def create_log(base_dir, category, content):
row = {"time": time.time(), "category": category, "content": content}
with oscar.context(base_dir) as context:
with oscar.command(context, "load") as command:
command.add_argument("table", "Log")
command.add_argument("values", oscar.to_json([row]))
command.execute()
def set_one(base_dir, config_name, value):
with oscar.context(base_dir) as context:
with oscar.command(context, "load") as command:
command.add_argument("table", "Config")
row = {"_key":"config_name","value":"value"}
command.add_argument("values", oscar.to_json([row]))
command.execute()
def put_all(base_dir, configs):
with oscar.context(base_dir) as context:
with oscar.command(context, "load") as command:
command.add_argument("table", "Config")
rows = map(lambda (x,y):{"_key":x,"value":oscar.to_json(y)}, configs.items())
command.add_argument("values", oscar.to_json(rows))
command.execute()
def consume(base_dir,
limit=100,
concurrency=1,
id_prefix=None,
utf8_check=False):
with oscar.context(base_dir) as context:
with oscar.command(context, "select") as command:
command.add_argument("table", "FileQueue")
command.add_argument("output_columns", "_id,_key,name")
if id_prefix:
command.add_argument(
"filter",
"_key @^ \"%s\"" % oscar.escape_for_groonga(id_prefix))
command.add_argument("sortby", "size")
command.add_argument("limit", str(limit))
rows = json.loads(command.execute())[0][2:]
jobs = map(lambda x: (base_dir, x[0], x[2].encode("utf-8"), utf8_check),
rows)
if concurrency > 1:
pool = multiprocessing.Pool(
concurrency, lambda: signal.signal(signal.SIGINT, signal.SIG_IGN))
try:
rst = sum(pool.map(add_file, jobs))
pool.close()
except KeyboardInterrupt:
pool.terminate()
finally:
pool.join()
else:
rst = sum(map(add_file, jobs))
return rst
def show_one(base_dir, config_name):
with oscar.context(base_dir) as context:
with oscar.command(context, "select") as command:
command.add_argument("table", "Config")
command.add_argument("filter", "_key == \"%s\"" % config_name)
rows = json.loads(command.execute())[0][2:]
if len(rows) > 0:
print rows[0]
def add_file(args):
base_dir, _id, filename, utf8_check = args
with oscar.context(base_dir) as context:
add.add_file(context, base_dir, filename, utf8_check)
with oscar.command(context, "delete") as command:
command.add_argument("table", "FileQueue")
command.add_argument("id", str(_id))
command.execute()
return 1
def process_file_event(share, event_mask, event_pathname):
if (event_mask & pyinotify.IN_CLOSE_WRITE) or (
event_mask & pyinotify.IN_MOVED_TO): # @UndefinedVariable
oscar.log.debug(u"Adding %s to %s" %
(event_pathname.decode("utf-8"), share.name))
with oscar.context(share.path) as context:
walk.enqueue(context, share.path, event_pathname)
elif (event_mask & pyinotify.IN_DELETE) or (
event_mask & pyinotify.IN_MOVED_FROM): # @UndefinedVariable
file_id = oscar.sha1(event_pathname)
oscar.log.debug(u"Removing %s from %s(%s)" %
(event_pathname.decode("utf-8"), share.name, file_id))
with oscar.context(share.path) as context:
with oscar.command(context, "delete") as command:
command.add_argument("table", "FileQueue")
command.add_argument("key", file_id)
command.execute()
with oscar.command(context, "delete") as command:
command.add_argument("table", "Files")
command.add_argument("key", file_id)
command.execute()
def cleanup(context, base_dir):
offset = 0
total = 1
while offset < total:
with oscar.command(context, "select") as command:
command.add_argument("table", "Files")
command.add_argument("output_columns", "_id,path,name")
command.add_argument("offset", str(offset))
result = json.loads(command.execute())
total = result[0][0][0]
rows=result[0][2:]
#oscar.log.debug("total:%d offset:%d" % (total, offset))
for row in rows:
_id,path,name = row[0],row[1].encode("utf-8"),row[2].encode("utf-8")
exact_filename = os.path.join(base_dir,path if path != "/" else "",name)
if not os.path.isfile(exact_filename):
oscar.log.info("Missing file: %s. remove from database" % exact_filename)
with oscar.command(context, "delete") as command:
command.add_argument("table", "Files")
command.add_argument("id", str(_id))
command.execute()
offset += len(rows)
def get(base_dir, config_name = None):
with oscar.context(base_dir) as context:
with oscar.command(context, "select") as command:
command.add_argument("table", "Config")
if config_name: command.add_argument("filter", "_key == \"%s\"" % config_name)
rows = json.loads(command.execute())[0][2:]
if config_name:
return json.loads(rows[0][2]) if len(rows) > 0 else None
#else
result = {}
for row in rows:
result[row[1]] = json.loads(row[2])
return result
def share_info(share_name):
share = oscar.get_share(share_name)
