southfox/miniflux
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions

Calculate reading time during feed processing

Browse source 
The goal is to speed up the user interface.

Detecting the language based on the content is pretty slow.
This commit is contained in:
Frédéric Guillot 2020-11-18 17:29:40 -08:00 committed by fguillot
parent b1c9977711
commit de7a613098
12 changed files with 84 additions and 50 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
client/core.go
View file

@ -141,6 +141,7 @@ type Entry struct {
Author string `json:"author"`
ShareCode string `json:"share_code"`
Starred bool `json:"starred"`
ReadingTime int `json:"reading_time"`
Enclosures Enclosures `json:"enclosures,omitempty"`
Feed *Feed `json:"feed,omitempty"`
}

2
database/migration.go
View file

@ -12,7 +12,7 @@ import (
"miniflux.app/logger"
)
const schemaVersion = 40
const schemaVersion = 41
// Migrate executes database migrations.
func Migrate(db *sql.DB) {

2
database/sql.go
View file

@ -203,6 +203,7 @@ alter table users add column entry_direction entry_sorting_direction default 'as
add column keeplist_rules text not null default ''
;
`,
"schema_version_41": `alter table entries add column reading_time int not null default 0;`,
"schema_version_5": `create table integrations (
user_id int not null,
pinboard_enabled bool default 'f',
@ -264,6 +265,7 @@ var SqlMapChecksums = map[string]string{
"schema_version_39": "b0f90b97502921d4681a07c64d180a91a0b4ccac7d3c1dbe30519ad6f1bf1737",
"schema_version_4": "216ea3a7d3e1704e40c797b5dc47456517c27dbb6ca98bf88812f4f63d74b5d9",
"schema_version_40": "6a8fec92399f853ed6817aff4cfa43255dce4c19afad796e41519d09de62105e",
"schema_version_41": "128e118ce61267ea1f6ae03b63a6d4734eae87e520b00e309ad083f1f6afdfe5",
"schema_version_5": "46397e2f5f2c82116786127e9f6a403e975b14d2ca7b652a48cd1ba843e6a27c",
"schema_version_6": "9d05b4fb223f0e60efc716add5048b0ca9c37511cf2041721e20505d6d798ce4",
"schema_version_7": "33f298c9aa30d6de3ca28e1270df51c2884d7596f1283a75716e2aeb634cd05c",

1
database/sql/schema_version_41.sql Normal file
View file

@ -0,0 +1 @@
alter table entries add column reading_time int not null default 0;

1
model/entry.go
View file

@ -33,6 +33,7 @@ type Entry struct {
Author string `json:"author"`
ShareCode string `json:"share_code"`
Starred bool `json:"starred"`
ReadingTime int `json:"reading_time"`
Enclosures EnclosureList `json:"enclosures,omitempty"`
Feed *Feed `json:"feed,omitempty"`
}

22
reader/processor/processor.go
View file

@ -5,8 +5,11 @@
package processor
import (
"math"
"regexp"
"strings"
"time"
"unicode/utf8"
"miniflux.app/config"
"miniflux.app/logger"
@ -16,6 +19,8 @@ import (
"miniflux.app/reader/sanitizer"
"miniflux.app/reader/scraper"
"miniflux.app/storage"
"github.com/rylans/getlang"
)
// ProcessFeedEntries downloads original web page for entries and apply filters.
@ -58,6 +63,7 @@ func ProcessFeedEntries(store *storage.Storage, feed *model.Feed) {
// The sanitizer should always run at the end of the process to make sure unsafe HTML is filtered.
entry.Content = sanitizer.Sanitize(entry.URL, entry.Content)
entry.ReadingTime = calculateReadingTime(entry.Content)
filteredEntries = append(filteredEntries, entry)
}
@ -108,7 +114,23 @@ func ProcessEntryWebPage(entry *model.Entry) error {
if content != "" {
entry.Content = content
entry.ReadingTime = calculateReadingTime(content)
}
return nil
}
func calculateReadingTime(content string) int {
sanitizedContent := sanitizer.StripTags(content)
languageInfo := getlang.FromString(sanitizedContent)
var timeToReadInt int
if languageInfo.LanguageCode() == "ko" || languageInfo.LanguageCode() == "zh" || languageInfo.LanguageCode() == "jp" {
timeToReadInt = int(math.Ceil(float64(utf8.RuneCountInString(sanitizedContent)) / 500))
} else {
nbOfWords := len(strings.Fields(sanitizedContent))
timeToReadInt = int(math.Ceil(float64(nbOfWords) / 265))
}
return timeToReadInt
}

