Is there a dynamic way to extract information from forms? - r

Question

I want to write a R-Script which allows me to extract information of MSG-Files (Email).
The emails are automated sign-up-mails from a Website. They are containing Information about the User (Forename, Surname, Email etc.). I try to extract the specific Information by using regex. The Problem is, that the order of fields may vary.
I use the msgxtractr-Library which works fine. The Output looks like this:
\r\n\r\nAnrede \r\n\r\nHerr\r\n\r\nVorname \r\n\r\nJames \r\n\r\nName \r\n\r\nBond \r\n\r\
To get the Information, i extract the text inbetween two text patterns ->(.*?)
Example:
"Vorname \r\n\r\n(.*?) \r\n\r\n"
library(msgxtractr) #usage
library(magrittr)
#------pfad setzen-----------------------------------------------------------
setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
#------Msg-Datei einlesen-----------------------------------------------------------
BALBLI = read_msg("MSG/Test2.msg")
#------Text zwischen 2 Pattern Extrahieren-----------------------------------
testAR = BALBLI[["body"]][["text"]] #Body aus MSG-Datei
patternVN= "Vorname \r\n\r\n(.*?) \r\n\r\n"
searchVN <- regmatches(testAR,regexec(patternVN,testAR))
Vorname = searchVN[[1]][2]
Vorname
I have been trying two Test-Cases:
1) Good Result:
> patternVN= "Vorname \r\n\r\n(.*?) \r\n\r\n"
> searchVN <- regmatches(testAR,regexec(patternVN,testAR))
> Vorname = searchVN[[1]][2]
> Vorname
[1] "James"
2) Bad Result:
> patternVN= "Vorname \r\n\r\n(.*?) \r\n\r\n"
> searchVN <- regmatches(testAR,regexec(patternVN,testAR))
> Vorname = searchVN[[1]][2]
> Vorname
[1] "John\r\n\r\nName"
In this Case it takes the Pattern after the Name.

Answer 1

I would try a completely different approach.
msg <- "\r\n\r\nAnrede \r\n\r\nHerr\r\n\r\nVorname \r\n\r\nJames \r\n\r\nName \r\n\r\nBond \r\n\r\n"
msg <- gsub("^\\s+", "", msg) # remove spaces at the beginning and end
msg <- gsub("\\s+$", "", msg)
words <- strsplit(msg, " *[\n\r]+ *")[[1]]
res <- as.list(words[seq(2, length(words), 2)])
names(res) <- words[seq(1, length(words), 2)]
Result
> res
$Anrede
[1] "Herr"
$Vorname
[1] "James"
$Name
[1] "Bond"