Diff
Logged in as anonymous

Differences From Artifact [da1a6d4590]:

To Artifact [ec9c851bcd]:


1
2
3
4
5
6
7
8
9
10
11
12
13
14
use anyhow::{anyhow, bail, Context, Result};
use async_std::task;
use chrono::DateTime;
use sqlx::{
	postgres::PgPoolOptions,
	Row,
};
use std::{
	borrow::Cow,
	collections::{
		BTreeMap,
		HashSet,
	},
	sync::{Arc, Mutex},



<
|
<
<







1
2
3

4


5
6
7
8
9
10
11
use anyhow::{anyhow, bail, Context, Result};
use async_std::task;
use chrono::DateTime;

use sqlx::postgres::PgPoolOptions;


use std::{
	borrow::Cow,
	collections::{
		BTreeMap,
		HashSet,
	},
	sync::{Arc, Mutex},
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184

185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278

279
280
281

282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
					set.insert(id.clone());
					id.clone()
				},
			}
		};
		let count = Arc::strong_count(&id);
		if count == 2 {
			let mut conn = self.pool.acquire().await
				.with_context(|| format!("Query queue fetch conn:\n{:?}", &self.pool))?;
			let row = sqlx::query("select source_id, channel_id, url, iv_hash, owner, url_re from rsstg_source where source_id = $1 and owner = $2")
				.bind(*id)
				.bind(owner)
				.fetch_one(&mut conn).await
				.with_context(|| format!("Query source:\n{:?}", &self.pool))?;
			drop(conn);

			let channel_id: i64 = row.try_get("channel_id")?;
			let url: &str = row.try_get("url")?;
			let iv_hash: Option<&str> = row.try_get("iv_hash")?;
			let url_re = match row.try_get("url_re")? {
				Some(x) => Some(sedregex::ReplaceCommand::new(x)?),
				None => None,
			};
			let destination = match real {
				true => telegram_bot::UserId::new(channel_id),
				false => telegram_bot::UserId::new(row.try_get("owner")?),
			};
			let mut this_fetch: Option<DateTime<chrono::FixedOffset>> = None;
			let mut posts: BTreeMap<DateTime<chrono::FixedOffset>, String> = BTreeMap::new();

			let response = self.http_client.get(url).send().await?;
			let status = response.status();
			let content = response.bytes().await?;
			match rss::Channel::read_from(&content[..]) {
				Ok(feed) => {
					for item in feed.items() {
						if let Some(link) = item.link() {
							let date = match item.pub_date() {
								Some(feed_date) => DateTime::parse_from_rfc2822(feed_date),
								None => DateTime::parse_from_rfc3339(&item.dublin_core_ext().unwrap().dates()[0]),
							}?;
							let url = link;
							posts.insert(date, url.to_string());
						}
					};
				},
				Err(err) => match err {
					rss::Error::InvalidStartTag => {
						let feed = atom_syndication::Feed::read_from(&content[..])
							.with_context(|| format!("Problem opening feed url:\n{}\n{}", &url, status))?;
						for item in feed.entries() {
							let date = item.published().unwrap();
							let url = item.links()[0].href();
							posts.insert(*date, url.to_string());
						};
					},
					rss::Error::Eof => (),
					_ => bail!("Unsupported or mangled content:\n{:?}\n{:#?}\n{:#?}\n", &url, err, status)
				}
			};
			for (date, url) in posts.iter() {
				let mut conn = self.pool.acquire().await
					.with_context(|| format!("Check post fetch conn:\n{:?}", &self.pool))?;
				let post_url: Cow<str> = match url_re {
					Some(ref x) => x.execute(url),
					None => url.into(),
				};
				let row = sqlx::query("select exists(select true from rsstg_post where url = $1 and source_id = $2) as exists;")
					.bind(&*post_url)
					.bind(*id)
					.fetch_one(&mut conn).await
					.with_context(|| format!("Check post:\n{:?}", &conn))?;
				let exists: bool = row.try_get("exists")?;
				if ! exists {
					if this_fetch.is_none() || *date > this_fetch.unwrap() {
						this_fetch = Some(*date);
					};
					self.tg.send( match iv_hash {
							Some(hash) => telegram_bot::SendMessage::new(destination, format!("<a href=\"https://t.me/iv?url={}&rhash={}\"> </a>{0}", &post_url, hash)),
							None => telegram_bot::SendMessage::new(destination, format!("{}", post_url)),
						}.parse_mode(telegram_bot::types::ParseMode::Html)).await
						.context("Can't post message:")?;
					sqlx::query("insert into rsstg_post (source_id, posted, url) values ($1, $2, $3);")
						.bind(*id)
						.bind(date)
						.bind(&*post_url)
						.execute(&mut conn).await
						.with_context(|| format!("Record post:\n{:?}", &conn))?;
					drop(conn);
					task::sleep(std::time::Duration::new(4, 0)).await;

