1
0
mirror of https://git.sr.ht/~cadence/bibliogram synced 2024-11-23 08:37:28 +00:00

Compare commits

...

3 Commits

Author SHA1 Message Date
Cadence Ember
a9c7923b2c
Sure, let's cache self_blocked_status again! 2022-07-25 01:59:06 +12:00
Cadence Ember
f20554ddfb
Disable Tor by default.
Feel free to experiment and see if it helps if you enable it.
2022-07-25 01:57:44 +12:00
Cadence Ember
c2d7aca1cb
Replace all profile fetching methods with IWeb
The previous HTML method is gone due to a page restructure. It was
able to consistently bypass Instagram's blocking.

The IWeb method has a few hundred uses per X time for selfhosters, and
a couple dozen uses per X time for servers. This will likely change in
the future. There is no known way to bypass Instagram's IWeb blocking.

Feel free to look for a way.

Further timeline pages are still blocked. The "next page" button
defaults to not automatically loading when scrolled, since it will
basically never work anyway. Users running personal instances may be
able to get a couple of uses out of it.
2022-07-25 01:57:44 +12:00
6 changed files with 59 additions and 145 deletions

View File

@ -28,65 +28,18 @@ async function fetchUser(username, context) {
} }
let mode = constants.allow_user_from_reel let mode = constants.allow_user_from_reel
if (mode === "preferForRSS") { if (mode === "iweb") {
if (context === constants.symbols.fetch_context.RSS) mode = "prefer" return fetchUserFromIWeb(username)
else mode = "onlyPreferSaved"
} }
if (context === constants.symbols.fetch_context.ASSISTANT) {
const saved = db.prepare("SELECT username, user_id, updated_version, biography, post_count, following_count, followed_by_count, external_url, full_name, is_private, is_verified, profile_pic_url FROM Users WHERE username = ?").get(username) throw new Error(`Your instance admin selected fetch mode ${mode}, which is now unsupported. Please use "iweb" instead (the default).`)
if (saved && saved.updated_version >= 2) {
return fetchUserFromSaved(saved)
} else {
return fetchUserFromHTML(username)
}
}
if (mode === "never") {
return fetchUserFromHTML(username)
}
if (mode === "prefer") {
const saved = db.prepare("SELECT username, user_id, updated_version, biography, post_count, following_count, followed_by_count, external_url, full_name, is_private, is_verified, profile_pic_url FROM Users WHERE username = ?").get(username)
if (saved && saved.updated_version >= 2) {
return fetchUserFromSaved(saved)
} else if (saved && saved.updated_version === 1) {
return fetchUserFromCombined(saved.user_id, saved.username)
} else {
return fetchUserFromHTML(username)
}
}
if (mode === "onlyPreferSaved") {
const saved = db.prepare("SELECT username, user_id, updated_version, biography, post_count, following_count, followed_by_count, external_url, full_name, is_private, is_verified, profile_pic_url FROM Users WHERE username = ?").get(username)
if (saved && saved.updated_version >= 2) {
return fetchUserFromSaved(saved)
} else {
mode = "fallback"
}
}
if (mode === "fallback") {
return fetchUserFromHTML(username).catch(error => {
if (error === constants.symbols.INSTAGRAM_DEMANDS_LOGIN || error === constants.symbols.RATE_LIMITED) {
const saved = db.prepare("SELECT username, user_id, updated_version, biography, post_count, following_count, followed_by_count, external_url, full_name, is_private, is_verified, profile_pic_url FROM Users WHERE username = ?").get(username)
if (saved && saved.updated_version === 1) {
return fetchUserFromCombined(saved.user_id, username)
} else if (saved && saved.updated_version >= 2) {
return fetchUserFromSaved(saved)
} else if (assistantSwitcher.enabled()) {
return assistantSwitcher.requestUser(username).catch(error => {
if (error === constants.symbols.NO_ASSISTANTS_AVAILABLE) throw constants.symbols.RATE_LIMITED
else throw error
})
}
}
throw error
})
}
throw new Error(`Selected fetch mode ${mode} was unmatched.`)
} }
/** /**
* @param {string} username * @param {string} username
* @returns {Promise<{user: import("./structures/User"), quotaUsed: number}>} * @returns {Promise<{user: import("./structures/User"), quotaUsed: number}>}
*/ */
function fetchUserFromHTML(username) { function fetchUserFromIWeb(username) {
const blockedCacheConfig = constants.caching.self_blocked_status.user_html const blockedCacheConfig = constants.caching.self_blocked_status.user_html
if (blockedCacheConfig) { if (blockedCacheConfig) {
if (history.store.has("user")) { if (history.store.has("user")) {
@ -99,56 +52,44 @@ function fetchUserFromHTML(username) {
let quotaUsed = 0 let quotaUsed = 0
return userRequestCache.getOrFetch("user/"+username, false, true, () => { return userRequestCache.getOrFetch("user/"+username, false, true, () => {
quotaUsed++ quotaUsed++
return switcher.request("user_html", `https://www.instagram.com/${username}/feed/`, async res => { const params = new URLSearchParams({username})
return switcher.request("user_html", `https://i.instagram.com/api/v1/users/web_profile_info/?${params}`, async res => {
if (res.status === 301) throw constants.symbols.ENDPOINT_OVERRIDDEN if (res.status === 301) throw constants.symbols.ENDPOINT_OVERRIDDEN
if (res.status === 302) throw constants.symbols.INSTAGRAM_DEMANDS_LOGIN if (res.status === 302) throw constants.symbols.INSTAGRAM_DEMANDS_LOGIN
if (res.status === 429) throw constants.symbols.RATE_LIMITED if (res.status === 429) throw constants.symbols.RATE_LIMITED
return res return res
}).then(async g => { }).then(async g => {
const res = await g.response() const res = await g.response()
if (res.status === 404) { const json = await g.json()
