library(tidyverse)
c2_count <-"/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case2/060325_c2_https_phab_comments.csv"
c2_input_df <- read.csv(c2_count , header = TRUE) 

c3_count <-"/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case3/060325_c3_http_phab_comments.csv"
c3_input_df <- read.csv(c3_count , header = TRUE) 

#sampling relevance for the text of the Task description
http_relevant <- function(text) {
  if (is.na(text)) {
    return(FALSE)
  }
  # expanded dictionary for relevancy: http, login, SSL, TLS, certificate
  words <- unlist(strsplit(text, "\\s+"))
  for (word in words) {
    lw <- tolower(word)
    if (!grepl("://", lw, fixed = TRUE)) {
      # http
      if (grepl("http", lw, fixed = TRUE)) {
        return(TRUE)
      }
      # login
      if (grepl("login", lw, fixed = TRUE)) {
        return(TRUE)
      }
      # ssl
      if (grepl("ssl", lw, fixed = TRUE)) {
        return(TRUE)
      }
      # tls
      if (grepl("tls", lw, fixed = TRUE)) {
        return(TRUE)
      }
      # cert (but not "certain")
      if (startsWith(lw, "cert") && !startsWith(lw, "certain")) {
        return(TRUE)
      }
    }
  }
  return(FALSE)
}

library(dplyr)


c1_count <-"/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case1/060325_c1_ve_phab_comments.csv"
c1_input_df <- read.csv(c1_count , header = TRUE) 
#C1: after 11-11-2012 before 10-1-2013
c1_desc_flags <- c1_input_df %>%
  filter(comment_type == "task_description") %>%
  mutate(time_flag = date_created >= 1352592000 & date_created <= 1380671999 ) |>
  select(TaskPHID, time_flag)

c1_flagged <- c1_input_df %>%
  left_join(c1_desc_flags, by = "TaskPHID")

c1_sampled <- c1_flagged |>
  filter(time_flag == TRUE)
sum(c1_sampled$comment_type == "task_description")


write.csv(c1_sampled, "/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case1/062725_c1_cleaned_phab.csv", row.names=FALSE)

# going off of the dateCreated of the Task
# C2: after 9-3-2011 before 11-27-2013 
c2_desc_flags <- c2_input_df %>%
  filter(comment_type == "task_description") %>%
  mutate(http_flag = sapply(task_title, http_relevant)) |>
  mutate(time_flag = date_created >= 1315008000 & date_created <= 1385596799) |>
  select(TaskPHID, http_flag, time_flag)

#mutate(http_flag = mapply(function(txt, ttl) http_relevant(txt) || http_relevant(ttl), comment_text, task_title)) %>%

c2_flagged <- c2_input_df %>%
  left_join(c2_desc_flags, by = "TaskPHID")

c2_sampled <- c2_flagged |>
  filter(http_flag == TRUE) |>
  filter(time_flag == TRUE)
sum(c2_sampled$comment_type == "task_description")

write.csv(c2_sampled, "/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case2/070125_c2_title_cleaned.csv", row.names=FALSE)
# C3: after 07-01-2013 before 10-01-2015
c3_desc_flags <- c3_input_df %>%
  filter(comment_type == "task_description") %>%
  mutate(http_flag = sapply(task_title, http_relevant)) |>
  mutate(time_flag = date_created >= 1372636800 & date_created <= 1443743999) |>
  select(TaskPHID, http_flag, time_flag)

c3_flagged <- c3_input_df %>%
  left_join(c3_desc_flags, by = "TaskPHID")

c3_sampled <- c3_flagged |>
  filter(http_flag == TRUE) |>
  filter(time_flag == TRUE)
sum(c3_sampled$comment_type == "task_description")

write.csv(c3_sampled, "/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case3/070125_c3_title_cleaned.csv", row.names=FALSE)