From eec7e72ce027b700e070d99d4a607250f571b9b0 Mon Sep 17 00:00:00 2001
From: Lightling <contact@lightling.xyz>
Date: Fri, 9 Feb 2024 22:28:41 -0500
Subject: [PATCH] handle /search API after running /media requests

---
 lib/dl.js         | 54 +++++++++++++++++++++++++++++++++++++++++------
 run-downloadDb.js |  9 +++++---
 2 files changed, 53 insertions(+), 10 deletions(-)

diff --git a/lib/dl.js b/lib/dl.js
index 36a68bb..feb2cd3 100644
--- a/lib/dl.js
+++ b/lib/dl.js
@@ -4,21 +4,44 @@ import { getArg } from './args.js';
 import { error, log } from './log.js';
 
 const ctx = 'getUser.js';
+const loggedDateRegex = new RegExp('[a-zA-Z0-9]+\-[0-9]+\-([0-9]{4})([0-9]{2})([0-9]{2})_([0-9]{2})([0-9]{2})([0-9]{2})', 'gm');
+
+/**
+ * Gets a date formatted as yyyy-mm-dd for use with `/search`
+ * from the last logged date (which is assumed to be the oldest retrieved image from a previous run)
+ * @param {string[]} logs 
+ */
+const getDateUrlFromLog = (logs) => {
+  // sometimes the logs get grouped into one single string, while others get individually logged
+  // may as well flatten it to a single string and test with regex to make consistent and avoid bugs
+  const flat = logs.flat().join('\n');
+  let result = loggedDateRegex.exec(flat), last = loggedDateRegex.exec(flat);
+  while (!!last) {
+    result = last;
+    last = loggedDateRegex.exec(flat);
+  }
+  let date = new Date(result[1], result[2], result[3], result[4], result[5], result[6], 0);
+  // to be safe, spring forward a day
+  date.setDate(date.getDate() + 1);
+
+  return `${date.getUTCFullYear()}-${date.getUTCMonth()}-${date.getUTCDate()}`;
+};
 
 /**
  * Runs {@link getUser} concurrently for many users
  * @param {{ user: string, logs: string[] }[]} userDb userDb to run {@link getUser} on
  * @param {number} threadMax max number of threads to run concurrently
  * @param {string} directory the directory to save the user media folders in
+ * @param {'media' | 'search'} mode whether to retrieve from the `/media` route or `/search` API
  * @returns {Promise<void>} promise which resolves once all threads for every user have finished
  */
-export const getMany = (userDb, threadMax, directory) => new Promise((resolve, reject) => {
+export const getMany = (userDb, threadMax, directory, mode) => new Promise((resolve, reject) => {
   let running = 0;
   let index = 0;
 
   const get = () => {
     const onFinish = (currentIndex) => {
-      log(ctx, `Finished ${userDb[currentIndex].user}/media`);
+      log(ctx, `Finished ${userDb[currentIndex].user} under ${mode} mode`);
       --running;
       get();
     }
@@ -26,8 +49,12 @@ export const getMany = (userDb, threadMax, directory) => new Promise((resolve, r
     while (running < threadMax && index < userDb.length) {
       ++running;
       let currentIndex = index++;
+      const modeParams = mode === 'media' ? 'media' : {
+        from: '2007-12-31',
+        to: getDateUrlFromLog(userDb[currentIndex].logs),
+      };
 
-      let proc = getUser(userDb[currentIndex].user, directory);
+      let proc = getUser(userDb[currentIndex].user, directory, modeParams);
       proc.stdout.on('data', data => {
         userDb[currentIndex].logs.push(data);
       });
@@ -47,10 +74,23 @@ export const getMany = (userDb, threadMax, directory) => new Promise((resolve, r
  * Retrieves gallery for specified user and saves to the specified parent path
  * @param {string} user the user to retrieve media from
  * @param {string} path the path to save the user's media folder in
+ * @param {'media' | { from: string, to: string }} modeParams depending on the mode:
+ *  - if `'media'`, will retrieve from /media
+ *  - if object, will retrieve from /search?f=live&src=typed_query&q=(from%3A`<USER>`)%20since%3A<FROM|yyyy-mm-dd>%20until%3A`<TO|yyyy-mm-dd>`
  * @returns {ChildProcess} the process that was {@link spawn spawned}
  */
-export const getUser = (user, path) => {
-  const url = `https://twitter.com/${user}/media`;
+export const getUser = (user, path, modeParams) => {
+  //"https://twitter.com/search?f=live&q=(from%3Aad_varg)%20until%3A2023-06-30%20since%3A2007-01-01&src=typed_query"
+  let url = `https://twitter.com/${user}/media`;
+  if (modeParams !== 'media') {
+    url = `https://twitter.com/search?f=live&src=typed_query&q=(from%3A${user})`;
+    if (!!modeParams.from) {
+      url += `%20since%3A${modeParams.from}`;
+    }
+    if (!!modeParams.to) {
+      url += `%20until%3A${modeParams.to}`;
+    }
+  }
   let args;
   try {
     args = getArg('args');
@@ -58,8 +98,8 @@ export const getUser = (user, path) => {
     log(ctx, 'No args being provided to gallery-dl');
   }
 
-  log(ctx, `python3 ~/.local/bin/gallery-dl -c ./config.json${!!args ? ' ' + args + ' ' : ' '}-d ${path} ${url}`);
-  const proc = spawn(`python3 ~/.local/bin/gallery-dl -c ./config.json${!!args ? ' ' + args + ' ' : ' '}-d ${path} ${url}`, { shell: true });
+  log(ctx, `python3 ~/.local/bin/gallery-dl -c ./config.json${!!args ? ' ' + args + ' ' : ' '}-d ${path} "${url}"`);
+  const proc = spawn(`python3 ~/.local/bin/gallery-dl -c ./config.json${!!args ? ' ' + args + ' ' : ' '}-d ${path} "${url}"`, { shell: true });
 
   proc.stdout.on('data', data => {
     log(ctx, data);
diff --git a/run-downloadDb.js b/run-downloadDb.js
index ccacb11..4583a76 100644
--- a/run-downloadDb.js
+++ b/run-downloadDb.js
@@ -39,10 +39,13 @@ const downloadDb = async () => {
     logs: [],
   }));
 
-  log(ctx, `Building db using <user>/media for ${processes.length} users`);
-  await getMany(processes, threadMax, directory);
+  log(ctx, `Downloading media using <user>/media for ${processes.length} users`);
+  await getMany(processes, threadMax, directory, 'media');
 
-  log(ctx, 'Building db using /search');
+  log(ctx, 'Downloading media using /search');
+  await getMany(processes, threadMax, directory, 'search');
+
+  log(ctx, 'Done');
 }
 
 downloadDb();