Add JSON, syslog, and Apache parsers
Some checks failed
Some checks failed
This commit is contained in:
@@ -1,70 +1,190 @@
|
|||||||
import json
|
"""JSON log parser."""
|
||||||
import re
|
|
||||||
from datetime import datetime
|
from datetime import datetime
|
||||||
from typing import Any, Optional
|
from typing import Any, Optional
|
||||||
|
|
||||||
from loglens.parsers.base import BaseParser, LogFormat, ParsedEntry
|
import orjson
|
||||||
|
|
||||||
|
from loglens.parsers.base import LogParser, ParsedLogEntry
|
||||||
|
|
||||||
|
|
||||||
class JSONParser(BaseParser):
|
class JSONParser(LogParser):
|
||||||
"""Parser for JSON log formats."""
|
"""Parser for JSON-formatted logs."""
|
||||||
|
|
||||||
def get_format(self) -> LogFormat:
|
format_name = "json"
|
||||||
return LogFormat.JSON
|
|
||||||
|
|
||||||
def parse(self, line: str) -> Optional[ParsedEntry]:
|
def __init__(self):
|
||||||
"""Parse a JSON log line."""
|
self.timestamp_fields = [
|
||||||
|
"@timestamp",
|
||||||
|
"timestamp",
|
||||||
|
"time",
|
||||||
|
"date",
|
||||||
|
"datetime",
|
||||||
|
"created_at",
|
||||||
|
"updated_at",
|
||||||
|
"log_time",
|
||||||
|
"event_time",
|
||||||
|
]
|
||||||
|
self.level_fields = ["level", "severity", "log_level", "priority", "levelname"]
|
||||||
|
self.message_fields = ["message", "msg", "log", "text", "content"]
|
||||||
|
self.logger_fields = ["logger", "logger_name", "name", "source"]
|
||||||
|
|
||||||
|
def can_parse(self, line: str) -> bool:
|
||||||
|
"""Check if line is valid JSON."""
|
||||||
|
line = line.strip()
|
||||||
|
if not line:
|
||||||
|
return False
|
||||||
|
if line.startswith("[") or line.startswith("{"):
|
||||||
try:
|
try:
|
||||||
data = json.loads(line.strip())
|
orjson.loads(line)
|
||||||
except json.JSONDecodeError:
|
return True
|
||||||
|
except orjson.JSONDecodeError:
|
||||||
|
pass
|
||||||
|
return False
|
||||||
|
|
||||||
|
def parse(self, line: str, line_number: int = 0) -> Optional[ParsedLogEntry]:
|
||||||
|
"""Parse a JSON log line."""
|
||||||
|
line = line.strip()
|
||||||
|
if not line:
|
||||||
return None
|
return None
|
||||||
|
|
||||||
if isinstance(data, list):
|
try:
|
||||||
return None
|
data = orjson.loads(line)
|
||||||
|
except orjson.JSONDecodeError as e:
|
||||||
if not isinstance(data, dict):
|
return ParsedLogEntry(
|
||||||
return None
|
raw_line=line,
|
||||||
|
message=f"JSON parse error: {str(e)}",
|
||||||
timestamp = self._extract_timestamp(data)
|
line_number=line_number,
|
||||||
level = self._extract_level(data)
|
severity="error",
|
||||||
message = self._extract_message(data)
|
|
||||||
|
|
||||||
return ParsedEntry(
|
|
||||||
raw_line=line.strip(),
|
|
||||||
format=self.get_format(),
|
|
||||||
timestamp=timestamp,
|
|
||||||
level=level,
|
|
||||||
message=message,
|
|
||||||
metadata=data,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
def _extract_timestamp(self, data: dict[str, Any]) -> Optional[str]:
|
entry = ParsedLogEntry(raw_line=line, line_number=line_number)
|
||||||
timestamp_fields = ["timestamp", "time", "@timestamp", "date", "created_at"]
|
|
||||||
for field in timestamp_fields:
|
if isinstance(data, dict):
|
||||||
|
entry.timestamp = self._extract_timestamp(data)
|
||||||
|
entry.level = self._extract_field(data, self.level_fields)
|
||||||
|
entry.message = self._extract_field(data, self.message_fields)
|
||||||
|
entry.logger = self._extract_field(data, self.logger_fields)
|
||||||
|
entry.extra = {
|
||||||
|
k: v
|
||||||
|
for k, v in data.items()
|
||||||
|
if k not in self.timestamp_fields
|
||||||
|
and k not in self.level_fields
|
||||||
|
and k not in self.message_fields
|
||||||
|
and k not in self.logger_fields
|
||||||
|
and not k.startswith("_")
|
||||||
|
}
|
||||||
|
elif isinstance(data, list):
|
||||||
|
entry.message = str(data)
|
||||||
|
entry.extra = {"array_length": len(data)}
|
||||||
|
|
||||||
|
return entry
|
||||||
|
|
||||||
|
def _extract_timestamp(self, data: dict[str, Any]) -> Optional[datetime]:
|
||||||
|
"""Extract timestamp from data dict."""
