Jira parsing

jira-parsing.R

library(tidyverse)
library(magrittr)
library(httr2)
library(rvest)
library(glue)
library(jsonlite)
library(stringi)
library(checkmate)
library(tictoc)
library(futile.logger)
library(data.table)

# tech credentials
user_id <- "ВАША УЧЕТКА"
user_pwd <- "ВАШ ПАРОЛЬ"
base_url <- "https://САЙТ ДЖИРА/rest/api/latest"
base_req <- httr2::request(base_url) %>%
  req_auth_basic(user_id, user_pwd)

current_sprint_tag <- "Sprint14"

flog.appender(appender.tee("jira.log"))
flog.threshold(INFO)
flog.info("Start batch processing")

# 1. получили список интересующих проектов ----------
resp <- base_req %>%
  req_url_path_append("project") %>%
  req_perform()

proj_df <- resp_body_json(resp) %>%
  bind_rows() %>%
  filter(stri_detect_fixed(name, "ВАШ МАРКЕР ПРОЕКТОВ")) %>%
  distinct(self, id, key, name)

# 2. Получили список всех Issues по выбранным проектам ----

getIssuesInProj <- function(project_tag){
  
  resp <- base_req %>%
    req_url_path_append("search") %>%
    req_url_query(jql = glue("project='{project_tag}'")) %>%
    req_url_query(maxResults = 1000) %>%
    req_perform()
  resp_body <- resp_body_string(resp)
  
  # убедимся, что количество записей меньше окна выдачи
  jqr::jq(resp_body, '.total') %T>%
    {flog.info(glue("Project '{project_tag}' has {.} issues"))} %>%
    {assertInt(as.integer(.), upper = 1000)}
  
  resp_body %>%
    jqr::jq('[.issues[] | . + .fields | del(.fields) | {id, key, 
     issuetype:(.issuetype.name), is_subtask:(.issuetype.subtask), created, 
     status:.status.name, summary, sprint_raw:(.customfield_10101[]? // ""), progress, project_name:(.project.name)}]') %>%
    fromJSON() %>%
    # вытащим маркер спринта 'name=YYYYY Sprint10 1-15.06,'
    mutate(sprint = stri_match_first_regex(sprint_raw, "name=([^,]+),")[, 2]) %>%
    select(-sprint_raw)
}

issues_df <- proj_df$name %>%
  map_dfr(getIssuesInProj) %>%
  # оставим только интересующий спринт
  filter(stri_detect_fixed(sprint, !!current_sprint_tag))

janitor::tabyl(issues_df, project_name)

# 3. Тащим ворклоги по задачам спринта ----
getWklogInIssue <- function(issue_tag){
  
  resp <- base_req %>%
    req_url_path_append("issue") %>%
    req_url_path_append(issue_tag) %>%
    req_url_path_append("worklog") %>%
    req_url_query(maxResults = 1000) %>%
    req_perform()
  resp_body <- resp_body_string(resp)
  
  # убедимся, что количество записей меньше окна выдачи
  crc_lst <- resp_body %>%
    jqr::jq('{maxResults, total, get_all:(.maxResults == .total)}') %>%
    fromJSON() %T>%
    {flog.info(glue("Issue '{issue_tag}' has {.$total} records"))}
  assertTRUE(crc_lst$get_all)
  
  resp_body %>%
    jqr::jq('[.worklogs[] | . + {author:.author.displayName} + {updateAuthor:.updateAuthor.displayName}]') %>%
    fromJSON()
}

wklog_df <- issues_df$key %>%
  map_dfr(getWklogInIssue) %>%
  # TODO: надо отфильтровать по временному диапазону спринта
  mutate(time_hr = timeSpentSeconds/3600) %>%
  mutate(timestamp = anytime::anytime(updated),
         date = lubridate::as_date(timestamp)) %>%
  select(date, author, updateAuthor, time_hr, issueId, timestamp)

# 4. Собираем воедино в формате отчета
report_df <- issues_df %>%
  select(key, summary, status, id, project_name) %>%
  left_join(wklog_df, by = c("id" = "issueId")) %>%
  filter(status %in% c("Done", "Cancelled")) %>%
  group_by(key, summary, project_name, status) %>%
  summarise(time_hr = sum(time_hr), author = stri_c(unique(author), collapse = "; ")) %>%
  ungroup()

# 5. Проверка корректности агрегации по задачам
as.data.table(report_df) %>%
  .[, .N, by = key] %>%
  {assertDataTable(.[N > 1], max.rows = 0)}
# 5. Сбрасываем в excel
report_df %>%
  openxlsx::write.xlsx(here::here("output", glue("{current_sprint_tag}.xlsx")), asTable = TRUE)