Files
forge/modules/sitemap/sitemap.go
zeekayandHanzo Dev 9a07b4d089 legal: Hanzo AI, Inc. copyright above the upstream notices
2978 files. The fork is 2026-07-15 (rebrand: gitea.dev -> github.com/hanzoai/git),
so 2026 is the year, and "Copyright 2026 Hanzo AI, Inc." was already the form
used by the handful of files that had ours.

Shape is now:

	// Copyright 2026 Hanzo AI, Inc. All rights reserved.
	// Copyright 2015 The Gogs Authors. All rights reserved.
	// Copyright 2019 The Gitea Authors. All rights reserved.
	// SPDX-License-Identifier: MIT

Ours on top, theirs retained verbatim. Retained because MIT requires it, and
separately because they are true — those authors did write that code.

Not renamed to "The Git Authors": Git is a different project under a different
license, and putting its name here would attribute this code to people who have
nothing to do with it. Adding our line is the fix; overwriting theirs would have
been both a licence violation and a false statement.

gofmt still reports import-order drift in some of these files. It predates this
change (verified against a stash) and is left alone rather than burying a
one-line-per-file legal change under 3000 files of unrelated reordering.

Build clean.

Co-authored-by: Hanzo Dev <[email protected]>
2026-07-27 13:49:52 -07:00

88 lines
2.1 KiB
Go

// Copyright 2026 Hanzo AI, Inc. All rights reserved.
// Copyright 2022 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package sitemap
import (
"bytes"
"encoding/xml"
"fmt"
"io"
"time"
)
const (
sitemapFileLimit = 50 * 1024 * 1024 // the maximum size of a sitemap file
urlsLimit = 50000
schemaURL = "http://www.sitemaps.org/schemas/sitemap/0.9"
urlsetName = "urlset"
sitemapindexName = "sitemapindex"
)
// URL represents a single sitemap entry
type URL struct {
URL string `xml:"loc"`
LastMod *time.Time `xml:"lastmod,omitempty"`
}
// Sitemap represents a sitemap
type Sitemap struct {
XMLName xml.Name
Namespace string `xml:"xmlns,attr"`
URLs []URL `xml:"url"`
Sitemaps []URL `xml:"sitemap"`
}
// NewSitemap creates a sitemap
func NewSitemap() *Sitemap {
return &Sitemap{
XMLName: xml.Name{Local: urlsetName},
Namespace: schemaURL,
}
}
// NewSitemapIndex creates a sitemap index.
func NewSitemapIndex() *Sitemap {
return &Sitemap{
XMLName: xml.Name{Local: sitemapindexName},
Namespace: schemaURL,
}
}
// Add adds a URL to the sitemap
func (s *Sitemap) Add(u URL) {
if s.XMLName.Local == sitemapindexName {
s.Sitemaps = append(s.Sitemaps, u)
} else {
s.URLs = append(s.URLs, u)
}
}
// WriteTo writes the sitemap to a response
func (s *Sitemap) WriteTo(w io.Writer) (int64, error) {
if l := len(s.URLs); l > urlsLimit {
return 0, fmt.Errorf("sitemap contains %d URLs, but only %d are allowed", l, urlsLimit)
}
if l := len(s.Sitemaps); l > urlsLimit {
return 0, fmt.Errorf("sitemap contains %d sub-sitemaps, but only %d are allowed", l, urlsLimit)
}
buf := bytes.NewBufferString(xml.Header)
encoder := xml.NewEncoder(buf)
defer encoder.Close()
if err := encoder.Encode(s); err != nil {
return 0, err
}
_ = encoder.Flush()
if err := buf.WriteByte('\n'); err != nil {
return 0, err
}
// FIXME: such limit is not right, the content has been written, it would have already caused OOM
if buf.Len() > sitemapFileLimit {
return 0, fmt.Errorf("sitemap has %d bytes, but only %d are allowed", buf.Len(), sitemapFileLimit)
}
return buf.WriteTo(w)
}