Skip to content

Latest commit

 

History

History
61 lines (39 loc) · 1.43 KB

README.md

File metadata and controls

61 lines (39 loc) · 1.43 KB
Raw

WC parser

This repo contains my solution to building a concurrent scraper which generates a top-k list of word frequencies.

Run the CLI

Execute the following from the root directory of the repo.\

go run cmd/scraper/main.go

You can also view help on commnands and flags (currently only the root command exists)

go run cmd/scraper/main.go --help

Note: when running in a terminal, the above will print stderr to terminal as well. To avoid seeing stderr at all, run

 go run cmd/scraper/main.go 2>/dev/null

Specify configuration

To specify configuration file, overriding the default, run

 go run cmd/scraper/main.go --config ./cmd/scraper/scrapeconfig.yaml

See the provided cmd/scraper/config.yaml for example Note: the .job_loader.page_cutoff parameter is set to 3, to avoid getting accidentally blocked during development.
For a real-world use case, either omit the attribute or set it to 0.

Specify log level

To set the log level, add a GO_LOG=<level> envvar when executing, e.g.

GO_LOG=debug go run cmd/scraper/main.go

When log level is unspecified, the default is "info".

Run Tests

Execute the following from the root directory of the repo.

go test ./...

Future improvements

  • Supprort multi-node architecture for large-scale scraping
  • Bloom filter to support larger dictionaries
  • Benchmarks
  • Test core scraper flow control