DrugLaws

# INSTALL THE BELOW PACKAGES IF NEEDED
#install.packages("pdftools")
#install.packages("remotes")
#remotes::install_github(c("ropensci/tabulizerjars", "ropensci/tabulizer"), INSTALL_opts = "--no-multiarch", dependencies = c("Depends", "Imports"))
library(tabulizer)
library(pdftools)
library(rvest)
library(tidyverse)
library(dplyr)

url = "https://www.ncsl.org/research/health/state-laws-and-legislation-related-to-biologic-medications-and-substitution-of-biosimilars.aspx"

webpage=read_html(url)
possible = webpage %>% 
  html_element(css = "#dnn_ctr84472_HtmlModule_lblContent > div > table.table.table-hover.bio") %>% 
  html_table()

possible = data.frame(possible)
possible=possible[-1, -c(10:17)]

View(possible)

# Mandatory 2014 ####################################

mandatory_2014 = webpage %>% 
  html_element(css = "#dnn_ctr84472_HtmlModule_lblContent > div > table:nth-child(15)") %>% 
  html_table()
mandatory_2014 = data.frame(mandatory_2014)

mandatory_2014 <- data.frame(Mandatory_substitution_2014 = trimws(scan(text = sub("\\s+\\(.*", "",gsub("(\\n+\\t+)", ",", mandatory_2014)), what="", sep=",")))

View(mandatory_2014)


# 2013 - 2019 ########################################
MintzOriginal = read.csv('https://raw.githubusercontent.com/melissaalejandro/Research-Assistant/main/State%20Legislation%20on%20Biosimilars2.csv')
Mintz = MintzOriginal

# delete row if col 2 has a blank value. 
Mintz = Mintz[Mintz$Substitution.Requirements != "", ]

# remove entire row if column 1 has the word 'State'
Mintz=Mintz[Mintz$State != "State", ]

#reset row numbers
Mintz= Mintz %>% data.frame(row.names = 1:nrow(.)) # original data only had 46 states (including PR) 

# delete PR
Mintz = Mintz[-34,]

#reset row numbers
Mintz= Mintz %>% data.frame(row.names = 1:nrow(.)) 

# Copy the yr each law went into effect into a new col
Mintz$Year = sub('.*(\\d{4}).*', '\\1', Mintz$State)

Mintz = Mintz[,c(1,4,2,3)]

# Keep only state name in column 1
mintz=Mintz
mintz[,1] = mintz %>% 
  select(State) %>% 
  mutate(State = str_extract(State, paste(state.name, collapse='|'))
  ) 

View(mintz)

url2 = "https://oppe.pharmacy.washington.edu/PracticumSite/forms/2019_Survey_of_Pharmacy_Law.pdf?-session=Students_Session%3A42F94F5D0a61a20754trv33D875D&fbclid=IwAR0qeK2tYmyI7T_8ict1Hnew9JxPkpt0bvajI3KL3IFDWg6JHNSSFWGlKY4"

out <- pdf_text(url2)
all_2019=as.data.frame(out[[93]],header=F)

all_2019=all_2019 %>%
  rename(x = `out[[93]]`) %>%
  mutate(x = stringr::str_split(x, '\\n\n',simplify = T)[2]) %>%
  mutate(x = strsplit(x, '\\n')) %>%
  unnest() %>%
  .[-c(1:3),] %>%
  separate(x, c("a","State", "State Drug Formulary","Two-line Rx Format","Permissive or Mandatory*","How to Prevent Substitution","Cost Savings Pass-on","Patient Consent**"), sep = "\\s{2,}") %>%
  select(-a)

View(all_2019)