Merge pull request #186 from karthik2804/file_list_with_globs

create site.pages with globs

Cargo.toml

@@ -24,6 +24,8 @@ spin-sdk = { git = "https://github.com/fermyon/spin", rev = "139c40967a75dbdd5d4
 toml = "0.5.9"
 wit-bindgen-rust = { git = "https://github.com/bytecodealliance/wit-bindgen", rev = "cb871cfa1ee460b51eb1d144b175b9aab9c50aba", default-features = false }
 regex = "1"
+walkdir = "2.4.0"
+glob = "0.3.1"
 
 [workspace]
 members = ["bart"]

src/content.rs

@@ -1,3 +1,5 @@
+use anyhow::bail;
+use glob::glob;
 use regex::Regex;
 use serde::{Deserialize, Serialize};
 use std::collections::{BTreeMap, HashMap};
@@ -167,6 +169,19 @@ pub fn all_pages(
     }
 }
 
+pub fn get_pages_by_glob(
+    dir: PathBuf,
+    globs: &Vec<String>,
+    show_unpublished: bool,
+) -> anyhow::Result<BTreeMap<String, PageValues>> {
+    let index_cache: IndexCache = pages_by_glob_load(dir, globs, show_unpublished)?;
+    Ok(index_cache.contents)
+}
+
+pub fn all_pages_load(dir: PathBuf, show_unpublished: bool) -> anyhow::Result<IndexCache> {
+    let files = all_files(dir)?;
+    get_index_cache_from_files(files, show_unpublished)
+}
 pub struct IndexCache {
     contents: BTreeMap<String, PageValues>,
     cache_expiration: Option<DateTime<Utc>>,
@@ -176,60 +191,25 @@ pub struct IndexCache {
 ///
 /// If show_unpublished is `true`, this will include pages that Bartholomew has determined are
 /// unpublished.
-pub fn all_pages_load(dir: PathBuf, show_unpublished: bool) -> anyhow::Result<IndexCache> {
-    let files = all_files(dir)?;
-    let mut contents = BTreeMap::new();
-    let mut contains_unpublished: bool = false;
-    let mut earliest_unpublished: Option<DateTime<Utc>> = None;
-    for f in files {
-        // Dotfiles should not be loaded.
-        if f.file_name()
-            .map(|f| f.to_string_lossy().starts_with('.'))
-            .unwrap_or(false)
-        {
-            eprintln!("Skipping dotfile {f:?}");
-            continue;
-        }
-        let raw_data = std::fs::read_to_string(&f)
-            .map_err(|e| anyhow::anyhow!("File is not string data: {:?}: {}", &f, e))?;
-        match raw_data.parse::<Content>() {
-            Ok(content) => {
-                if show_unpublished || content.published {
-                    contents.insert(f.to_string_lossy().to_string(), content.into());
-                } else {
-                    // find earliest unpublished article to save timestamp to refresh cache
-                    let article_date = content.head.date;
-                    match contains_unpublished {
-                        true => {
-                            if match earliest_unpublished {
-                                Some(val) => article_date.map(|d| d <= val).unwrap_or(true),
-                                _ => false,
-                            } {
-                                earliest_unpublished = article_date;
-                            }
-                        }
-                        false => {
-                            if let Some(val) = article_date {
-                                if val > Utc::now() {
-                                    earliest_unpublished = article_date;
-                                    contains_unpublished = true;
-                                }
-                            };
-                        }
-                    }
+pub fn pages_by_glob_load(
+    dir: PathBuf,
+    glob_patterns: &Vec<String>,
+    show_unpublished: bool,
+) -> anyhow::Result<IndexCache> {
+    let mut files: Vec<PathBuf> = Vec::new();
+    for glob_pattern in glob_patterns {
+        let full_pattern = format!("{}{}", dir.to_string_lossy(), glob_pattern);
+        let full_path = PathBuf::from(&full_pattern);
+        for entry in glob(full_path.to_str().unwrap_or_default())? {
+            match entry {
+                Ok(path) => files.push(path),
+                Err(e) => {
+                    bail!("Failed to read file glob \"{glob_pattern}\": {e}")
                 }
             }
-            Err(e) => {
-                // If a parse fails, don't take down the entire site. Just skip this piece of content.
-                eprintln!("File {:?}: {}", &f, e);
-                continue;
-            }
         }
     }
-    Ok(IndexCache {
-        contents,
-        cache_expiration: earliest_unpublished,
-    })
+    get_index_cache_from_files(files, show_unpublished)
 }
 
