Lines of
src/core.rs
from check-in 61899b5618
that are changed by the sequence of edits moving toward
check-in 1bd041d00f:
1: use crate::{
2: command,
3: sql::Db,
4: };
5:
6: use std::{
7: borrow::Cow,
8: collections::{
9: BTreeMap,
10: HashSet,
11: },
12: num::TryFromIntError,
13: sync::{
14: Arc,
15: Mutex
16: },
17: };
18:
19: use anyhow::{
20: bail,
21: Result,
22: };
23: use async_std::task;
24: use chrono::DateTime;
25: use frankenstein::{
26: client_reqwest::Bot,
27: methods::{
28: GetUpdatesParams,
29: SendMessageParams
30: },
31: types::{
32: AllowedUpdate,
33: MessageEntityType,
34: User,
35: },
36: updates::UpdateContent,
61899b5618 2025-04-26 37: AsyncTelegramApi,
61899b5618 2025-04-26 38: ParseMode,
39: };
40: use thiserror::Error;
41:
42: #[derive(Error, Debug)]
43: pub enum RssError {
44: // #[error(transparent)]
45: // Tg(#[from] TgError),
46: #[error(transparent)]
47: Int(#[from] TryFromIntError),
48: }
49:
50: #[derive(Clone)]
51: pub struct Core {
52: owner_chat: i64,
53: pub tg: Bot,
54: pub me: User,
55: pub db: Db,
56: sources: Arc<Mutex<HashSet<Arc<i32>>>>,
57: http_client: reqwest::Client,
58: }
59:
60: impl Core {
61: pub async fn new(settings: config::Config) -> Result<Core> {
62: let owner_chat = settings.get_int("owner")?;
63: let api_key = settings.get_string("api_key")?;
64: let tg = Bot::new(&api_key);
65:
66: let mut client = reqwest::Client::builder();
67: if let Ok(proxy) = settings.get_string("proxy") {
68: let proxy = reqwest::Proxy::all(proxy)?;
69: client = client.proxy(proxy);
70: }
71: let http_client = client.build()?;
72: let me = tg.get_me().await?;
73: let me = me.result;
74: let core = Core {
75: tg,
76: me,
77: owner_chat,
78: db: Db::new(&settings.get_string("pg")?)?,
79: sources: Arc::new(Mutex::new(HashSet::new())),
80: http_client,
81: };
82: let mut clone = core.clone();
83: task::spawn(async move {
84: loop {
85: let delay = match &clone.autofetch().await {
86: Err(err) => {
87: if let Err(err) = clone.send(format!("š {err:?}"), None, None).await {
88: eprintln!("Autofetch error: {err:?}");
89: };
90: std::time::Duration::from_secs(60)
91: },
92: Ok(time) => *time,
93: };
94: task::sleep(delay).await;
95: }
96: });
97: Ok(core)
98: }
99:
100: pub async fn stream(&mut self) -> Result<()> {
101: let mut offset: i64 = 0;
102: let mut params = GetUpdatesParams {
103: offset: None,
104: limit: Some(100),
105: timeout: Some(300),
106: allowed_updates: Some(vec![AllowedUpdate::Message]),
107: };
108: loop {
109: let updates = self.tg.get_updates(¶ms).await?.result;
110: if updates.is_empty() {
111: offset = 0;
112: params.offset = None;
113: continue;
114: }
115: for update in updates {
116: if i64::from(update.update_id) >= offset {
117: offset = i64::from(update.update_id) + 1;
118: params.offset = Some(offset);
119: }
120: if let UpdateContent::Message(msg) = update.content {
121: if let Some(text) = msg.text {
122: if let Some(entities) = msg.entities {
123: let chars: Vec<u16> = text.encode_utf16().collect();
124: for entity in entities {
125: if entity.type_field == MessageEntityType::BotCommand && entity.offset != 0 {
126: bail!("commands should be at message start");
127: };
128: let cmd = String::from_utf16_lossy(&chars[entity.offset as usize..entity.length as usize]);
129: let words: Vec<&str> = text.split_whitespace().collect();
61899b5618 2025-04-26 130: match cmd.as_ref() {
61899b5618 2025-04-26 131: "/check" | "/clean" | "/enable" | "/delete" | "/disable" => { command::command(self, msg.chat.id, words).await? },
61899b5618 2025-04-26 132: "/start" => { command::start(self, msg.chat.id).await?; },
61899b5618 2025-04-26 133: "/list" => { command::list(self, msg.chat.id).await?; },
61899b5618 2025-04-26 134: "/add" | "/update" => { command::update(self, msg.chat.id, words).await?; },
61899b5618 2025-04-26 135: any => {
61899b5618 2025-04-26 136: self.send(format!("\\#error\n```\nUnknown command: {any}\n```"),
61899b5618 2025-04-26 137: Some(msg.chat.id),
61899b5618 2025-04-26 138: Some(ParseMode::MarkdownV2)
61899b5618 2025-04-26 139: ).await?;
61899b5618 2025-04-26 140: },
61899b5618 2025-04-26 141: };
142: };
143: };
144: };
145: };
146: }
147: }
148: }
149:
150: pub async fn send <S>(&self, msg: S, target: Option<i64>, mode: Option<ParseMode>) -> Result<()>
151: where S: Into<String> {
152: let msg = msg.into();
153:
154: let mode = mode.unwrap_or(ParseMode::Html);
155: let target = target.unwrap_or(self.owner_chat);
156: let send = SendMessageParams::builder()
