week 3 Assignment

By Brian Weinfeld

February 15, 2018

Regulax Expressions with Stringr

3. Copy the introductory example. The vector names stores the extracted names.

raw.data <-"555-1239Moe Szyslak(636) 555-0113Burns, C. Montgomery555-6542Rev. Timothy Lovejoy555 8904Ned Flanders636-555-3226Simpson, Homer5553642Dr. Julius Hibbert"
#from textbook
names <- unlist(str_extract_all(raw.data, '[[:alpha:]., ]{2,}'))
names

## [1] "Moe Szyslak"          "Burns, C. Montgomery" "Rev. Timothy Lovejoy"
## [4] "Ned Flanders"         "Simpson, Homer"       "Dr. Julius Hibbert"

• Use the tools of this chapter to rearrange the vector so that all elements conform to the standard first_name last_name.

corrected_names <- str_replace_all(names, '([:graph:]+?), ([:print:]+)', '\\2 \\1')
corrected_names

## [1] "Moe Szyslak"          "C. Montgomery Burns"  "Rev. Timothy Lovejoy"
## [4] "Ned Flanders"         "Homer Simpson"        "Dr. Julius Hibbert"

I was uncertain how to properly format Mr. Burns’ name to follow the assignment. His first name is Charles (C) but as it’s an abbreviation he would likely want Montgomery used as his first name. Ultimately, I decided to keep both ala M. Night Shyamalan.

• Construct a logical vector indicating whether a character has a title (ie, Rev and Dr).

has_title <- str_detect(corrected_names, '[:alpha:]{2,}\\.')
has_title

## [1] FALSE FALSE  TRUE FALSE FALSE  TRUE

• Construct a logical vector indicating whether a character has a second name.

str_detect(corrected_names, '([:print:]+? ){2}[:print:]+') & !has_title

## [1] FALSE  TRUE FALSE FALSE FALSE FALSE

4. Describe the type of strings that conform to the following regular expressions and construct an example that is matched by the regular expression.

• [0-9]+\$

test <- c('200$', "Hello do you have 350$?", "1$! That is too much!")
str_match(test, '[0-9]+\\$')

##      [,1]  
## [1,] "200$"
## [2,] "350$"
## [3,] "1$"

This matches any string that contains a string of digits of any length ending in a dollar sign. Apparently some parts of Europe and Canada typically place the dollar sign after the digits, so perhaps this is trying to match those situations

• \b[a-z]{1,4}\b

test2 <- c('what', "we're wishing you well", "Best wishes with you")
str_match(test2, '\\b[a-z]{1,4}\\b')

##      [,1]  
## [1,] "what"
## [2,] "we"  
## [3,] "with"

This matches any string that contains a 1 to 4 lowercase characters using only the standard English 26 characters. This may not grab the entire word as in the case of accented characters or possessives and contractions.

• .*?\.txt$

test3 <- c('hello.txt', '#(*$&@#($*&#.txt', 'The file you want is called name.txt')
str_match(test3, '.*?\\.txt$')

##      [,1]                                  
## [1,] "hello.txt"                           
## [2,] "#(*$&@#($*&#.txt"                    
## [3,] "The file you want is called name.txt"

This matches the entirety of any string that concludes with the four characters ‘.txt’. This may be attempting to find text file names but as in the example above, grabs extra likely unneeded information.

• \d{2}/\d{2}/\d{4}

test4 <- c('10/15/2000', '33/99/9999', '100/10/100000')
str_match(test4, '\\d{2}/\\d{2}/\\d{4}')

##      [,1]        
## [1,] "10/15/2000"
## [2,] "33/99/9999"
## [3,] "00/10/1000"

This seems to be attempting to match a date found in a string. 2 digits, a slash, 2 more digits, a slash and finally 4 digits. However, this will end up matching non-date like strings and also can match string formated as dates with invalid dates (for example, month 33). It will also miss single month/day combinations

• <(.+?)>.+?</\1>

test5 <- c('<hello> alsifejaslfij </hello>', '<p>This is an <b>important</b> example</p>')
str_match(test5, '<(.+?)>.+?</\\1>')

##      [,1]                                         [,2]   
## [1,] "<hello> alsifejaslfij </hello>"             "hello"
## [2,] "<p>This is an <b>important</b> example</p>" "p"

This string seems like it would be used when processing html. It captures the first tag that it finds (identified by being inside the < and > symbols) and then looks for a matching closing bracket further on in the text. This could be useful if you wanted to extract all the tables on a webpage for example.

9. The following code hides a secret message. Crack it with R and regular expressions.

code <- 'clcopCow1zmstc0d87wnkig7OvdicpNuggvhryn92Gjuwczi8hqrfpRxs5Aj5dwpn0TanwoUwisdij7Lj8kpf03AT5Idr3coc0bt7yczjatOaootj55t3Nj3ne6c4Sfek.r1w1YwwojigOd6vrfUrbz2.2bkAnbhzgv4R9i05zEcrop.wAgnb.SqoU65fPa1otfb7wEm24k6t3sR9zqe5fy89n6Nd5t9kc4fE905gmc4Rgxo5nhDk!gr'
code %>% str_match_all('[A-Z.!]') %>% unlist() %>% str_replace('\\.', ' ') %>% paste(collapse='')

## [1] "CONGRATULATIONS YOU ARE A SUPERNERD!"

I began by exploring the data looking for patterns. I thought perhaps the puzzle was similar to a puzzle in the daily newspaper where each letter was a substitution for another letter. I created a frequency table but saw no pattern.
Next, I used regular expressions to try to find repeated patterns in the data but found none.
Next, I examined the digits in the code with regular expressions again looking for a pattern.
Finally, I isolated the capital letters. This led to my breakthrough. The remaining code was written from there to clean up the output.