41
storage/entry.go
View file

@ -75,11 +75,11 @@ func (s *Storage) UpdateEntryContent(entry *model.Entry) error {
UPDATE
entries
SET
content=$1
content=$1, reading_time=$2
WHERE
id=$2 AND user_id=$3
id=$3 AND user_id=$4
`
_, err = tx.Exec(query, entry.Content, entry.ID, entry.UserID)
_, err = tx.Exec(query, entry.Content, entry.ReadingTime, entry.ID, entry.UserID)
if err != nil {
tx.Rollback()
return fmt.Errorf(`store: unable to update content of entry #%d: %v`, entry.ID, err)
@ -106,9 +106,35 @@ func (s *Storage) UpdateEntryContent(entry *model.Entry) error {
func (s *Storage) createEntry(tx *sql.Tx, entry *model.Entry) error {
query := `
INSERT INTO entries
(title, hash, url, comments_url, published_at, content, author, user_id, feed_id, changed_at, document_vectors)
(
title,
hash,
url,
comments_url,
published_at,
content,
author,
user_id,
feed_id,
reading_time,
changed_at,
document_vectors
)
VALUES
($1, $2, $3, $4, $5, $6, $7, $8, $9, now(), setweight(to_tsvector(substring(coalesce($1, '') for 1000000)), 'A') || setweight(to_tsvector(substring(coalesce($6, '') for 1000000)), 'B'))
(
$1,
$2,
$3,
$4,
$5,
$6,
$7,
$8,
$9,
$10,
now(),
setweight(to_tsvector(substring(coalesce($1, '') for 1000000)), 'A') || setweight(to_tsvector(substring(coalesce($6, '') for 1000000)), 'B')
)
RETURNING
id, status
`
@ -123,6 +149,7 @@ func (s *Storage) createEntry(tx *sql.Tx, entry *model.Entry) error {
entry.Author,
entry.UserID,
entry.FeedID,
entry.ReadingTime,
).Scan(&entry.ID, &entry.Status)
if err != nil {
@ -154,9 +181,10 @@ func (s *Storage) updateEntry(tx *sql.Tx, entry *model.Entry) error {
comments_url=$3,
content=$4,
author=$5,
reading_time=$6,
document_vectors = setweight(to_tsvector(substring(coalesce($1, '') for 1000000)), 'A') || setweight(to_tsvector(substring(coalesce($4, '') for 1000000)), 'B')
WHERE
user_id=$6 AND feed_id=$7 AND hash=$8
user_id=$7 AND feed_id=$8 AND hash=$9
RETURNING
id
`
@ -167,6 +195,7 @@ func (s *Storage) updateEntry(tx *sql.Tx, entry *model.Entry) error {
entry.CommentsURL,
entry.Content,
entry.Author,
entry.ReadingTime,
entry.UserID,
entry.FeedID,
entry.Hash,

2
storage/entry_query_builder.go
View file

@ -226,6 +226,7 @@ func (e *EntryQueryBuilder) GetEntries() (model.Entries, error) {
e.content,
e.status,
e.starred,
e.reading_time,
f.title as feed_title,
f.feed_url,
f.site_url,
@ -284,6 +285,7 @@ func (e *EntryQueryBuilder) GetEntries() (model.Entries, error) {
&entry.Content,
&entry.Status,
&entry.Starred,
&entry.ReadingTime,
&entry.Feed.Title,
&entry.Feed.FeedURL,
&entry.Feed.SiteURL,

