##############
## Packages ##
##############
library(readxl)
library(scales)
library(knitr)
library(tidyverse)
library(aveytoolkit)
 
###############
## Variables ##
###############
data_dir <- "plotBible"
OTRefFile <- file.path(data_dir, "OldTestamentReference.xlsx")
NTRefFile <- file.path(data_dir, "NewTestamentReference.xlsx")
abbrevFile <- file.path(data_dir, "BibleAbbreviations.xlsx")
LectSundaysFile <- file.path(data_dir, "Lectionary-SundaysAndFeasts.xlsx")
 
###############
## Functions ##
###############
source(file.path(data_dir, "helperFunctions.R"))

########################
## Read in references ##
########################
testaments <- c("OT", "NT")
abbrev <- map(testaments, ~read_excel(abbrevFile, sheet = .x))
names(abbrev) <- testaments
 
## Old Testament
OTsections <- c("Torah", "Historical", "Wisdom", "MajorProphets", "MinorProphets")
OTref <- map_df(OTsections, function(OTsection) {
    read_excel(OTRefFile, sheet = OTsection, na = ".") %>%
        mutate(Section = OTsection)
}) %>%
    rename(Book = `Book Name`) %>%
    gather(matches("[0-9]+"), key = "Chapter", value = "Verses") %>%
    filter(!is.na(Verses)) %>%
    select(Section, Book, Chapter, Verses) %>%
    mutate(Book = factor(Book,            # Order books in factor levels
                         levels = abbrev$OT$Name),
           Abbrv = abbrev$OT$Abbreviation[match(Book, abbrev$OT$Name)],
           Chapter = as.numeric(Chapter)) %>%
    arrange(Book) %>%
    mutate(Abbrv = factor(Abbrv, levels = unique(Abbrv)))
 
## Munge so that reference contains 1 row for every verse in the order they
## appear in the Old Testament with a `Pos` column to denote the position of the
## verse.
OTref2Pos <- OTref %>%
    arrange(Book, Chapter) %>%
    mutate(Verse = map(Verses, function(x) 1:x)) %>%
    unnest() %>%
    mutate(Chapter_Verse = paste(Chapter, Verse, sep = ':')) %>%
    mutate(Pos = 1:n()) %>%
    mutate(Testament = "Old")    
 
## New Testament
NTsections <- c("Gospels", "NT")
NTref <- map_df(NTsections, function(NTsection) {
    read_excel(NTRefFile, sheet = NTsection, na = ".") %>%
        mutate(Section = NTsection) %>%
        mutate_at(vars(matches("[0-9]+")),
                  funs(as.integer(str_replace(., fixed("*"), ""))))
}) %>%
    rename(Abbrv = `Book  Name`) %>%
    gather(matches("[0-9]+"), key = "Chapter", value = "Verses") %>%
    filter(!is.na(Verses)) %>%
    select(Section, Abbrv, Chapter, Verses) %>%
    mutate(Abbrv = factor(Abbrv,
                         levels = abbrev$NT$Abbreviation),
           Book = abbrev$NT$Name[match(Abbrv, abbrev$NT$Abbreviation)],
           Chapter = as.numeric(Chapter)) %>%
    arrange(Abbrv) %>%
    mutate(Book = factor(Book, levels = unique(Book)))
 
## Munge so that reference contains 1 row for every verse in the order they
## appear in the New Testament with a `Pos` column to denote the position of the
## verse.
NTref2Pos <- NTref %>%
    arrange(Book, Chapter) %>%
    mutate(Verse = map(Verses, function(x) 1:x)) %>%
    unnest() %>%
    mutate(Chapter_Verse = paste(Chapter, Verse, sep = ':')) %>%
    mutate(Pos = 1:n()) %>%
    mutate(Testament = "New")
 
ref2Pos <- bind_rows(OTref2Pos, NTref2Pos) %>%
    mutate(Pos = 1:n())

ref2Pos %>%
    select(Testament, Section, Book, Chapter, Verse, Pos) %>%
    head() %>%
    knitr::kable()

ref2Pos %>%
    select(Testament, Section, Book, Chapter, Verse, Pos) %>%
    tail() %>%
    knitr::kable()

###################################################
## Read in the Lectionary for Sundays and Feasts ##
###################################################
readings <- c("OT", "Psalm", "NT", "Gospel")
LectSundays <- map(readings, ~read_excel(LectSundaysFile, sheet = .x))
names(LectSundays) <- readings
 
SunLect <- bind_rows(LectSundays) %>%
    tbl_df() %>%
    separate(LectNum_Year, c("LectNum", "Year"), sep = '-') %>%
    mutate(YearA = grepl("A", Year),
           YearB = grepl("B", Year),
           YearC = grepl("C", Year))

