import Language
|
|
|
|
| 12 |
self.main_config = main_config
|
| 13 |
self.repo_path = repo_path
|
| 14 |
self.rag_path_ext= rag_path_ext
|
| 15 |
+
|
| 16 |
+
#when files are present in dir
|
| 17 |
def loaders(self):
|
| 18 |
# Load
|
| 19 |
try :
|
|
|
|
| 32 |
)
|
| 33 |
documents = loader.load()
|
| 34 |
print(len(documents))
|
|
|
|
| 35 |
except Exception as e:
|
| 36 |
print(e)
|
| 37 |
|
| 38 |
return documents
|
| 39 |
|
| 40 |
+
|
| 41 |
+
# which files are uploading in streamlit
|
| 42 |
+
def code_loader_and_splitters(self):
|
| 43 |
+
documents = []
|
| 44 |
+
docs=[]
|
| 45 |
+
try :
|
| 46 |
+
code_files = self.repo_path
|
| 47 |
+
langlist = dict(self.main_config['Langlist'])
|
| 48 |
+
langlist = eval(langlist["lang_exts"]) # Convert the string to a dictionary
|
| 49 |
+
if code_files:
|
| 50 |
+
for code_file in code_files:
|
| 51 |
+
with tempfile.NamedTemporaryFile(delete=False) as tmp_file:
|
| 52 |
+
tmp_file.write(code_file.read())
|
| 53 |
+
|
| 54 |
+
lang_ext = str(code_file.name.split(".")[-1])
|
| 55 |
+
lang_ext = str('.'+lang_ext)
|
| 56 |
+
tmp_file_path = tmp_file.name+lang_ext
|
| 57 |
+
tmp_file_path=tmp_file_path.replace("\\","/")
|
| 58 |
+
os.rename(tmp_file.name,tmp_file_path )
|
| 59 |
+
loader = GenericLoader.from_filesystem(
|
| 60 |
+
tmp_file_path,
|
| 61 |
+
glob="**/*",
|
| 62 |
+
suffixes=[lang_ext],
|
| 63 |
+
exclude=["**/non-utf8-encoding.py"],
|
| 64 |
+
parser=LanguageParser(parser_threshold=500),
|
| 65 |
+
)
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
selected_lang =""
|
| 69 |
+
for lang, ext in langlist.items():
|
| 70 |
+
if ext.lower() == lang_ext:
|
| 71 |
+
selected_lang = lang
|
| 72 |
+
print(f"selected lang: {selected_lang}") # For debugging
|
| 73 |
+
|
| 74 |
+
if selected_lang =="":
|
| 75 |
+
raise ValueError("Issue in evaluating techctack from extension")
|
| 76 |
+
documents.extend(loader.load())
|
| 77 |
+
|
| 78 |
+
splitter = RecursiveCharacterTextSplitter.from_language(
|
| 79 |
+
language=selected_lang, chunk_size=2000, chunk_overlap=200
|
| 80 |
+
)
|
| 81 |
+
docs.extend(splitter.split_documents(documents))
|
| 82 |
+
|
| 83 |
+
# Clean up temporary file
|
| 84 |
+
os.remove(tmp_file_path)
|
| 85 |
+
|
| 86 |
+
except Exception as e:
|
| 87 |
+
raise ValueError(f"Problem Occured in loader and splitter : {e}")
|
| 88 |
+
|
| 89 |
+
return docs
|
| 90 |
+
|
| 91 |
+
|
| 92 |
|
app.py
CHANGED
|
@@ -1,3 +1,4 @@
|
|
|
|
|
| 1 |
from langchain_text_splitters import TextSplitter
|
| 2 |
import streamlit as st
|
| 3 |
import pandas as pd
|
|
@@ -32,13 +33,17 @@ main_config.read(config_file_path)
|
|
| 32 |
#Process Code Repo
|
| 33 |
def process_code_repo(main_config,db_option,repo_path,groq_api_key,selected_model, rag_path_ext,query):
|
| 34 |
if db_option == "Qdrant":
|
| 35 |
-
# LOAD
|
| 36 |
-
obj_loaders = GenericLoaders(main_config,repo_path, rag_path_ext)
|
| 37 |
-
documents = obj_loaders.loaders()
|
| 38 |
|
| 39 |
-
# TRANSFORM
|
| 40 |
-
obj_code_splitters = CodeSplitters(main_config,repo_path,rag_path_ext)
|
| 41 |
-
docs = obj_code_splitters.code_splitters(documents)
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
|
| 43 |
#EMBEDDINGS and stored into VDB
|
| 44 |
obj_qdr = QdrantVDB(main_config,repo_path,groq_api_key,selected_model, rag_path_ext,query)
|
|
@@ -133,17 +138,29 @@ db_option = st.sidebar.selectbox("Select Vector DB:", ["Qdrant", "Chroma"])
|
|
| 133 |
st.sidebar.header("RAG Document Format")
|
| 134 |
file_format = st.sidebar.selectbox("Select Input Document Format:", ["Code Repo", "PDF", "CSV"])
|
| 135 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 136 |
|
| 137 |
if file_format == "Code Repo":
|
| 138 |
-
st.sidebar.header("Select Language")
|
| 139 |
-
lang_options = ["java", "python", "csharp"] # Replace with your actual models
|
| 140 |
-
selected_lang = st.sidebar.selectbox("Select Model:", lang_options)
|
| 141 |
-
if selected_lang == "java":
|
