完善错误处理

README.md

@@ -0,0 +1,46 @@
+# exloli
+
+从 E 站下载指定关键词的画(ben)廊(zi)并上传到 Telegraph 并发布到 Telegram 频道
+
+## 用法
+
+1. 创建 Telegram Channel, 并设为公开 (私有 Channel 需要手动获取 chat id
+2. 创建 Telegram Bot, 记录 TOKEN, 并拉进 Channel
+3. 创建 Telegraph 账号, 记录 TOKEN. 创建方法: 访问 `https://api.telegra.ph/createAccount?short_name={}&author_name={}&author_url={}`
+4. 在当前目录下建立 config.toml, 然后运行 exloli
+
+模板如下
+
+```toml
+# 日志等级, 可选 INFO, DEBUG, ERROR
+log_level = "INFO"
+# 抓取线程
+threads_num = "4"
+
+[exhentai]
+# E 站用户名
+username = "username"
+# E 站密码
+password = "password"
+# 搜索关键词
+keyword = "female:lolicon language:Chinese"
+
+[telegraph]
+# telegraph 账号 token
+access_token = "TOKEN"
+# 作者名称
+author_name = "exloli"
+# 作者地址(通常为频道链接)
+author_url = "https://t.me/exlolicon"
+
+[telegram]
+# telegram 频道 ID, 公共频道直接 @+频道名, 私有频道需要需要获取数字格式的 id
+channel_id = "@exlolicon"
+# 机器人 token
+token = "TOKEN"
+
+```
+
+第一次启动将会默认从前两天的画廊开始抓取, 
+抓取完一本本子后将会在当前目录下生成 LAST_TIME 文件,
+下次抓取会一直抓取到这个时间

src/config.rs

@@ -4,6 +4,8 @@ use std::{fs::File, io::Read, path::Path};
 
 #[derive(Debug, Deserialize)]
 pub struct Config {
+    pub log_level: Option<String>,
+    pub threads_num: Option<String>,
     pub exhentai: ExHentai,
     pub telegraph: Telegraph,
     pub telegram: Telegram,

src/exhentai.rs

@@ -7,41 +7,25 @@ use reqwest::{
     Client, ClientBuilder, StatusCode,
 };
 
-lazy_static! {
-    static ref HEADERS: HeaderMap = {
+macro_rules! set_header {
+    ($($k:ident => $v:expr), *) => {{
         let mut headers = HeaderMap::new();
-        headers.insert(
-            header::ACCEPT,
-            HeaderValue::from_static(
-                "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
-            ),
-        );
-        headers.insert(
-            header::ACCEPT_ENCODING,
-            HeaderValue::from_static("gzip, deflate, br"),
-        );
-        headers.insert(
-            header::ACCEPT_LANGUAGE,
-            HeaderValue::from_static("zh-CN,en-US;q=0.7,en;q=0.3"),
-        );
-        headers.insert(header::CACHE_CONTROL, HeaderValue::from_static("max-age=0"));
-        headers.insert(header::DNT, HeaderValue::from_static("1"));
-        headers.insert(header::HOST, HeaderValue::from_static("exhentai.org"));
-        headers.insert(
-            header::REFERER,
-            HeaderValue::from_static("https://exhentai.org/"),
-        );
-        headers.insert(
-            header::UPGRADE_INSECURE_REQUESTS,
-            HeaderValue::from_static("1"),
-        );
-        headers.insert(
-            header::USER_AGENT,
-            HeaderValue::from_static(
-                "Mozilla/5.0 (X11; Linux x86_64; rv:66.0) Gecko/20100101 Firefox/66.0",
-            ),
-        );
+        $(headers.insert(header::$k, HeaderValue::from_static($v));)*
         headers
+    }};
+}
+
+lazy_static! {
+    static ref HEADERS: HeaderMap = set_header!{
+        ACCEPT => "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
+        ACCEPT_ENCODING => "gzip, deflate, br",
+        ACCEPT_LANGUAGE => "zh-CN,en-US;q=0.7,en;q=0.3",
+        CACHE_CONTROL => "max-age=0",
+        DNT => "1",
+        HOST => "exhentai.org",
+        REFERER => "https://exhentai.org/",
+        UPGRADE_INSECURE_REQUESTS => "1",
+        USER_AGENT => "Mozilla/5.0 (X11; Linux x86_64; rv:66.0) Gecko/20100101 Firefox/66.0"
     };
 }
 
