summaryrefslogtreecommitdiff
path: root/lib/hocr/hocr.go
diff options
context:
space:
mode:
authorNick White <git@njw.name>2019-01-25 09:55:55 +0000
committerNick White <git@njw.name>2019-01-25 09:55:55 +0000
commit30c088b90e7b6a25d93cbdad7564ff063e62afd3 (patch)
treec47d7bc086a076cfe5e702628c4e5e3b1eab1aa5 /lib/hocr/hocr.go
parent1c17766952bdcd6f7d31d0fa1a2e504b1aa4f14a (diff)
Reorganisation and cleanup
Diffstat (limited to 'lib/hocr/hocr.go')
-rw-r--r--lib/hocr/hocr.go79
1 files changed, 79 insertions, 0 deletions
diff --git a/lib/hocr/hocr.go b/lib/hocr/hocr.go
new file mode 100644
index 0000000..0d10819
--- /dev/null
+++ b/lib/hocr/hocr.go
@@ -0,0 +1,79 @@
+package hocr
+
+import (
+ "encoding/xml"
+ "regexp"
+ "strconv"
+ "strings"
+)
+
+type Hocr struct {
+ Lines []OcrLine `xml:"body>div>div>p>span"`
+}
+
+type OcrLine struct {
+ Class string `xml:"class,attr"`
+ Id string `xml:"id,attr"`
+ Title string `xml:"title,attr"`
+ Words []OcrWord `xml:"span"`
+ Text string `xml:",chardata"`
+}
+
+type OcrWord struct {
+ Class string `xml:"class,attr"`
+ Id string `xml:"id,attr"`
+ Title string `xml:"title,attr"`
+ Chars []OcrChar `xml:"span"`
+ Text string `xml:",chardata"`
+}
+
+type OcrChar struct {
+ Class string `xml:"class,attr"`
+ Id string `xml:"id,attr"`
+ Title string `xml:"title,attr"`
+ Chars []OcrChar `xml:"span"`
+ Text string `xml:",chardata"`
+}
+
+// Returns the confidence for a word based on its x_wconf value
+func wordConf(s string) (float64, error) {
+ re, err := regexp.Compile(`x_wconf ([0-9.]+)`)
+ if err != nil {
+ return 0.0, err
+ }
+ conf := re.FindStringSubmatch(s)
+ return strconv.ParseFloat(conf[1], 64)
+}
+
+func boxCoords(s string) ([4]int, error) {
+ var coords [4]int
+ re, err := regexp.Compile(`bbox ([0-9]+) ([0-9]+) ([0-9]+) ([0-9]+)`)
+ if err != nil {
+ return coords, err
+ }
+ coordstr := re.FindStringSubmatch(s)
+ for i := range coords {
+ c, err := strconv.Atoi(coordstr[i+1])
+ if err != nil {
+ return coords, err
+ }
+ coords[i] = c
+ }
+ return coords, nil
+}
+
+func noText(s string) bool {
+ t := strings.Trim(s, " \n")
+ return len(t) == 0
+}
+
+func Parse(b []byte) (Hocr, error) {
+ var hocr Hocr
+
+ err := xml.Unmarshal(b, &hocr)
+ if err != nil {
+ return hocr, err
+ }
+
+ return hocr, nil
+}