def GetPageText(pg):
dl = fitz.DisplayList()
dv = fitz.Device(dl)
pg.run(dv, fitz.Identity)
ts = fitz.TextSheet()
tp = fitz.TextPage()
rect = pg.bound()
dl.run(fitz.Device(ts, tp), fitz.Identity, rect)
return tp.extractJSON()
ln = ln.next # we create a transformation matrix here zoom = int(sys.argv[3]) rotate = int(sys.argv[4]) trans = fitz.Matrix(zoom / 100.0, zoom / 100.0).preRotate(rotate) ''' here we introduce the display list, which provides caching-mechanisms to reduce parsing of a page. first, we need to create a display list hand it over to a list device and then populate the display list by running the page through that device, with transformation applied ''' mediabox = page.rect dl = fitz.DisplayList(mediabox) dv = fitz.Device(dl) page.run(dv, trans) # get the page size, and then apply the transformation rect = mediabox.transform(trans) # create a pixmap with RGB as colorspace and bounded by irect pm = fitz.Pixmap(fitz.Colorspace(fitz.CS_RGB), rect.round()) # clear it with 0xff white pm.clearWith(0xff) # fitz.Device(pm, None) is a device for drawing # we run the display list above through this drawing device # with area provided dl.run(fitz.Device(pm, None), fitz.Identity, rect)
#!/usr/bin/env python
import fitz
f = "sdw_2015_06.pdf"
d = fitz.Document(f)
seiten = d.pageCount
for seite in range(seiten):
print "=============== processing page", seite, " ==============="
pg = d.loadPage(seite)
dl = fitz.DisplayList()
print "ok: dl = fitz.DisplayList()"
dv = fitz.Device(dl)
print "ok: dv = fitz.Device(dl)"
pg.run(dv, fitz.Identity)
print "ok: pg.run(dv, fitz.Identity)"
ts = fitz.TextSheet()
print "ok: ts = fitz.TextSheet()"
tp = fitz.TextPage()
print "ok: tp = fitz.TextPage()"
rect = pg.bound()
dl.run(fitz.Device(ts, tp), fitz.Identity, rect)
print "ok: dl.run(fitz.Device(ts, tp), fitz.Identity, rect)"