Fix CSV parsing to use proper csv.reader() instead of naive string splitting

[Copilot]

The parse_csv_raw() function was using naive string splitting to parse CSV headers, which doesn't handle CSV edge cases properly. This caused incorrect parsing when CSV files contained quoted fields with commas, escaped quotes, or other standard CSV formatting.

Problem

The current implementation used:

column_names = header.strip().split(col_delimiter)

This fails for CSV files like:

f1,"field with, comma",f3,"field with ""quotes"""

The naive splitting would incorrectly parse this as 6 columns instead of 4:

['f1', '"field with', ' comma"', 'f3', '"field with ""quotes"""']

Solution

Updated the function to use Python's standard csv module:

column_names = list(csv.reader([header.strip()]))[0]

This properly handles all CSV edge cases and parses the same header correctly as:

['f1', 'field with, comma', 'f3', 'field with "quotes"']

Changes

• Fixed parse_csv_raw() function in outrank/core_utils.py (line 396)
• Added comprehensive test case test_parse_csv_with_quoted_fields() in tests/data_io_test.py

The fix is surgical and minimal - only one line changed. The csv module was already imported and used elsewhere in the codebase (parse_ob_csv_line() function), so this change brings consistency across all CSV parsing methods.

All existing tests pass and the new test case verifies that complex CSV files with quoted fields are now parsed correctly.

Fixes #103.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.