				};
				posted += 1;
			};
			posts.clear();
		};
		let mut conn = self.pool.acquire().await
			.with_context(|| format!("Update scrape fetch conn:\n{:?}", &self.pool))?;
		sqlx::query("update rsstg_source set last_scrape = now() where source_id = $1;")
			.bind(*id)
			.execute(&mut conn).await
			.with_context(|| format!("Update scrape:\n{:?}", &conn))?;
		Ok(format!("Posted: {}", &posted).into())
	}

	pub async fn delete<S>(&self, source_id: &i32, owner: S) -> Result<Cow<'_, str>>
	where S: Into<i64> {
		let owner = owner.into();

		let mut conn = self.pool.acquire().await
			.with_context(|| format!("Delete fetch conn:\n{:?}", &self.pool))?;
		match sqlx::query("delete from rsstg_source where source_id = $1 and owner = $2;")
			.bind(source_id)
			.bind(owner)
			.execute(&mut conn).await
			.with_context(|| format!("Delete source rule:\n{:?}", &self.pool))?
			.rows_affected() {
			0 => { Ok("No data found found.".into()) },
			x => { Ok(format!("{} sources removed.", x).into()) },
		}
	}

	pub async fn clean<S>(&self, source_id: &i32, owner: S) -> Result<Cow<'_, str>>
	where S: Into<i64> {
		let owner = owner.into();

		let mut conn = self.pool.acquire().await
			.with_context(|| format!("Clean fetch conn:\n{:?}", &self.pool))?;
		match sqlx::query("delete from rsstg_post p using rsstg_source s where p.source_id = $1 and owner = $2 and p.source_id = s.source_id;")
			.bind(source_id)
			.bind(owner)
			.execute(&mut conn).await
			.with_context(|| format!("Clean seen posts:\n{:?}", &self.pool))?
			.rows_affected() {
			0 => { Ok("No data found found.".into()) },
			x => { Ok(format!("{} posts purged.", x).into()) },
		}
	}

	pub async fn enable<S>(&self, source_id: &i32, owner: S) -> Result<&str>
	where S: Into<i64> {
		let owner = owner.into();

		let mut conn = self.pool.acquire().await
			.with_context(|| format!("Enable fetch conn:\n{:?}", &self.pool))?;
		match sqlx::query("update rsstg_source set enabled = true where source_id = $1 and owner = $2")
			.bind(source_id)
			.bind(owner)
			.execute(&mut conn).await
			.with_context(|| format!("Enable source:\n{:?}", &self.pool))?
			.rows_affected() {
			1 => { Ok("Source enabled.") },
			0 => { Ok("Source not found.") },
			_ => { Err(anyhow!("Database error.")) },
		}
	}

	pub async fn disable<S>(&self, source_id: &i32, owner: S) -> Result<&str>
	where S: Into<i64> {
		let owner = owner.into();

		let mut conn = self.pool.acquire().await
			.with_context(|| format!("Disable fetch conn:\n{:?}", &self.pool))?;
		match sqlx::query("update rsstg_source set enabled = false where source_id = $1 and owner = $2")
			.bind(source_id)
			.bind(owner)
			.execute(&mut conn).await
			.with_context(|| format!("Disable source:\n{:?}", &self.pool))?
			.rows_affected() {
			1 => { Ok("Source disabled.") },
			0 => { Ok("Source not found.") },
			_ => { Err(anyhow!("Database error.")) },
		}
	}

	pub async fn update<S>(&self, update: Option<i32>, channel: &str, channel_id: i64, url: &str, iv_hash: Option<&str>, url_re: Option<&str>, owner: S) -> Result<&str>
	where S: Into<i64> {
		let owner = owner.into();

		let mut conn = self.pool.acquire().await
			.with_context(|| format!("Update fetch conn:\n{:?}", &self.pool))?;

		match match update {
				Some(id) => {
					sqlx::query("update rsstg_source set channel_id = $2, url = $3, iv_hash = $4, owner = $5, channel = $6, url_re = $7 where source_id = $1").bind(id)

				},
				None => {
					sqlx::query("insert into rsstg_source (channel_id, url, iv_hash, owner, channel, url_re) values ($1, $2, $3, $4, $5, $6)")

