Importing and redefining keywords to search facebook posts

### Workflow for importing and processing the keywords###

### The code was developed as part of the Ancient Identities in Modern Britain (IARH) project
### It imports and redefines the keywords that are going to be used to search Facebook posts
### Project: Ancient Identities in Modern Britain (IARH); ancientidentities.org
### Author: Marta Krzyzanska

# Set working directory
setwd("~path")

# Require xlsx package. If you have not installed it already, run install.packages() first.

require(xlsx)

###Part 1: Importing the keywords

#load the list of keywords
keywords <- read.xlsx("#Enter_file_name.xlsx",1)

#Make a separate lists of the keywords to look for in the pages and in the posts

#Substract one from the number of unique keywords if one of them is NA
pkn <- length(unique(keywords[[1]])) - 1
pageKeywords <-as.character(unique(keywords[[1]])[1:pkn])

pokn <- length(unique(keywords[[2]]))-1
postKeywords <- as.character(unique(keywords[[2]])[1:pokn])

###Part 2: Redefining keywords
#Adds spaces and interpunction with spaces to keywords that are shorter than 5 letters and to the word "Penda".
#so that the search function will not pick them up if they are not a separate word.
#Word "Penda" was included as it frequently features as part of another word.
#Word "celt" wa excluded as it is not likely to form a part of another word in English and is one of the most relavant keywords.

j=length(postKeywords)+1
k=j
i=1
while(i < k){
if ((nchar(postKeywords[i])<5 && postKeywords[i] != "Celt") || postKeywords[i] == "Penda" ){
postKeywords[j] <- paste(" ",postKeywords[i],"\\. ",sep="")
j=j+1
postKeywords[j] <- paste(" ",postKeywords[i],"\\, ",sep="")
j=j+1
postKeywords[j] <- paste(" ",postKeywords[i],"\\! ",sep="")
j=j+1
postKeywords[j] <- paste(" ",postKeywords[i],"\\? ",sep="")
j=j+1
postKeywords[j] <- paste(" ",postKeywords[i],"\\; ",sep="")
j=j+1
postKeywords[j] <- paste(" ",postKeywords[i],"\\: ",sep="")
postKeywords[i] <-paste(" ",postKeywords[i]," ",sep="")
j=j+1
i=i+1}else{i=i+1}}


######Part 2: gets the data that features the keywords
### my_posts is a list of lists of posts extracted from pages
i=1
j=length(my_posts)+1
while (i<j){
if(length(my_posts[[i]])>0){
print (i)
my_posts[[i]] <- searchPosts(my_posts[[i]],postKeywords)
my_posts[[i]] <- subset(my_posts[[i]], my_posts[[i]]$keywords != "" )
i=i+1}else{
i=i+1}
}