throw constants.symbols.NOT_FOUND // require down here or have to deal with require loop. require cache will take care of it anyway.
} else { // User -> Timeline -> TimelineEntry -> collectors -/> User
const text = await g.text() const User = require("./structures/User")
// require down here or have to deal with require loop. require cache will take care of it anyway. const user = new User(json.data.user)
// User -> Timeline -> TimelineEntry -> collectors -/> User history.report("user", true)
const User = require("./structures/User") // sure, cache the user info. why not.
const result = extractSharedData(text) if (constants.caching.db_user_id) {
if (result.status === constants.symbols.extractor_results.SUCCESS) { const existing = db.prepare("SELECT created, updated_version FROM Users WHERE username = ?").get(user.data.username)
const sharedData = result.value db.prepare(
const user = new User(sharedData.entry_data.ProfilePage[0].graphql.user) "REPLACE INTO Users (username, user_id, created, updated, updated_version, biography, post_count, following_count, followed_by_count, external_url, full_name, is_private, is_verified, profile_pic_url) VALUES "
history.report("user", true) +"(@username, @user_id, @created, @updated, @updated_version, @biography, @post_count, @following_count, @followed_by_count, @external_url, @full_name, @is_private, @is_verified, @profile_pic_url)"
if (constants.caching.db_user_id) { ).run({
const existing = db.prepare("SELECT created, updated_version FROM Users WHERE username = ?").get(user.data.username) username: user.data.username,
db.prepare( user_id: user.data.id,
"REPLACE INTO Users (username, user_id, created, updated, updated_version, biography, post_count, following_count, followed_by_count, external_url, full_name, is_private, is_verified, profile_pic_url) VALUES " created: existing && existing.updated_version === constants.database_version ? existing.created : Date.now(),
+"(@username, @user_id, @created, @updated, @updated_version, @biography, @post_count, @following_count, @followed_by_count, @external_url, @full_name, @is_private, @is_verified, @profile_pic_url)" updated: Date.now(),
).run({ updated_version: constants.database_version,
username: user.data.username, biography: user.data.biography || null,
user_id: user.data.id, post_count: user.posts || 0,
created: existing && existing.updated_version === constants.database_version ? existing.created : Date.now(), following_count: user.following || 0,
updated: Date.now(), followed_by_count: user.followedBy || 0,
updated_version: constants.database_version, external_url: user.data.external_url || null,
biography: user.data.biography || null, full_name: user.data.full_name || null,
post_count: user.posts || 0, is_private: +user.data.is_private,
following_count: user.following || 0, is_verified: +user.data.is_verified,
followed_by_count: user.followedBy || 0, profile_pic_url: user.data.profile_pic_url
external_url: user.data.external_url || null, })
full_name: user.data.full_name || null,
is_private: +user.data.is_private,
is_verified: +user.data.is_verified,
profile_pic_url: user.data.profile_pic_url
})
}
return user
} else if (result.status === constants.symbols.extractor_results.AGE_RESTRICTED) {
// I don't like this code.
history.report("user", true)
throw constants.symbols.extractor_results.AGE_RESTRICTED
} else {
throw result.status
}
} }
return user
}).catch(error => { }).catch(error => {
if (error === constants.symbols.INSTAGRAM_DEMANDS_LOGIN || error === constants.symbols.RATE_LIMITED) { if (error === constants.symbols.INSTAGRAM_DEMANDS_LOGIN || error === constants.symbols.RATE_LIMITED) {
history.report("user", false, error) history.report("user", false, error)
@ -239,36 +180,6 @@ function fetchUserFromCombined(userID, username) {
}) })
} }
function fetchUserFromSaved(saved) {
let quotaUsed = 0
return userRequestCache.getOrFetch("user/"+saved.username, false, true, async () => {
// require down here or have to deal with require loop. require cache will take care of it anyway.
// ReelUser -> Timeline -> TimelineEntry -> collectors -/> ReelUser
const ReelUser = require("./structures/ReelUser")
const user = new ReelUser({
username: saved.username,
id: saved.user_id,
biography: saved.biography,
edge_follow: {count: saved.following_count},
edge_followed_by: {count: saved.followed_by_count},
external_url: saved.external_url,
full_name: saved.full_name,
is_private: !!saved.is_private,
is_verified: !!saved.is_verified,
profile_pic_url: saved.profile_pic_url
})
// Add first timeline page
if (!user.timeline.pages[0]) {
const {result: page, fromCache} = await fetchTimelinePage(user.data.id, "")
if (!fromCache) quotaUsed++
user.timeline.addPage(page)
}
return user
}).then(user => {
return {user, quotaUsed}
})
}
/** /**
* @param {string} userID * @param {string} userID
* @param {string} after * @param {string} after
@ -548,6 +459,5 @@ module.exports.timelineEntryCache = timelineEntryCache
module.exports.getOrFetchShortcode = getOrFetchShortcode module.exports.getOrFetchShortcode = getOrFetchShortcode
module.exports.updateProfilePictureFromReel = updateProfilePictureFromReel module.exports.updateProfilePictureFromReel = updateProfilePictureFromReel
module.exports.history = history module.exports.history = history
module.exports.fetchUserFromSaved = fetchUserFromSaved
module.exports.assistantSwitcher = assistantSwitcher module.exports.assistantSwitcher = assistantSwitcher
module.exports.verifyUserPair = verifyUserPair module.exports.verifyUserPair = verifyUserPair

