Extract Questions from XML.Rmd

---
title: "Text Analysis 10"
author: "Junyan Yao"
date: "12/17/2017"
output: html_document
---

```{r}
library(XML)
library(xml2)
```

```{r}
setwd("/Users/YaoJunyan/Documents/Text-Analysis/problem")

temp <- list.files()
myfiles<- lapply (temp, xmlParse)

myfiles[[1]]

readfiles= lapply (temp, read_xml)

readfiles[1]

```

extract the pattern


Some problems have images inserted. Some are not;
When they insert the image, the problem questions is the second line;
If they don't have image inserted, the first line is the question

```{r}
img<-rep(NA,122)
for( i in 1:length(myfiles)){
  img[i]<- grepl("img", readfiles[[i]])
}


scriptype<-rep(NA,122)

for( i in 1:length(myfiles)){
  scriptype[i]<- grepl("system_path", readfiles[[i]])
}

```

Get the questions content for these questions without images

```{r}
xml_children((readfiles[[1]]))[2]
xml_children((readfiles[[3]]))[2]
xml_children((readfiles[[7]]))[2]
```
get the content for all questions
```{r}
for (i in 1:length(readfiles)){
  if (img[i]==TRUE){
    c[i]<- xml_children((readfiles[[i]]))[2]
  }
  else{
    c[i]<-xml_children((readfiles[[i]]))[1]
  }
}


#second try
d<-rep(NA,122)
for (i in 1:length(readfiles)){
  if (img[i]==TRUE){
    d[i]<- xml_children((readfiles[[i]]))[3]
  }
  else{
    d[i]<-xml_children((readfiles[[i]]))[2]
  }
}



df<-data.frame(question1=(as.character(c)))
df2<-data.frame(question2=(as.character(d)))


```