mirror of
https://codeberg.org/scip/epuppy.git
synced 2025-12-17 12:31:02 +01:00
b50c6acff0
- Use antchfx/xmlquery for easier XML parsing. No more regexp wrangling and the result is much more reliable over a variety of ebooks. Much good. - fix chapter selection, look for `<?xml[...]` which is much more reliable - add option `-x` to dump the XML ebook source for debugging
58 lines
1.2 KiB
Go
58 lines
1.2 KiB
Go
package epub
|
||
|
||
import (
|
||
"regexp"
|
||
"strings"
|
||
|
||
"github.com/antchfx/xmlquery"
|
||
)
|
||
|
||
var (
|
||
cleanentitles = regexp.MustCompile(`&.+;`)
|
||
empty = regexp.MustCompile(`(?s)^[\s ]*$`)
|
||
newlines = regexp.MustCompile(`[\r\n]+`)
|
||
)
|
||
|
||
// Content nav-point content
|
||
type Content struct {
|
||
Src string `xml:"src,attr" json:"src"`
|
||
Empty bool
|
||
Body string
|
||
Title string
|
||
XML []byte
|
||
}
|
||
|
||
func (c *Content) String(content []byte) error {
|
||
// parse XML, look for title and <p>.*</p> stuff
|
||
doc, err := xmlquery.Parse(
|
||
strings.NewReader(
|
||
cleanentitles.ReplaceAllString(string(content), " ")))
|
||
if err != nil {
|
||
panic(err)
|
||
}
|
||
|
||
// extract the title
|
||
for _, item := range xmlquery.Find(doc, "//title") {
|
||
c.Title = strings.TrimSpace(item.InnerText())
|
||
}
|
||
|
||
// extract all paragraphs, ignore any formatting and re-fill the
|
||
// paragraph, that is, we replaces all newlines inside with one
|
||
// space.
|
||
txt := strings.Builder{}
|
||
for _, item := range xmlquery.Find(doc, "//p") {
|
||
if !empty.MatchString(item.InnerText()) {
|
||
txt.WriteString(newlines.ReplaceAllString(item.InnerText(), " ") + "\n\n")
|
||
}
|
||
}
|
||
|
||
c.Body = strings.TrimSpace(txt.String())
|
||
c.XML = content
|
||
|
||
if len(c.Body) == 0 {
|
||
c.Empty = true
|
||
}
|
||
|
||
return nil
|
||
}
|