View File

@ -26,11 +26,11 @@ let constants = {
// Things that server owners _could_ change if they want to. // Things that server owners _could_ change if they want to.
tor: { tor: {
enabled: true, // If false, everything else in this block has no effect. enabled: false, // If false, everything else in this block has no effect.
password: null, // If `null`, Bibliogram will run its own Tor process instead. password: null, // If `null`, Bibliogram will run its own Tor process instead.
port: 9051, // If a password is provided, Bibliogram will connect to Tor on this port. (This is ignored when running its own Tor process.) port: 9051, // If a password is provided, Bibliogram will connect to Tor on this port. (This is ignored when running its own Tor process.)
for: { for: {
user_html: true, user_html: false,
timeline_graphql: false, timeline_graphql: false,
post_graphql: false, post_graphql: false,
reel_graphql: false reel_graphql: false
@ -41,7 +41,7 @@ let constants = {
// change this to `true` to serve it, which will make extensions like Privacy Badger automatically whitelist the domain. // change this to `true` to serve it, which will make extensions like Privacy Badger automatically whitelist the domain.
does_not_track: false, does_not_track: false,
allow_user_from_reel: "fallback", // one of: "never", "fallback", "prefer", "onlyPreferSaved", "preferForRSS" allow_user_from_reel: "iweb", // legacy. this must be "iweb" now.
proxy_media: { // Whether to proxy media (images, videos, thumbnails) through Bibliogram. This is strongly recommended to protect user privacy. If proxy is turned off, some browser content blockers may break all images since they are served from Facebook domains. proxy_media: { // Whether to proxy media (images, videos, thumbnails) through Bibliogram. This is strongly recommended to protect user privacy. If proxy is turned off, some browser content blockers may break all images since they are served from Facebook domains.
image: true, image: true,
video: true, video: true,
@ -155,7 +155,7 @@ let constants = {
replaceEmptyWithDefault: false replaceEmptyWithDefault: false
},{ },{
name: "infinite_scroll", name: "infinite_scroll",
default: "normal", default: "off",
boolean: false, boolean: false,
replaceEmptyWithDefault: true replaceEmptyWithDefault: true
},{ },{
@ -223,12 +223,12 @@ let constants = {
csrf_time: 60*60*1000, csrf_time: 60*60*1000,
self_blocked_status: { self_blocked_status: {
user_html: { user_html: {
enabled: false, enabled: true,
time: 15*60*1000 time: 2*60*60*1000
}, },
timeline_graphql: { timeline_graphql: {
enabled: false, enabled: true,
time: 24*60*60*1000 time: 2*60*60*1000
} }
}, },
db_user_id: true, db_user_id: true,
@ -239,7 +239,7 @@ let constants = {
// Instagram uses this stuff. This shouldn't be changed, except to fix a bug that hasn't yet been fixed upstream. // Instagram uses this stuff. This shouldn't be changed, except to fix a bug that hasn't yet been fixed upstream.
external: { external: {
reel_query_hash: "c9100bf9110dd6361671f113dd02e7d6", reel_query_hash: "c9100bf9110dd6361671f113dd02e7d6",
timeline_query_hash: "e769aa130647d2354c40ea6a439bfc08", timeline_query_hash: "69cba40317214236af40e7efa697781d",
timeline_query_hash_2: "42323d64886122307be10013ad2dcc44", // https://github.com/rarcega/instagram-scraper/blob/dc022081dbefc81500c5f70cce5c70cfd2816e3c/instagram_scraper/constants.py#L30 timeline_query_hash_2: "42323d64886122307be10013ad2dcc44", // https://github.com/rarcega/instagram-scraper/blob/dc022081dbefc81500c5f70cce5c70cfd2816e3c/instagram_scraper/constants.py#L30
shortcode_query_hash: "2b0673e0dc4580674a88d426fe00ea90", shortcode_query_hash: "2b0673e0dc4580674a88d426fe00ea90",
igtv_query_hash: "bc78b344a68ed16dd5d7f264681c4c76", igtv_query_hash: "bc78b344a68ed16dd5d7f264681c4c76",

