extract detail from pdf

oml/item/handlers.py

@@ -93,6 +93,24 @@ class EpubHandler(OMLHandler):
                 self.set_header('Content-Type', content_type)
                 self.write(z.read(filename))
 
+class CropHandler(OMLHandler):
+
+    def get(self, id, page, left, top, right, bottom):
+        from media.pdf import crop
+        with db.session():
+            item = Item.get(id)
+            path = item.get_path()
+            print(path, page, left, top, right, bottom)
+            data = crop(path, page, left, top, right, bottom)
+            if data:
+                self.set_header('Content-Type', 'image/jpeg')
+                self.set_header('Content-Length', str(len(data)))
+                self.write(data)
+                return
+        self.set_status(404)
+        return
+
+
 def serve_static(handler, path, mimetype, include_body=True, disposition=None):
     handler.set_header('Content-Type', mimetype)
     size = os.stat(path).st_size

oml/media/pdf.py

@@ -10,6 +10,7 @@ from glob import glob
 from datetime import datetime
 
 from PyPDF2 import PdfFileReader
+from PIL import Image
 import ox
 
 import settings
@@ -24,13 +25,13 @@ def cover(pdf):
     else:
         return page(pdf, 1)
 
-def ql_cover(pdf):
+def ql_cover(pdf, size=1024):
     tmp = tempfile.mkdtemp()
     cmd = [
         'qlmanage',
         '-t',
         '-s',
-        '1024',
+        str(size),
         '-o',
         tmp,
         pdf
@@ -48,7 +49,7 @@ def ql_cover(pdf):
     shutil.rmtree(tmp)
     return data
 
-def page(pdf, page):
+def page(pdf, page, size=1024):
     tmp = tempfile.mkdtemp()
     if sys.platform == 'win32':
         pdf = get_short_path_name(pdf)
@@ -57,7 +58,7 @@ def page(pdf, page):
         pdf,
         '-jpeg',
         '-f', str(page), '-l', str(page),
-        '-scale-to', '1024', '-cropbox',
+        '-scale-to', str(size), '-cropbox',
         os.path.join(tmp, 'page')
     ]
     if sys.platform == 'win32':
@@ -79,6 +80,47 @@ def page(pdf, page):
     shutil.rmtree(tmp)
     return data
 
+def crop(pdf, page, left, top, right, bottom):
+    size = 2048
+    tmp = tempfile.mkdtemp()
+    if sys.platform == 'win32':
+        pdf = get_short_path_name(pdf)
+    cmd = [
+        'pdftocairo',
+        pdf,
+        '-jpeg',
+        '-f', str(page), '-l', str(page),
+        '-scale-to', str(size), '-cropbox',
+        os.path.join(tmp, 'page')
+    ]
+    if sys.platform == 'win32':
+        startupinfo = subprocess.STARTUPINFO()
+        startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW
+        startupinfo.wShowWindow = subprocess.SW_HIDE
+        p = subprocess.Popen(cmd, close_fds=True, startupinfo=startupinfo)
+    else:
+        p = subprocess.Popen(cmd, close_fds=True)
+    p.wait()
+    image = glob('%s/*' % tmp)
+    if image:
+        image = image[0]
+        crop = [int(p) for p in (left, top, right, bottom)]
+        print(crop)
+        img = Image.open(image).crop(crop)
+        img.save(image)
+        with open(image, 'rb') as fd:
+            data = fd.read()
+    else:
+        logger.debug('pdftocairo %s %s', pdf, ' '.join(cmd))
+        data = None
+    shutil.rmtree(tmp)
+    return data
+    
+
+
+
+
+
 '''
 def page(pdf, page):
     image = tempfile.mkstemp('.jpg')[1]
@@ -281,3 +323,4 @@ def extract_isbn(text):
     isbns = find_isbns(text)
     if isbns:
         return isbns[0]
+

oml/server.py

@@ -7,11 +7,13 @@ import signal
 import time
 
 from tornado.ioloop import IOLoop
-from tornado.web import StaticFileHandler, Application
+import tornado.web
+from tornado.web import Application
 
 from cache import Cache
 from item.handlers import EpubHandler, ReaderHandler, FileHandler
 from item.handlers import OMLHandler, UploadHandler
+from item.handlers import CropHandler
 from item.icons import IconHandler
 import db
 import node.server
@@ -29,6 +31,12 @@ import logging
 
 logger = logging.getLogger(__name__)
 
+class StaticFileHandler(tornado.web.StaticFileHandler):
+    def get_content_type(self):
+        if self.request.path.split('?')[0].endswith('.mjs'):
+            return 'application/javascript'
+        return super().get_content_type()
+
 class MainHandler(OMLHandler):
 
     def get(self, path):
@@ -126,6 +134,7 @@ def run():
         (r'/(.*?)/get/', FileHandler, {
             'attachment': True
         }),
+        (r'/(.*)/2048p(\d*),(\d*),(\d*),(\d*),(\d*).jpg', CropHandler),
         (r'/(.*)/(cover|preview)(\d*).jpg', IconHandler),
     ]
     handlers = common_handlers + [