Log Analyzer

A tool that takes a server log file and produces a useful diagnostic report. Available as both a Python CLI and an interactive web dashboard.

Live web dashboard → https://kamer-heap.github.io/log-analyzer/web_ui.html

---

Two ways to use it

	CLI (analyze.py)	Web dashboard (web_ui.html)
Input	File path argument	Paste text or upload a file
Output	Terminal report	Interactive charts + filters
Requires	Python 3.8+	Any browser — no install
Best for	Scripts, CI, large files	Exploring, sharing results

---

CLI quickstart

1. Generate a test log file

python scripts/generate_logs.py --seed 42 --lines 500 --output sample_logs.log

On Windows, set the encoding first:

$env:PYTHONIOENCODING="utf-8"
python scripts/generate_logs.py --seed 42 --lines 500 --output sample_logs.log

2. Run the analyzer

python analyze.py sample_logs.log

With options:

python analyze.py sample_logs.log --top 10       # show top 10 slowest endpoints
python analyze.py sample_logs.log --json          # output as JSON
python analyze.py sample_logs.log --no-color      # plain text (for piping)

3. Run the test suite

python test_manual.py

All 35 tests should pass. Edge case log files used by the tests are in test_logs/.

---

What the output includes

• Request summary — total lines parsed, valid vs malformed count, error rate
• Status code distribution — breakdown of 2xx / 3xx / 4xx / 5xx
• Slowest endpoints — top N endpoints by average response time
• Traffic by hour — request volume across the day
• Anomaly flags — repeated 401s (brute-force), 5xx spikes, slow outliers
• Malformed line report — count of skipped lines, never silently dropped

---

Log formats supported

The analyzer handles a deliberately messy mix of real-world formats:

# Space-separated (assessment spec format)
2024-03-15T14:23:01Z 192.168.1.42 GET /api/users 200 142ms

# JSON-formatted lines (mixed in mid-file)
{"timestamp":"2024-03-15T14:23:05Z","method":"GET","path":"/health","status_code":200,"response_time":"142ms"}

Timestamp variants handled: ISO 8601, slash date (2024/03/15), human month (15-Mar-2024), Unix epoch Response time variants: 142ms, 0.142s, bare number 142 Status code variants: standard codes, - placeholder, missing field Bad lines: blank, truncated, stack traces, non-UTF-8, JSON fragments — all skipped with count

---

Web dashboard

Open web_ui.html in any browser — no server or install needed.

Features:

• Drop a .log file or paste lines directly
• Click status bar segments or hour bars to filter the request table
• Click any IP address to filter by that client
• Slowest endpoints ranked by average response time
• Anomaly detection: brute-force attempts, 5xx spikes, slow outliers

Or use the live hosted version: https://kamer-heap.github.io/log-analyzer/web_ui.html

---

Screenshots

CLI report

Web dashboard

Filtered view (click any status segment, hour bar, or IP to filter)

---

Project structure

log_analyzer/
├── analyze.py                    ← CLI analyzer (main tool)
├── web_ui.html                   ← standalone browser dashboard (no install)
├── test_manual.py                ← 35-test edge case verifier
├── README.md
├── ANSWERS.md                    ← assessment questions answered
├── scripts/
│   └── generate_logs.py          ← test log generator with --seed and --lines flags
├── screenshots/
│   ├── terminal_report.png       ← CLI report screenshot
│   ├── web_ui_dashboard.png      ← web dashboard screenshot
│   └── web_ui_filtered.png       ← filtered view screenshot
└── test_logs/
    ├── edge35_empty.log          ← empty file edge case
    ├── edge37_all_malformed.log  ← 100% malformed lines edge case
    └── edge38_no_final_newline.log ← no trailing newline edge case

sample_logs.log is not committed — generate it fresh with python scripts/generate_logs.py

---

Requirements

• Python 3.8 or later
• No external packages — standard library only (re, json, argparse, collections, datetime)
• Windows: set $env:PYTHONIOENCODING="utf-8" before running to enable Unicode bar characters in the terminal report

---

Running against your own log files

The tool accepts any file path — it does not assume a specific filename, line count, or set of values:

python analyze.py /path/to/your/server.log
python analyze.py /path/to/your/server.log --top 20 --json > report.json