summaryrefslogtreecommitdiffci
path: root/internal
diff refs
from: back
to: back
| flip
diff options
context:
space:
mode:
Diffstat (limited to 'internal')
-rw-r--r--internal/blog/b2.go10
-rw-r--r--internal/blog/s3.go148
2 files changed, 149 insertions, 9 deletions
diff --git a/internal/blog/b2.go b/internal/blog/b2.go
index 8a4d363..0caa0ba 100644
--- a/internal/blog/b2.go
+++ b/internal/blog/b2.go
@@ -70,15 +70,17 @@ func (c *B2Client) Scan(prefix string) ([]*Page, error) {
return nil, fmt.Errorf("failed to parse published time metadata field: %w", err)
}
- link := obj.Name()
- fileName := link[strings.LastIndex(link, "/")+1 : strings.LastIndex(link, ".")]
+ linkParts := strings.Split(obj.Name(), "/")
+ nameParts := strings.Split(linkParts[len(linkParts)-1], ".")
+ fileName := strings.Join(nameParts[:len(linkParts)-1], ".")
+
tags := strings.Split(attrs.Info["tags"], ",")
slices.Sort(tags)
- lang, _ := strings.CutPrefix(link[0:strings.Index(link, "/")], c.prefix)
+ lang, _ := strings.CutPrefix(linkParts[0], c.prefix)
filePaths = append(filePaths, &Page{
- Link: link,
+ Link: obj.Name(),
FileName: fileName,
Lang: lang,
ModifiedTime: attrs.LastModified,
diff --git a/internal/blog/s3.go b/internal/blog/s3.go
index 5ee896a..bbd5238 100644
--- a/internal/blog/s3.go
+++ b/internal/blog/s3.go
@@ -3,9 +3,15 @@ package blog
import (
"context"
"encoding/json"
+ "errors"
"fmt"
"io"
+ "log/slog"
+ "net/url"
+ "slices"
"strings"
+ "sync"
+ "time"
"github.com/SayaAndy/saya-today-web/config"
"github.com/SayaAndy/saya-today-web/internal/frontmatter"
@@ -13,6 +19,7 @@ import (
awsconfig "github.com/aws/aws-sdk-go-v2/config"
"github.com/aws/aws-sdk-go-v2/credentials"
"github.com/aws/aws-sdk-go-v2/service/s3"
+ s3types "github.com/aws/aws-sdk-go-v2/service/s3/types"
)
const s3ScanConcurrency = 32
@@ -60,23 +67,38 @@ func NewS3Client(cfg *config.StorageConfig) (Client, error) {
}
func (c *S3Client) Scan(prefix string) ([]*Page, error) {
+ pages, err := c.scanFromIndex(prefix)
+ if err == nil {
+ return pages, nil
+ }
+
+ var nsk *s3types.NoSuchKey
+ if !errors.As(err, &nsk) {
+ return nil, err
+ }
+
+ slog.Warn("index.json missing, falling back to listing", slog.String("prefix", c.prefix))
+ return c.scanByListing(prefix)
+}
+
+func (c *S3Client) scanFromIndex(prefix string) ([]*Page, error) {
out, err := c.s3cl.GetObject(context.Background(), &s3.GetObjectInput{
Bucket: aws.String(c.bucketName),
Key: aws.String(IndexFileName),
})
if err != nil {
- return nil, fmt.Errorf("get %s: %w", IndexFileName, err)
+ return nil, fmt.Errorf("get index.json: %w", err)
}
defer out.Body.Close()
raw, err := io.ReadAll(out.Body)
if err != nil {
- return nil, fmt.Errorf("read %s: %w", IndexFileName, err)
+ return nil, fmt.Errorf("read index.json: %w", err)
}
var idx Index
if err := json.Unmarshal(raw, &idx); err != nil {
- return nil, fmt.Errorf("unmarshal %s: %w", IndexFileName, err)
+ return nil, fmt.Errorf("unmarshal index.json: %w", err)
}
wantLang := ""
@@ -98,7 +120,9 @@ func (c *S3Client) Scan(prefix string) ([]*Page, error) {
if !strings.HasPrefix(e.Link, fullPrefix) {
continue
}
- fileName := e.Link[strings.LastIndex(e.Link, "/")+1 : strings.LastIndex(e.Link, ".")]
+ linkParts := strings.Split(e.Link, "/")
+ nameParts := strings.Split(linkParts[len(linkParts)-1], ".")
