From 787d63fc5d13c6250bd33da5a8e1eadbe86188cd Mon Sep 17 00:00:00 2001
From: Nick White <git@njw.name>
Date: Tue, 8 Oct 2019 15:37:07 +0100
Subject: Continue separating the repository; remove all but preproc, and move
 integralimg package under it

---
 lib/hocr/hocr.go | 129 -------------------------------------------------------
 1 file changed, 129 deletions(-)
 delete mode 100644 lib/hocr/hocr.go

(limited to 'lib/hocr/hocr.go')

diff --git a/lib/hocr/hocr.go b/lib/hocr/hocr.go
deleted file mode 100644
index dcd0494..0000000
--- a/lib/hocr/hocr.go
+++ /dev/null
@@ -1,129 +0,0 @@
-package hocr
-
-import (
-	"encoding/xml"
-	"errors"
-	"io/ioutil"
-	"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
-}
-
-func GetText(hocrfn string) (string, error) {
-	var s string
-
-	file, err := ioutil.ReadFile(hocrfn)
-	if err != nil {
-		return s, err
-	}
-
-	h, err := Parse(file)
-	if err != nil {
-		return s, err
-	}
-
-
-	for _, l := range h.Lines {
-		s += getLineText(l)
-	}
-	return s, nil
-}
-
-func GetAvgConf(hocrfn string) (float64, error) {
-	file, err := ioutil.ReadFile(hocrfn)
-	if err != nil {
-		return 0, err
-	}
-
-	h, err := Parse(file)
-	if err != nil {
-		return 0, err
-	}
-
-	var total, num float64
-	for _, l := range h.Lines {
-		for _, w := range l.Words {
-			c, err := wordConf(w.Title)
-			if err != nil {
-				return 0, err
-			}
-			total += c
-			num++
-		}
-	}
-	if num == 0 {
-		return 0, errors.New("No words found")
-	}
-	return total / num, nil
-}
-- 
cgit v1.2.1-24-ge1ad