refactor: improve fallback page path checking and logging (#522)

* refactor: check fallback page path and add log debug if not
* refactor: improve fallback page docs and command description

docs/content/configuration/command-line-arguments.md

@@ -35,7 +35,7 @@ Options:
       --page404 <PAGE404>
           HTML file path for 404 errors. If the path is not specified or simply doesn't exist then the server will use a generic HTML error message. If a relative path is used then it will be resolved under the root directory [env: SERVER_ERROR_PAGE_404=] [default: ./404.html]
       --page-fallback <PAGE_FALLBACK>
-          HTML file path that is used for GET requests when the requested path doesn't exist. The fallback page is served with a 200 status code, useful when using client routers. If the path is not specified or simply doesn't exist then this feature will not be active [env: SERVER_FALLBACK_PAGE=] [default: ]
+          A HTML file path (not relative to the root) used for GET requests when the requested path doesn't exist. The fallback page is served with a 200 status code, useful when using client routers. If the path doesn't exist then the feature is not activated [env: SERVER_FALLBACK_PAGE=] [default: ]
   -g, --log-level <LOG_LEVEL>
           Specify a logging level in lower case. Values: error, warn, info, debug or trace [env: SERVER_LOG_LEVEL=] [default: error]
   -c, --cors-allow-origins <CORS_ALLOW_ORIGINS>
@@ -89,7 +89,7 @@ Options:
       --log-forwarded-for [<LOG_FORWARDED_FOR>]
           Log the X-Forwarded-For header for remote IP information [env: SERVER_LOG_FORWARDED_FOR=] [default: false] [possible values: true, false]
       --trusted-proxies <TRUSTED_PROXIES>
-          A comma separated list of IP addresses to accept the X-Forwarded-For header from. Empty means trust all IPs [env: SERVER_TRUSTED_PROXIES] [default: ""]
+          List of IPs to use X-Forwarded-For from. The default is to trust all [env: SERVER_TRUSTED_PROXIES=]
       --redirect-trailing-slash [<REDIRECT_TRAILING_SLASH>]
           Check for a trailing slash in the requested directory URI and redirect permanently (308) to the same path with a trailing slash suffix if it is missing [env: SERVER_REDIRECT_TRAILING_SLASH=] [default: true] [possible values: true, false]
       --ignore-hidden-files [<IGNORE_HIDDEN_FILES>]

docs/content/configuration/environment-variables.md

@@ -48,7 +48,7 @@ HTML file path for 50x errors. If the path is not specified or simply doesn't ex
 If a relative path is used then it will be resolved under the root directory. Default `./50x.html`
 
 ### SERVER_FALLBACK_PAGE
-HTML file path that is used for `GET` requests when the requested path doesn't exist. The fallback page is served with a `200` status code, useful when using client routers (e.g. `React Router``). If the path is not specified or simply doesn't exist then this feature will not be active.
+A HTML file path (not relative to the root) used for `GET` requests when the requested path doesn't exist. The fallback page is served with a `200` status code, useful when using client routers. If the path doesn't exist then the feature is not activated.
 
 ### SERVER_THREADS_MULTIPLIER
 The number of worker threads multiplier will be multiplied by the number of system CPUs using the formula: `worker threads = number of CPUs * n` where `n` is the value that changes here. When the multiplier value is 0 or 1 then the `number of CPUs` is used. The number of worker threads result should be a number between 1 and 32,768 though it is advised to keep this value on the smaller side. Default one thread per core.

docs/content/features/error-pages.md

@@ -19,7 +19,12 @@ static-web-server \
 
 ## Fallback Page for use with Client Routers
 
-An HTML file path that is used for `GET` requests when the requested path doesn't exist. The fallback page is served with a `200` status code, useful when using client routers like `React Router` or similar. If the path is not specified or simply doesn't exist then this feature will not be active.
+It is possible to provide a HTML file to be used as fallback page when `GET` request paths dont exist.
+The fallback page will be served with a `200` status code, useful when using client routers like `React Router` or similar.
+If the path is not specified or simply doesn't exist then this feature will not be activated.
+
+!!! info "The fallback page path is not relative to the root"
+    The fallback page is an independent path, so provide a valid relative or absolute path.
 
 It can be set with the `SERVER_FALLBACK_PAGE` environment variable or with the CLI argument `--page-fallback`.
 

src/fallback_page.rs

@@ -14,13 +14,22 @@ use std::path::Path;
 use crate::{handler::RequestHandlerOpts, helpers, http_ext::MethodExt, Error};
 
 /// Initializes fallback page processing
-pub(crate) fn init(path: &Path, handler_opts: &mut RequestHandlerOpts) {
-    handler_opts.page_fallback = helpers::read_bytes_default(path);
+pub(crate) fn init(file_path: &Path, handler_opts: &mut RequestHandlerOpts) {
+    let found = file_path.is_file();
+    if found {
+        handler_opts.page_fallback =
+            String::from_utf8_lossy(&helpers::read_bytes_default(file_path))
+                .trim()
+                .as_bytes()
+                .to_owned();
+    } else {
+        tracing::debug!("fallback page path not found or not a regular file");
+    }
 
     server_info!(
         "fallback page: enabled={}, value=\"{}\"",
-        !handler_opts.page_fallback.is_empty(),
-        path.to_string_lossy()
+        found,
+        file_path.display()
     );
 }
 

src/settings/cli.rs

@@ -116,7 +116,7 @@ pub struct General {
     #[cfg(feature = "fallback-page")]
     #[cfg_attr(docsrs, doc(cfg(feature = "fallback-page")))]
     #[arg(long, default_value = "", value_parser = value_parser_pathbuf, env = "SERVER_FALLBACK_PAGE")]
-    /// HTML file path that is used for GET requests when the requested path doesn't exist. The fallback page is served with a 200 status code, useful when using client routers. If the path is not specified or simply doesn't exist then this feature will not be active.
+    /// A HTML file path (not relative to the root) used for GET requests when the requested path doesn't exist. The fallback page is served with a 200 status code, useful when using client routers. If the path doesn't exist then the feature is not activated.
     pub page_fallback: PathBuf,
 
     #[arg(long, short = 'g', default_value = "error", env = "SERVER_LOG_LEVEL")]