View File

@ -4,7 +4,12 @@ const SavedRequestManager = require("./saved_requests/manager")
const constants = require("../constants") const constants = require("../constants")
const userAgent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.169 Safari/537.36" const userAgent = "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:102.0) Gecko/20100101 Firefox/102.0"
const headers = {
"User-Agent": userAgent,
"X-IG-App-ID": 936619743392459 // needed for profile iweb to work
}
const backendStatusLineMap = new Map([ const backendStatusLineMap = new Map([
["node-fetch", "NF "], ["node-fetch", "NF "],
@ -23,16 +28,12 @@ function request(url, options = {}, settings = {}) {
if (constants.request_backend === "node-fetch") { if (constants.request_backend === "node-fetch") {
return new NodeFetch(url, Object.assign({ return new NodeFetch(url, Object.assign({
headers: { headers,
"User-Agent": userAgent
},
redirect: "manual" redirect: "manual"
}, options)) }, options))
} else if (constants.request_backend === "got") { } else if (constants.request_backend === "got") {
return new Got(url, Object.assign({ return new Got(url, Object.assign({
headers: { headers,
"User-Agent": userAgent
},
followRedirect: false, followRedirect: false,
throwHttpErrors: false throwHttpErrors: false
}, options)) }, options))

View File

@ -47,11 +47,14 @@ class NextPage extends FreezeWidth {
this.controller = controller this.controller = controller
this.clicked = false this.clicked = false
this.nextPageNumber = +this.element.getAttribute("data-page") this.nextPageNumber = +this.element.getAttribute("data-page")
this.auto = this.element.getAttribute("data-auto")
this.attribute("href", "javascript:void(0)") this.attribute("href", "javascript:void(0)")
this.event("click", event => this.onClick(event)) this.event("click", event => this.onClick(event))
this.observer = new IntersectionObserver(entries => this.onIntersect(entries), {rootMargin: "0px", threshold: intersectionThreshold}) this.observer = new IntersectionObserver(entries => this.onIntersect(entries), {rootMargin: "0px", threshold: intersectionThreshold})
this.observer.observe(this.element) if (this.auto !== "off") {
this.observer.observe(this.element)
}
} }
onClick(event) { onClick(event) {

View File

@ -10,6 +10,7 @@ mixin next_page_button(user, selectedTimeline, url, type)
data-username=(user.data.username) data-username=(user.data.username)
data-type=type data-type=type
data-loading-text=ll.next_page_button_loading data-loading-text=ll.next_page_button_loading
data-auto=settings.infinite_scroll
)#next-page.next-page= ll.next_page_button )#next-page.next-page= ll.next_page_button
else else
div div

View File

@ -102,9 +102,8 @@ html(dir=ll.meta_direction, lang=settings.language)
+checkbox("spa", ll.fast_navigation, ll.t_enabled, false) +checkbox("spa", ll.fast_navigation, ll.t_enabled, false)
+select("infinite_scroll", ll.infinite_scroll, true, [ +select("infinite_scroll", ll.infinite_scroll, false, [
{value: "normal", text: ll.t_normal}, {value: "normal", text: ll.t_normal},
{value: "eager", text: ll.t_eager},
{value: "off", text: ll.t_manual} {value: "off", text: ll.t_manual}
]) ])