renun/koreader
1
0
Fork 0
mirror of https://github.com/koreader/koreader.git synced 2025-08-10 00:52:38 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

add document language option in config dialog

Browse Source 
Document language setting is used for better text extraction and OCR.
This commit is contained in:
chrox
2013-07-20 02:46:51 +08:00
parent 167e6ca0aa
commit 1808c7e5c3
7 changed files with 51 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
frontend/document/koptinterface.lua
View File

@@ -62,6 +62,7 @@ function KoptInterface:createContext(doc, pageno, bbox)
kc:setDefectSize(doc.configurable.defect_size)
kc:setLineSpacing(doc.configurable.line_spacing)
kc:setWordSpacing(doc.configurable.word_spacing)
kc:setLanguage(doc.configurable.doc_language)
kc:setBBox(bbox.x0, bbox.y0, bbox.x1, bbox.y1)
if Dbg.is_on then kc:setDebug() end
return kc
@@ -129,6 +130,13 @@ function KoptInterface:getReflewTextBoxes(doc, pageno)
local cached = Cache:check(kctx_hash)
if cached then
local kc = self:waitForContext(cached.kctx)
--kc:setDebug()
local lang = doc.configurable.doc_language
if lang == "chi_sim" or lang == "chi_tra" or
lang == "jpn" or lang == "kor" then
kc:setCJKChar()
end
kc:setLanguage(lang)
local fullwidth, fullheight = kc:getPageDim()
local boxes = kc:getWordBoxes(0, 0, fullwidth, fullheight)
Cache:insert(hash, CacheItem:new{ rfpgboxes = boxes })
@@ -146,6 +154,12 @@ function KoptInterface:getTextBoxes(doc, pageno)
local kc_hash = "kctx|"..doc.file.."|"..pageno
local kc = KOPTContext.new()
kc:setDebug()
local lang = doc.configurable.doc_language
if lang == "chi_sim" or lang == "chi_tra" or
lang == "jpn" or lang == "kor" then
kc:setCJKChar()
end
kc:setLanguage(lang)
local page = doc._document:openPage(pageno)
page:getPagePix(kc)
local fullwidth, fullheight = kc:getPageDim()
@@ -167,6 +181,7 @@ function KoptInterface:getReflewOCRWord(doc, pageno, rect)
local dummy = KOPTContext.new()
Cache:insert(ocrengine, OCREngine:new{ ocrengine = dummy })
end
self.ocr_lang = doc.configurable.doc_language
local bbox = doc:getPageBBox(pageno)
local context_hash = self:getContextHash(doc, pageno, bbox)
local hash = "rfocrword|"..context_hash..rect.x..rect.y..rect.w..rect.h
@@ -197,6 +212,7 @@ function KoptInterface:getOCRWord(doc, pageno, rect)
local dummy = KOPTContext.new()
Cache:insert(ocrengine, OCREngine:new{ ocrengine = dummy })
end
self.ocr_lang = doc.configurable.doc_language
local hash = "ocrword|"..doc.file.."|"..pageno..rect.x..rect.y..rect.w..rect.h
local cached = Cache:check(hash)
if not cached then