Document extractor service (#15665)
* Document extractor service * Fixing vendor modules * Addressing PR Review comments * Some small simplifications * Fixing a linter complain * simplifying a bit the code using package variables Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
Этот коммит содержится в:
коммит произвёл
GitHub
родитель
04ef5c682e
Коммит
8d5be2d657
98
vendor/code.sajari.com/docconv/xml.go
сгенерированный
поставляемый
Обычный файл
98
vendor/code.sajari.com/docconv/xml.go
сгенерированный
поставляемый
Обычный файл
@@ -0,0 +1,98 @@
|
||||
package docconv
|
||||
|
||||
import (
|
||||
"bytes"
|
||||
"encoding/xml"
|
||||
"fmt"
|
||||
"io"
|
||||
)
|
||||
|
||||
// ConvertXML converts an XML file to text.
|
||||
func ConvertXML(r io.Reader) (string, map[string]string, error) {
|
||||
meta := make(map[string]string)
|
||||
cleanXML, err := Tidy(r, true)
|
||||
if err != nil {
|
||||
return "", nil, fmt.Errorf("tidy error: %v", err)
|
||||
}
|
||||
result, err := XMLToText(bytes.NewReader(cleanXML), []string{}, []string{}, true)
|
||||
if err != nil {
|
||||
return "", nil, fmt.Errorf("error from XMLToText: %v", err)
|
||||
}
|
||||
return result, meta, nil
|
||||
}
|
||||
|
||||
// XMLToText converts XML to plain text given how to treat elements.
|
||||
func XMLToText(r io.Reader, breaks []string, skip []string, strict bool) (string, error) {
|
||||
var result string
|
||||
|
||||
dec := xml.NewDecoder(r)
|
||||
dec.Strict = strict
|
||||
for {
|
||||
t, err := dec.Token()
|
||||
if err != nil {
|
||||
if err == io.EOF {
|
||||
break
|
||||
}
|
||||
return "", err
|
||||
}
|
||||
|
||||
switch v := t.(type) {
|
||||
case xml.CharData:
|
||||
result += string(v)
|
||||
case xml.StartElement:
|
||||
for _, breakElement := range breaks {
|
||||
if v.Name.Local == breakElement {
|
||||
result += "\n"
|
||||
}
|
||||
}
|
||||
for _, skipElement := range skip {
|
||||
if v.Name.Local == skipElement {
|
||||
depth := 1
|
||||
for {
|
||||
t, err := dec.Token()
|
||||
if err != nil {
|
||||
// An io.EOF here is actually an error.
|
||||
return "", err
|
||||
}
|
||||
|
||||
switch t.(type) {
|
||||
case xml.StartElement:
|
||||
depth++
|
||||
case xml.EndElement:
|
||||
depth--
|
||||
}
|
||||
|
||||
if depth == 0 {
|
||||
break
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
return result, nil
|
||||
}
|
||||
|
||||
// XMLToMap converts XML to a nested string map.
|
||||
func XMLToMap(r io.Reader) (map[string]string, error) {
|
||||
m := make(map[string]string)
|
||||
dec := xml.NewDecoder(r)
|
||||
var tagName string
|
||||
for {
|
||||
t, err := dec.Token()
|
||||
if err != nil {
|
||||
if err == io.EOF {
|
||||
break
|
||||
}
|
||||
return nil, err
|
||||
}
|
||||
|
||||
switch v := t.(type) {
|
||||
case xml.StartElement:
|
||||
tagName = string(v.Name.Local)
|
||||
case xml.CharData:
|
||||
m[tagName] = string(v)
|
||||
}
|
||||
}
|
||||
return m, nil
|
||||
}
|
||||
Ссылка в новой задаче
Block a user