LexLogger API
LexLogger uses a builder pattern — chain methods together and call .log() at the end. Every method returns the logger instance, so you can chain as many as you need.
Core Methods
add_text(text: str)
Adds a plain text paragraph.
LexLogger().add_text("Processing complete.").log()add_heading(text: str, level: int = 1)
Adds a Markdown heading (levels 1–6).
LexLogger().add_heading("Summary", level=2) \
.add_text("All items processed.") \
.log()add_table(headers: list, rows: list)
Adds a Markdown table.
LexLogger().add_table(
["Name", "Amount"],
[["Alice", "500"], ["Bob", "750"]]
).log()add_dataframe(df: pd.DataFrame)
Renders a Pandas DataFrame as a Markdown table.
import pandas as pd
df = pd.DataFrame({"Q": ["Q1", "Q2"], "Revenue": [100000, 120000]})
LexLogger().add_dataframe(df).log()add_code(code: str, language: str = "")
Adds a fenced code block.
import json
config = {"rate": 0.19}
LexLogger().add_code(json.dumps(config, indent=2), language="json").log()log()
Writes the accumulated content to the database. Always call this last.
If you forget to call
.log(), nothing is written. This is the most common mistake.
Context-Aware Logging
LexLogger automatically resolves:
Context | How |
|---|---|
Calculation ID | Links to the current |
Model Instance | Identifies which model is executing |
Parent/Child | Links nested calculations hierarchically |
You never need to pass context manually.
Nested Calculations
When a parent calculation triggers a child, wrap the child execution in model_logging_context to preserve the log hierarchy:
from lex.audit_logging.utils.ModelContext import model_logging_context
class ParentCalculation(CalculationModel):
def calculate(self):
LexLogger().add_text("Starting parent").log()
child = CalculateNAV.objects.filter(quarter=self.quarter).first()
with model_logging_context(child):
child.is_calculated = "IN_PROGRESS"
child.save()
LexLogger().add_text("Child finished.").log()See Logging for more examples and usage patterns.
Was this helpful?
Your feedback shapes what we write next.