+ fileName := strings.Join(nameParts[:len(nameParts)-1], ".")
pages = append(pages, &Page{
Link: e.Link,
FileName: fileName,
@@ -112,14 +136,128 @@ func (c *S3Client) Scan(prefix string) ([]*Page, error) {
Thumbnail: e.Thumbnail,
Tags: e.Tags,
Geolocation: e.Geolocation,
+ Medley: e.Medley,
+ MedleyPart: e.MedleyPart,
},
})
}
}
-
return pages, nil
}
+func (c *S3Client) scanByListing(prefix string) ([]*Page, error) {
+ fullPrefix := c.prefix + prefix
+
+ type candidate struct {
+ key string
+ lastModified time.Time
+ }
+ var candidates []candidate
+
+ paginator := s3.NewListObjectsV2Paginator(c.s3cl, &s3.ListObjectsV2Input{
+ Bucket: aws.String(c.bucketName),
+ Prefix: aws.String(fullPrefix),
+ })
+ for paginator.HasMorePages() {
+ output, err := paginator.NextPage(context.Background())
+ if err != nil {
+ return nil, fmt.Errorf("list S3 objects: %w", err)
+ }
+ for _, obj := range output.Contents {
+ key := aws.ToString(obj.Key)
+ if !strings.HasSuffix(key, ".md") {
+ continue
+ }
+ candidates = append(candidates, candidate{key, aws.ToTime(obj.LastModified)})
+ }
+ }
+
+ pages := make([]*Page, len(candidates))
+ sem := make(chan struct{}, s3ScanConcurrency)
+ var wg sync.WaitGroup
+ var firstErr error
+ var errMu sync.Mutex
+
+ for i, cand := range candidates {
+ sem <- struct{}{}
+ wg.Go(func() {
+ defer func() { <-sem }()
+
+ head, err := c.s3cl.HeadObject(context.Background(), &s3.HeadObjectInput{
+ Bucket: aws.String(c.bucketName),
+ Key: aws.String(cand.key),
+ })
+ if err != nil {
+ errMu.Lock()
+ if firstErr == nil {
+ firstErr = fmt.Errorf("head S3 object %s: %w", cand.key, err)
+ }
+ errMu.Unlock()
+ return
+ }
+
+ if head.ContentType == nil || !strings.Contains(*head.ContentType, "text/markdown") {
+ return
+ }
+ meta := head.Metadata
+ if meta["title"] == "" {
+ return
+ }
+
+ publishedTime, err := time.Parse(time.RFC3339, meta["published-time"])
+ if err != nil {
+ errMu.Lock()
+ if firstErr == nil {
+ firstErr = fmt.Errorf("failed to parse published time metadata field: %w", err)
+ }
+ errMu.Unlock()
+ return
+ }
+
+ linkParts := strings.Split(cand.key, "/")
+ nameParts := strings.Split(linkParts[len(linkParts)-1], ".")
+ fileName := strings.Join(nameParts[:len(nameParts)-1], ".")
+ lang, _ := strings.CutPrefix(linkParts[0], c.prefix)
+
+ tags := strings.Split(meta["tags"], ",")
+ slices.Sort(tags)
+
+ title, _ := url.QueryUnescape(meta["title"])
+ shortDescription, _ := url.QueryUnescape(meta["short-description"])
+ thumbnail, _ := url.QueryUnescape(meta["thumbnail"])
+
+ pages[i] = &Page{
+ Link: cand.key,
+ FileName: fileName,
+ Lang: lang,
+ ModifiedTime: cand.lastModified,
+ Metadata: &frontmatter.Metadata{
+ Title: title,
+ ShortDescription: shortDescription,
+ ActionDate: meta["action-date"],
+ PublishedTime: publishedTime,
+ Thumbnail: thumbnail,
+ Tags: tags,
+ Geolocation: meta["geolocation"],
+ },
+ }
+ })
+ }
+ wg.Wait()
+
+ if firstErr != nil {
+ return nil, firstErr
+ }
+
+ out := pages[:0]
+ for _, p := range pages {
+ if p != nil {
+ out = append(out, p)
+ }
+ }
+ return out, nil
+}
+
func (c *S3Client) ReadAll(path string) ([]byte, error) {
output, err := c.s3cl.GetObject(context.Background(), &s3.GetObjectInput{
Bucket: aws.String(c.bucketName),
generated by cgitext v1.3.1+17 (git 2.54.0) at 2026-10-22 10:23:29 +0000