LIHKGr

The goal of LIHKGr is to scrape text data on the LIHKG, the Hong Kong version of Reddit, for analysis. LIHKG has gained popularity in 2016 and become a popular research data source during recent years. LIHKG is currently protected by Google's reCAPTCHA, this package currently builds on RSelenium and adopts a semi-manual approach to bypass it.

Installation

devtools::install_github("justinchuntingho/LIHKGr")

Instructions

lihkgr.R contains all the required functions. Please install the following packages: RSelenium, raster, magrittr rvest, and purrr. Follow the following workflow:

Step 1: Create a scraper

For RSelenium to work, you need to specify the browser. If you are using Chrome, you need to also specify the version. For example,create_lihkg(browser = "chrome", chromever = "83.0.4103.39"). If a version is not supplied, by default it will run the most recent version. To see Chrome version currently sourced run binman::list_versions("chromedriver").

## Creating a Firefox instance with a random port.

lihkg <- create_lihkg(browser = "firefox", port = sample(10000:60000, 1), verbose = FALSE)

Step 2: Scrape

# It can accept a single post id
lihkg$scrape(2091171)

# Or a vector
lihkg$scrape(1610753:1610755)

# Another way to do it
postids <- c(1610753, 2091171)
lihkg$scrape(postids)

Step 2.1: If any post id cannot be scraped, retry

lihkg$retry()

Step 3: Get / Save the data

To obtain the dataframe:

lihkg$bag

To save as .RDS:

lihkg$save("lihkg.RDS")

If you don't want to save the data as RDS, you can just save the bag as any format you like. It is just a regular data frame / tibble：

rio::export(lihkg$bag, "lihkg.xlsx")

Step 4: Destroy the scraper

lihkg$finalize()

Contributors

• Justin Chun-ting Ho
• Nick H. K. Or
• Chung-hong Chan
• Elgar Teo

Citation

Ho, J.C. & Or, N.H.K. (2020). LIHKGr. An application for scraping LIHKG. Source code and releases available at https://github.com/justinchuntingho/LIHKGr.