scip/epuppy
1
0
Fork 0
mirror of https://codeberg.org/scip/epuppy.git synced 2025-12-18 04:51:01 +01:00
Code Issues Packages Projects Releases Wiki Activity

fix XML parsing (#2)

Browse Source 
- 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
This commit is contained in:
T.v.Dein
2025-10-16 18:57:05 +02:00
committed by GitHub
parent 90d30cb3e1
commit b50c6acff0
13 changed files with 143 additions and 71 deletions
Download Patch File Download Diff File Expand all files Collapse all files

48
pkg/epub/content.go
View File

@@ -1,19 +1,16 @@
package epub
import (
"encoding/xml"
"fmt"
"regexp"
"strings"
"github.com/antchfx/xmlquery"
)
var (
cleantitle = regexp.MustCompile(`(?s)<head>.*</head>`)
cleanmarkup = regexp.MustCompile(`<[^<>]+>`)
cleanentities = regexp.MustCompile(`&.+;`)
cleancomments = regexp.MustCompile(`/*.*/`)
cleanspace = regexp.MustCompile(`^\s*`)
cleanh1 = regexp.MustCompile(`<h[1-6].*</h[1-6]>`)
cleanentitles = regexp.MustCompile(`&.+;`)
empty = regexp.MustCompile(`(?s)^[\s ]*$`)
newlines = regexp.MustCompile(`[\r\n]+`)
)
// Content nav-point content
@@ -26,25 +23,30 @@ type Content struct {
}
func (c *Content) String(content []byte) error {
title := Title{}
err := xml.Unmarshal(content, &title)
// parse XML, look for title and <p>.*</p> stuff
doc, err := xmlquery.Parse(
strings.NewReader(
cleanentitles.ReplaceAllString(string(content), " ")))
if err != nil {
if !strings.HasPrefix(err.Error(), "XML syntax error") {
return fmt.Errorf("XML parser error %w", err)
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.Title = strings.TrimSpace(title.Content)
txt := cleantitle.ReplaceAllString(string(content), "")
txt = cleanh1.ReplaceAllString(txt, "")
txt = cleanmarkup.ReplaceAllString(txt, "")
txt = cleanentities.ReplaceAllString(txt, " ")
txt = cleancomments.ReplaceAllString(txt, "")
txt = strings.TrimSpace(txt)
c.Body = cleanspace.ReplaceAllString(txt, "")
c.Body = strings.TrimSpace(txt.String())
c.XML = content
if len(c.Body) == 0 {