allow for human readable console logs, but defaults to JSON on file logs.

src/auto_archiver/core/orchestrator.py

@@ -15,7 +15,7 @@ import traceback
 from copy import copy
 
 from rich_argparse import RichHelpFormatter
-from auto_archiver.utils.custom_logger import logger
+from auto_archiver.utils.custom_logger import format_for_human_readable_console, logger
 import requests
 
 from auto_archiver.utils.misc import random_str
@@ -348,7 +348,9 @@ Here's how that would look: \n\nsteps:\n  extractors:\n  - [your_extractor_name_
                 sys.stderr,
                 level=use_level,
                 catch=True,
-                format="<level>{level}</level>: <fg #64FFDA>{message}</fg #64FFDA> {extra[serialize_no_message]}",
+                format="<level>{extra[serialized]}</level>"
+                if logging_config.get("format", "").lower() == "json"
+                else format_for_human_readable_console(),
             )
 
             rotation = logging_config["rotation"]

src/auto_archiver/utils/custom_logger.py

@@ -2,13 +2,21 @@ from loguru import logger
 import json
 
 
+def extract_location(record, short=False):
+    """Extracts the file name, function name, and line number from the log record."""
+    if short:
+        return f"{record['file'].name}:{record['line']}"
+    return f"{record['file'].name}:{record['function']}:{record['line']}"
+
+
 def extract_log_data(record):
     subset = {
         "level": record["level"].name,
         "time": record["time"].isoformat(timespec="seconds"),
     }
-    subset["loc"] = f"{record['file'].name}:{record['function']}:{record['line']}"
+    subset["loc"] = extract_location(record)
 
+    # This is where logger.contextualize() parameters can be added to the output
     for extra_key in ["trace", "url", "worksheet", "row"]:
         if extra_val := record.get("extra", {}).get(extra_key):
             subset[extra_key] = extra_val
@@ -19,9 +27,14 @@ def extract_log_data(record):
     return subset
 
 
-def serialize_no_message(record):
+def serialize_for_console(record):
     subset = extract_log_data(record)
     subset.pop("message", None)
+    subset.pop("level", None)
+    subset.pop("loc", None)
+    subset.pop("time", None)
+    if not subset:
+        return ""
     return json.dumps(subset, ensure_ascii=False)
 
 
@@ -31,7 +44,16 @@ def serialize(record):
 
 def patching(record):
     record["extra"]["serialized"] = serialize(record)
-    record["extra"]["serialize_no_message"] = serialize_no_message(record)
+    record["extra"]["serialize_for_console"] = serialize_for_console(record)
+
+
+def format_for_human_readable_console():
+    return (
+        "<green>{time:YYYY-MM-DD HH:mm:ss.SSS}</green> | "
+        "<level>{level: <8}</level> | "
+        "<cyan>{file}</cyan>:<cyan>{function}</cyan>:<cyan>{line}</cyan> | "
+        "{extra[serialize_for_console]} <level>{message}</level>"
+    )
 
 
 logger = logger.patch(patching)