# Method 1

baseurl <- 'http://www.stat.cmu.edu/~cshalizi/statcomp/hw/09/'
urls <- paste(baseurl, 'page', 1:4, '.html', sep = '')
page <- lapply(urls, readLines, warn = FALSE)
page <- lapply(page, paste, collapse = '\n')
html <- do.call(paste, c(page, list(sep = '\n')))

pat <- paste(
    '<tr style="cursor: pointer;">\\s*',
    '<td class="rank">(\\d+)</td>\\s*',
    '<td class="company">\\s*',
    '<img src="[^"]*" alt="[^"]*">\\s*',
    '<h3><a rel="[^"]*"></a>([^<]+)</h3>\\s*',
    '</td>\\s*',
    '<td class="worth">\\$(\\d+\\.?\\d*)\\s*B</td>\\s*',
    '<td>([^<]+)</td>\\s*',
    '<td>([^<]+)</td>\\s*',
    '<td>([^<]+)</td>\\s*',
    '</tr>',
sep = '')

# First pass: extract disjoint cases 
m <- gregexpr(pat, html, ignore.case = TRUE)
x <- regmatches(html, m)
x <- do.call(c, x)

# Second pass: extract capture groups
m <- regexec(pat, x, ignore.case = TRUE)
wealthy <- regmatches(x, m)

# Put it all together
library(plyr)
df <- ldply(wealthy, function(x) data.frame(
        rank = as.numeric(x[2]), 
        name = x[3],
        net.worth = as.numeric(x[4]),
        age = as.numeric(x[5]),
        residence = x[6],
        source = x[7],
stringsAsFactors = FALSE))



# Method 2
raw400 <- readLines('http://www.forbes.com/forbes-400/ajax/loadList')

parseone <- function(html) {
    pat <- paste(
        '"(\\w+)"\\s*:\\s*([^,]*)'
    )

    # First pass: extract disjoint cases 
    m <- gregexpr(pat, html, ignore.case = TRUE)
    x <- regmatches(html, m)
    x <- do.call(c, x)

    # Second pass: extract capture groups
    m <- regexec(pat, x, ignore.case = TRUE)
    x <- regmatches(x, m)

    return(x)
}

pat <- paste(
    '"rank":(\\d+),',
    '"firstName":"([^"]*)",',
    '"lastName":"([^"]*)",',
    '"finalWorth":(\\d+\\.?\\d*),',
    '"age":(\\d+|\\-),',
    '"city":"([^"]*)",',
    '"state":"([^"]*)",',
    '"source":"([^"]*)",',
    '"industries":\\[([^\\]+)\\],',
sep = '.*')

# First pass: Extract disjoint cases --- Billionaires are separated by },{
split400 <- strsplit(raw400, '\\},\\{')[[1]]

# Second pass: Extract capture groups
m <- regexec(pat, split400, ignore.case = TRUE)
wealthy <- regmatches(split400, m)

# Put it all together
library(plyr)
df <- ldply(wealthy, function(x) data.frame(
        rank = as.numeric(x[2]), 
        name = paste(x[3], x[4]),
        net.worth = as.numeric(x[5]),
        age = as.numeric(x[6]),
        residence = paste(x[7], ', ', x[8], sep = ''),
        source = x[9],
        industries = gsub('"', '', x[10]), # Extra
stringsAsFactors = FALSE))