if share == None: return "Share not found", 404
path = oscar.remove_preceding_slash(flask.request.args.get("path") or "")
if not os.path.isdir(share.real_path(path)):
return "Dir not found", 404
check_access_credential(share)
if path != "" and not path.endswith("/"): path = path + "/"
with oscar.context(share.real_path("/")) as context:
with oscar.command(context, "select") as command:
command.add_argument("table", "Files")
if path != "": command.add_argument("filter", "path @^ \"%s\"" % oscar.escape_for_groonga(path))
command.add_argument("limit", "0")
count = json.loads(command.execute())[0][0][0]
with oscar.command(context, "select") as command:
command.add_argument("table", "FileQueue")
command.add_argument("limit", "0")
queued = json.loads(command.execute())[0][0][0]
return flask.jsonify({"share_name":share_name,"count":count,"queued":queued,"eden":is_eden(flask.request)})
def search(context, path, keyword, offset=None, limit=None):
path = oscar.remove_preceding_slash(path)
if path != "" and not path.endswith("/"): path = path + "/"
with oscar.command(context, "select") as command:
command.add_argument("table", "Files")
command.add_argument("output_columns", "_key,path,name,mtime,size,fulltext.title,snippet_html(path),snippet_html(name),snippet_html(fulltext.content)")
command.add_argument("match_columns", "name*10||fulltext.title*10||fulltext.content*5||path_ft")
if path != "": command.add_argument("filter", "path @^ \"%s\"" % oscar.escape_for_groonga(path))
command.add_argument("query", keyword)
command.add_argument("sortby", "-_score")
command.add_argument("command_version", "2")
if offset: command.add_argument("offset", str(offset))
if limit: command.add_argument("limit", str(limit))
result = json.loads(command.execute())
return {
"count":result[0][0][0],
"rows":map(lambda row:{"key":row[0],"path":row[1],"name":row[2],"mtime":row[3],"size":row[4],"title":row[5],"snippets":{"path":row[6],"name":row[7],"content":row[8]}}, result[0][2:])
}
def get_log(base_dir, category=None, offset=None, limit=None):
with oscar.context(base_dir) as context:
with oscar.command(context, "select") as command:
command.add_argument("table", "Log")
command.add_argument("output_columns", "time,category,content")
if category:
command.add_argument(
"filter",
"category == \"%s\"" % oscar.escape_for_groonga(category))
if offset: command.add_argument("offset", str(offset))
if limit: command.add_argument("limit", str(limit))
command.add_argument("sortby", "-time")
result = json.loads(command.execute())
return {
"count":
result[0][0][0],
"rows":
map(
lambda row: {
"time": row[0],
"category": row[1],
"content": row[2]
}, result[0][2:])
}
def add_file(context, base_dir, filename, utf8_check=False):
filename = oscar.remove_preceding_slash(filename)
exact_filename = os.path.join(base_dir, filename)
if not os.path.isfile(exact_filename):
oscar.log.error("File %s does not exist" % exact_filename)
return False
stat = os.stat(exact_filename)
if stat.st_size < 50000000: # 50MB以上は大きすぎる
file_hash = calc_file_hash(exact_filename)
oscar.log.debug("File hash: %s" % file_hash)
with oscar.command(context, "select") as command:
command.add_argument("table", "Fulltext")
command.add_argument("output_columns", "_id")
command.add_argument("filter", "_key == '%s'" % file_hash)
num_hits = json.loads(command.execute())[0][0][0]
if num_hits == 0: # まだ登録されてない場合
extractor = extract.get_extractor(exact_filename)
if extractor:
try:
title, text = extractor(exact_filename)
except Exception as e:
oscar.log.exception("extractor")
log.create_log(
base_dir, "extract",
u"%s (%s): %s" % (filename.decode("utf-8"), file_hash,
e.message.decode("utf-8")))
else:
if utf8_check: utf8_check_by_iconv(text)
if len(
text
) > 3000000: # 3MB以上のテキストは切り捨てる(snippetつきで検索しようとしたときにgroongaが落ちるため)
text = text.decode("utf-8")[0:1000000].encode("utf-8")
row = {"_key": file_hash, "title": title, "content": text}
with oscar.command(context, "load") as command:
command.add_argument("table", "Fulltext")
command.add_argument("values", oscar.to_json([row]))
command.execute()
else:
oscar.log.debug("%s is too large (%d). the content is ignored" %
(filename, stat.st_size))
file_hash = ""
# select Files --filter 'name @^ \"walbrix\"'
path = os.path.dirname(filename)
if not path.endswith('/'): path += '/'
row = {
"_key": oscar.sha1(filename),
"path": path,
"path_ft": path,
"name": os.path.basename(filename),
"mtime": stat.st_mtime,
"size": stat.st_size,
"fulltext": file_hash
}
oscar.log.info("Adding: %s" % exact_filename)
with oscar.command(context, "load") as command:
command.add_argument("table", "Files")
command.add_argument("values", oscar.to_json([row]))
command.execute()
return True
