Example use of ldply for web scraping

ldply_example

# Using ldply returns a clean data frame, so avoids 
# the lapply + do.call(rbind, ...) idiom
# Also, the .progress argument is very useful

df <- ldply(c(2002:2014, 52, 26, 13, 4, 1), function(x) {
        sprintf("http://distrowatch.com/index.php?dataspan=%d", x) %>%
        html() %>% 
        html_nodes(xpath = 
        "//table[@class = 'News' and @style = 'direction: ltr'][2]") %>%
        .[[1]] %>%
        html_table() %>% 
        .[-(1:2), -1] %>% 
        setNames(c("distro", "npd")) %>% 
        mutate(npd = npd %>% str_replace(",", "") %>% as.numeric(),
               time = x) %>% 
        tbl_df()
}, .progress = "text")

head(df)
#     distro npd time
# 1 Mandrake 473 2002
# 2  Red Hat 453 2002
# 3   Gentoo 326 2002
# 4   Debian 311 2002
# 5 Sorcerer 253 2002
# 6     SuSE 250 2002

Gah. That's freaking useful and elegant. Thanks for posting that.