 /// Fetch a list of paths to every file in the directory
@@ -434,6 +414,64 @@ impl FromStr for Content {
     }
 }
 
+fn get_index_cache_from_files(
+    files: Vec<PathBuf>,
+    show_unpublished: bool,
+) -> Result<IndexCache, anyhow::Error> {
+    let mut contents = BTreeMap::new();
+    let mut contains_unpublished: bool = false;
+    let mut earliest_unpublished: Option<DateTime<Utc>> = None;
+    for f in files {
+        // Dotfiles should not be loaded.
+        if f.file_name()
+            .map(|f| f.to_string_lossy().starts_with('.'))
+            .unwrap_or(false)
+        {
+            eprintln!("Skipping dotfile {f:?}");
+            continue;
+        }
+        let raw_data = std::fs::read_to_string(&f)
+            .map_err(|e| anyhow::anyhow!("File is not string data: {:?}: {}", &f, e))?;
+        match raw_data.parse::<Content>() {
+            Ok(content) => {
+                if show_unpublished || content.published {
+                    contents.insert(f.to_string_lossy().to_string(), content.into());
+                } else {
+                    // find earliest unpublished article to save timestamp to refresh cache
+                    let article_date = content.head.date;
+                    match contains_unpublished {
+                        true => {
+                            if match earliest_unpublished {
+                                Some(val) => article_date.map(|d| d <= val).unwrap_or(true),
+                                _ => false,
+                            } {
+                                earliest_unpublished = article_date;
+                            }
+                        }
+                        false => {
+                            if let Some(val) = article_date {
+                                if val > Utc::now() {
+                                    earliest_unpublished = article_date;
+                                    contains_unpublished = true;
+                                }
+                            };
+                        }
+                    }
+                }
+            }
+            Err(e) => {
+                // If a parse fails, don't take down the entire site. Just skip this piece of content.
+                eprintln!("File {:?}: {}", &f, e);
+                continue;
+            }
+        }
+    }
+    Ok(IndexCache {
+        contents,
+        cache_expiration: earliest_unpublished,
+    })
+}
+
 #[cfg(test)]
 mod test {
     use super::*;

src/template.rs

@@ -4,10 +4,14 @@ use {
     std::path::PathBuf,
 };
 
+use std::{collections::HashMap, fs::File, io::Read};
+
 use crate::rhai_engine::custom_rhai_engine_init;
 
 use super::content::{Content, Head};
+use anyhow::bail;
 use serde::{Deserialize, Serialize};
+use walkdir::WalkDir;
 
 /// The name of the default template.
 /// This will be resolved to $TEMPLATE_DIR/$DEFAULT_TEMPLATE.hbs
@@ -91,11 +95,17 @@ impl From<Content> for PageValues {
         }
     }
 }
+#[cfg(feature = "server")]
+#[derive(Serialize, Deserialize, Default, Debug)]
+pub struct TemplateMeta {
+    read_pages_glob: Option<Vec<String>>,
+}
 
 /// Renderer can execute a handlebars template and render the results into HTML.
 #[cfg(feature = "server")]
 pub struct Renderer<'a> {
     pub template_dir: PathBuf,
+    pub template_meta: HashMap<String, TemplateMeta>,
     pub theme_dir: Option<PathBuf>,
     pub script_dir: PathBuf,
     pub content_dir: PathBuf,
@@ -117,9 +127,11 @@ impl<'a> Renderer<'a> {
         // Create custom rhai engine and assign to handlebars
         let rhai_engine = custom_rhai_engine_init();
         handlebars.set_engine(rhai_engine);
+        let template_meta = HashMap::new();
 