@@ -174,8 +158,7 @@ impl ExHentai {
             .into_text()
             .unwrap()[0]
             .split(' ')
-            .skip(1)
-            .next()
+            .nth(1)
             .unwrap()
             .to_owned();
         debug!("评分: {}", rating);
@@ -197,6 +180,7 @@ impl ExHentai {
             .xpath(r#"//table[@class="ptt"]//td[last()]/a/@href"#)?
             .into_text()
         {
+            debug!("下一页: {:?}", next_page);
             let mut response = self.client.get(&next_page.swap_remove(0)).send()?;
             html = parse_html(response.text()?)?;
             img_pages.extend(
@@ -210,7 +194,9 @@ impl ExHentai {
 
     /// 根据图片页面的 URL 获取图片的真实地址
     pub fn get_image_url(&self, url: &str) -> Result<String, Error> {
+        debug!("获取图片真实地址");
         let mut response = self.client.get(url).send()?;
+        debug!("状态码: {}", response.status());
         let html = parse_html(response.text()?)?;
         Ok(html
             .xpath(r#"//img[@id="img"]/@src"#)?

src/main.rs

@@ -15,11 +15,12 @@ use rayon::prelude::*;
 use std::{
     fs,
     io::{Read, Write},
-    path,
     sync::{
         atomic::{AtomicU32, Ordering::SeqCst},
         Arc,
     },
+    thread::sleep,
+    time,
 };
 
 mod config;
@@ -28,8 +29,7 @@ mod telegram;
 mod telegraph;
 mod xpath;
 
-fn run() -> Result<(), Error> {
-    let config = Config::new("config.toml")?;
+fn run(config: &Config) -> Result<(), Error> {
     info!("登录中...");
     let bot = Bot::new(&config.telegram.token);
     let exhentai = ExHentai::new(&config.exhentai.username, &config.exhentai.password)?;
@@ -40,18 +40,18 @@ fn run() -> Result<(), Error> {
         exhentai.search(&config.exhentai.keyword, page).ok()
     });
 
-    let last_time = if path::Path::new("./LAST_TIME").exists() {
+    let last_time = if std::path::Path::new("./LAST_TIME").exists() {
         let mut s = String::new();
         fs::File::open("./LAST_TIME")?.read_to_string(&mut s)?;
         s.parse::<DateTime<Local>>()?
     } else {
-        // 默认从一天前开始
-        Local::now() - Duration::days(1)
+        // 默认从两天前开始
+        Local::now() - Duration::days(2)
     };
+    debug!("截止时间: {:?}", last_time);
 
     let galleries = galleries
         .flatten()
-        .into_iter()
         // FIXME: 由于时间只精确到分钟, 此处存在极小的忽略掉本子的可能性
         .take_while(|gallery| gallery.post_time > last_time)
         .collect::<Vec<Gallery>>();
@@ -73,12 +73,21 @@ fn run() -> Result<(), Error> {
                 let now = i.load(SeqCst);
                 info!("第 {} / {} 张图片", now + 1, img_pages.len());
                 i.store(now + 1, SeqCst);
-                exhentai
-                    .get_image_url(url)
-                    .and_then(|img_url| upload_by_url(&img_url))
-                    .map(|result| result[0].src.to_owned())
+                loop {
+                    let img_url = exhentai
+                        .get_image_url(url)
+                        .and_then(|img_url| upload_by_url(&img_url))
+                        .map(|result| result["src"].to_string());
+                    match img_url {
+                        Ok(v) => break Ok(v),
+                        Err(e) => {
+                            error!("获取图片地址失败: {}", e);
+                            sleep(time::Duration::from_secs(10));
+                        },
+                    }
+                }
             })
-            .collect::<Result<Vec<_>, _>>()?;
+            .collect::<Result<Vec<String>, Error>>()?;
         gallery.img_urls.extend(img_urls);
 