| 142 |
-
|
| 143 |
-
elif selected_lang == "python":
|
| 144 |
-
|
| 145 |
-
elif selected_lang == "csharp":
|
| 146 |
-
|
|
|
|
| 147 |
|
| 148 |
elif file_format == "PDF":
|
| 149 |
rag_path_ext = ".pdf"
|
|
@@ -152,7 +169,8 @@ elif file_format == "CSV":
|
|
| 152 |
rag_path_ext = ".csv"
|
| 153 |
|
| 154 |
|
| 155 |
-
|
|
|
|
| 156 |
|
| 157 |
|
| 158 |
st.sidebar.header("Groq API Key")
|
|
|
|
| 1 |
+
from typing import Sequence
|
| 2 |
from langchain_text_splitters import TextSplitter
|
| 3 |
import streamlit as st
|
| 4 |
import pandas as pd
|
|
|
|
| 33 |
#Process Code Repo
|
| 34 |
def process_code_repo(main_config,db_option,repo_path,groq_api_key,selected_model, rag_path_ext,query):
|
| 35 |
if db_option == "Qdrant":
|
| 36 |
+
# # LOAD
|
| 37 |
+
# obj_loaders = GenericLoaders(main_config,repo_path, rag_path_ext)
|
| 38 |
+
# documents = obj_loaders.loaders()
|
| 39 |
|
| 40 |
+
# # TRANSFORM
|
| 41 |
+
# obj_code_splitters = CodeSplitters(main_config,repo_path,rag_path_ext)
|
| 42 |
+
# docs = obj_code_splitters.code_splitters(documents)
|
| 43 |
+
|
| 44 |
+
#LOAD and SPLIT
|
| 45 |
+
obj_loaders_splitter = GenericLoaders(main_config,repo_path, rag_path_ext)
|
| 46 |
+
docs = obj_loaders_splitter.code_loader_and_splitters()
|
| 47 |
|
| 48 |
#EMBEDDINGS and stored into VDB
|
| 49 |
obj_qdr = QdrantVDB(main_config,repo_path,groq_api_key,selected_model, rag_path_ext,query)
|
|
|
|
| 138 |
st.sidebar.header("RAG Document Format")
|
| 139 |
file_format = st.sidebar.selectbox("Select Input Document Format:", ["Code Repo", "PDF", "CSV"])
|
| 140 |
|
| 141 |
+
# Check if the section exists
|
| 142 |
+
if not main_config.has_section('Supported_Extensions'):
|
| 143 |
+
raise ValueError("Section 'Supported_Extensions' not found in config.ini")
|
| 144 |
+
|
| 145 |
+
# Extract extensions from the config dictionary
|
| 146 |
+
extensions = main_config['Supported_Extensions']['extensions']
|
| 147 |
+
supported_extensions = eval(extensions)
|
| 148 |
+
accepted_types = [f"{ext}" for ext in supported_extensions]
|
| 149 |
+
|
| 150 |
+
# Print the list of extensions
|
| 151 |
+
print(accepted_types)
|
| 152 |
|
| 153 |
if file_format == "Code Repo":
|
| 154 |
+
# st.sidebar.header("Select Language")
|
| 155 |
+
# lang_options = ["java", "python", "csharp"] # Replace with your actual models
|
| 156 |
+
# selected_lang = st.sidebar.selectbox("Select Model:", lang_options)
|
| 157 |
+
# if selected_lang == "java":
|
| 158 |
+
# rag_path_ext = ".java"
|
| 159 |
+
# elif selected_lang == "python":
|
| 160 |
+
# rag_path_ext = ".py"
|
| 161 |
+
# elif selected_lang == "csharp":
|
| 162 |
+
# rag_path_ext = ".cs"
|
| 163 |
+
rag_path_ext = accepted_types
|
| 164 |
|
| 165 |
elif file_format == "PDF":
|
| 166 |
rag_path_ext = ".pdf"
|
|
|
|
| 169 |
rag_path_ext = ".csv"
|
| 170 |
|
| 171 |
|
| 172 |
+
|
| 173 |
+
uploaded_files = st.file_uploader(f"Upload files", type=rag_path_ext,accept_multiple_files=True,)
|
| 174 |
|
| 175 |
|
| 176 |
st.sidebar.header("Groq API Key")
|
config.ini
CHANGED
|
@@ -1,2 +1,13 @@
|
|
| 1 |
[Langlist]
|
| 2 |
-
lang_exts = {"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
[Langlist]
|
| 2 |
+
lang_exts = {"c": ".c", "cpp": ".cpp", "cobol": ".cbl", "elixir": ".ex",
|
| 3 |
+
"go": ".go", "haskell": ".hs", "html": ".html", "java": ".java",
|
| 4 |
+
"js": ".js", "kotlin": ".kt", "lua": ".lua", "latex": ".tex",
|
| 5 |
+
"markdown": ".md", "perl": ".pl", "php": ".php", "proto": ".proto",
|
| 6 |
+
"python": ".py", "rst": ".rst", "ruby": ".rb", "rust": ".rs",
|
| 7 |
+
"scala": ".scala", "sol": ".sol", "swift": ".swift", "csharp": ".cs"}
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
[Supported_Extensions]
|
| 11 |
+
extensions = [".cpp", ".go", ".java", ".kt", ".js", ".ts", ".php", ".proto", ".py", ".rst",
|
| 12 |
+
".rb", ".rs", ".scala", ".swift", ".md", ".tex", ".html", ".sol", ".cs",
|
| 13 |
+
".cob", ".c", ".lua", ".pl", ".hs", ".ex"]
|