         Renderer {
             template_dir,
+            template_meta,
             theme_dir,
             script_dir,
             content_dir,
@@ -147,8 +159,26 @@ impl<'a> Renderer<'a> {
             self.handlebars
                 .register_templates_directory(".hbs", templates)?;
         }
-        self.handlebars
-            .register_templates_directory(".hbs", &self.template_dir)?;
+        for entry in WalkDir::new(&self.template_dir)
+            .into_iter()
+            .filter_map(|e| e.ok())
+        {
+            if entry.file_type().is_file() && entry.path().extension().map_or(false, |e| e == "hbs")
+            {
+                let filename = entry.file_name().to_str().unwrap().replace(".hbs", "");
+                let res = parse_hbs_template(entry);
+                match res {
+                    Ok((template_meta, body)) => {
+                        self.template_meta
+                            .insert(filename.to_owned(), template_meta);
+                        self.handlebars.register_template_string(&filename, body)?;
+                    }
+                    Err(err) => {
+                        eprintln!("Error reading template {}: {}", filename, err);
+                    }
+                }
+            }
+        }
         Ok(())
     }
 
@@ -220,11 +250,14 @@ impl<'a> Renderer<'a> {
                 pages: match &info.index_site_pages {
                     Some(templates) => {
                         if templates.contains(&tpl) {
-                            crate::content::all_pages(
-                                self.content_dir.clone(),
-                                self.show_unpublished,
-                                self.disable_cache,
-                            )?
+                            let mut glob_pattern: Option<&Vec<String>> = None;
+                            let default_meta = TemplateMeta::default();
+                            let template_meta =
+                                self.template_meta.get(&tpl).unwrap_or(&default_meta);
+                            if let Some(val) = &template_meta.read_pages_glob {
+                                glob_pattern = Some(val);
+                            }
+                            self.create_site_pages_index(glob_pattern, &tpl)?
                         } else {
                             BTreeMap::new()
                         }
@@ -249,6 +282,32 @@ impl<'a> Renderer<'a> {
     fn register_helpers(&mut self) {
         handlebars_sprig::addhelpers(&mut self.handlebars)
     }
+    fn create_site_pages_index(
+        &self,
+        glob_pattern: Option<&Vec<String>>,
+        tpl: &str,
+    ) -> anyhow::Result<BTreeMap<String, PageValues>> {
+        match glob_pattern {
+            Some(pattern) => {
+                let pages = crate::content::get_pages_by_glob(
+                    self.content_dir.clone(),
+                    pattern,
+                    self.show_unpublished,
+                );
+                match pages {
+                    Ok(val) => Ok(val),
+                    Err(err) => {
+                        bail!("Error parsing glob in template \"{tpl}\": {err}")
+                    }
+                }
+            }
+            None => crate::content::all_pages(
+                self.content_dir.clone(),
+                self.show_unpublished,
+                self.disable_cache,
+            ),
+        }
+    }
 }
 
 /*
@@ -295,3 +354,27 @@ pub fn error_values(title: &str, msg: &str) -> PageValues {
         published: true,
     }
 }
+
+fn read_file(file_path: &str) -> anyhow::Result<String> {
+    let mut file = File::open(file_path)?;
+    let mut contents = String::new();
+    file.read_to_string(&mut contents)?;
+    Ok(contents)
+}
+
+#[cfg(feature = "server")]
+fn parse_hbs_template(entry: walkdir::DirEntry) -> anyhow::Result<(TemplateMeta, String)> {
+    let path = entry.path();
+    match read_file(path.to_str().unwrap()) {
+        Ok(contents) => {
+            let doc = contents.replace("\r\n", "\n");
+            let (toml_text, body) = doc.split_once("\n---\n").unwrap_or(("", &doc));
+            let toml_text = toml_text.trim_start_matches("---").trim();
+            let template_meta = toml::from_str(toml_text)?;
+            Ok((template_meta, body.to_owned()))
+        }
+        Err(err) => {
+            bail!("Failed to parse hbs template \"entry\": {err}")
+        }
+    }
+}