         let content = gallery
@@ -88,13 +97,23 @@ fn run() -> Result<(), Error> {
             .collect::<Vec<_>>()
             .join(",");
         info!("发布文章");
-        let article_url = publish_article(
-            &config.telegraph.access_token,
-            &gallery.title,
-            &config.telegraph.author_name,
-            &config.telegraph.author_url,
-            &format!("[{}]", content),
-        )?;
+
+        let article_url = loop {
+            let result = publish_article(
+                &config.telegraph.access_token,
+                &gallery.title,
+                &config.telegraph.author_name,
+                &config.telegraph.author_url,
+                &format!("[{}]", content),
+            );
+            match result {
+                Ok(v) => break v,
+                Err(e) => {
+                    eprintln!("发布文章失败: {}", e);
+                    sleep(time::Duration::from_secs(10));
+                }
+            }
+        };
         info!("文章地址: {}", article_url);
         bot.send_message(
             &config.telegram.channel_id,
@@ -111,10 +130,31 @@ fn run() -> Result<(), Error> {
 }
 
 fn main() {
+    let config = Config::new("config.toml").unwrap_or_else(|e| {
+        eprintln!("配置文件解析失败:\n{}", e);
+        std::process::exit(1);
+    });
+
+    // 设置相关环境变量
+    if let Some(log_level) = config.log_level.as_ref() {
+        std::env::set_var("RUST_LOG", format!("exloli={}", log_level));
+    }
+    if let Some(threads_num) = config.threads_num.as_ref() {
+        std::env::set_var("RAYON_NUM_THREADS", threads_num);
+    }
+
     env_logger::init();
 
-    match run() {
-        Ok(()) => (),
-        Err(e) => eprintln!("{}", e),
+    loop {
+        match run(&config) {
+            Ok(()) => {
+                info!("任务完成!");
+                return;
+            }
+            Err(e) => {
+                error!("任务出错: {}", e);
+                sleep(time::Duration::from_secs(60));
+            },
+        }
     }
 }

src/telegraph.rs

@@ -1,18 +1,11 @@
 use failure::Error;
+use json::JsonValue;
 use reqwest::{multipart::Form, Client, StatusCode};
-use serde::Deserialize;
 use std::io;
 use tempfile::NamedTempFile;
 
-/// 图片上传结果
-#[derive(Debug, Deserialize)]
-pub struct UploadResult {
-    /// 图片 URL, 为相对 "telegra.ph" 的地址
-    pub src: String,
-}
-
 /// 通过 URL 上传图片至 telegraph
-pub fn upload_by_url(url: &str) -> Result<Vec<UploadResult>, Error> {
+pub fn upload_by_url(url: &str) -> Result<JsonValue, Error> {
     let client = Client::new();
     // 下载图片
     debug!("下载图片: {}", url);
@@ -27,10 +20,10 @@ pub fn upload_by_url(url: &str) -> Result<Vec<UploadResult>, Error> {
         .post("https://telegra.ph/upload")
         .multipart(form)
         .send()?;
-    let json = response.json()?;
-    debug!("结果: {:?}", json);
+    let json = json::parse(&response.text()?)?;
+    debug!("结果: {}", json);
 
-    Ok(json)
+    Ok(json[0].clone())
 }
 
 /// 发布文章, 返回文章地址
@@ -58,6 +51,9 @@ pub fn publish_article(
         return Err(format_err!("{}", text));
     }
     let json = json::parse(&text)?;
+    if json["result"]["url"] == JsonValue::Null {
+        return Err(format_err!("发布文章失败: {}", json));
+    }
     Ok(json["result"]["url"].to_string())
 }