|
||||||
|
for field in self.timestamp_fields:
|
||||||
if field in data:
|
if field in data:
|
||||||
|
value = data[field]
|
||||||
|
if isinstance(value, (int, float)):
|
||||||
|
return datetime.fromtimestamp(value)
|
||||||
|
elif isinstance(value, str):
|
||||||
|
try:
|
||||||
|
return datetime.fromisoformat(value.replace("Z", "+00:00"))
|
||||||
|
except ValueError:
|
||||||
|
pass
|
||||||
|
return None
|
||||||
|
|
||||||
|
def _extract_field(self, data: dict[str, Any], fields: list[str]) -> Optional[str]:
|
||||||
|
"""Extract first matching field from data."""
|
||||||
|
for field in fields:
|
||||||
|
if field in data and data[field] is not None:
|
||||||
value = data[field]
|
value = data[field]
|
||||||
if isinstance(value, str):
|
if isinstance(value, str):
|
||||||
return value
|
return value
|
||||||
elif isinstance(value, (int, float)):
|
return str(value)
|
||||||
return datetime.fromtimestamp(value).isoformat()
|
|
||||||
return None
|
return None
|
||||||
|
|
||||||
def _extract_level(self, data: dict[str, Any]) -> Optional[str]:
|
def parse_batch(self, lines: list[str]) -> list[ParsedLogEntry]:
|
||||||
level_fields = ["level", "severity", "log_level", "levelname", "status"]
|
"""Parse multiple lines, handling multi-line JSON."""
|
||||||
for field in level_fields:
|
results = []
|
||||||
if field in data:
|
buffer = ""
|
||||||
value = data[field]
|
line_number = 0
|
||||||
if isinstance(value, str):
|
|
||||||
return value.lower()
|
|
||||||
return None
|
|
||||||
|
|
||||||
def _extract_message(self, data: dict[str, Any]) -> str:
|
for line in lines:
|
||||||
message_fields = ["message", "msg", "text", "content", "error", "reason"]
|
line_number += 1
|
||||||
for field in message_fields:
|
line_stripped = line.strip()
|
||||||
if field in data:
|
|
||||||
value = data[field]
|
if not line_stripped:
|
||||||
if isinstance(value, str):
|
continue
|
||||||
return value
|
|
||||||
elif isinstance(value, dict):
|
if buffer:
|
||||||
return json.dumps(value)
|
buffer += line_stripped
|
||||||
return str(data)
|
else:
|
||||||
|
buffer = line_stripped
|
||||||
|
|
||||||
|
try:
|
||||||
|
data = orjson.loads(buffer)
|
||||||
|
entry = self._create_entry_from_data(data, line, line_number)
|
||||||
|
results.append(entry)
|
||||||
|
buffer = ""
|
||||||
|
except orjson.JSONDecodeError:
|
||||||
|
if line_stripped.startswith("{") or line_stripped.startswith("["):
|
||||||
|
if line_stripped.endswith("}") or line_stripped.endswith("]"):
|
||||||
|
results.append(
|
||||||
|
ParsedLogEntry(
|
||||||
|
raw_line=line,
|
||||||
|
message="Invalid JSON",
|
||||||
|
line_number=line_number,
|
||||||
|
severity="error",
|
||||||
|
)
|
||||||
|
)
|
||||||
|
buffer = ""
|
||||||
|
elif buffer.endswith("}") or buffer.endswith("]"):
|
||||||
|
try:
|
||||||
|
data = orjson.loads(buffer)
|
||||||
|
entry = self._create_entry_from_data(data, buffer, line_number)
|
||||||
|
results.append(entry)
|
||||||
|
except orjson.JSONDecodeError:
|
||||||
|
results.append(
|
||||||
|
ParsedLogEntry(
|
||||||
|
raw_line=buffer,
|
||||||
|
message="Invalid JSON",
|
||||||
|
line_number=line_number,
|
||||||
|
severity="error",
|
||||||
|
)
|
||||||
|
)
|
||||||
|
buffer = ""
|
||||||
|
elif len(buffer) > 10000:
|
||||||
|
results.append(
|
||||||
|
ParsedLogEntry(
|
||||||
|
raw_line=buffer[:100] + "...",
|
||||||
|
message="JSON too large to parse",
|
||||||
|
line_number=line_number,
|
||||||
|
severity="error",
|
||||||
|
)
|
||||||
|
)
|
||||||
|
buffer = ""
|
||||||
|
|
||||||
|
return results
|
||||||
|
|
||||||
|
def _create_entry_from_data(self, data: Any, raw_line: str, line_number: int) -> ParsedLogEntry:
|
||||||
|
"""Create ParsedLogEntry from parsed JSON data."""
|
||||||
|
entry = ParsedLogEntry(raw_line=raw_line, line_number=line_number)
|
||||||
|
|
||||||
|
if isinstance(data, dict):
|
||||||
|
entry.timestamp = self._extract_timestamp(data)
|
||||||
|
entry.level = self._extract_field(data, self.level_fields)
|
||||||
|
entry.message = self._extract_field(data, self.message_fields)
|
||||||
|
entry.logger = self._extract_field(data, self.logger_fields)
|
||||||
|
entry.extra = {
|
||||||
|
k: v
|
||||||
|
for k, v in data.items()
|
||||||
|
if k not in self.timestamp_fields
|
||||||
|
and k not in self.level_fields
|
||||||
|
and k not in self.message_fields
|
||||||
|
and k not in self.logger_fields
|
||||||
|
}
|
||||||
|
else:
|
||||||
|
entry.message = str(data)
|
||||||
|
|
||||||
|
return entry
|
||||||
|
|||||||
Reference in New Issue
Block a user