improve docs and adapt hugo changes

README.md

@@ -1,20 +1,21 @@
 # hugoext
 
-Utility to parse the hugo config file and create the same file structure for content through an
+Utility to parse the hugo config file and recreate the same file structure for content files through
-arbitrary output pipe extension.
+an arbitrary output pipe extension for processing.
 
 Hugo parses primarily markdown files and go templates. The initial motivation for this utility was
-to enable the same tools to publish a gemlog version of the same blog to make it accessible as
+to enable the same tools to publish a Gemlog version of a hugo blog to make it accessible as Gemtext
-gemtext via the Gemini protocol alongside the web html version.
+via the Gemini protocol.
 
-**NOTE**: this is minimal and only has one use case for now, many edge cases may not be covered.
+**NOTE**: this is rather minimal and only has one use case for now, many edge cases may not be covered.
 
 Features
 - reads hugo `.toml` file for section output formats
 - supports an arbitrary document processor, any program that supports UNIX pipes
 - ugly urls
 - section listings
-- with drafts
+- supports drafts
+- composable with other tools
 
 To illustrate what this program does, run the following in the hugo directory.
 
@@ -89,3 +90,11 @@ publish: build
 
 The output directory for both hugo and hugoext is `./public`. It's ok to mix the two into the same
 file tree as each directory will contain an `index.html` and an `index.gmi` file.
+
+## Acknowledgements & Attribution
+
+For config parsing and compatibility, this repo uses the latest hugo source tree as the functions are all
+exported. For markdown and metadata parsing of hugo content files, I've extracted some code from
+[hugo@v0.49.2](https://github.com/gohugoio/hugo/tree/v0.49.2/) tree and made them importable in the
+local [hugo](./hugo) package. The version had required frontmatter, metadata and section permalink
+parsers still available in well isolated functions, so I didn't need to recreate them.

file_tree.go

@@ -20,44 +20,46 @@ type File struct {
 	Parent      string
 	Name        string
 	Extension   string
+	Draft       bool
 
-	Draft bool
+	Metadata hugo.PageMetadata
-
+	Body     []byte
-	Body    []byte
+	NewBody  []byte
-	NewBody []byte
 }
 
-func parse(fullpath string) ([]byte, *hugo.Content, error) {
+func parsePage(fullpath string) (hugo.Page, error) {
 	file, err := os.Open(fullpath)
 	if err != nil {
-		return nil, nil, err
+		return nil, err
 	}
 	defer file.Close()
 
 	page, err := hugo.ReadFrom(file)
 	if err != nil {
-		return nil, nil, err
+		return nil, err
 	}
+	return page, nil
+}
+
+func parseMetadata(page hugo.Page) (*hugo.PageMetadata, error) {
 	meta, err := page.Metadata()
 	if err != nil {
-		return nil, nil, err
+		return nil, err
 	}
 	c := NewContentFromMeta(meta)
-	body := page.FrontMatter()
-	body = append(body, '\n')
-	body = append(body, page.Content()...)
 
-	return body, c, nil
+	return c, nil
 }
 
 func destinationPath(file *File, pattern string) error {
-	body, c, err := parse(file.Source)
+	p, err := parsePage(file.Source)
 	if err != nil {
 		return err
 	}
+
+	// create content
+	c, err := parseMetadata(p)
 	c.Filepath = file.Name
-	file.Body = body
-	file.Draft = c.Draft
 
 	if file.Parent != "." {
 		link, err := hugo.PathPattern(pattern).Expand(c)
@@ -69,6 +71,10 @@ func destinationPath(file *File, pattern string) error {
 		file.Destination = strings.TrimLeft(file.Name, "_")
 	}
 
+	file.Draft = c.Draft
+	file.Metadata = *c
+	file.Body = p.Body()
+
 	return nil
 }
 

page_metadata.go

@@ -6,8 +6,8 @@ import (
 	"github.com/n0x1m/hugoext/hugo"
 )
 
-func NewContentFromMeta(meta map[string]interface{}) *hugo.Content {
+func NewContentFromMeta(meta map[string]interface{}) *hugo.PageMetadata {
-	return &hugo.Content{
+	return &hugo.PageMetadata{
 		Title:      stringFromInterface(meta["title"]),
 		Slug:       stringFromInterface(meta["slug"]),
 		Summary:    stringFromInterface(meta["summary"]),