simon987/od-database
1
0
Fork 0
mirror of https://github.com/simon987/od-database.git synced 2025-04-19 18:36:44 +00:00
Code Issues Projects Releases Wiki Activity

Bug when directory is empty with new file upload

Browse Source
This commit is contained in:
Simon 2018-07-17 18:24:05 -04:00
parent 898ffcf410
commit 8ef1d36c9d
1 changed files with 15 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
crawl_server/task_manager.py
View File

@ -48,21 +48,22 @@ class TaskManager:
logger.info("Uploading file list in small chunks") logger.info("Uploading file list in small chunks")
filename = "./crawled/" + str(task_result.website_id) + ".json" filename = "./crawled/" + str(task_result.website_id) + ".json"
CHUNK_SIZE = 1000000 * 10 CHUNK_SIZE = 1000000 * 10
with open(filename) as f: if os.path.exists(filename):
chunk = f.read(CHUNK_SIZE) with open(filename) as f:
while chunk:
payload = {
"token": config.API_TOKEN,
"website_id": task_result.website_id
}
files = {
"file_list": chunk
}
r = requests.post(config.SERVER_URL + "/task/upload", data=payload, files=files)
logger.info("RESPONSE: " + r.text)
chunk = f.read(CHUNK_SIZE) chunk = f.read(CHUNK_SIZE)
while chunk:
payload = {
"token": config.API_TOKEN,
"website_id": task_result.website_id
}
files = {
"file_list": chunk
}
r = requests.post(config.SERVER_URL + "/task/upload", data=payload, files=files)
logger.info("RESPONSE: " + r.text)
chunk = f.read(CHUNK_SIZE)
payload = { payload = {
"token": config.API_TOKEN, "token": config.API_TOKEN,