				},
			}
			.bind(channel_id)
			.bind(url)
			.bind(iv_hash)
			.bind(owner)
			.bind(channel)
			.bind(url_re)
			.execute(&mut conn).await {
			Ok(_) => Ok(match update {
				Some(_) => "Channel updated.",
				None => "Channel added.",
			}),
			Err(sqlx::Error::Database(err)) => {
				match err.downcast::<sqlx::postgres::PgDatabaseError>().routine() {
					Some("_bt_check_unique", ) => {







<
<
|
<
<
|
<
<
<
<
<
<
<
<
<
<

|
|




|


















|







|



<
<
|
|


|
<
<
|
<
<
|
|
|
|
|
|
|
|
|
|
<
<
<
|
<
<
|
>





<
<
|
<
|
<







<
<
|
<
<
<
<
|









<
<
|
<
<
<
<
|









<
<
|
<
<
<
<
|










<
<
|
<
<
<
<
|










<
<
<


|
>


|
>

|
<
<
<
<
<
<
<







92
93
94
95
96
97
98


99


100










101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138


139
140
141
142
143


144


145
146
147
148
149
150
151
152
153
154



155


156
157
158
159
160
161
162


163

164

165
166
167
168
169
170
171


172




173
174
175
176
177
178
179
180
181
182


183




184
185
186
187
188
189
190
191
192
193


194




195
196
197
198
199
200
201
202
203
204
205


206




207
208
209
210
211
212
213
214
215
216
217



218
219
220
221
222
223
224
225
226
227







228
229
230
231
232
233
234
					set.insert(id.clone());
					id.clone()
				},
			}
		};
		let count = Arc::strong_count(&id);
		if count == 2 {


			let source = sqlx::query!("select source_id, channel_id, url, iv_hash, owner, url_re from rsstg_source where source_id = $1 and owner = $2",


				*id, owner).fetch_one(&mut self.pool.acquire().await?).await?;










			let destination = match real {
				true => telegram_bot::UserId::new(source.channel_id),
				false => telegram_bot::UserId::new(source.owner),
			};
			let mut this_fetch: Option<DateTime<chrono::FixedOffset>> = None;
			let mut posts: BTreeMap<DateTime<chrono::FixedOffset>, String> = BTreeMap::new();

			let response = self.http_client.get(&source.url).send().await?;
			let status = response.status();
			let content = response.bytes().await?;
			match rss::Channel::read_from(&content[..]) {
				Ok(feed) => {
					for item in feed.items() {
						if let Some(link) = item.link() {
							let date = match item.pub_date() {
								Some(feed_date) => DateTime::parse_from_rfc2822(feed_date),
								None => DateTime::parse_from_rfc3339(&item.dublin_core_ext().unwrap().dates()[0]),
							}?;
							let url = link;
							posts.insert(date, url.to_string());
						}
					};
				},
				Err(err) => match err {
					rss::Error::InvalidStartTag => {
						let feed = atom_syndication::Feed::read_from(&content[..])
							.with_context(|| format!("Problem opening feed url:\n{}\n{}", &source.url, status))?;
						for item in feed.entries() {
							let date = item.published().unwrap();
							let url = item.links()[0].href();
							posts.insert(*date, url.to_string());
						};
					},
					rss::Error::Eof => (),
					_ => bail!("Unsupported or mangled content:\n{:?}\n{:#?}\n{:#?}\n", &source.url, err, status)
				}
			};
			for (date, url) in posts.iter() {


				let post_url: Cow<str> = match source.url_re {
					Some(ref x) => sedregex::ReplaceCommand::new(x)?.execute(&source.url),
					None => url.into(),
				};
				if let Some(exists) = sqlx::query!("select exists(select true from rsstg_post where url = $1 and source_id = $2) as exists;",


					&post_url, *id).fetch_one(&mut self.pool.acquire().await?).await?.exists {


					if ! exists {
						if this_fetch.is_none() || *date > this_fetch.unwrap() {
							this_fetch = Some(*date);
						};
						self.tg.send( match &source.iv_hash {
								Some(hash) => telegram_bot::SendMessage::new(destination, format!("<a href=\"https://t.me/iv?url={}&rhash={}\"> </a>{0}", &post_url, hash)),
								None => telegram_bot::SendMessage::new(destination, format!("{}", post_url)),
							}.parse_mode(telegram_bot::types::ParseMode::Html)).await
							.context("Can't post message:")?;
						sqlx::query!("insert into rsstg_post (source_id, posted, url) values ($1, $2, $3);",



