mirror of
https://github.com/arc53/DocsGPT.git
synced 2025-11-29 16:43:16 +00:00
(feat:attachment) upload single file
This commit is contained in:
@@ -313,18 +313,18 @@ def sync_worker(self, frequency):
|
||||
for key in ["total_sync_count", "sync_success", "sync_failure"]
|
||||
}
|
||||
|
||||
def attachment_worker(self, directory, saved_files, user):
|
||||
def attachment_worker(self, directory, file_info, user):
|
||||
"""
|
||||
Process and store attachments without vectorization.
|
||||
Process and store a single attachment without vectorization.
|
||||
|
||||
Args:
|
||||
self: Reference to the instance of the task.
|
||||
directory (str): Base directory for storing files.
|
||||
saved_files (list): List of dictionaries with folder and filename info.
|
||||
file_info (dict): Dictionary with folder and filename info.
|
||||
user (str): User identifier.
|
||||
|
||||
Returns:
|
||||
dict: Information about processed attachments.
|
||||
dict: Information about processed attachment.
|
||||
"""
|
||||
import datetime
|
||||
import os
|
||||
@@ -334,71 +334,63 @@ def attachment_worker(self, directory, saved_files, user):
|
||||
db = mongo["docsgpt"]
|
||||
attachments_collection = db["attachments"]
|
||||
|
||||
file_entries = []
|
||||
total_tokens = 0
|
||||
total_files = len(saved_files)
|
||||
job_name = file_info["folder"]
|
||||
logging.info(f"Processing attachment: {job_name}", extra={"user": user, "job": job_name})
|
||||
|
||||
job_name = saved_files[0]["folder"] if saved_files else "attachment_job"
|
||||
logging.info(f"Processing attachments: {job_name}", extra={"user": user, "job": job_name})
|
||||
self.update_state(state="PROGRESS", meta={"current": 10})
|
||||
|
||||
for idx, file_info in enumerate(saved_files):
|
||||
|
||||
progress = int(((idx + 1) / total_files) * 100)
|
||||
self.update_state(state="PROGRESS", meta={"current": progress})
|
||||
folder_name = file_info["folder"]
|
||||
filename = file_info["filename"]
|
||||
|
||||
base_dir = os.path.join(directory, user, "attachments", folder_name)
|
||||
file_path = os.path.join(base_dir, filename)
|
||||
|
||||
logging.info(f"Processing file: {file_path}", extra={"user": user, "job": job_name})
|
||||
|
||||
if not os.path.exists(file_path):
|
||||
logging.warning(f"File not found: {file_path}", extra={"user": user, "job": job_name})
|
||||
return {"error": "File not found"}
|
||||
|
||||
try:
|
||||
reader = SimpleDirectoryReader(
|
||||
input_files=[file_path]
|
||||
)
|
||||
|
||||
folder_name = file_info["folder"]
|
||||
filename = file_info["filename"]
|
||||
documents = reader.load_data()
|
||||
|
||||
base_dir = os.path.join(directory, user, "attachments", folder_name)
|
||||
file_path = os.path.join(base_dir, filename)
|
||||
self.update_state(state="PROGRESS", meta={"current": 50})
|
||||
|
||||
logging.info(f"Processing file: {file_path}", extra={"user": user, "job": job_name})
|
||||
|
||||
if not os.path.exists(file_path):
|
||||
logging.warning(f"File not found: {file_path}", extra={"user": user, "job": job_name})
|
||||
continue
|
||||
|
||||
try:
|
||||
reader = SimpleDirectoryReader(
|
||||
input_files=[file_path]
|
||||
)
|
||||
if documents:
|
||||
content = documents[0].text
|
||||
token_count = num_tokens_from_string(content)
|
||||
|
||||
documents = reader.load_data()
|
||||
file_path_relative = f"{user}/attachments/{folder_name}/{filename}"
|
||||
|
||||
if documents:
|
||||
content = documents[0].text
|
||||
token_count = num_tokens_from_string(content)
|
||||
total_tokens += token_count
|
||||
|
||||
file_entries.append({
|
||||
"path": f"{user}/attachments/{folder_name}/{filename}",
|
||||
"content": content,
|
||||
"token_count": token_count
|
||||
})
|
||||
logging.info(f"Successfully processed {filename} with {token_count} tokens",
|
||||
extra={"user": user, "job": job_name})
|
||||
except Exception as e:
|
||||
logging.error(f"Error processing file {filename}: {e}",
|
||||
extra={"user": user, "job": job_name}, exc_info=True)
|
||||
|
||||
if file_entries:
|
||||
attachment_id = attachments_collection.insert_one({
|
||||
"user": user,
|
||||
"files": file_entries,
|
||||
"total_tokens": total_tokens,
|
||||
"date": datetime.datetime.now(),
|
||||
}).inserted_id
|
||||
|
||||
logging.info(f"Stored attachment with ID: {attachment_id}",
|
||||
extra={"user": user, "job": job_name})
|
||||
|
||||
return {
|
||||
"attachment_id": str(attachment_id),
|
||||
"files": [{"filename": fe["filename"], "folder": fe["folder"], "path": fe["path"]} for fe in file_entries],
|
||||
"total_tokens": total_tokens,
|
||||
"file_contents": [{"filename": fe["filename"], "token_count": fe["token_count"]} for fe in file_entries]
|
||||
}
|
||||
else:
|
||||
logging.warning("No files were successfully processed",
|
||||
extra={"user": user, "job": job_name})
|
||||
return {"error": "No files were successfully processed"}
|
||||
attachment_id = attachments_collection.insert_one({
|
||||
"user": user,
|
||||
"path": file_path_relative,
|
||||
"content": content,
|
||||
"token_count": token_count,
|
||||
"date": datetime.datetime.now(),
|
||||
}).inserted_id
|
||||
|
||||
logging.info(f"Stored attachment with ID: {attachment_id}",
|
||||
extra={"user": user, "job": job_name})
|
||||
|
||||
self.update_state(state="PROGRESS", meta={"current": 100})
|
||||
|
||||
return {
|
||||
"attachment_id": str(attachment_id),
|
||||
"filename": filename,
|
||||
"folder": folder_name,
|
||||
"path": file_path_relative,
|
||||
"token_count": token_count
|
||||
}
|
||||
else:
|
||||
logging.warning("No content was extracted from the file",
|
||||
extra={"user": user, "job": job_name})
|
||||
return {"error": "No content was extracted from the file"}
|
||||
except Exception as e:
|
||||
logging.error(f"Error processing file {filename}: {e}",
|
||||
extra={"user": user, "job": job_name}, exc_info=True)
|
||||
return {"error": f"Error processing file: {str(e)}"}
|
||||
|
||||
Reference in New Issue
Block a user