Add -update and -dry-run options to hashtags command

hashtags_cmd.go

@@ -5,15 +5,26 @@ import (
 	"flag"
 	"fmt"
 	"github.com/google/subcommands"
+	"github.com/gomarkdown/markdown"
+	"github.com/gomarkdown/markdown/ast"
+	"github.com/hexops/gotextdiff"
+	"github.com/hexops/gotextdiff/myers"
+	"github.com/hexops/gotextdiff/span"
 	"io"
 	"os"
+	"regexp"
 	"sort"
+	"strings"
 )
 
 type hashtagsCmd struct {
+	update bool
+	dryRun bool
 }
 
 func (cmd *hashtagsCmd) SetFlags(f *flag.FlagSet) {
+	f.BoolVar(&cmd.update, "update", false, "create and update hashtag pages")
+	f.BoolVar(&cmd.dryRun, "dry-run", false, "only report the changes it would make")
 }
 
 func (*hashtagsCmd) Name() string     { return "hashtags" }
@@ -25,6 +36,9 @@ func (*hashtagsCmd) Usage() string {
 }
 
 func (cmd *hashtagsCmd) Execute(_ context.Context, f *flag.FlagSet, _ ...interface{}) subcommands.ExitStatus {
+	if cmd.update {
+		return hashtagsUpdateCli(os.Stdout, cmd.dryRun)
+	}
 	return hashtagsCli(os.Stdout)
 }
 
@@ -57,3 +71,138 @@ func hashtagsCli(w io.Writer) subcommands.ExitStatus {
 
 	return subcommands.ExitSuccess
 }
+
+// hashtagsUpdateCli runs the hashtags command on the command line and creates and updates the hashtag pages in the
+// current directory. That is, pages in subdirectories are skipped! It is used here with an io.Writer for easy testing.
+func hashtagsUpdateCli(w io.Writer, dryRun bool) subcommands.ExitStatus {
+	index.load()
+	// no locking necessary since this is for the command-line
+	namesMap := make(map[string]string)
+	for hashtag, docids := range index.token {
+		if len(docids) <= 5 {
+			if dryRun {
+				fmt.Fprintf(w, "Skipping #%s because there are not enough entries (%d)\n", hashtag, len(docids))
+			}
+			continue
+		}
+		title, ok := namesMap[hashtag]
+		if (!ok) {
+			title = hashtagName(namesMap, hashtag, docids)
+			namesMap[hashtag] = title
+		}
+		pageName := strings.ReplaceAll(title, " ", "_")
+		h, err := loadPage(pageName)
+		original := ""
+		new := false
+		if err != nil {
+			new = true
+			h = &Page{Name: pageName, Body: []byte("# " + title + "\n\n#" + pageName + "\n\nBlog posts:\n\n")}
+		} else {
+			original = string(h.Body)
+		}
+		for _, docid := range docids {
+			name := index.documents[docid]
+			if strings.Contains(name, "/") {
+				continue
+			}
+			p, err := loadPage(name)
+			if err != nil {
+				fmt.Fprintf(w, "Loading %s: %s\n", name, err)
+				return subcommands.ExitFailure
+			}
+			if !p.IsBlog() {
+				continue
+			}
+			p.handleTitle(false)
+			if p.Title == "" {
+				p.Title = p.Name
+			}
+			esc := nameEscape(p.Base())
+			link := "* [" + p.Title + "](" + esc + ")\n"
+			// I guess & used to get escaped and now no longer does
+			re := regexp.MustCompile(`(?m)^\* \[[^\]]+\]\(` + strings.ReplaceAll(esc, "&", "(&|%26)") + `\)\n`)
+			addLinkToPage(h, link, re)
+		}
+		// only save if something changed
+		if string(h.Body) != original {
+			if dryRun {
+				if new {
+					fmt.Fprintf(w, "Creating %s.md\n", title)
+				} else {
+					fmt.Fprintf(w, "Updating %s.md\n", title)
+				}
+				fn := h.Name + ".md"
+				edits := myers.ComputeEdits(span.URIFromPath(fn), original, string(h.Body))
+				diff := fmt.Sprint(gotextdiff.ToUnified(fn + "~", fn, original, edits))
+				fmt.Fprint(w, diff)
+			} else {
+				err = h.save()
+				if err != nil {
+					fmt.Fprintf(w, "Saving hashtag %s failed: %s", hashtag, err)
+					return subcommands.ExitFailure
+				}
+			}
+		}
+	}
+	return subcommands.ExitSuccess
+}
+
+// Go through all the documents in the same directory and look for hashtag matches in the rendered HTML in order to
+// determine the most likely capitalization.
+func hashtagName (namesMap map[string]string, hashtag string, docids []docid) string {
+	candidate := make(map[string]int)
+	var mostPopular string
+	for _, docid := range docids {
+		name := index.documents[docid]
+		if strings.Contains(name, "/") {
+			continue
+		}
+		p, err := loadPage(name)
+		if err != nil {
+			continue
+		}
+		// parsing finds all the hashtags
+		parser, _ := wikiParser()
+		doc := markdown.Parse(p.Body, parser)
+		ast.WalkFunc(doc, func(node ast.Node, entering bool) ast.WalkStatus {
+			if entering {
+				switch v := node.(type) {
+				case *ast.Link:
+					for _, attr := range v.AdditionalAttributes {
+						if attr == `class="tag"` {
+							tagName := []byte("")
+							ast.WalkFunc(v, func(node ast.Node, entering bool) ast.WalkStatus {
+								if entering && node.AsLeaf() != nil {
+									tagName = append(tagName, node.AsLeaf().Literal...)
+								}
+								return ast.GoToNext
+							})
+							tag := string(tagName[1:])
+							if strings.EqualFold(hashtag, strings.ReplaceAll(tag, " ", "_")) {
+								_, ok := candidate[tag]
+								if ok {
+									candidate[tag] += 1
+								} else {
+									candidate[tag] = 1
+								}
+							}
+						}
+					}
+				}
+			}
+			return ast.GoToNext
+		})
+		count := 0
+		for key, val := range candidate {
+			if val > count {
+				mostPopular = key
+				count = val
+			}
+		}
+		// shortcut
+		if count >= 5 {
+			return mostPopular
+		}
+	}
+	return mostPopular
+}

