Merge pull request #1 from Marclass/rag-arbitrary-files

Allow any file to be used for RAG.
This commit is contained in:
Marclass 2024-01-19 10:44:53 -07:00 committed by GitHub
commit e2edbedede
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 27 additions and 43 deletions

View file

@ -144,37 +144,21 @@ def store_doc(
# "https://www.gutenberg.org/files/1727/1727-h/1727-h.htm" # "https://www.gutenberg.org/files/1727/1727-h/1727-h.htm"
print(file.content_type) print(file.content_type)
if file.content_type not in [
"application/pdf", text_xml=["xml"]
"text/plain",
"text/csv",
"text/xml",
"text/x-python",
"text/css",
"application/vnd.openxmlformats-officedocument.wordprocessingml.document",
"application/octet-stream",
"application/x-javascript",
]:
raise HTTPException(
status_code=status.HTTP_400_BAD_REQUEST,
detail=ERROR_MESSAGES.FILE_NOT_SUPPORTED,
)
text_xml=["text/xml"]
octet_markdown=["md"] octet_markdown=["md"]
octet_plain=[ known_source_ext=[
"go", "py", "java", "sh", "bat", "ps1", "cmd", "js", "go", "py", "java", "sh", "bat", "ps1", "cmd", "js",
"css", "cpp", "hpp","h", "c", "cs", "sql", "log", "ini", "css", "cpp", "hpp","h", "c", "cs", "sql", "log", "ini",
"pl" "pm", "r", "dart", "dockerfile", "env", "php", "hs", "pl" "pm", "r", "dart", "dockerfile", "env", "php", "hs",
"hsc", "lua", "nginxconf", "conf", "m", "mm", "plsql", "perl", "hsc", "lua", "nginxconf", "conf", "m", "mm", "plsql", "perl",
"rb", "rs", "db2", "scala", "bash", "swift", "vue", "svelte" "rb", "rs", "db2", "scala", "bash", "swift", "vue", "svelte"
] ]
docx_type="application/vnd.openxmlformats-officedocument.wordprocessingml.document"
known_doc_ext=["doc","docx"]
file_ext=file.filename.split(".")[-1].lower() file_ext=file.filename.split(".")[-1].lower()
if file.content_type == "application/octet-stream" and file_ext not in (octet_markdown + octet_plain): known_type=True
raise HTTPException(
status_code=status.HTTP_400_BAD_REQUEST,
detail=ERROR_MESSAGES.FILE_NOT_SUPPORTED,
)
try: try:
filename = file.filename filename = file.filename
file_path = f"{UPLOAD_DIR}/{filename}" file_path = f"{UPLOAD_DIR}/{filename}"
@ -188,27 +172,22 @@ def store_doc(
collection_name = calculate_sha256(f)[:63] collection_name = calculate_sha256(f)[:63]
f.close() f.close()
if file.content_type == "application/pdf": if file_ext=="pdf":
loader = PyPDFLoader(file_path) loader = PyPDFLoader(file_path)
elif ( elif (file.content_type ==docx_type or file_ext in known_doc_ext):
file.content_type
== "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
):
loader = Docx2txtLoader(file_path) loader = Docx2txtLoader(file_path)
elif file_ext=="csv":
elif file.content_type == "text/csv":
loader = CSVLoader(file_path) loader = CSVLoader(file_path)
elif file.content_type in text_xml: elif file_ext in text_xml:
loader=UnstructuredXMLLoader(file_path) loader=UnstructuredXMLLoader(file_path)
elif file.content_type == "text/plain" or file.content_type.find("text/")>=0: elif file_ext in known_source_ext or file.content_type.find("text/")>=0:
loader = TextLoader(file_path) loader = TextLoader(file_path)
elif file.content_type == "application/octet-stream": elif file_ext in octet_markdown:
if file_ext in octet_markdown: loader = UnstructuredMarkdownLoader(file_path)
loader = UnstructuredMarkdownLoader(file_path) else:
if file_ext in octet_plain:
loader = TextLoader(file_path)
elif file.content_type == "application/x-javascript":
loader = TextLoader(file_path) loader = TextLoader(file_path)
known_type=False
data = loader.load() data = loader.load()
result = store_data_in_vector_db(data, collection_name) result = store_data_in_vector_db(data, collection_name)
@ -218,6 +197,7 @@ def store_doc(
"status": True, "status": True,
"collection_name": collection_name, "collection_name": collection_name,
"filename": filename, "filename": filename,
"known_type":known_type,
} }
else: else:
raise HTTPException( raise HTTPException(

View file

@ -173,7 +173,8 @@
) { ) {
uploadDoc(file); uploadDoc(file);
} else { } else {
toast.error(`Unsupported File Type '${file['type']}'.`); toast.error(`Unknown File Type '${file['type']}', but accepting and treating as plain text`);
uploadDoc(file);
} }
} else { } else {
toast.error(`File not found.`); toast.error(`File not found.`);
@ -308,8 +309,9 @@
uploadDoc(file); uploadDoc(file);
filesInputElement.value = ''; filesInputElement.value = '';
} else { } else {
toast.error(`Unsupported File Type '${file['type']}'.`); toast.error(`Unknown File Type '${file['type']}', but accepting and treating as plain text`);
inputFiles = null; uploadDoc(file);
filesInputElement.value = '';
} }
} else { } else {
toast.error(`File not found.`); toast.error(`File not found.`);

View file

@ -73,7 +73,8 @@
) { ) {
uploadDoc(file); uploadDoc(file);
} else { } else {
toast.error(`Unsupported File Type '${file['type']}'.`); toast.error(`Unknown File Type '${file['type']}', but accepting and treating as plain text`);
uploadDoc(file);
} }
} else { } else {
toast.error(`File not found.`); toast.error(`File not found.`);
@ -153,7 +154,8 @@
) { ) {
uploadDoc(file); uploadDoc(file);
} else { } else {
toast.error(`Unsupported File Type '${file['type']}'.`); toast.error(`Unknown File Type '${file['type']}', but accepting and treating as plain text`);
uploadDoc(file);
} }
inputFiles = null; inputFiles = null;