test GHA bot

.github/workflows/bot_test.yml

@@ -0,0 +1,31 @@
+# .github/workflows/test.yml
+name: Test the bot with slack
+on: [push]
+
+jobs:
+  my_job:
+    runs-on: ubuntu-latest
+
+    steps:
+        - name: Checkout repo
+          uses: actions/checkout@v3
+
+        - name: Update apt cache
+          run: sudo apt-get update
+
+        - name: Install cURL Headers
+          run: sudo apt-get install libcurl4-openssl-dev
+
+        - name: Setup R
+          uses: r-lib/actions/setup-r@v2
+          with:
+              r-version: '4.3'
+
+        - name: Install dependencies
+          run: | 
+               install.packages(c('httr', 'jsonlite', 'purrr', 'dplyr', 'readr', 
+               'stringr', 'rlist', 'slackr', 'stringdist','googledrive','base64enc'))
+          shell: Rscript {0}  
+
+        - name: test bot
+          run: Rscript bot_test.r ${{secrets.CREDENTIALS}} ${{secrets.STATEFILE_GDRIVE}} ${{secrets.SLACK_BOT_TOKEN}} ${{secrets.F1000AUTH}}

bot_test.R

@@ -0,0 +1,158 @@
+suppressPackageStartupMessages({
+    library(httr)
+    library(jsonlite)
+    library(purrr)
+    library(dplyr)
+    library(stringr)
+    library(rlist)
+    library(slackr)
+    library(stringdist)
+    library(googledrive)
+    library(readr)
+})
+
+
+# get command line secret inputs
+args <- commandArgs(trailingOnly = TRUE)
+
+# 1st: credentials for Google Drive
+credentials <- args[[1]]
+
+# File path on google drive: 
+file_path <- args[[2]]
+
+# Slack bot token:
+bottoken <- args[[3]]
+
+# Sciwheel authentication token:
+f1000auth <- args[[4]]
+
+
+
+# Load the state from Drive
+# authenticate to Google drive
+drive_auth(path = rawToChar(base64enc::base64decode(credentials)), email = "[email protected]" )
+
+# read the file from google drive
+webhooks <- drive_read_string(file = file_path) %>% read_csv()
+
+
+slackr_setup(token = bottoken)
+users <- slackr_users()
+
+# returns last date
+lastDates <- webhooks %>% pmap_dbl(\(...) {
+    current <- data.frame(...)
+
+    templastDate <- current$lastDate
+
+    resp <- GET(
+        paste0(
+            "https://sciwheel.com/extapi/work/references?projectId=",
+            current$projectId, "&sort=addedDate:desc"
+        ),
+        add_headers(Authorization = paste("Bearer", f1000auth))
+    )
+    print(paste(
+        Sys.time(), "channel", current$channel, "GET status",
+        resp$status_code
+    ))
+
+    if (resp$status_code == 200) {
+        refs <- content(resp)$results %>%
+            list.filter(f1000AddedDate > templastDate) %>%
+            rev()
+
+        blocks <- refs %>% map(\(r) {
+            noteResp <- GET(
+                paste0("https://sciwheel.com/extapi/work/references/", r$id, "/notes?"),
+                add_headers(Authorization = paste("Bearer", f1000auth))
+            )
+
+            comments <- list.filter(content(noteResp), is.null(highlightText))
+
+            note <- ifelse(noteResp$status_code == 200 & length(comments) > 0,
+                           paste0(list.sort(comments, created)[[1]]$comment, "\n"), ""
+            ) %>% str_replace_all("@\\w+", \(name){
+                match <- amatch(
+                    str_remove(name, "@") %>%
+                        tolower(),
+                    users %>%
+                        pull(display_name_normalized) %>%
+                        str_remove_all("\\s") %>%
+                        tolower(),
+                    method = "jw",
+                    weight = c(d = 0.1, s = 1, i = 1, t = 1),
+                    p = 0.1,
+                    maxDist = 0.5)
+
+                if (is.na(match)) {
+                    return(name)
+                } else {
+                    paste0("<@", users %>% pluck("id", match), ">")
+                }
+            })
+
+            details <- paste0(
+                note,
+                r$authorsText, ". <https://sciwheel.com/work/#/items/",
+                r$id, "/detail?collection=", current$projectId, "|", r$title, "> ",
+                r$journalName, ". ", r$publishedYear,
+                " - added by: ", r$f1000AddedBy,
+                ifelse(length(r$f1000Tags) > 0,
+                       paste0(" - tags: ", paste0(r$f1000Tags, collapse = " ")),
+                       ""
+                )
+            )
+
+            if (nchar(details) > 3000) {
+                details <- paste0(
+                    "Full information too long to display. ",
+                    "<https://sciwheel.com/work/#/items/",
+                    r$id, "/detail?collection=", current$projectId, "|", r$title, "> ",
+                    " - added by: ", r$f1000AddedBy
+                )
+            }
+
+            list(type = "section", text = list(type = "mrkdwn", text = details))
+        })
+
+        if (length(blocks) > 0) {
+            webhook <- as.character(current$webhook)
+
+            # one paper at a time with wait in between so we don't get rate limited
+            alive <- TRUE
+            respcodes <- blocks %>% map2_int(
+                refs %>% map(~ .x$f1000AddedDate),
+                \(block, addedDate) {
+                    if (alive) {
+                        Sys.sleep(1.05)
+                        content <- toJSON(list(blocks = list(block)), auto_unbox = TRUE)
+                        outcome <- POST(url = webhook, content_type_json(), body = content)
+                        if (outcome$status_code == 200) {
+                            if (addedDate > templastDate) {
+                                templastDate <<- addedDate
+                            }
+                        } else {
+                            alive <<- FALSE
+                        }
+                        outcome$status_code
+                    } else {
+                        -1
+                    }
+                }
+            )
+
+            print(paste(Sys.time(), "POST status", paste(respcodes, collapse = ", ")))
+        }
+    }
+    templastDate
+})
+
+if (sum(lastDates > webhooks$lastDate) > 0) {
+    webhooks$lastDate <- lastDates
+
+    # write the file back to google drive
+    webhooks %>% write_csv("./temp.csv")
+    drive_update(file = file_path, media = "./temp.csv")
+}