157: .chat_id(target)
158: .text(msg)
159: .parse_mode(mode)
160: .build();
61899b5618 2025-04-26 161: self.tg.send_message(&send).await?;
162: Ok(())
163: }
164:
165: pub async fn check (&mut self, id: &i32, owner: i64, real: bool) -> Result<String> {
166: let mut posted: i32 = 0;
167: let mut conn = self.db.begin().await?;
168:
169: let id = {
170: let mut set = self.sources.lock().unwrap();
171: match set.get(id) {
172: Some(id) => id.clone(),
173: None => {
174: let id = Arc::new(*id);
175: set.insert(id.clone());
176: id.clone()
177: },
178: }
179: };
180: let count = Arc::strong_count(&id);
181: if count == 2 {
182: let source = conn.get_source(*id, owner).await?;
183: conn.set_scrape(*id).await?;
184: let destination = match real {
185: true => source.channel_id,
186: false => source.owner,
187: };
188: let mut this_fetch: Option<DateTime<chrono::FixedOffset>> = None;
189: let mut posts: BTreeMap<DateTime<chrono::FixedOffset>, String> = BTreeMap::new();
190:
191: let response = self.http_client.get(&source.url).send().await?;
192: let status = response.status();
193: let content = response.bytes().await?;
194: match rss::Channel::read_from(&content[..]) {
195: Ok(feed) => {
196: for item in feed.items() {
197: if let Some(link) = item.link() {
198: let date = match item.pub_date() {
199: Some(feed_date) => DateTime::parse_from_rfc2822(feed_date),
200: None => DateTime::parse_from_rfc3339(&item.dublin_core_ext().unwrap().dates()[0]),
201: }?;
202: let url = link;
203: posts.insert(date, url.to_string());
204: }
205: };
206: },
207: Err(err) => match err {
208: rss::Error::InvalidStartTag => {
209: match atom_syndication::Feed::read_from(&content[..]) {
210: Ok(feed) => {
211: for item in feed.entries() {
212: let date = item.published().unwrap();
213: let url = item.links()[0].href();
214: posts.insert(*date, url.to_string());
215: };
216: },
217: Err(err) => {
218: bail!("Unsupported or mangled content:\n{:?}\n{err:#?}\n{status:#?}\n", &source.url)
219: },
220: }
221: },
222: rss::Error::Eof => (),
223: _ => bail!("Unsupported or mangled content:\n{:?}\n{err:#?}\n{status:#?}\n", &source.url)
224: }
225: };
226: for (date, url) in posts.iter() {
227: let post_url: Cow<str> = match source.url_re {
228: Some(ref x) => sedregex::ReplaceCommand::new(x)?.execute(url),
229: None => url.into(),
230: };
231: if let Some(exists) = conn.exists(&post_url, *id).await? {
232: if ! exists {
233: if this_fetch.is_none() || *date > this_fetch.unwrap() {
234: this_fetch = Some(*date);
235: };
236: self.send( match &source.iv_hash {
237: Some(hash) => format!("<a href=\"https://t.me/iv?url={post_url}&rhash={hash}\"> </a>{post_url}"),
238: None => format!("{post_url}"),
239: }, Some(destination), Some(ParseMode::Html)).await?;
240: conn.add_post(*id, date, &post_url).await?;
241: };
242: };
243: posted += 1;
244: };
245: posts.clear();
246: };
247: Ok(format!("Posted: {posted}"))
248: }
249:
250: async fn autofetch(&mut self) -> Result<std::time::Duration> {
251: let mut delay = chrono::Duration::minutes(1);
252: let now = chrono::Local::now();
253: let mut conn = self.db.begin().await?;
254: for row in conn.get_queue().await? {
255: if let Some(next_fetch) = row.next_fetch {
256: if next_fetch < now {
257: if let (Some(owner), Some(source_id)) = (row.owner, row.source_id) {
258: let mut clone = Core {
259: owner_chat: owner,
260: ..self.clone()
261: };
262: task::spawn(async move {
263: if let Err(err) = clone.check(&source_id, owner, true).await {
264: if let Err(err) = clone.send(&format!("š {err:?}"), None, None).await {
265: eprintln!("Check error: {err:?}");
266: // clone.disable(&source_id, owner).await.unwrap();
267: };
268: };
269: });
270: }
271: } else if next_fetch - now < delay {
272: delay = next_fetch - now;
273: }
274: }
275: };
276: Ok(delay.to_std()?)
277: }
278:
279: pub async fn list (&mut self, owner: i64) -> Result<String> {
280: let mut reply: Vec<Cow<str>> = vec![];
281: reply.push("Channels:".into());
282: let mut conn = self.db.begin().await?;
283: for row in conn.get_list(owner).await? {
284: reply.push(format!("\n\\#ļøā£ {} \\*ļøā£ `{}` {}\nš `{}`", row.source_id, row.channel,
285: match row.enabled {
286: true => "š enabled",
287: false => "ā disabled",
288: }, row.url).into());
289: if let Some(hash) = &row.iv_hash {
290: reply.push(format!("IV: `{hash}`").into());
291: }
292: if let Some(re) = &row.url_re {
293: reply.push(format!("RE: `{re}`").into());
294: }
295: };
296: Ok(reply.join("\n"))
297: }
298: }