week9_assignment

library(XML2R)

## Warning: package 'XML2R' was built under R version 3.1.3

## Loading required package: XML

## Warning: package 'XML' was built under R version 3.1.3

library(htmltools)

## Warning: package 'htmltools' was built under R version 3.1.3

library(htmlTable)

## Warning: package 'htmlTable' was built under R version 3.1.3

library(RJSONIO) 

## Warning: package 'RJSONIO' was built under R version 3.1.3

library(jsonlite)

## Warning: package 'jsonlite' was built under R version 3.1.3

## 
## Attaching package: 'jsonlite'
## 
## The following objects are masked from 'package:RJSONIO':
## 
##     fromJSON, toJSON
## 
## The following object is masked from 'package:utils':
## 
##     View

html <- readLines(“http://www.gutenberg.org/files/17185/17185-h/17185-h.htm”)

head(html, n=200)

# html file

url1 <- "https://github.com/jameyletherton/week9_assignment/blob/f12a908cd9ca8abd7dda725192c776f46f4d4c5c/book.html"

tables1 <- readHTMLTable(url1,header = T)

head(tables1)

## named list()

str(tables1)

##  Named list()

url <- "file:///C:/Users/jamey/Documents/IS%20607/book.html"
tables <- readHTMLTable(url,header = T)

head(tables)

## $`NULL`
##                                                                   Title
## 1 "Stories about the Instinct of Animals, Their Characters, and Habits"
##           Author Illustrator         eBook URL
## 1 Thomas Bingley T. Landseer www.gutenberg.net

str(tables)

## List of 1
##  $ NULL:'data.frame':    1 obs. of  4 variables:
##   ..$ Title      : Factor w/ 1 level "\"Stories about the Instinct of Animals, Their Characters, and Habits\"": 1
##   ..$ Author     : Factor w/ 1 level "Thomas Bingley": 1
##   ..$ Illustrator: Factor w/ 1 level "T. Landseer": 1
##   ..$ eBook URL  : Factor w/ 1 level "www.gutenberg.net": 1

#xml file

#doc1 <- xmlTreeParse("https://github.com/jameyletherton/week9_assignment/blob/master/book.xml") ; failed

#class(doc1)

doc <- xmlTreeParse("C://Users/jamey/Documents/IS 607/book.xml")


class(doc)

## [1] "XMLDocument"         "XMLAbstractDocument"

root <- xmlRoot(doc)

child <- xmlChildren(root)

root

## <book category="WEB">
##  <title lang="en">XQuery Kick Start</title>
##  <author>James McGovern</author>
##  <author>Per Bothner</author>
##  <author>Kurt Cagle</author>
##  <author>James Linn</author>
##  <author>Vaidyanathan Nagarajan</author>
##  <year>2003</year>
##  <price>49.99</price>
## </book>

child

## $title
## <title lang="en">XQuery Kick Start</title>
## 
## $author
## <author>James McGovern</author>
## 
## $author
## <author>Per Bothner</author>
## 
## $author
## <author>Kurt Cagle</author>
## 
## $author
## <author>James Linn</author>
## 
## $author
## <author>Vaidyanathan Nagarajan</author>
## 
## $year
## <year>2003</year>
## 
## $price
## <price>49.99</price>
## 
## attr(,"class")
## [1] "XMLNodeList"

sapply(child, xmlValue)

##                    title                   author                   author 
##      "XQuery Kick Start"         "James McGovern"            "Per Bothner" 
##                   author                   author                   author 
##             "Kurt Cagle"             "James Linn" "Vaidyanathan Nagarajan" 
##                     year                    price 
##                   "2003"                  "49.99"

 book2 <- "C://Users/jamey/Documents/IS 607/book.json"

bookjason <- readLines(book2)
str(bookjason)

##  chr [1:3] "" ...

readbook2 <- fromJSON(bookjason)
readbook2

## $Title
## [1] " Everyday Italian"
## 
## $Author
## [1] "Giada De Laurentiis"
## 
## $Year
## [1] "2005"
## 
## $Price
## [1] "$30.00"