6
template/common.go
View file

@ -242,10 +242,10 @@ SOFTWARE.
<li>
<time datetime="{{ isodate .entry.Date }}" title="{{ isodate .entry.Date }}">{{ elapsed .user.Timezone .entry.Date }}</time>
</li>
{{ if .user.ShowReadingTime }}
{{ if and .user.ShowReadingTime (gt .entry.ReadingTime 0) }}
<li>
<span>
{{ plural "entry.estimated_reading_time" (timeToRead .entry.Content) (timeToRead .entry.Content) }}
{{ plural "entry.estimated_reading_time" .entry.ReadingTime .entry.ReadingTime }}
</span>
</li>
{{ end }}
@ -523,7 +523,7 @@ var templateCommonMapChecksums = map[string]string{
"feed_list": "931e43d328a116318c510de5658c688cd940b934c86b6ec82a472e1f81e020ae",
"feed_menu": "318d8662dda5ca9dfc75b909c8461e79c86fb5082df1428f67aaf856f19f4b50",
"icons": "9a41753778072f286216085d8712495e2ccca20c7a24f5c982775436a3d38579",
"item_meta": "eb72c6e2a924759af20b8ef41f2ce7495aedc053181c2e5ca1b063f9410c58b0",
"item_meta": "56ab09d7dd46eeb2e2ee11ddcec0c157a5832c896dbd2887d9e2b013680b2af6",
"layout": "65767e7dbebe1f7ed42895ecd5a737b0693e4a2ec35e84e3e391f462beb11977",
"pagination": "7b61288e86283c4cf0dc83bcbf8bf1c00c7cb29e60201c8c0b633b2450d2911f",
"settings_menu": "e2b777630c0efdbc529800303c01d6744ed3af80ec505ac5a5b3f99c9b989156",

3
template/engine.go
View file

@ -65,9 +65,6 @@ func (e *Engine) Render(name, language string, data interface{}) []byte {
"plural": func(key string, n int, args ...interface{}) string {
return printer.Plural(key, n, args...)
},
"timeToRead": func(content string) int {
return timeToRead(content)
},
})
var b bytes.Buffer

21
template/functions.go
View file

@ -11,19 +11,16 @@ import (
"net/mail"
"strings"
"time"
"unicode/utf8"
"miniflux.app/config"
"miniflux.app/http/route"
"miniflux.app/locale"
"miniflux.app/model"
"miniflux.app/proxy"
"miniflux.app/reader/sanitizer"
"miniflux.app/timezone"
"miniflux.app/url"
"github.com/gorilla/mux"
"github.com/rylans/getlang"
)
type funcMap struct {
@ -94,9 +91,6 @@ func (f *funcMap) Map() template.FuncMap {
"plural": func(key string, n int, args ...interface{}) string {
return ""
},
"timeToRead": func(content string) int {
return 0
},
}
}
@ -195,18 +189,3 @@ func formatFileSize(b int64) string {
return fmt.Sprintf("%.1f %ciB",
float64(b)/float64(div), "KMGTPE"[exp])
}
func timeToRead(content string) int {
sanitizedContent := sanitizer.StripTags(content)
languageInfo := getlang.FromString(sanitizedContent)
var timeToReadInt int
if languageInfo.LanguageCode() == "ko" || languageInfo.LanguageCode() == "zh" || languageInfo.LanguageCode() == "jp" {
timeToReadInt = int(math.Ceil(float64(utf8.RuneCountInString(sanitizedContent)) / 500))
} else {
nbOfWords := len(strings.Fields(sanitizedContent))
timeToReadInt = int(math.Ceil(float64(nbOfWords) / 265))
}
return timeToReadInt
}

4
template/html/common/item_meta.html
View file

@ -7,10 +7,10 @@
<li>
<time datetime="{{ isodate .entry.Date }}" title="{{ isodate .entry.Date }}">{{ elapsed .user.Timezone .entry.Date }}</time>
</li>
{{ if .user.ShowReadingTime }}
{{ if and .user.ShowReadingTime (gt .entry.ReadingTime 0) }}
<li>
<span>
{{ plural "entry.estimated_reading_time" (timeToRead .entry.Content) (timeToRead .entry.Content) }}
{{ plural "entry.estimated_reading_time" .entry.ReadingTime .entry.ReadingTime }}
</span>
</li>
{{ end }}