Merge branch 'ignore' into 'main'

exclude files from site

See merge request larswirzenius/riki!25

3 files changed, 101 insertions, 23 deletions

diff --git a/riki.md b/riki.md
index fa982b2..62093ff 100644
--- a/riki.md
+++ b/riki.md
@@ -445,3 +445,38 @@ then file output/index.html contains "<img src="img.jpg""
 ~~~{#jpeg .file}
 This is a dummy JPEG image.
 ~~~
+
+## Source file tree
+
+### Listing source files
+
+_Requirement: source files can be listed._
+
+~~~scenario
+given an installed riki
+given file site/index.mdwn from empty
+given file site/img.jpg from empty
+when I run riki list site
+then stdout contains "img.jpg"
+then stdout contains "index.mdwn"
+~~~
+
+### Exclude unusual files
+
+_Requirement: files and directories that aren't meant to be part of
+the site content should be excluded._
+
+~~~scenario
+given an installed riki
+given file site/index.mdwn from empty
+given file site/img.jpg from empty
+given file site/.git/HEAD from empty
+given file site/index.mdwn~ from empty
+given file site/#index.mdwn# from empty
+when I run riki list site
+then stdout contains "img.jpg"
+then stdout contains "index.mdwn"
+then stdout doesn't contain ".git"
+then stdout doesn't contain "index.mdwn~"
+then stdout doesn't contain "#index.mdwn#"
+~~~
diff --git a/src/bin/riki.rs b/src/bin/riki.rs
index 9749e3b..1455b45 100644
--- a/src/bin/riki.rs
+++ b/src/bin/riki.rs
@@ -32,6 +32,7 @@ fn real_main() -> Result<(), SiteError> {
     let args = Args::parse();
     match args.command {
         Command::Build(cmd) => cmd.run()?,
+        Command::List(cmd) => cmd.run()?,
     }
 
     info!("riki ends OK");
@@ -47,6 +48,7 @@ struct Args {
 #[derive(Parser)]
 enum Command {
     Build(Build),
+    List(List)
 }
 
 #[derive(Parser)]
@@ -92,3 +94,22 @@ impl Build {
         Ok(())
     }
 }
+
+#[derive(Parser)]
+struct List {
+    srcdir: PathBuf,
+}
+
+impl List {
+    fn run(&self) -> Result<(), SiteError> {
+        let srcdir = canonicalize(&self.srcdir)?;
+        let mut site = Site::new(&srcdir, &srcdir);
+        site.scan()?;
+        let mut filenames = site.included_files().to_vec();
+        filenames.sort_unstable();
+        for filename in filenames {
+            println!("{}", filename.display());
+        }
+        Ok(())
+    }
+}
diff --git a/src/site.rs b/src/site.rs
index 1a58801..48de58d 100644
--- a/src/site.rs
+++ b/src/site.rs
@@ -2,7 +2,7 @@ use crate::error::SiteError;
 use crate::page::{MarkdownPage, UnprocessedPage, WikitextPage};
 use crate::parser::WikitextParser;
 use crate::token::TokenPatterns;
-use crate::util::{make_path_absolute, make_path_relative_to, make_relative_link, join_subpath};
+use crate::util::{join_subpath, make_path_absolute, make_path_relative_to, make_relative_link};
 use log::{debug, info, trace};
 use std::collections::HashMap;
 use std::path::{Path, PathBuf};
@@ -13,6 +13,7 @@ pub struct Site {
     unprocessed_pages: Vec<UnprocessedPage>,
     markdown_pages: Vec<MarkdownPage>,
     files: Vec<PathBuf>,
+    included_files: Vec<PathBuf>,
     patterns: TokenPatterns,
     srcdir: PathBuf,
     destdir: PathBuf,
@@ -20,6 +21,9 @@ pub struct Site {
 }
 
 impl Site {
+    const EXCLUDE_SUBSTRINGS: &'static [&'static str] = &[".git"];
+    const EXCLUDE_ENDS: &'static [&'static str] = &[".git", "~", "#"];
+
     pub fn new<P>(srcdir: P, destdir: P) -> Self
     where
         P: AsRef<Path>,
@@ -29,6 +33,7 @@ impl Site {
             unprocessed_pages: vec![],
             markdown_pages: vec![],
             files: vec![],
+            included_files: vec![],
             patterns: TokenPatterns::default(),
             srcdir: srcdir.as_ref().into(),
             destdir: destdir.as_ref().into(),
@@ -38,6 +43,8 @@ impl Site {
 
     pub fn scan(&mut self) -> Result<(), SiteError> {
         for filename in Self::all_files(&self.srcdir)? {
+            self.included_files
+                .push(make_path_relative_to(&self.srcdir, &filename));
             if Self::is_markdown(&filename) {
                 let page = WikitextPage::read(&self.srcdir, &filename)?;
                 self.add_wikitextpage(page);
@@ -104,6 +111,10 @@ impl Site {
         &self.files
     }
 