							*id, date, &post_url).execute(&mut self.pool.acquire().await?).await?;


						task::sleep(std::time::Duration::new(4, 0)).await;
					};
				};
				posted += 1;
			};
			posts.clear();
		};


		sqlx::query!("update rsstg_source set last_scrape = now() where source_id = $1;",

			*id).execute(&mut self.pool.acquire().await?).await?;

		Ok(format!("Posted: {}", &posted).into())
	}

	pub async fn delete<S>(&self, source_id: &i32, owner: S) -> Result<Cow<'_, str>>
	where S: Into<i64> {
		let owner = owner.into();



		match sqlx::query!("delete from rsstg_source where source_id = $1 and owner = $2;",




			source_id, owner).execute(&mut self.pool.acquire().await?).await?.rows_affected() {
			0 => { Ok("No data found found.".into()) },
			x => { Ok(format!("{} sources removed.", x).into()) },
		}
	}

	pub async fn clean<S>(&self, source_id: &i32, owner: S) -> Result<Cow<'_, str>>
	where S: Into<i64> {
		let owner = owner.into();



		match sqlx::query!("delete from rsstg_post p using rsstg_source s where p.source_id = $1 and owner = $2 and p.source_id = s.source_id;",




			source_id, owner).execute(&mut self.pool.acquire().await?).await?.rows_affected() {
			0 => { Ok("No data found found.".into()) },
			x => { Ok(format!("{} posts purged.", x).into()) },
		}
	}

	pub async fn enable<S>(&self, source_id: &i32, owner: S) -> Result<&str>
	where S: Into<i64> {
		let owner = owner.into();



		match sqlx::query!("update rsstg_source set enabled = true where source_id = $1 and owner = $2",




			source_id, owner).execute(&mut self.pool.acquire().await?).await?.rows_affected() {
			1 => { Ok("Source enabled.") },
			0 => { Ok("Source not found.") },
			_ => { Err(anyhow!("Database error.")) },
		}
	}

	pub async fn disable<S>(&self, source_id: &i32, owner: S) -> Result<&str>
	where S: Into<i64> {
		let owner = owner.into();



		match sqlx::query!("update rsstg_source set enabled = false where source_id = $1 and owner = $2",




			source_id, owner).execute(&mut self.pool.acquire().await?).await?.rows_affected() {
			1 => { Ok("Source disabled.") },
			0 => { Ok("Source not found.") },
			_ => { Err(anyhow!("Database error.")) },
		}
	}

	pub async fn update<S>(&self, update: Option<i32>, channel: &str, channel_id: i64, url: &str, iv_hash: Option<&str>, url_re: Option<&str>, owner: S) -> Result<&str>
	where S: Into<i64> {
		let owner = owner.into();




		match match update {
				Some(id) => {
					sqlx::query!("update rsstg_source set channel_id = $2, url = $3, iv_hash = $4, owner = $5, channel = $6, url_re = $7 where source_id = $1",
						id, channel_id, url, iv_hash, owner, channel, url_re).execute(&mut self.pool.acquire().await?).await
				},
				None => {
					sqlx::query!("insert into rsstg_source (channel_id, url, iv_hash, owner, channel, url_re) values ($1, $2, $3, $4, $5, $6)",
						channel_id, url, iv_hash, owner, channel, url_re).execute(&mut self.pool.acquire().await?).await
				},
			} {







			Ok(_) => Ok(match update {
				Some(_) => "Channel updated.",
				None => "Channel added.",
			}),
			Err(sqlx::Error::Database(err)) => {
				match err.downcast::<sqlx::postgres::PgDatabaseError>().routine() {
					Some("_bt_check_unique", ) => {
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325

326
327
328
329
330
331
332
333
334
335
336

337
338

339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
				bail!("Sorry, unknown error:\n{:#?}\n", err);
			},
		}
	}

	async fn autofetch(&self) -> Result<std::time::Duration> {
		let mut delay = chrono::Duration::minutes(1);
		let mut conn = self.pool.acquire().await
			.with_context(|| format!("Autofetch fetch conn:\n{:?}", &self.pool))?;
		let now = chrono::Local::now();
		let mut queue = sqlx::query("select source_id, next_fetch, owner from rsstg_order natural left join rsstg_source where next_fetch < now() + interval '1 minute';")
			.fetch_all(&mut conn).await?;
		for row in queue.iter() {
			let source_id: i32 = row.try_get("source_id")?;
			let owner: i64 = row.try_get("owner")?;
			let next_fetch: DateTime<chrono::Local> = row.try_get("next_fetch")?;
			if next_fetch < now {

				let clone = Core {
					owner_chat: telegram_bot::UserId::new(owner),
					..self.clone()
				};
				task::spawn(async move {
					if let Err(err) = clone.check(&source_id, owner, true).await {
						if let Err(err) = clone.send(&format!("šŸ›‘ {:?}", err), None, None).await {
							eprintln!("Check error: {}", err);
						};
					};
				});

