6. Disclosure Text — Reading Narrative Disclosures with sections

Read the text hidden behind the numbers. DartLab horizontalizes all disclosure text into sections. Every section of an annual report is organized into a topic × period matrix, so you can compare any subject side by side across periods.

Setup

import dartlab
import polars as pl

c = dartlab.Company("005930")

Viewing the Full Structure with sections

c.sections

sections is the complete disclosure map of a company. All sections are horizontalized into a topic × period matrix. The blockType column distinguishes text from tables, and blockOrder preserves the original order.

│ chapter │ topic            │ blockType │ blockOrder │ 2024   │ 2023   │ ...
│ I       │ companyOverview  │ text      │ 0          │ 서술문 │ 서술문 │
│ I       │ companyOverview  │ table     │ 1          │ 테이블 │ 테이블 │
│ II      │ businessOverview │ text      │ 0          │ 서술문 │ 서술문 │

Checking the Topic List

c.topics

Key topic examples:

Available topics vary by company. Use c.topics to check the full list for a given stock.

Opening a Topic with show()

show() extracts the relevant topic from sections and presents it block by block.

# Block index — see what blocks are available
c.show("businessOverview")

After checking the block numbers in the index, retrieve the desired block.

# Block 0: Text DataFrame (original text by period)
c.show("businessOverview", 0)

# Block 1: Table DataFrame (items × periods matrix)
c.show("businessOverview", 1)

Text blocks display original text columns side by side for each period, and table blocks return a horizontalized DataFrame in an items × periods format.

Filtering sections with Polars

Since sections is a Polars DataFrame, you can filter it freely.

# Text blocks only for a specific topic
text_rows = c.sections.filter(
    (pl.col("topic") == "businessOverview") & (pl.col("blockType") == "text")
)

# Table blocks only for a specific topic
table_rows = c.sections.filter(
    (pl.col("topic") == "businessOverview") & (pl.col("blockType") == "table")
)

# Reading original text for a specific period
row = c.sections.filter(
    (pl.col("topic") == "companyOverview") & (pl.col("blockType") == "text")
)
print(row["2024"][0])  # 2024 text

Exploring by Period

# List of available periods
c.sections.periods()

# Sorted newest first
c.sections.ordered()

periods() returns the full list of periods available for the company. ordered() reorders the columns so the most recent period comes first.

Detecting Changes with diff()

Track text changes between two periods. Quickly identify which years saw significant changes in disclosure text.

# Change rate summary by topic
c.diff()

# Period-by-period change history for a specific topic
c.diff("businessOverview")

# Detailed line-by-line comparison between two periods
c.diff("businessOverview", "2023", "2024")

Years with high change rates may indicate business restructuring, new business ventures, or added risk factors. An efficient analysis workflow is to first identify change points with diff(), then read the original text for that period with show().

Going Deeper with the source Namespace

c.sections is a merged view. To access the raw data at a deeper level, use the c.docs namespace.

# Pure docs source view (before finance/report merge)
c.docs.sections

# Original block retrieval (for search and AI)
c.docs.retrievalBlocks

# LLM context slice
c.docs.contextSlices

# K-IFRS notes
c.docs.notes

Practical Example: Tracking Risk Factor Changes

Risk factors are updated every year, so tracking changes helps you quickly spot new risks.

# Risk factors block index
c.show("riskFactors")

# Open text block
c.show("riskFactors", 0)

# Check change rates by year
c.diff("riskFactors")

# Detailed comparison 2023→2024
c.diff("riskFactors", "2023", "2024")

Source Tracking

Check where the data for a topic originated.

c.trace("businessOverview")

trace() returns metadata such as the source (docs/finance/report), number of periods, and whether text or tables are present for that topic.

Next Steps

• Sections Guide — Detailed sections structure
• 7. Advanced Analysis — K-IFRS notes, tangible assets, affiliates
• 8. Cross-Company Comparison — Sector, insight, ranking