+    pub fn included_files(&self) -> &[PathBuf] {
+        &self.included_files
+    }
+
     pub fn input_filename(&self, filename: &Path) -> Result<PathBuf, SiteError> {
         Ok(join_subpath(&self.srcdir, filename))
     }
@@ -114,14 +125,35 @@ impl Site {
 
     fn all_files(root: &Path) -> Result<Vec<PathBuf>, SiteError> {
         let mut files = vec![];
-        for e in WalkDir::new(root) {
+        for e in WalkDir::new(root)
+        {
             let e = e.map_err(|err| SiteError::WalkDir(root.to_path_buf(), err))?;
             let path = e.path();
-            files.push(path.to_path_buf());
+            if Self::is_excluded(path) {
+                debug!("exclude {}", path.display());
+            } else {
+                debug!("include {}", path.display());
+                files.push(path.to_path_buf());
+            }
         }
         Ok(files)
     }
 
+    fn is_excluded(path: &Path) -> bool {
+        let path = path.to_string_lossy();
+        for pat in Self::EXCLUDE_ENDS {
+            if path.ends_with(pat) {
+                return true;
+            }
+        }
+        for pat in Self::EXCLUDE_SUBSTRINGS {
+            if path.contains(pat) {
+                return true;
+            }
+        }
+        false
+    }
+
     fn is_markdown(path: &Path) -> bool {
         if let Some(ext) = path.extension() {
             ext == "mdwn"
@@ -171,7 +203,11 @@ impl Site {
         // Does a sibling page or file exist?
         if let Some(parent) = page.parent() {
             let path = parent.join(target);
-            trace!("checking for sibling in {}: {}", parent.display(), path.display());
+            trace!(
+                "checking for sibling in {}: {}",
+                parent.display(),
+                path.display()
+            );
             if let Some(path) = self.pages.get(path.as_path()) {
                 trace!("sibling page exists: {}", path.display());
                 return Ok(path.into());
@@ -215,9 +251,7 @@ impl PageSet {
     }
 
     fn normalize(path: &Path) -> String {
-        path.to_str()
-            .expect("path is UTF8")
-            .to_lowercase()
+        path.to_str().expect("path is UTF8").to_lowercase()
     }
 }
 
@@ -265,10 +299,7 @@ mod test {
         let mut site = Site::new(".", ".");
         site.add_wikitextpage(page("/foo/yo"));
         site.process().unwrap();
-        assert_eq!(
-            site.resolve("/foo/bar", "yo").unwrap(),
-            Path::new("yo")
-        );
+        assert_eq!(site.resolve("/foo/bar", "yo").unwrap(), Path::new("yo"));
     }
 
     #[test]
@@ -276,10 +307,7 @@ mod test {
         let mut site = Site::new(".", ".");
         site.add_wikitextpage(page("/foo/yo"));
         site.process().unwrap();
-        assert_eq!(
-            site.resolve("/foo/bar", "YO").unwrap(),
-            Path::new("yo")
-        );
+        assert_eq!(site.resolve("/foo/bar", "YO").unwrap(), Path::new("yo"));
     }
 
     #[test]
@@ -287,10 +315,7 @@ mod test {
         let mut site = Site::new(".", ".");
         site.add_wikitextpage(page("/foo/bar/yo"));
         site.process().unwrap();
-        assert_eq!(
-            site.resolve("/foo/bar", "yo").unwrap(),
-            Path::new("bar/yo")
-        );
+        assert_eq!(site.resolve("/foo/bar", "yo").unwrap(), Path::new("bar/yo"));
     }
 
     #[test]
@@ -299,10 +324,7 @@ mod test {
         site.add_wikitextpage(page("/foo/bar/yo"));
         site.add_wikitextpage(page("/foo/yo"));
         site.process().unwrap();
-        assert_eq!(
-            site.resolve("/foo/bar", "yo").unwrap(),
-            Path::new("bar/yo")
-        );
+        assert_eq!(site.resolve("/foo/bar", "yo").unwrap(), Path::new("bar/yo"));
     }
 
     #[test]