A website to scrape? There's a simple way.

Shared:

• Decoupled architecture with Azure Queue Storage
• Executable as a CLI with a standalone binary
• Idenpotent operations that can be run in parallel
• Scraped content is stored in Azure Blob Storage

Scraper:

• Avoid re-scrape a page if it hasn't changed
• Block ads to lower network costs with The Block List Project
• Explore pages in depth by detecting links and de-duplicating them
• Extract markdown content from a page with html2text
• Load dynamic JavaScript content with Playwright
• Preserve anonymity with a random user agent, random viewport size and no client hints headers
• Show progress with a status command
• Track progress of total network usage
• Enhance anonymity with proxies
• Respect robots.txt

Indexer:

• AI Search index is created automatically
• Chunk markdown while keeping the content coherent
• Embed chunks with OpenAI embeddings
• Indexed content is semantically searchable with Azure AI Search

Download the latest release from the releases page. Binaries are available for Linux, macOS and Windows.

For configuring the CLI (including authentication to the backend services), use environment variables, a .env file or command line options.

Application must be run with Python 3.12 or later. If this version is not installed, an easy way to install it is pyenv.

# Download the source code
git clone https://github.com/clemlesne/scrape-it-now.git
# Move to the directory
cd scrape-it-now
# Run install scripts
make install dev
# Run the CLI
scrape-it-now --help

Basic usage:

export AZURE_STORAGE_CONNECTION_STRING=xxx
scrape-it-now scrape run https://nytimes.com

Example output:

❯ Starting scraping job 7yz91ma
Queued 71/71 links for referrer https://www.google.com/search (1)
3 workers started
Browser chromium launched
Processing new messages
...
Queued 15/28 links for referrer https://www.nytimes.com/2024/08/15/business/economy/kamala-harris-inflation-price-gouging.html (2)
Scraped https://www.nytimes.com/2024/08/15/business/economy/kamala-harris-inflation-price-gouging.html (2)

Most frequent options are:

For documentation on all available options, run:

scrape-it-now scrape run --help

Basic usage:

export AZURE_STORAGE_CONNECTION_STRING=xxx
scrape-it-now scrape status [job_name]

Example output:

❯ {"created_at":"2024-08-16T15:33:06.602922Z","last_updated":"2024-08-16T16:17:51.571136Z","network_used_mb":5.650620460510254,"processed":1263,"queued":3120}

Most frequent options are:

For documentation on all available options, run:

scrape-it-now scrape status --help

Basic usage:

export AZURE_OPENAI_API_KEY=xxx
export AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME=xxx
export AZURE_OPENAI_EMBEDDING_DIMENSIONS=xxx
export AZURE_OPENAI_EMBEDDING_MODEL_NAME=xxx
export AZURE_OPENAI_ENDPOINT=xxx
export AZURE_SEARCH_API_KEY=xxx
export AZURE_SEARCH_ENDPOINT=xxx
export AZURE_STORAGE_CONNECTION_STRING=xxx
scrape-it-now index run [job_name]

Example output:

❯ Starting indexing job 7yz91ma
5 workers started
Processing new messages
...
434b227 chunked into 6 parts
434b227 is indexed
f001b3e chunked into 86 parts
f001b3e is already indexed

Most frequent options are:

For documentation on all available options, run:

scrape-it-now index run --help

graph LR
  cli["CLI"]
  web["Website"]

  subgraph "Azure Queue Storage"
    to_chunk["To chunk"]
    to_scrape["To scrape"]
  end

  subgraph "Azure Blob Storage"
    subgraph "Container"
      job["job"]
      scraped["scraped"]
      state["state"]
    end
  end

  cli -- 1. Pull message --> to_scrape
  cli -- 2. Get cache --> scraped
  cli -- 3. Browse --> web
  cli -- 4. Update cache --> scraped
  cli -- 5. Push state --> state
  cli -- 6. Add message --> to_scrape
  cli -- 7. Add message --> to_chunk
  cli -- 8. Update state --> job

graph LR
  ai_search["Azure AI Search"]
  cli["CLI"]
  embeddings["Azure OpenAI Embeddings"]

  subgraph "Azure Queue Storage"
    to_chunk["To chunk"]
  end

  subgraph "Azure Blob Storage"
    subgraph "Container"
      scraped["scraped"]
    end
  end

  cli -- 1. Pull message --> to_chunk
  cli -- 2. Get cache --> scraped
  cli -- 3. Chunk --> cli
  cli -- 4. Embed --> embeddings
  cli -- 5. Push to search --> ai_search

To configure easily the CLI, source environment variables from a .env file. For example, for the --azure-storage-connection-string option:

AZURE_STORAGE_CONNECTION_STRING=xxx

For arguments that accept multiple values, use a space-separated list. For example, for the --whitelist option:

WHITELIST=learn\.microsoft\.com,^/(?!en-us).*,^/[^/]+/answers/,^/[^/]+/previous-versions/ go\.microsoft\.com,.*

Browser binaries are automatically downloaded or updated at each run. Browser is Chromium and it is not configurable (feel free to open an issue if you need another browser), it weights around 450MB.

The cache directoty depends on the operating system:

• ~/.config/scrape-it-now (Unix)
• ~/Library/Application Support/scrape-it-now (macOS)
• C:\Users\<user>\AppData\Roaming\scrape-it-now (Windows)