Parsing google alerts using gmail and app script

Question

I have google alerts set up and receive alert emails which look like this:

I would like to parse the alert and save it in a google sheet with the format: alert_name,publication_date, title,news source

Not sure where to start any help/guidance appreciated Thanks

Answer 1

Although I share the opinion that @Ruben has given in the comments, I think this topic is interesting and could help other users to save and manage their Google Alerts.

Code.gs

/* Retrieve all news from googlealert source */
const threads = GmailApp.search('from:(googlealerts-noreply@google.com)')
/* Our sheet for save the news */
const sS = SpreadsheetApp.openById(SS_ID).getSheetByName('Google Alerts')
/* Control the already added answers */
let addedNews = []
try { addedNews = sS.getRange('A1:A' + sS.getLastRow()).getValues().flat() } catch (err) { }

function parseContent() {
  const totalNews = []
  /* Maybe add a control system for remove the threads already saved */
  threads.forEach((th) => {
    const msgs = th.getMessages()
    msgs.forEach((msg) => {
      /* Divide the content in new lines an parse the content */
      const body = msg.getPlainBody().split('\n')
      /* Extract the filter name eg === News - 2 new results for [python] === */
      const filterName = body.slice(0, 1)[0].match(/\[(.*?)\]/)[1]
      const date = msg.getDate()
      /* Remove the unnecessary lines  */
      const cleanedBody = body.slice(1, -11)
      /* Detect the news via empty new lines "\r" */
      const newsIdxs = cleanedBody.reduce((pre, curr, idx) => {
        curr === "\r" && pre.push(idx)
        return pre
      }, [])
      newsIdxs.forEach((v, idx, arr) => {
        if (idx == arr.length - 1) return
        /* From one empty line to the nex */
        const parsedNew = cleanedBody.slice(v + 1, arr[idx + 1])
        /* Simply extracted from the first line */
        const title = parsedNew[0].split('|')[0]
        /* Last line and between <> */
        const url = parsedNew[parsedNew.length - 1].match(/<(.*?)>/)[1]
        /* Extracted from the URL rather than the title due variability  */
        const source = url.match(/url=https:\/\/(.*?)\//)[1]
        totalNews.push({ title, url, date, source, filterName })
      })
    })
  })
  totalNews.forEach((nw) => {
    /* Hash the object for preventing adding already present  */
    const id = hashCode(Object.values(nw).toString())
    if (addedNews.includes(id)) return
    sS.appendRow([id, ...Object.values(nw)])
  })
}

/* Extracted from here https://stackoverflow.com/questions/7616461/generate-a-hash-from-string-in-javascript */
const hashCode = s => s.split('').reduce((a, b) => { a = ((a << 5) - a) + b.charCodeAt(0); return a & a }, 0)

Results

---

Note 1: This script is an approximation of the problem, and has only been tested for News related alerts.

Note 2: Thanks to @DalmTo for the pseudo-code, it has helped me to approach the problem more quickly.

Note 3: The hashCode function has been extracted from here

Note 4: I have decided to take an approach using RegExp due to the use of getPlainBody(), but I think that in this case, using a library that allows parsing HTML with getBody() would be easier to implement.