stdf-reader（标准文件阅读器）

STDF Reader Skill

This skill provides access to the STDF Reader CLI tool for parsing and analyzing semiconductor STDF test data files.

Prerequisites

Install via pip (python version 3.9+):

pip install stdf-reader

Use the virtual environment for running any commands:

source .venv/bin/activate

Available Commands

1. Convert STDF to CSV (most common)

Parse one or more STDF files into a CSV log for analysis:

stdf-reader convert-csv <stdf_file(s)> [-o output_name]

Options:

• --ignore-tnum — Ignore test number (match by name only)
• --ignore-chnum — Ignore channel number in test name
• --no-merge — Output separate CSV per file (default: merge into one)

Example:

stdf-reader convert-csv input.stdf
# Output: input.stdf_csv_log.csv

stdf-reader convert-csv lot1.stdf lot2.stdf -o merged_output
# Output: merged_output_csv_log.csv

2. Convert STDF to XLSX

Parse STDF file into a formatted Excel table (shows all record types):

stdf-reader convert-xlsx <stdf_file>

3. Convert Diagnosis STDF to ASCII

Convert STDF V4-2007.1 files with STR/PSR diagnosis records to ASCII CSV:

stdf-reader convert-diag <stdf_file>

4. Extract Single Record Type

Extract a specific STDF record type (DTR, GDR, TSR) to CSV:

stdf-reader extract-record <stdf_file> --type DTR

5. Generate Analysis Report

Generate comprehensive XLSX report from CSV (includes data statistics, bin summary, wafer map):

stdf-reader report <csv_file(s)> [-o output.xlsx]

6. Generate Correlation Report

Compare means across 2+ STDF files merged in a single CSV:

stdf-reader correlation <csv_file(s)> [-o output.xlsx]

7. Generate Site-to-Site Correlation

Compare test data across different test sites:

stdf-reader s2s <csv_file(s)> [-o output.xlsx] [--cherry-pick "1,3,5"]

Options:

• --cherry-pick "1,3,5" — Site numbers to pick, comma-separated

8. Generate PDF Charts

Generate PDF with trendline + histogram charts for selected tests:

# By regex pattern
stdf-reader pdf <csv_file> --regex "IDD.*"

# All tests
stdf-reader pdf <csv_file> --all

# Specific tests
stdf-reader pdf <csv_file> --tests "210 - IDD_Static <> curr"

Options:

• --no-limits — Don't plot failure limit lines
• --group-by-file — Group trends by file instead of by site

9. Extract Sub-CSV

Extract only specific tests from a CSV file:

stdf-reader extract-tests <csv_file> --regex "VDD.*"

10. List Tests

List all test instances in a CSV file:

stdf-reader list-tests <csv_file>
stdf-reader list-tests <csv_file> --filter "IDD.*"

11. Transpose CSV

Swap rows and columns in a CSV file:

stdf-reader transpose <csv_file>

Typical Workflow

1. Convert STDF to CSV: convert-csv input.stdf
2. List available tests: list-tests input.stdf_csv_log.csv
3. Report: report input.stdf_csv_log.csv
4. PDF charts: pdf input.stdf_csv_log.csv --regex "pattern"

Output File Naming

• convert-csv → {input}_csv_log.csv
• convert-xlsx → {input}_excel.xlsx
• convert-diag → {input}_diag_log.csv
• extract-record → {input}_{TYPE}_Rec.csv
• report → {base}_analysis_report_{timestamp}.xlsx
• correlation → {base}_correlation_report_{timestamp}.xlsx
• s2s → {base}_s2s_correlation_table{timestamp}.xlsx
• pdf → {input}_results.pdf

All outputs default to the same directory as the input unless -o is specified.