dblpsurvey

A quick & fast survey tool with a grep-friendly text file generated from dblp database

Usage

$ dblpsurvey [-k] [-d] [keyword...]

Options:

• -k: Remove DBLP keys from the output
• -d: Remove DOI URLs from the output
• keyword: Used as initial keywords when specified

When running dblpsurvey, you can select your favorite lines if you have installed incremental search tools such as peco. The results are pasted to the clipboard with pbcopy.

Prerequisites

• Basic commands: bash, curl, gzip, gunzip, realpath, perl, and make
• for the main scripts: ruby
  • (optional) for a fast XML processing: nokogiri
• for expanding XML entities: xmllint in libxml2
• for search: fzf, peco, or grep
• (optional) for pasting to the clipboard: pbcopy, xsel, or putclip

Installation

$ git clone https://github.com/sh5i/dblpsurvey.git
$ cd dblpsurvey
$ cp config.yaml.sample config.yaml
# (Edit config.yaml as you like)
$ make
$ sudo make install   # this just does: ln -s $(realpath ./dblpsurvey) /usr/local/bin/

The make first downloads the DBLP XML database file from https://dblp.org/ and generates a smaller XML based on the preference specified by config.yaml. Then, the extracted XML will be converted to a simple text, each line represents a DBLP entry (<article> or <inproceedings>). Such a text file is suitable for the grep-based search.

Example of config.yaml

journals:
  # Enumerate your favorite journals in the DBLP world.
  # Only the <article>s of ID "journals/(journal ID)/*" survive.
  - tse
  - tosem

conferences:
  # Enumerate your favorite conferences in the DBLP world.
  # Only the <inproceedings>s of ID "conf/(conference ID)/*" survive.
  - icse
  - sigsoft
  - kbse

year:
  # Only the entries whose publishing year is in [lower, upper] survive.
  lower: 2005
  upper: 2100