man/oddmu-hashtags.1

@@ -5,20 +5,31 @@
 .nh
 .ad l
 .\" Begin generated content:
-.TH "ODDMU-HASHTAGS" "1" "2024-08-29"
+.TH "ODDMU-HASHTAGS" "1" "2025-08-09"
 .PP
 .SH NAME
 .PP
-oddmu-hashtags - count the hashtags used
+oddmu-hashtags - work with hashtags
 .PP
 .SH SYNOPSIS
 .PP
 \fBoddmu hashtags\fR
 .PP
+\fBoddmu hashtags -update\fR [\fB-dry-run\fR]
+.PP
 .SH DESCRIPTION
 .PP
-The "hashtags" subcommand counts all the hashtags used and lists them, separated
-by a TAB character.\&
+By default, the "hashtags" subcommand counts all the hashtags used and lists
+them, separated by a TAB character.\&
+.PP
+With the \fB-update\fR flag, the hashtag pages are update with links to all the blog
+pages having the corresponding tag.\& This only necessary when migrating a
+collection of Markdown files.\& Ordinarily, Oddmu maintains the hashtag pages
+automatically.\& When writing pages offline, use \fIoddmu-notify\fR(1) to update the
+hashtag pages.\&
+.PP
+Use the \fB-dry-run\fR flag to see what would change with the \fB-update\fR flag without
+actually changing any files.\&
 .PP
 .SH EXAMPLES
 .PP
@@ -30,6 +41,22 @@ oddmu hashtags | head -n 11
 .fi
 .RE
 .PP
+See what kind of changes Oddmu would suggest:
+.PP
+.nf
+.RS 4
+oddmu hashtags -update -dry-run
+.fi
+.RE
+.PP
+And then do it:
+.PP
+.nf
+.RS 4
+oddmu hashtags -update
+.fi
+.RE
+.PP
 .SH SEE ALSO
 .PP
 \fIoddmu\fR(1)

man/oddmu-hashtags.1.txt

@@ -2,16 +2,27 @@ ODDMU-HASHTAGS(1)
 
 # NAME
 
-oddmu-hashtags - count the hashtags used
+oddmu-hashtags - work with hashtags
 
 # SYNOPSIS
 
 *oddmu hashtags*
 
+*oddmu hashtags -update* [*-dry-run*]
+
 # DESCRIPTION
 
-The "hashtags" subcommand counts all the hashtags used and lists them, separated
-by a TAB character.
+By default, the "hashtags" subcommand counts all the hashtags used and lists
+them, separated by a TAB character.
+
+With the *-update* flag, the hashtag pages are update with links to all the blog
+pages having the corresponding tag. This only necessary when migrating a
+collection of Markdown files. Ordinarily, Oddmu maintains the hashtag pages
+automatically. When writing pages offline, use _oddmu-notify_(1) to update the
+hashtag pages.
+
+Use the *-dry-run* flag to see what would change with the *-update* flag without
+actually changing any files.
 
 # EXAMPLES
 
@@ -21,6 +32,18 @@ List the top 10 hashtags. This requires 11 lines because of the header line.
 oddmu hashtags | head -n 11
 ```
 
+See what kind of changes Oddmu would suggest:
+
+```
+oddmu hashtags -update -dry-run
+```
+
+And then do it:
+
+```
+oddmu hashtags -update
+```
+
 # SEE ALSO
 
 _oddmu_(1)

man/oddmu.1

@@ -5,7 +5,7 @@
 .nh
 .ad l
 .\" Begin generated content:
-.TH "ODDMU" "1" "2025-07-16"
+.TH "ODDMU" "1" "2025-08-09"
 .PP
 .SH NAME
 .PP
@@ -384,7 +384,7 @@ Oddmu running as a webserver:
 .PP
 .PD 0
 .IP \(bu 4
-\fIoddmu-hashtags\fR(1), on how to count the hashtags used
+\fIoddmu-hashtags\fR(1), on working with hashtags
 .IP \(bu 4
 \fIoddmu-html\fR(1), on how to render a page
 .IP \(bu 4

man/oddmu.1.txt

@@ -315,7 +315,7 @@ If you run Oddmu as a web server:
 If you run Oddmu as a static site generator or pages offline and sync them with
 Oddmu running as a webserver:
 
-- _oddmu-hashtags_(1), on how to count the hashtags used
+- _oddmu-hashtags_(1), on working with hashtags
 - _oddmu-html_(1), on how to render a page
 - _oddmu-list_(1), on how to list pages and titles
 - _oddmu-links_(1), on how to list the outgoing links for a page