			} else if next_fetch - now < delay {
				delay = next_fetch - now;

			}
		};
		queue.clear();
		Ok(delay.to_std()?)
	}

	pub async fn list<S>(&self, owner: S) -> Result<String>
	where S: Into<i64> {
		let owner = owner.into();

		let mut reply: Vec<Cow<str>> = vec![];
		let mut conn = self.pool.acquire().await
			.with_context(|| format!("List fetch conn:\n{:?}", &self.pool))?;
		reply.push("Channels:".into());
		let rows = sqlx::query("select source_id, channel, enabled, url, iv_hash, url_re from rsstg_source where owner = $1 order by source_id")
			.bind(owner)
			.fetch_all(&mut conn).await?;
		for row in rows.iter() {
			let source_id: i32 = row.try_get("source_id")?;
			let username: &str = row.try_get("channel")?;
			let enabled: bool = row.try_get("enabled")?;
			let url: &str = row.try_get("url")?;
			let iv_hash: Option<&str> = row.try_get("iv_hash")?;
			let url_re: Option<&str> = row.try_get("url_re")?;
			reply.push(format!("\n\\#ļøāƒ£ {} \\*ļøāƒ£ `{}` {}\nšŸ”— `{}`", source_id, username,  
				match enabled {
					true  => "šŸ”„ enabled",
					false => "ā›” disabled",
				}, url).into());
			if let Some(hash) = iv_hash {
				reply.push(format!("IV: `{}`", hash).into());
			}
			if let Some(re) = url_re {
				reply.push(format!("RE: `{}`", re).into());
			}
		};
		Ok(reply.join("\n"))
	}
}







<
<

|
|

<
<
|
|
>
|
|
|
|
|
|
|
|
|
|
|
>
|
|
>











<
<

|
<
|

<
<
<
<
<
<
|
|


|
|


|






246
247
248
249
250
251
252


253
254
255
256


257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285


286
287

288
289






290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
				bail!("Sorry, unknown error:\n{:#?}\n", err);
			},
		}
	}

	async fn autofetch(&self) -> Result<std::time::Duration> {
		let mut delay = chrono::Duration::minutes(1);


		let now = chrono::Local::now();
		let mut queue = sqlx::query!(r#"select source_id, next_fetch as "next_fetch: DateTime<chrono::Local>", owner from rsstg_order natural left join rsstg_source where next_fetch < now() + interval '1 minute';"#)
			.fetch_all(&mut self.pool.acquire().await?).await?;
		for row in queue.iter() {


			if let Some(next_fetch) = row.next_fetch {
				if next_fetch < now {
					if let (Some(owner), Some(source_id)) = (row.owner, row.source_id) {
						let clone = Core {
							owner_chat: telegram_bot::UserId::new(owner),
							..self.clone()
						};
						task::spawn(async move {
							if let Err(err) = clone.check(&source_id, owner, true).await {
								if let Err(err) = clone.send(&format!("šŸ›‘ {:?}", err), None, None).await {
									eprintln!("Check error: {}", err);
								};
							};
						});
					}
				} else if next_fetch - now < delay {
					delay = next_fetch - now;
				}
			}
		};
		queue.clear();
		Ok(delay.to_std()?)
	}

	pub async fn list<S>(&self, owner: S) -> Result<String>
	where S: Into<i64> {
		let owner = owner.into();

		let mut reply: Vec<Cow<str>> = vec![];


		reply.push("Channels:".into());
		let rows = sqlx::query!("select source_id, channel, enabled, url, iv_hash, url_re from rsstg_source where owner = $1 order by source_id",

			owner).fetch_all(&mut self.pool.acquire().await?).await?;
		for row in rows.iter() {






			reply.push(format!("\n\\#ļøāƒ£ {} \\*ļøāƒ£ `{}` {}\nšŸ”— `{}`", row.source_id, row.channel,  
				match row.enabled {
					true  => "šŸ”„ enabled",
					false => "ā›” disabled",
				}, row.url).into());
			if let Some(hash) = &row.iv_hash {
				reply.push(format!("IV: `{}`", hash).into());
			}
			if let Some(re) = &row.url_re {
				reply.push(format!("RE: `{}`", re).into());
			}
		};
		Ok(reply.join("\n"))
	}
}