SunLect %>%
    select(Reading, LectNum, Year, Day) %>%
    head() %>%
    kable()

ParseFull(SunLect$Reading[1])

ParseFull(SunLect$Reading[2])

## Apply parser on the full Sunday Lectionary
res <- SunLect %>%
    mutate(Pos = map(Reading, function(x) {
        ## cat(x, sep = '\n') # debugging
        ParseFull(x) %>%
        left_join(ref2Pos, by = c(start = "Chapter_Verse", Abbrv = "Abbrv")) %>%
        left_join(ref2Pos, by = c(end = "Chapter_Verse", Abbrv = "Abbrv")) %>%
        rowwise() %>%
            do(data.frame(Pos = if (any(is.na(.))) {
                                    NA
                                } else {
                                    .$Pos.x:.$Pos.y
                                }
                          )) %>%
            pull(Pos)}))
 
## Combine the results with the reference
comb_dat <- res %>%
    unnest(Pos) %>%
    left_join(ref2Pos, by = "Pos")

comb_dat %>%
    select(Reading, Section, Book, Chapter, Verse, Pos) %>%
    filter(Reading == "Isa 63:16b-17, 19b; 64:2-7") %>%
    kable()

## Get all section lengths
sections <- ref2Pos %>%
    select(Section, Abbrv) %>%
    distinct() %>%
    group_by(Section) %>%
    mutate(Abbrv = list(Abbrv)) %>%
    distinct() %>%
    deframe()
sectPos <- map_int(names(sections), function(sect) {
    abbrvs <- sections[[sect]]
    lastAbbrv <- abbrvs[length(abbrvs)]
    maxPos <- ref2Pos %>%
        filter(Abbrv == lastAbbrv) %>%
        pull(Pos) %>%
        max()
    return(maxPos)
})
names(sectPos) <- names(sections)
sectPos2 <- c(0, sectPos[-length(sectPos)])
names(sectPos2) <- names(sectPos)
sectLength <- sectPos - sectPos2
 
## Tweak Section Labels for Plot
sectLabels <- names(sections) %>%
    paste0(., " (V=", comma(sectLength[.]), ")") %>%
    str_replace("Prophets", " Prophets") %>%
    str_replace("NT", "New Testament") %>%
    setNames(names(sections))
 
## Generate plotting data
dat <- comb_dat %>%
    filter(!is.na(Pos)) %>%
    select(Section, Abbrv, Chapter, Verse, Pos) %>%
    mutate(Pos = (Pos - sectPos2[Section]) / sectLength[Section]) %>%
    mutate(SectionLabel = factor(sectLabels[Section],
                                 levels = sectLabels)) %>%
    mutate(Label = paste0(Abbrv, " ", Chapter, ":", Verse)) %>%
    distinct()      # ignore how many times something appeared (read 2 or 3 times)
book_dat <- ref2Pos %>%
    mutate(Abbrv = factor(Abbrv, levels = unique(Abbrv))) %>%
    mutate(Pos = (Pos - sectPos2[Section]) / sectLength[Section]) %>%
    mutate(SectionLabel = factor(sectLabels[Section],
                                 levels = sectLabels)) %>%
    group_by(SectionLabel, Abbrv) %>%
    summarize(Pos = min(Pos)) %>%
    ungroup() %>%
    group_by(SectionLabel) %>%
    mutate(y = rep(c(0, 0.33, 0.66, 1), length.out = n())) %>%
    ungroup()
 
## Create the plot
gg <- ggplot(dat, x = 0, y = 1) +
    geom_vline(aes(xintercept = Pos, color = Abbrv), show.legend = FALSE) +
    geom_label(data = book_dat, aes(x = Pos, y = y, label = Abbrv),
               hjust = 0, size = 3.5) +    
    scale_x_continuous(labels = percent) +    
    scale_y_continuous(labels = NULL, breaks = NULL) +
    xlab("") +
    ylab("") +
    ggtitle(paste("Bible Coverage of Catholic Lectionary for",
                  "Sundays and Major Feasts", sep = '\n')) +
    labs(caption = "Three Year Cycle") +
    facet_grid(SectionLabel ~ ., scales = "fixed") +
    getBaseTheme() +
    theme(strip.text = element_text(size = 10, face = "bold"))
plot(gg)

GetCoverage <- function(abbrv) {
    tot <- ref2Pos %>%
        filter(Abbrv == abbrv) %>%
        select(Chapter_Verse) %>%
        distinct() %>%
        nrow()
    cov <- dat %>%
        filter(Abbrv == abbrv) %>%
        select(Pos) %>%
        distinct() %>%
        nrow()
    return(cov/tot)
}